Professional Documents
Culture Documents
PRESS gam
Principles of |
Fermentation
Technology
Wie
P F STANBURY
~ and A WHITAKER
Pe tN Tn, ie
Principles of
Fermentation Technology
An inspection copy of any book published in the Pergamon International Library will gladly be
sent to academic staff without obligation for their consideration for course adoption or
recommendation. Copies may be retained for a period of 60 days from receipt and returned if not
suitable. When a particular title is adopted or recommended for adoption for class use and the
recommendation results in a sale of 12 or more copies, the inspection copy may be retained with
our compliments. The Publishers will be pleased to receive suggestions for revised editions and new
titles to be published in this important International Library.
Other Pergamon publications of related interest
Books
CHEMICAL ENGINEERING
Coulson,J.M. & Richardson,J. F.
FERMENTATION PRODUCTS
(Advances in Biotechnology, Volumes 1-3)
Moo-Young, M.
BIOTECHNOLOGY ADVANCES
PETER F. STANBURY
B.Sce-WESe.,, DiL.G:
and
ALLAN WHITAKER
M'Ses PLD HAR CSPD LE
Division of Biological and Environmental Sciences
Hatfield Polytechnic, Hertfordshire, UK
PERGAMON
©PRESS
OXFORD: NEW YORK : TORONTO: SYDNEY - PARIS - FRANKFURT
U.K. Pergamon Press Ltd., Headington Hill Hall,
Oxford OX3 OBW, England
U.S.A Pergamon Press Inc., Maxwell House, Fairview Park,
Elmsford, New York 10523, U.S.A.
CANADA Pergamon Press Canada Ltd., Suite 104,
150 Consumers Road, Willowdale, Ontario M2J 1P9, Canada
AUSTRALIA Pergamon Press (Aust.) Pty. Ltd., P.O. Box 544,
Potts Point, N.S.W. 2011, Australia
FRANCE Pergamon Press SARL, 24 rue des Ecoles,
75240 Paris, Cedex 05, France
FEDERAL REPUBLIC Pergamon Press GmbH, Hammerweg 6,
OF GERMANY D-6242 Kronberg-Taunus, Federal Republic of Germany
Stanbury, Peter F.
Principles of fermentation technology.
(Pergamon international library ofscience,
technology, engineering, and social studies)
Includes index.
1. Industrial microbiology.
I. Whitaker, Allan.
II. Title. III. Series.
QR53.S73 1984 660’.62 83-25665
British Library Cataloguing in Publication Data
Stanbury, Peter F.
Principles of fermentation technology.
—(Pergamon international library)
1. Industrial microbiology
I. Title II. Whitaker, Allan
660'.62 QR53
ISBN 0-08-024400-9
ISBN 0-08-024406-8 Pbk
— net Pap
cathe el paces fawoe Ai hd ' © Lf am
one) 0 atatecad
ean
©e ada
: @ 4 7 : mm ci n é : ne _
fae) | tig ‘ A2reow
ot as :7 ——
a 7 Q : hin ee
Sa ; - 2 ere tl mane 5 dae on
- = : ¥ " 7 hd Me a | Hoa
5 STERILIZATION 91
Introduction 9]
Medium sterilization 92
The design of batch sterilization processes 96
Calculation of the Del factor during heating and cooling 98
Calculation of the holding time at constant temperature (121°C) 98
Richards’ rapid method for the design of sterilization cycles 99
The scale up of batch sterilization processes 99
Methods of batch sterilization 100
The design of continuous sterilization processes 100
Sterilization of the fermenter 103
Sterilization of the feeds 104
Sterilization of air 104
The theory of fibrous filters 104
Filter design 106
References 107
xiii
Contents
220
1] EFFLUENT TREATMENT
220
Introduction
220
Dissolved oxygen concentration as an indicator of water quality
221
Factory surveys
pape
The strengths of fermentation effluents
Treatment and disposal of effluents
222
225
Disposal
223
Seas and rivers
223
Lagoons
Spray irrigation
Pasa
Well disposal 224
Disposal of effluents to sewers 224
Treatment processes 224
Physical treatment 224
Chemical treatment 22
Biological treatment Zao
Aerobic processes 225
Trickling filters yo55)
Towers 226
Rotating discs 226
Rotating drums 227
Activated sludge 227
Anaerobic treatment 228
Anaerobic digestion 228
Anaerobic filters 228
By-products 228
Distilleries 229
Breweries 229
Amino acid wastes 229
References 229
Index 247
XIV
CHAPTER 1
An Introduction to
Fermentation Processes
THE term ‘fermentation’ is derived from the Latin large scale for very many years and was the first
verb, fervere, to boil, thus describing the appearance ‘industrial’ process for the production of amicrobial
of the action of yeast on extracts of fruit or malted metabolite. Thus, industrial microbiologists have
grain. The boiling appearance is due to the produc- extended the term fermentation to describe any
tion of carbon dioxide bubbles caused by the process for the production of product by the mass
anaerobic catabolism of the sugars present in the culture of amicro-organism. Brewing and the pro-
extract. However, fermentation has come to have duction of organic solvents may be described as
_ different meanings to biochemists and to industrial fermentations in both senses of the word but the
microbiologists. Its biochemical meaning relates to description ofan aerobic process as a fermentation is
the generation of energy by the catabolism of organic obviously using the term in the broader, micro-
compounds, whereas its meaning in industrial biological, context and it is in this sense that the
microbiology tends to be much broader. term is used in this book.
The catabolism of sugars is an oxidative process
which results in the production of reduced pyridine THE RANGE OF FERMENTATION
nucleotides which must be reoxidized for the process PROCESSES
to continue. Under aerobic conditions, reoxidation
of reduced pyridine nucleotide occurs by electron There are four major groups of commercially
transfer, via the cytochrome system, with oxygen important fermentations:
acting as the terminal electron acceptor. However, (i) Those that produce microbial cells (or
under anaerobic conditions, reduced pyridine biomass) as the product. '
nucleotide oxidation is coupled with the reduction (11) Those that produce microbial enzymes.
of an organic compound, which is often a subsequent (111) Those that produce microbial metabolites.
product of the catabolic pathway. In the case of the (iv) Those that modify a compound which is
action of yeast on fruit or grain extracts, NADH is added to the fermentation—the trans-
regenerated by the reduction of pyruvic acid to formation processes.
ethanol. Different microbial taxa are capable of The historical development ofthese processes will
reducing pyruvate to a wide range of end products, be considered in a later section ofthis chapter, but it
as illustrated in Fig. 1.1. Thus, the term fermenta- is first necessary to include a brief description ofthe
tion has been used in a strict biochemical sense to four groups.
mean an energy-generating process in which organic
compounds act as both electron donors and terminal Microbial biomass
electron acceptors. ‘
The production of alcohol by the action of yeast The commercial production of microbial biomass
on malt or fruit extracts has been carried out on a may be subdivided into two major processes; the
Principles of Fermentation Technology
Acetate
ACETYL CoA
co ~|-2
oH oo 2H
Acrylate ae Lactate Acetaldehyde ae Ethanol
Acetolactate
Oxaloacetate
2H | FLLCO
B Acetoin
Malate
F,G \- 2H
saa 420
Fumarate a 2,3- Butanediol
|‘ Se ee,
Formate + Acetyl CoA + Hz + CO2
Ss
D Ethanol
Succinate ; }
Ho CO,
co I
Acetate
Acetoacety! CoA
Propionate 60, i GN
!so-Propanol Butanol
production ofyeast to be used in the baking industry microbial enzymes have the enormous advantage of
and the production of microbial cells to be used as being able to be produced in large quantities by
human or animal food (single-cell protein). Bakers’ established fermentation techniques. Also, it is
yeast has been produced on a large scale since the infinitely easier to improve the productivity of a
early 1900s and yeast was produced as human food microbial system compared with a plant or animal
in Germany during the First World War. However, one. The uses to which microbial enzymes have
it was not until the 1960s that the production of been put are summarized in Table 1.1, from which
microbial biomass as a source of food protein was it may be seen that the majority of applications are
explored to any great depth. Asa result of this work, in the food and related industries. Enzyme produc-
reviewed briefly in Chapter 2, a few very large-scale tion is closely controlled in micro-organisms and in
continuous processes have been established in order to improve productivity these controls may
recent years using a variety ‘of different carbon have to be exploited or modified. Such control
sources.. systems as induction may be exploited by including
Microbial enzymes inducers in the medium, whereas feedback repres-
sion may be removed by mutation and selection
Enzymes have been produced commercially from techniques. These aspects are discussed in Chapters
plant, animal and microbial sources. However, 4 and 3, respectively.
2
An Introduction to Fermentation Processes
grow at a constant, maximum rate and this period is TaBLe 1.2. Some primary products of microbial metabolism and their
commercial significance
known as the log, or exponential, phase. Eventually,
growth ceases and the cells enter the so-called sta- Primary
tionary phase. After a further period of time the metabolite Commercial significance
viable cell number declines as the culture enters the Ethanol ‘Active ingredient’ in alcoholic beverages.
death phase. As well as this kinetic description of Used as a motor-car fuel when blended
growth, the behaviour of a culture may also be with petroleum
Citric acid Various uses in the food industry
described according to the products which it pro-
Acetone and Solvents
duces during the various stages of the growth curve. butanol
During the log phase of growth the products pro- Glutamicacid Flavour enhancer
Lysine Feed supplement
duced are essential to the growth of the cells and
Nucleotides Flavour enhancers
include amino acids, nucleotides, proteins, nucleic Polysaccharides Applications in the food industry
acids, lipids, carbohydrates, etc. These products are Enhanced oil recovery
Vitamins Feed supplements
referred to as the primary products of metabolism —
and the phase in which they are produced (equiva-
lent to the log, or exponential, phase) as the
trophophase (Bu’Lock et al., 1965). elaborated from the intermediates and products of
Many products of primary metabolism are of primary metabolism. Although the primary bio-
considerable economic importance and have been synthetic routes illustrated in Fig. 1.2 are common
produced by fermentation, as illustrated in Table to the vast majority of micro-organisms, each secon-
1.2. The synthesis of primary metabolites by wild- dary product would be synthesized by only a very
type micro-organisms is such that their production few different microbial species. Thus, Fig. 1.2 is a
is sufficient to meet the requirements of the representation of the secondary metabolism exhi-
organism. Thus, it is the task of the industrial bited by a very wide range of different micro-
microbiologist to modify the wild-type organism organisms. Also, not all micro-organisms undergo
and to provide cultural conditions to improve the secondary metabolism—it is common amongst the
productivity of these compounds. This aspect is filamentous bacteria and fungi and the sporing bac-
considered in Chapter 3. teria but it is not found, for example, in the Entero-
During the stationary phase some microbial cul- bacteriaceae. Thus, the taxonomic distribution of
tures synthesize compounds which are not produced secondary metabolism is quite different from that of
during the trophophase and which do not appear to primary metabolism. The physiological role of sec-
have any obvious function in cell metabolism. These ondary metabolism in the producer cells has been
compounds are referred to as the secondary pro- the subject of considerable debate, but the import-
ducts of metabolism and the phase in which they are ance of these metabolites to the fermentation indus-
produced (equivalent to the stationary phase) as the try is the affects they have on organisms other than
idiophase (Bu’Lock et al., 1965). It is important to those that produce them. Many secondary metabo-
realize that secondary metabolism may occur in lites havé antimicrobial activity, others are specific
continuous Cultures at low growth rates and, thus, it enzyme inhibitors, some are growth promoters and
is a property of slow-growing, as well as non-grow- many have pharmacological properties. Thus, the
ing, cells. When it is appreciated that micro- products of secondary metabolism have formed the
organisms grow at relatively low growth rates in basis of anumber of fermentation processes. As is
their natural environments, it is tempting to suggest the case for primary metabolites, wild-type micro-
that it is the idiophase state that prevails in nature organisms tend to produce only low concentrations
rather than the trophophase, which may be more a of secondary metabolites, their synthesis being con-
property of micro-organisms in culture. The inter- trolled by induction, catabolite repression and feed-
relationships between primary and_ secondary back systems. The techniques which have been
metabolism are illustrated in Fig. 1.2, from which it developed to improve secondary metabolite produc-
may be seen that secondary metabolites tend to be tion are considered in Chapters 3 and 4.
4
An Introduction to Fermentation Processes
CO,
Malonate (C3) = Acetate (C2) ——® Mevalonate (Cg) Isopenteny! pyrophosphate (C,)
oR De
Secondary TCA cycle
metabolites fn
Oxaloacetate (C4) Citrate (Cg)
Fic. 1.2. The interrelationships between primary and secondary metabolism. Primary catabolic routes are shown in heavy lines and
secondary products are italicized (Turner, 1972).
5
Principles of Fermentation Technology
TaBLe 1.3. The stages in the chronological development of the fermentation industry
Bakers’ yeast, Steel vessels of pH electrodes Batch and fed- Virtually nil Virtually Pure cultures
glycerol, citric up to 2000 with off-line batch nil used
acid, lactic hectalitres for control. systems
acid and acetone Temperature
2 acetone/ butanol. Air control
1900-1940 —_butanol spargers used
for bakers’
yeast.
Mechanical
stirring used in
small vessels
Rusates Les De ee es A a ee eee ———————————————————
Penicillin, Mechanically SterilizablepH - Batchand fed- Very im- Becomes Mutation and
streptomycin, aerated and oxygen batch portant common selection
other vessels, electrodes. Use common. programmes
antibiotics, operated of control Continuous essential
3 gibberelin, aseptically— loops which culture intro-
1940-date amino acids, true were later duced in
nucleotides, fermenters computerized brewing and
transforma- for some
tions, enzymes primary
metabolites
Single-cell Pressure cycle Use of Continuous Very im- Very im- Genetic
protein using and pressure computer- culture + portant portant engineering
hydrocarbon jet vessels linked control medium of productic
4 and other developed to loops recycle strains
1960—date feedstocks overcome gas
and heat-
exchange
problems
‘Foreign’ com- Fermenters Control and May be Very im- Very im- Introduction of
pounds, not developed in sensors batch, portant portant foreign genes
5 normally pro- stages 3 and 4 developed in fed-batch or into micro-
1979-date duced by stages 3 and 4 continuous bial hosts
microbial cells, using genetic
e.g. insulin engineering
interferon techniques
was Pasteur who eventually convinced the scientific developed sophisticated techniques for the produc-
world ofthe obligatory role of these micro-organisms tion of starter cultures. However, use of pure cul-
in the process. During the late 1800s Hansen started tures did not spread to the British ale breweries and
his pioneer work at the Carlsberg brewery and it is true to say that many ofthe small, traditional,
developed methods for isolating and propagating ale-producing breweries still use mixed yeast cul-
single yeast cells to produce pure cultures and tures at the present time.
6
An Introduction to Fermentation Processes
Vinegar was originally produced by leaving wine fermentation was susceptible to contamination in
in shallow bowls or partially filled barrels which was the early stages by aerobic bacteria, and by acid
slowly oxidized to vinegar by the development of a producing anaerobic ones once anaerobic condi-
natural flora. The appreciation of the importance of tions had been established in the later stages of the
air in the process eventually led to the development process. The fermenters employed were vertical
of the ‘generator’ which consisted ofa vessel packed cylinders with hemispherical tops and bottoms con-
with an inert material (such as coke, charcoal and structed from mild steel. They could be steam
various types of wood shavings) over which the wine sterilized under pressure and were constructed to
or beer was allowed to trickle. The vinegar generator minimize the possibility of contamination. Two-
may be considered as the first ‘aerobic’ fermenter to thousand-hectalitre fermenters were commissioned
be developed. By the late 1800s to early 1900s the which presented the problems of inoculum develop-
initial medium was being pasteurized and inocu- ment and the maintenance of aseptic conditions
lated with 10% good vinegar to make it acidic, and during the inoculation procedure. The techniques
therefore resistant to contamination, as well as pro- developed for the production of these organic sol-
viding a good inoculum (Bioletti, 1921). Thus, by vents were a major advance in fermentation tech-
the beginning ofthe twentieth century the concepts nology and paved the way for the successful intro-
of process control were well established in both the duction of aseptic aerobic processes in the 1940s.
brewing and vinegar industries. The third stage in the development ofthe fermen-
Between the years 1900 and 1940 the main new tation industry arose as a result of the wartime need
products were yeast biomass, glycerol, citric acid, to produce penicillin in submerged culture under
lactic acid and acetone and butanol. Probably the aseptic conditions. The production of penicillin is
most important advances during this period were an aerobic process which is very vulnerable to con-
the developments in the bakers’ yeast and organic tamination. Thus, although the knowledge gained
solvent fermentations. The production of bakers’ from the solvent fermentations was exceptionally
yeast is an aerobic process and it was soon recog- valuable, the problems of sparging a culture with
nized that the rapid growth of yeast cells in a rich large volumes of sterile air and mixing a highly
wort led to oxygen depletion in the medium which, viscous broth had to be overcome. Also, unlike the
in turn, resulted in ethanol production at the solvent fermentations, penicillin was synthesized in
expense of biomass formation. The problem was very small quantities by the initial isolates and this
minimized by restricting the initial wort concentra- resulted in the establishment of strain-improvement
tion such that the growth of the cells was limited by programmes which became a dominant feature of
the availability of the carbon source rather than the industry in subsequent years. Process develop-
oxygen. Subsequent growth of the culture was then ment was also aided by the introduction of pilot-
controlled by adding further wort in small amounts. plant facilities which enabled the testing of new
This technique is now called fed-batch culture and techniques on a semi-production scale. The develop-
is widely used in the fermentation industry to avoid ment of a large-scale extraction process for the
conditions of oxygen limitation. The aeration of recovery of penicillin was another major advance at
these early yeast cultures was also improved by the this time. This was probably the stage when the
introduction of air through sparging tubes which most significant changes in fermentation technology
could also be steam cleaned (de Becze and Liebman, took place resulting in the establishment of many
1944). new processes over the period, including other anti-
The development of the acetone—butanol ferment- biotics, vitamins, gibberellin, amino acids, enzymes
ation during the First World War by the pioneering and steroid transformations.
efforts of Weizmann led to the establishment ofthe In the early 1960s the decisions of a number of
first truly aseptic fermentation. All the processes multi-national companies to investigate the produc-
discussed so far could be conducted with relatively tion of microbial biomass as a source offeed protein
little contamination provided that a good inoculum led to a number of developments which may be
was used and reasonable standards of hygiene regarded as the fourth stage in the progress of the
employed. However, the anaerobic acetone—butanol industry. The largest mechanically stirred fermen-
7d
Principles of Fermentation Technology
tation vessels developed during stage 3 were in the production of insulin and interferon. Thus, the
range 80,000 to 150,000 dm’°. However, the rela- range of potential commercial microbial products
tively low selling price of microbial protein necessi- has been extended outside the spectrum of com-
tated its production in much larger quantities than pounds normally produced by micro-organisms.
other fermentation products. Also, hydrocarbons Furthermore, the productivity of conventional
were considered as potential carbon sources which microbial products may be increased using the
would result in increased oxygen demands by these techniques of genetic manipulation. It appears to be
fermentations (see Chapters 4 and 9). These widely believed that these methods of genetic
requirements led to the development ofthe pressure manipulation will revolutionize the fermentation
jet and pressure cycle fermenters which eliminated industry and give rise to a large number of new
the need for mechanical stirring (see Chapter 7). processes. However, it should be remembered that
Another feature of these potential processes was the exploitation of these new advances depends on
that they would have to be operated continuously if © the technology of mass cell culture which has
they were to be economic. At this time batch and evolved from the yeast and solvent fermentations
fed-batch processes were common in the industry via the antibiotic fermentations to the large-scale
but the technique of growing an organism continu- continuous biomass processes.
ously by adding fresh medium to the vessel and
removing culture fluid had only been applied to a
THE COMPONENT PARTS OF A
very limited extent on a large scale. The brewing
FERMENTATION PROCESS
industry was also investigating the potential of con-
tinuous culture at this time but its application in
that industry was short-lived. Several companies Regardless of the type of fermentation (with the
have persevered in the biomass field and a few possible exception of some transformation pro-
commercial processes are coming to fruition, of cesses) an established process may be divided into
which the most developed is that of ICI which six basic component parts:
utilizes a continuous 3,000,000-dm? pressure cycle
(i) The formulation of medium to be used in
fermenter (Smith, 1981). The operation of extremely
culturing the process organism during the
large continuous fermenters for time periods in
development of the inoculum and in the
excess of 100 days presents a considerable aseptic
production fermenter.
operation problem, far greater than that faced by
(ii) The sterilization of the medium, fermenters
the antibiotic industry in the 1940s. The aseptic
and ancillary equipment.
operation of fermenters of this type has been
(i111) The production of an active, pure culture in
achieved as a result of the high standards of fer-
sufficient quantity to inoculate the pro-
menter construction, the continuous sterilization of
duction vessel.
feed streams and the utilization of computer systems
(iv) The growth of the organism in the pro-
to control the sterilization and operation cycles,
duction fermenter under optimum condi-
thus minimizing the possibility of human error.
tions for product formation.
The fifth stage in the progress of the industry has
(v) The extraction of the product and its puri-
been initiated by the developments in the in vitro
fication.
genetic manipulation of organisms, commonly
(vi) The disposal of effluents produced by the
known as genetic engineering. The techniques of
process.
genetic engineering not only allow the transfer of
genes between unrelated organisms but also enable The interrelationships between the six component
the extremely precise alteration of the genome of an parts are illustrated in Fig. 1.3. However, one must
organism. Thus, microbial cells may be endowed also visualize the research and development pro-
with the ability to produce compounds normally gramme which is designed to gradually improve the
associated with higher cells. These higher cell pro- overall efficiency of the fermentation. Before a fer-
ducts synthesized by microbial cells may form the mentation process is established a producer
basis of new fermentation processes, for example the organism has to be isolated, modified such that it
8
An Introduction to Fermentation Processes
INOCULUM DEVELOPMENT
: Biomass
Production fermenter |
U Serio ye oot
i LE
Stock Shake Seed > Culture fluid > SEPARATION
culture flask fermenter |
yA Cell-free
MEDIUM STERILIZATION PRODUCT a supernatant
t | EXTRACTION
MEDIUM
DIU F ORMULATION
LAT ea
PRODUCT “
EFFLUENT
PURIFICATION TREATMENT
Medium raw materials
PRODUCT PACKAGING
produces the desired product in commercial quan- cesses are discussed in Chapter 12. Throughout the
tities, its cultural requirements determined and the book examples are drawn from a very wide range of
' plant designed accordingly. Also, the extraction fermentations to illustrate the applications of the
process has to be established. The development techniques being discussed but it has not been
programme would involve the continual improve; attempted to give detailed considerations ofspecific
ment ofthe process organism, the culture media an processes as this is well covered elsewhere, for
the extraction process. * ‘example in the Economic Microbiology series,
The subsequent chapters ofthis book consider the edited by Rose. We hope that the approach adopted
basic principles underlying the component parts of in this book will give the reader an understanding of
a fermentation. Chapter 2 considers growth, com- the basic principles underlying the techniques used
prehension of which is crucial to understanding for the large-scale production of microbial products.
many aspects of the process, other than simply the
_growth ofthe organism in the production fermenter.
The isolation and improvement of commercial REFERENCES
strains is considered in Chapter 3 and the design of
media in Chapter 4. The sterilization of the medium, Broverti, F. T. (1921) The manufacture of vinegar. In Microbiol-
fermenters and air is considered in Chapter 5 and 0g), pp. 636-648 (Editor Marshall, C. E.). Churchill, Lon-
don.
the techniques for the development ofinocula are
Borne,J. T. P. (1982) Enzyme production. In Prescott and Dunne’s
discussed in Chapter 6. Chapters 7, 8 and 9 consider Industrial Microbiology, 4th edition, pp. 634-708 (Editor
the fermenter as an environment for the culture of Reed, G.). MacMillan, New York.
Bu’Locx,J. D., Hamiiron, D., Hutme, M. A., Powe tt, A. J.,
micro-organisms; Chapter 7 considers the design
SHEPHERD, D., SMALLEY, H. M. and Situ, G. N. (1965)
and construction offermenters, Chapter 8 discusses Metabolic development and secondary biosynthesis in
the instrumentation involved in monitoring and Penicillium urticae. Can.J.Micro. 11, 765-778.
Corran, H. S. (1975) A History of Brewing. David and Charles,
maintaining a controlled environment in a fer- Newton Abbott.
menter while the provision of oxygen to a culture is Dawes, I. and Larcg, P. J. (1982) Class | reactions: Supply of
investigated in Chapter 9. The recovery offermenta- carbon skeletons. In Biochemistry of Bacterial Growth, pp.
125-158 (Editors Mandelstam, J., McQuillen, K. and
tion products is dealt with in Chapter 10 and the Dawes, I.). Blackwell, Oxford.
processes is covered in Chap-
disposal ofeffluents of bE Becze, and Liesmann, A. J. (1944) Aeration in the production
ter 1]. Finally, the economics of fermentation pro- of compressed yeast. Jnd. Eng. Chem. 36, 882-890.
Principles of Fermentation Technology
Hastincs, J. J. H. (1971) Development of the fermentation Rose, A. H. (1979) Economic Microbiology, Vol. 3, Secondary Products
industries in Great Britain. Adv. Appl. Microbiol. 16, 1-45. of Metabolism. Academic Press, London.
Kresuicu, K. (1982) Practical applications of microbial transfor- Rose, A. H. (1979b) Economic Microbiology, Vol. 4, Microbial
mations in the synthesis of natural compounds or analogues. Biomass. Academic Press, London.
In Overproduction of Microbial Metabolites, pp. 345-362 Smiru, S. R. L. (1981) Some aspects of ICI’s single cell protein
(Editors Krumphanzl, V., Sikyta, B. and Vanek, Z.). process. In Microbial Growth on C, Compounds, pp. 342-348
Academic Press, London. (Editor Dalton, H.). Heyden, London.
Rose, A. H. (1977) Economic Microbiology, Vol. 1, Alcoholic Bever- Turner, W. B. (1971) Fungal Metabolites. Academic Press, Lon-
ages. Academic Press, London. don.
Rosg, A. H. (1978) Economic Microbiology, Vol. 2, Primary Products
of Metabolism. Academic Press, London.
10
CAC? Taba
11
Principles of Fermentation Technology
TABLE 2.2. Some representative values of K, for a range of micro-organisms and substrates
tration has declined to a very low level. Thus, the subsequent to the exponential phase in which secon-
deceleration phase for such a culture would be short. dary metabolites were synthesized. However, it is
However, if the organism has a low affinity for the now obvious that the culture conditions may be
substrate (a high K, value) the growth rate will be manipulated to induce secondary metabolism dur-
deleteriously affected at a relatively high substrate ing logarithmic growth, for example by the use of
concentration. Thus, the deceleration phase for such carbon sources which support a reduced maximum
a culture would be relatively long. Typical values of growth rate. This aspect is considered in more detail
K, for a range of organisms and substrates are shown in Chapter 4.
in Table 2.2, from which it may be seen that such Pirt (1975) has discussed the kinetics of product
values are usually very small. formation by microbial cultures in terms of growth-
The stationary phase in batch culture is that point linked products and non-growth-linked products.
where the growth rate has declined to zero. How- Growth-linked may be considered equivalent to
ever, as Bull (1974) pointed out, the stationary primary metabolites which are synthesized by grow-
phase is a misnomer in terms of the physiology of the ing cells and non-growth-linked may be considered
organism, as the population is still metabolically equivalent to secondary metabolites. The formation
‘active during this phase and produces products of agrowth-linked product may be described by the
called secondary metabolites, which are not pro- equation:
duced during the exponential phase. Bull suggested
that this phase be termed the maximum population cae pX (2.6)
phase. The metabolic activity of the stationary
phase has been recognized in the physiological where fp is the concentration of product
descriptions of microbial growth presented by Bor- and q, is the specific rate of product formation.
row ef al. (1961) and Bu’Lock (1965). Borrow et al. Also, the product formation is related to biomass
investigated the biosynthesis of gibberellic acid by production by the equation:
Gibberella fujikuroi and divided the growth of the
organism into several phases: dp _
ae ~ F p/x (2.7)
Combining equations (2.6) and (2.9): The net change in cell concentration over a time
period may be expressed as:
Gp =e Ypj" be (2.10)
by the culture is balanced by the loss ofcells from the where x is the steady-state cell concentration
vessel. The flow of medium into the vessel is related and _ § is the steady-state residual substrate con-
to the volume ofthe vessel by the term dilution rate, centration.
D, defined as: Equation (2.18) explains the mechanism whereby
F D controls w. Cell growth will result in the depletion
DE V (2.11) of substrate until the residual substrate concentra-
tion equals the substrate concentration which will
where F is the flow rate support the growth rate equal to the dilution rate. If
and Vis the volume. substrate is depleted below the level that-will sup-
Thus, D is expressed in the units h7!. port the relevant growth rate, cells will be washed
14
Microbial Growth Kinetics
7)
=oS)
5
5
”
oO eh
ss
o-=
° Ze
oc
“ o
roi)
pes
a 0
+r Cc
ov oO
55 ve
&% Oo So
concentration
>§ TH
oh
TO o
oO Sade
2
Ww
23 n
Steady
biomass
concentration
state Steady
substrate
residual
state
Dilution rate
Dilution rate
Fic. 2.4. The effect of dilution rate on the steady-state biomass
and residual substrate concentrations in a chemostat culture of a Steady state biomass concentration
micro-organism with a low K, value for the limiting substrate, Steady state residual substrate concentration
compared with the initial substrate concentration.
Steady-state biomass concentration. Fic. 2.5. The effect of dilution rate on the steady-state biomass
— — — Steady-state residual substrate concentration. and residual substrate concentrations in a chemostat of a micro-
organism with a high K, value for the limiting substrate, com-
pared with the initial substrate concentration.
out at a rate greater than they are being produced The kinetic characteristics of an organism (and,
and s will increase resulting in an increase in the therefore, its behaviour in a chemostat) are
growth rate and the balance would be restored. described by the numerical values of the constants
Thus, the system is a self-balancing one. Y, Minax and K,. The value of Y affects the steady-state
The continuous culture system which has been biomass concentration; the value of 1,,,, affects the
described is termed a chemostat because the growth maximum dilution rate that may be employed and
rate of the culture is controlled by its chemical the value of K, affects the residual substrate concen-
environment, that is, the availability of one limiting tration (and, hence, the biomass concentration) and
component in the medium. An alternative type of also the maximum dilution rate that may be used.
continuous culture is termed the turbidostat where Fig. 2.4 illustrates the continuous culture behaviour
the concentration of cells in the culture is kept ofa hypothetical bacterium with a low K, value for
constant by controlling the flow of medium such the limiting substrate, compared with the initial
that the turbidity of the culture is kept within cer- limiting substrate concentration. With increasing
tain, narrow limits. This may be achieved by dilution rate, the residual substrate concentration
monitoring the biomass with a photoelectric cell increases only slightly until D approaches p,,,.. when
and feeding the signal to a pump supplying medium § increases significantly. The dilution rate at which x
to the culture such that the pump is switched on if equals zero (that is, the cells have been washed out
the biomass exceeds the set point and is switched off of the system) is termed the critical dilution rate
if the biomass falls below the set point. Systems (Di) and is given by the equation:
other than turbidity may be used to monitor the
=a Bmax SR P
biomass concentration, such as CO, concentation, Crit, ae ‘e ae Ke (2.19)
in which case it would be more correct to term the
culture a biostat. The chemostat is the more Thus, D,,;, is affected by the constants wmax and K,
commonly used system because it has the advantage and the variable, Sp, the larger Sp the closer is D.,;,
over the biostat of not requiring complex control tO Pinax- However, m,,,, May never be achieved in a
systems to maintain a steady state. However, the simple chemostat as substrate limited conditions
biostat may be advantageous in continuous enrich- always prevail.
ment culture in avoiding the total washout of the Figure 2.5 illustrates the continuous culture
culture in its early stages and this aspect is discussed behaviour of a hypothetical bacterium with a high
in Chapter 3. K, for the limiting substrate compared with the
15
Principles of Fermentation Technology
Medium inlet
Culture effluent
Culture effluent
concentration to effluent
collection or to
subsequent stages
Steady
substrate
residual
state
Steady
concentration
biomass
state
Dilution rate
effluent stream is less concentrated than in where R),,,<) is the output of the culture in terms of
the vessel. cell concentration per hour,
(ii) Subjecting the effluent stream to a biomass Xmax 1S the maximum cell concentration
separation process, such as sedimentation or achieved at stationary phase,
centrifugation, and returning a portion of oe is the initial cell concentration at inocu-
the concentrated biomass to the growth ves- lation,
sel. F is the time during which the organism
The reader is referred to Pirt (1975) for a full kinetic grows at Mma,
description of both types of feedback systems, but andj is the time during which the organism
the net effects of feedback may be summarized as: is not growing at M,,,, and includes the
(i) Biomass concentration in the vessel is lag phase, the deceleration phase and
increased. the periods of batching, sterilizing and
(ii) The increased biomass concentration results harvesting.
in a decrease in the residual substrate com- The productivity of a continuous culture may be
pared with a simple chemostat. represented as:
(ii) The maximum output of biomass and pro-
ducts is increased. Resa = Dx{1= a (2.21)
de
(iv) The critical dilution rate (the dilution rate at
which washout occurs) is increased. where R,,,,, is the output of the culture in terms of
Biomass feedback may be advantageous in cell concentration per hour,
obtaining better substrate utilization of very dilute t;; is the time period prior to the establish-
substrate streams, for example in brewing and ment of a steady state and includes
‘effluent treatment. Also, the feedback of biomass vessel preparation, sterilization and
may improve stability in a system where mixed operation in batch culture prior to con-
substrates of varying concentration are used, as in tinuous operation,
effluent treatment. and — if is the time period during which steady-
state conditions prevail.
The term Dx increases with increasing dilution rate
The application of continuous until it reaches a maximum value, after which any
culture in industrial processes further increase in D results in a decrease in Dx, as
illustrated in Fig. 2.8. Thus, maximum productivity
of biomass may be achieved by the use of the
The fermentation industry appears to have been
dilution rate giving the highest value of Dx.
fairly reluctant to adopt continuous culture as a
production technique despite claims that it has
many advantages over batch culture. Hospodka
(1966) considered that continuous culture was
superior to batch culture in productivity, uniformity
of operation and ease in automation but was more
susceptible to contamination.
PRODUCTIVITY
The productivity of a culture system may be
described as the output of biomass per unit time of (Dx)
productivity
Biomass / D crit
the fermentation. Thus, the productivity of abatch
culture may be represented as: Dilution rate
, max oe
: 0 2.20 Fic. 2.8. The effect of dilution rate on biomass productivity in
Weestth ~ steady-state continuous culture.
ti + bij
17
Principles of Fermentation Technology
The output of abatch fermentation described by concentration and toxin concentration should
equation (2.20) is an average over the period of the remain constant throughout the fermentation.
fermentation and, because the rate of biomass pro- Thus, once the culture is established the demands of
duction is dependent on initial biomass [(dx/dt) = the fermentation, in terms ofprocess control, should
ux], the vast proportion of the biomass is produced be constant. In a batch fermentation, the demands
towards the end of the fermentation. Thus, produc- of the culture vary during the fermentation—at the
tivity in batch culture is at its maximum only beginning, the oxygen demand is low but towards
towards the end of the process. For a continuous the end of the fermentation the demand is high due
culture operating at the optimum dilution rate, to the high biomass and the increased viscosity of
under steady-state conditions, the productivity will the broth. Also, the amount of cooling required will
be constant and always maximum. Thus, the pro- increase during the process, as will the degree of pH
ductivity of the continuous system must be greater control. In a continuous process oxygen demand,
than the batch. A continuous system may be oper- cooling requirements and pH control should remain
ated for a very long time period (several weeks or constant. Thus, the use of continuous culture should
months) so that the negative contribution of the allow for the easier introduction of process automa-
unproductive time,t;;;, to productivity would be min- tion.
imal. However, a batch culture may be operated for A batch process requires periods of intensive
only a limited time period so that the negative labour during medium preparation, sterilization,
contribution of the time, ¢,;,, would be very signifi- batching and harvesting but relatively little during
cant, especially when it is remembered that the the fermentation itself. However, a continuous pro-
batch culture would have to be re-established many cess results in a more constant labour demand in
times during the time-course ofa continuous run. _that medium is supplied continuously sterilized (see
Thus, the superior productivity of biomass by a Chapter 5), the product is continuously extracted
continuous culture, compared with a batch culture, and the relative time spent on equipment prepara-
is due to the maintenance of maximum output tion and sterilization is very small.
conditions throughout the fermentation and the
insignificance of the non-productive period SUSCEPTIBILITY OF CONTINUOUS
associated with a long-running continuous process. FERMENTATIONS TO CONTAMINATION
Theoretically, a fermentation to produce a product, Because the duration of a continuous fermenta-
other than biomass, should also be more productive tion is very much longer than a batch one there is a
in continuous culture than in batch because a con- greater probability of a contaminating organism
tinuous culture may be operated at the dilution rate entering the continuous process. However, it should
which maintains product output at its maximum, be remembered that continuous culture is a highly
whereas in batch culture product formation may be selective system and that the contaminant would
a transient phenomenon during the fermentation. have to be better suited to the environment than the
Thus, continuous fermentations may be operated at culture organism before it could become established.
lower volumes than corresponding batch processes Thus, the contaminant would have to grow at a
and still achieve the same productivities, resulting specific growth rate equal to, or greater than, the
in lower construction, installation and maintenance dilution rate operating in the chemostat. If the
costs. This aspect is considered in more detail in contaminant were to grow at the same rate as the
Chapter 12. dilution rate then it would coexist with the process
organism; whereas if the contaminant were to grow
UNIFORMITY OF OPERATION AND at a higher rate then it would displace the process
EASE OF AUTOMATION organism from the vessel.
As discussed previously, a continuous culture is a Contamination is a problem of fermenter design,
self-regulating system in which the growth rate of construction and operation which should be over-
the culture will adjust to the dilution rate imposed come and, therefore, its importance should not be
upon it. Thus, a steady state may be achieved where overemphasized. However, the highly selective
cell concentration, substrate concentration, product nature of continuous culture may give rise to prob-
18
Microbial Growth Kinetics
lems of strain degeneration, termed by Calcott (iv) The use of the technique to generate basic
(1981) as ‘contamination from inside’ which may information about the process organism
not be solved by the design of more ‘secure’ fermen- which may be used to improve the industrial
ters. A continuous culture operates for long time batch or fed-batch process.
periods and, hence, the probability of mutations
occurring is greater than in batch culture. Any CONTINUOUS CULTURE AND BREWING
mutant that is better adapted to the cultural condi- The brewing industry in the United Kingdom has
tions may displace the original ‘parental’ type, as had a relatively brief ‘courtship’ with continuous
would a better adapted contaminant. In a process culture. Two types of continuous brewing have been
for the production of biomass this property of con- used:
tinuous culture may be considered an advantage
(i) The cascade or multistage system.
because the system would constantly select for the
(11) The tower system.
most efficient variant. However, if the process
involves the production of a metabolite by a highly Hough et al. (1976) described the cascade system
mutated or genetically engineered organism the utilized at Watneys’ Mortlake brewery in London.
selective nature of continuous culture could be a The process utilized three vessels, the first two for
major disadvantage. fermentation and the third for separation of the
Industrial strains have been modified to produce yeast biomass. The specific gravity of the wort was
metabolites in concentrations far in excess of that reduced from 1040 to 1019 in the first vessel and
which they require. Thus, most industrial strains from 1019 to 1011 in the second vessel. The residence
may be described as extremely inefficient because time for the system was 15 to 20 hours, using worts
they produce certain metabolites in such great in the specific gravity range of 1035 to 1040, and it
excess. Variants of the industrial strains which do could be run continuously for 3 months. However, it
not produce as much product would be more ef- is believed that the system was abandoned due to
ficient in continuous culture than the parental types problems of excessive biomass production. This
and, should such variants arise by back mutation or process appears to have been widely used in New
loss of genetic elements, displacement of the paren- Zealand with greater success (Kirsop, 1982) but
tal types would occur. The genetic modification ofa apparently newer installations are of the batch type.
producing organism to stabilize its performance in A typical tower fermenter for brewing is illus-
continuous culture is discussed in Chapter 3. trated in Fig. 2.9. The system is partially closed in
In view ofthe considerable number of advantages that relatively little yeast leaves the fermenter due to
of continuous-culture over batch-culture systems it the highly flocculent nature of the strains employed.
seems surprising that the technique has found rela- Wort is introduced into the base of the tower and
tively little application in the fermentation industry. passes through a porous plug of yeast. As the wort
Pirt (1975) claimed that this is a reflection of the rises through the vessel it is progressively fermented
highly empirical nature of the original batch pro- and leaves the fermenter via a yeast-separation
cesses and lack of knowledge of the crucial control- zone, which is twice the diameter of the rest of the
ling conditions. However, in recent years the adop- tower. Hough (1976) described the protocol
tion of fed-batch culture has resulted in the manifes- employed in the establishment of a yeast plug in the
tation of many of the advantages of continuous tower and the subsequent operation conditions.
processes using the original batch-culture equip- Prior to the fermentation, the tower is thoroughly
ment. The fed-batch technique is considered in cleaned and steam sterilized, aseptic operation
more detail in a later section of this chapter. The being more important for the continuous process
application of continuous culture may be sum- than the batch one. The vessel is filled partially with
marized as: sterile wort and inoculated with a laboratory cul-
ture. The initial stages are designed to encourage
(i) Application in brewing. high biomass production by the periodical addition
(ii) Application in biomass production. of wort over about a 9-day period. A porous plug of
(iii) Application in strain isolation and selection. yeast develops at the base of the tower. The flow rate
19
_ PFT-C
Principles of Fermentation Technology
Sight glass
CONTINUOUS CULTURE AND BIOMASS
PRODUCTION
Microbial biomass which is produced for human
or animal consumption is referred to as single cell
Fic. 2.9. A schematic representation ofa tower fermenter for the protein (SCP). Although yeast was produced as
brewing ofbeer (Royston, 1966). food on a large scale in Germany during the First
World War (Laskin, 1977) the concept of utilizing
of the wort is then gradually increased over a further microbial biomass as food was not thoroughly inves-
9 to 12 days, by which time an approximate steady tigated until the 1960s. Since the 1960s, a large
state may be achieved. Following the establishment number ofindustrial companies have explored the
of a high biomass in the fermenter the system is potential of producing SCP from a wide range of
operated such that the wort is converted to beer with carbon sources. Almost without exception, these
the formation of approximately the same amount of investigations have been based on the use of continu-
yeast as would be produced in the batch process. ous Culture as the growth technique.
The beer produced during the 3-week start-up Continuous culture is the ideal method for the
period is usually below specification and would production of microbial biomass. The superior pro-
have to be blended with high-quality beer. Thus, ductivity of the technique, compared with batch
more than 3 months’ continuous operation is neces- culture, may be fully exploited and the problem of
sary to compensate for the initial losses of the pro- strain degeneration is not as significant as in the
cess. production of microbial metabolites. The selective
The major advantage of the continuous-tower pressure in the chemostat would tend to work to the
process was that the wort residence time could be advantage of the industrialist producing SCP, in
reduced from about | week to 4 to 8 hours as that the most efficient strain of the organism would
compared with the batch system. However, the be selected. The development of SCP processes has
development of the cylindro-conical vessel (initially generated considerable research into large-scale
described by Nathan in the 1930s, but not intro- chemostat design and the behaviour ofthe produc-
duced until the 1970s) led to the shortening of the tion organism in these very large vessels. Many
batch fermentation time to approximately 48 hours. ‘novel’ fermenters have been designed for SCP pro-
20
Microbial Growth Kinetics
cesses and these are considered in more detail in improvement of micro-organisms. The use of con-
Chapter 7. tinuous culture in this context is considered in Chap-
A very wide range of carbon sources have been ter 3, from which it may be seen that continuous
investigated for the production of SCP. Whey has enrichment culture offers considerable advantages
been used as a carbon source for biomass production over batch enrichment techniques and that continu-
since the 1940s (Meyrath and Bayer, 1979) and ous culture may be used very successfully to select
modern processes utilize continuous-culture sys- strains producing higher yields of certain microbial
tems, often based on the air-lift principle (see Chap- enzymes.
ter 7). Such fermentations have been shown to be
economic in that they provide a high-grade feed APPLICATION OF CONTINUOUS CULTURE
product, whilst removing an, otherwise, trouble- IN THE GENERATION OF BASIC
some waste product of the cheese industry (Meyrath INFORMATION ON COMMERCIAL ORGANISMS
and Bayer, 1979). Cellulose has been investigated The chemostat is an extremely good tool for the
extensively as a potential carbon source for SCP investigation ofthe basic physiology and biochemis-
production and this work has been reviewed by try of a micro-organism. The major experimental
Clayton et al. (1979). The major difficulty associated advantages of the chemostat over batch culture are:
with the use of cellulose as a substrate is its recalcit- (i) A steady state is obtained.
rant nature. Other carbohydrates have formed the (ii) The growth rate of the culture may be
basis of several biomass processes and one of the limited by the dilution rate.
most recent to reach fruition is the process of Rank, (11) The growth rate of the organism is ulti-
Hovis, Macdougall involving the production offun- mately limited by the limiting substrate con-
gal biomass for human consumption from wheat, centration so it is possible to investigate the
_ maize or potato starch. The advantage of fungal effect of different limiting substrates on the
biomass is that it may be processed to give a textured behaviour of the organism.
protein (Forage and Righelato, 1979). An enormous
amount of research has been conducted into the use Thus, experiments may be performed under more
of hydrocarbons as sources of carbon for biomass stringent conditions in continuous culture than in
processes; the hydrocarbons investigated being batch culture and information obtained from such
methane, methanol and n-alkanes. A large number experiments may be used to improve the industrial
of commercial firms were involved in this research batch or fed-batch process.
field but very few have created viable, commercial
processes based on SCP production from hydrocar- FED-BATCH CULTURE
bons because of the economic difficulties involved.
At the start of this research, hydrocarbons were Yoshida et al. (1973) introduced the term fed-
relatively cheap but, following the 1973 Middle East batch culture to describe batch cultures which are
War, oil prices escalated and transformed the fed continuously, or sequentially, with medium,
economic basis of biomass production from pet- without the removal of culture fluid. Thus, the
roleum sources. Imperial Chemical Industries have volume of the culture increases with time. The
been successful in developing a commercial process following description of the kinetics of fed-batch
for production of bacterial biomass from methanol culture is based on the discussion by Pirt (1975).
at an annual rate of 54,000 to 70,000 tonnes. The Consider a batch culture in which growth is
process utilizes a novel air-lift fermenter, of 3000 m° limited by the concentration of one substrate; the
capacity, and is the first commercial process to biomass at any point in time will be described by the
produce SCP from methanol (King, 1982). equation:
X= te AY (Ss) (2.22)
APPLICATION OF CONTINUOUS CULTURE
IN STRAIN ISOLATION AND IMPROVEMENT where x, is the biomass concentration after time, ¢
The fiercely’ selective nature of the chemostat hours,
makes it an excellent tool for the isolation and and_ x, is the inoculum concentration.
21
Principles of Fermentation Technology
Xmax
~ YS, (2.23) 0.20 0.8
a
If, at the time when x« = x,,,x, a medium feed is E 0.15 0.6 —
32) |
started, such that the dilution rate is less than ,,,.; Ix =
virtually all the substrate will be consumed as fast as 2 a
& 0.10 0.4
it enters the culture, thus: ln
X
FSR Y (2.24) 0.05 0.2
22
Microbial Growth Kinetics
Peppler, 1973). Thus, the organism was grown in an organic acids when the respiration rate exceeds the
initially weak medium to which additional medium aeration capacity of the fermenter. Both systems
was added at a rate less than that at which the appear to work well in controlling feed rates during
organism could use it. Thus, this process fulfils the the rapid-growth phase.
criteria stipulated in Pirt’s (1975) kinetic description During the production phase of the penicillin
for the establishment ofa quasi-steady state, that is, fermentation the feed rates utilized should limit the
a substrate limited culture and the use ofa feed rate growth rate and oxygen consumption such that a
equivalent to a dilution rate less than m,,,,. It is now high rate of penicillin synthesis is achieved, and
recognized that bakers’ yeast is very sensitive to free sufficient dissolved oxygen is available in the
glucose and respiratory activity may be repressed at medium. The control factor in this phase is normally
a concentration of about 5 wg dm”? (Crabtree, dissolved oxygen because pH is less responsive to
1929). Thus, a high glucose concentration represses the effect of dissolved oxygen on penicillin synthesis
respiratory activity as well as giving rise to a high than on growth. As the fed-batch process proceeds,
biomass, the oxygen demand of which cannot be then the total biomass, viscosity and oxygen demand
met. In modern fed-batch processes for yeast pro- increase until, eventually, the fermentation is oxy-
duction the feed of molasses is under strict control gen limited. However, limitation may be delayed by
based on the automatic measurement of traces of reducing the feed rate as the fermentation progresses
ethanol in the exhaust gas of the fermenter. and this may be achieved by the use of computer-
Although such systems may result in low growth controlled systems. In some fed-batch fermentations
rates, the biomass yield is near the theoretical a high rate of penicillin production may still be
obtainable (Fiechter, 1982). occurring when the fermenter is full and the only
The penicillin fermentation provides an excellent way to maintain productivity would be to remove
example ofthe use of feed systems in the production some of the batch and refeed. Queener and Swartz
ofa secondary metabolite. The fermentation may be discussed this approach and raised the following
divided into two phases—the ‘rapid-growth’ phase, points:
during which the culture grows at Myx, and the
(i) Unused nutrient is lost in the withdrawal,
‘slow-growth’ or ‘production’ phase. Glucose feeds
and cells are discarded, which are then un-
may be used to control the metabolism of the
available to contribute to further synthesis.
organism during both phases. During the rapid-
(ii) Periodic dilution and removal results in a
growth phase an excess of glucose causes an
more dilute, and larger broth volume to be
accumulation of acid and a biomass oxygen demand
extracted.
greater than the aeration capacity of the fermenter,
(iii) Toxic products may be excreted into the
whereas glucose starvation may result in the organic
medium at a rate greater than they are
nitrogen in the medium being used as a carbon
diluted and withdrawn, resulting in an
source resulting in a high pH and inadequate
accumulation of toxins and the eventual
biomass formation (Queener and Swartz, 1979).
toxin limitation of the fermentation.
The accumulation of hexose may be prevented by
(iv) Nutrients other than those incorporated into
the use of a slowly hydrolysed carbohydrate such as
the feed medium may become limiting.
lactose in simple batch culture (Matelova, 1976).
(v) The enrichment of non-producing variants
However, according to Queener and Swartz, consid-
may occur.
erable increases in productivity have been achieved
by the use of computer controlled feeding ofglucose Thus, the use of the ‘withdraw and fill’ technique
during the rapid growth phase, such that the dissol- must be considered in the light of the above limita-
ved oxygen or pH is maintained within certain tions. The use of the technique to control the viscos-
limits. Both control parameters essentially measure ity of a culture is considered in Chapter 9.
the same activity in that both oxygen consumption Many enzymes are subject to catabolite repres-
and pH will fall when glucose is in excess, due to an sion, where enzyme synthesis is prevented by the
increased respiration rate and the accumulation of presence of rapidly utilized carbon sources
23
Principles of Fermentation Technology
(Aunstrup et al., 1979). It is obvious that this Catcort, P. H. (1981) The construction and operation of con-
tinuous cultures. In Continuous Culture of Cells, vol. 1,
phenomenon must be avoided in enzyme fermenta-
pp. 13-26 (Editor Calcott, P. H.). CRC Press, Boca Raton.
tions and fed-batch culture is the major technique Crayton, D., Cattramn, L. D. and CLemmMer, J. E. (1979)
used to achieve this. Concentrated medium is fed to Biomass from cellulosic materials. In Economic Microbiology,
4. Microbial Biomass, pp. 208-270 (Editor Rose, A. H.).
the culture such that the carbon source does not
Academic Press, London.
reach the threshold for catabolite repression. For Crastree, H. G. (1929) Observations on the carbohydrate
example, Waki e¢ al. (1982) controlled the produc- metabolism of tumours. Biochem.J.23, 536-545.
tion of cellulase by Trichoderma reesei in fed-batch DostraLek, M., Haccstrom, C. and Mou, N. (1972) Optimisa-
tion of biomass production from methanol. Fermentation
culture utilizing CO, production as the control fac- Technology Today, Proc. TV Int. Fermentation Symp., pp.
tor. 497-511 (Editor, Terui, G.).
Eacon, R. G. (1961) Pseudomonas natriegens, a marine bacterium
with a generation time ofless than ten minutes./. Bact. 83,
USE OF FED-BATCH CULTURE IN THE 736-737.
AVOIDANCE OF TOXIC EFFECTS OF FrecuTer, A. (1982) Regulatory aspects in yeast metabolism. In
MEDIUM COMPONENTS Advances in Biotechnology, 1. Scientific and Engineering Principles,
Fed-batch culture may be used in circumstances pp. 261-267 (Editors Moo-Young, M., Robinson, C. W. and
Vezina, C.). Pergamon Press, Toronto.
where a component of the medium is toxic to the Forace, A. J. and RicHeLato, R. C. (1979) Biomass from
fermentation organism. The penicillin fermentation carbohydrates. In Economic Microbiology, +. Microbial Biomass,
provides an excellent example of this situation in pp. 289-314 (Editor Rose, A. H.). Academic Press, London.
Harrison, D. E. F. (1973) Studies on the affinity
of methanol and
that sodium phenylacetate, a precursor of the methane utilizing bacteria for their carbon substrates. /.
penicillin molecule provided during the fermenta- Appl. Bact. 36, 309-314.
tion, is toxic to Pencillium chrysogenum. Thus, sodium Harte, M. J. and Wess, F. C. (1967) Utilisation of mixed sugars
in continuous fermentations. Biotech. Bioeng. 9, 205-221.
phenylacetate is kept at sub-inhibitory levels by ‘Hospopka, J. (1966) Industrial applications of continuous fer-
feeding it to the culture either in small increments or mentation. In Theoretical and Methodological Basis ofContinuous
as a continuous feed (Queener and Swartz, 1979). Culture of Microorganisms, pp. 493-645 (Editors Malek, I. and
Fencl, Z.) Academic Press, New York.
The feed system also minimizes the hydroxylation of Hoven, J. S., Keevir, C. W., Maric, V., PHtuiskirk, G. and
the precursor and, thus, maximizes its incorporation Younc, T. W. (1976) Continuous culture in brewing. In
into the product. In the glutamic acid fermentation, Continuous Culture, 6. Applications and New Fields, pp. 226-238
(Editors Dean, A.C. R., Ellwood, D. C., Evans, C. G. T. and
slow feeding has made it possible to use substrates Melling, J.). Ellis Horwood, Chichester.
which would be inhibitory to both growth and Kirsop, B. H. (1982) Developments in beer fermentation. Topics
glutamate production if used in high concentrations in Enzyme and Fermentation Biotechnology, 6, 79-131 (Editor
Wiseman, A.). Ellis Horwood, Chichester.
in traditional batch processes. Kine, P. P. (1982) Biotechnology: an industrial view.J. Chem.
Tech. Biotechnol. 32, 2-8.
Lasky, A. I. (1977) Single cell protein. Ann. ReportsonFermentation
REFERENCES Processes, 1, 151-180 (Editor Perlman, D.). Academic Press,
New York.
Mateova, V. (1976) Utilization of carbon sources during
AunstrupP, K., ANDRESEN, O., Fatcu, E. A. and Nretsen, T. K. penicillin biosynthesis. Folia Microbiologica, 21, 208-209.
(1979) Production of microbial enzymes. In Microdial Meyratu, J. and Bayer, K. (1979) Biomass from whey. In
Technology, vol. | (2nd edition), pp. 282-309 (Editors Pep- Economic Microbiology, 4. Microbial Biomass, pp. 208-270.
pler, H. J.and Perlman, D.). Academic Press, New York. (Editor Rose, A. H.). Academic Press, New York.
Bazin, M. J. (1981) Theory of continuous culture. In Continuous Monon, J. (1942) Reserches sur les Croissances des Cultures Bacter-
Culture ofCells, vol. 1, pp. 27-62 (Editor Calcott, P. H.). CRC tennes, 2nd edition. Hermann and Cie, Paris.
Press, Boca Raton. Pirt, S. J. (1975) Principles of Microbe and Cell Cultivation.
Borrow, A., JEFFERYS, E. G., Kesset, R. H. J., Luoyp, E. C., Blackwell, Oxford.
Lioyp, P. B. and Nixon, I. S. (1961) Metabolism of Gib- Pirt, S. J. and Kurowsxi, W. M. (1970) An extension of the
berella fujikuroi in stirred culture. Can._J. Micro. '7, 227-276. theory of the chemostat with feedback of organisms. Its
Butt, A. T. (1974) Microbial growth. In Companion to Biochemistry, experimental realisation with a yeast culture.J.Gen. Micro.
Selected Topics for Further Study, pp. 415-442 (Editors Bull, A. 63, 357-366.
T., Lagnado, J. R., Thomas, J. O. and Tiron, K. F.). Queener, S. and Swartz, R. (1979) Penicillins: Biosynthetic
Longman, London. and semisynthetic. In Economic Microbiology, 3. Secondary
Bu’Lock,J. D., Hamitron, D., Hume, M. A., PowELt, ASIs Products of Metabolism, pp. 35-123 (Editor Rose, A. H.).
SHEPHERD, D., SMattey, H. M. and Smirn, G. N. (1965) Academic Press, London.
Metabolic development and secondary biosynthesis in Reep, G. and Pepper, H. J. (1973) Yeast Technology, pp. 664668.
Penicillium urticae. Can.J.Micro. 11, 765-778. Avi, Westport.
24
Microbial Growth Kinetics
Royston, M. G. (1966) Tower fermentation of beer. Process Young, M., Robinson, C. W. and Vezina, C.). Pergamon
Biochem. 1(4), 215-221. Press, Toronto.
SHEHATA, T. E. and Narr, A. G. (1971)J. Bact. 107, 210-225. Wuirtaker, A, (1980) Fed-batch culture. Proc. Biochem. 15(4),
Trinci, A. P. J. (1969) A kinetic study of growth of Aspergillus 10-15.
nidulans and other fungi.J.Gen. Micro. 57, 11-24. Yosuipa, F., YAMANE, T. and Nakamoto, K. (1973) Fed-batch
Waki, T., Luca, K. and Icuikawa, K. (1982) Production of hydrocarbon fermentations with colloidal emulsion feed.
cellulase in fed-batch culture. In Advances in Biotechnology, 1. Biotech. Bioeng. 15, 257-270.
Scientific and Engineering Principles, pp. 359-364 (Editors Moo-
25
CHAPTER 3
TaBLe 3.1. Major culture collections then be applied to the assessment of natural isolates.
—_—_—————————————— eee
27
Principles of Fermentation Technology
p= Bmax$ é
al
Keefer an
28
The Isolation, Preservation and Improvement of Industrial Micro-organisms
(1972) and Harrison (1978) have discussed such rates. The two-stage chemostat described by Harri-
procedures for the isolation of organisms to be used son (1978) is very similar to Johnson’s (1972) proce-
for biomass production. Johnson emphasized the dure. The first stage of the system was used as a
importance of using the carbon source to be continuous inoculum for the second stage and con-
employed in the subsequent commercial process as sisted of a large bottle containing a basic medium
the sole source of organic carbon in the enrichment inoculated with a soil infusion. Continuous inocula-
medium and that the medium should be carbon tion was employed until an increasing absorbance
limited. The inclusion of other organic carbon was observed in the second stage.
sources, such as vitamins or yeast extract, may The use of continuous enrichment culture has
result in the isolation of strains adapted to using frequently resulted in the selection of stable, mixed
these, rather than the principal carbon source, as cultures presumably based on some form of symbio-
energy sources. The isolation of an organism capa- tic relationship. It is extremely unlikely that such
ble of growth on a simple medium should also form mixed, stable systems could be isolated by batch
the basis of a cheaper commercial process and techniques so that the adoption of continuous
should be more resistant to contamination—a major enrichment may result in the development ofnovel,
consideration in the design of acommercial continu- mixed culture fermentations. Harrison et al. (1972)
ous process. The use of as high as possible an isolated a mixed culture using methane as the car-
isolation temperature should also result in the isola- bon source in a continuous enrichment and
tion ofa strain presenting minimal cooling problems demonstrated that the mixture contained one
in the subsequent process. methylotroph and a number of non-methylotrophic
The main difficulty in using a continuous-enrich- symbionts. The performance of the methylotroph in
ment process is the washout of the inoculum before pure culture was invariably poorer than the mixture,
an adapted culture is established. Johnson (1972) in terms of growth rate, yield and culture stability.
suggested the isolation process should be started in Continuous enrichment has also been used for the
batch culture using a 20% inoculum and as soon as isolation of organisms to be used in systems other
growth is observed, the culture should be transferred than biomass production; Rowley and Bull (1977)
to fresh medium and the subsequent purification used the technique to isolate an Arthrobacter sp.
and stabilization of the enrichment performed in producing a yeast lysing enzyme complex. Hsieh
continuous culture. The continuous system should and Munnecke (1972) used the technique to
be periodically inoculated with soil or sewage which acclimatize soil isolates to the detoxification of high
may not only be a source of potential isolates but concentrations of parathion.
should also ensure that the dominant flora is
extremely resistant to contamination. THE USE OF SOLIDIFIED MEDIA
Harrison (1978) proposed two solutions to the Solidified media have been used for the isolation
problem of early washout in continuous isolation of certain enzyme producers and these techniques
processes: The first uses a turbidostat and the sec- usually involve the use ofa selective medium incor-
ond uses a two-stage chemostat. A turbidostat is a porating the substrate of the enzyme which encour-
continuous-flow system provided with a photoelec- ages the growth ofthe producing types. Aunstrup ef
tric cell to determine the turbidity of the culture and al. (1972) isolated species of Bacillus producing
maintain the turbidity between set points by initiat- alkaline proteases. Soils of various pHs were used as
ing or terminating the addition of medium. Thus, the initial inoculum and, to a certain extent, the
washout is avoided as the medium supply will be number of producers isolated correlated with the
switched off if the biomass falls below the lower fixed alkalinity of the soil sample. The soil samples were
point. The use ofa turbidostat will result in selection pasteurized to eliminate non-sporulating organisms
on the basis of maximum specific growth rate as it and then spread on to the surface of agar media at
operates at high levels of limiting substrate. Thus, pH 9-10, containing a dispersion of an insoluble
although the use of the turbidostat removes the protein. Colonies which produced a clear zone due
danger of washout it is not as flexible a system as the to the digestion of the insoluble protein were taken
chemostat which may be used at a range ofdilution to be alkaline protease producers. The size of the
29
Principles of Fermentation Technology
clearing zone could not be used quantitatively to the inhibition ofthe test organism(s). Alternatively,
select high producers as there was not an absolute the microbial isolate may be grown in liquid culture
correlation between the size of the clearing zone and and the cell-free broth tested for antimicrobial
the production of alkaline protease in submerged activity. The use of liquid medium in initial screen-
culture. However, this example demonstrates the ing tests has the advantage of avoiding any
importance of choice ofstarting material, the use of anomalies in the relation of results of agar tests to
a selective force in the isolation and the incorpora- submerged culture studies but takes up considerably
tion ofa preliminary diagnostic test, albeit of limited more space and facilities. Nisbet (1982) gives a
use. detailed account of the use of solid and liquid media
screening tests.
Zahner (1978) discussed the use oftest organisms
Isolation methods not utilizing selection of
to detect the production of new antibiotics. Antibio-
the desired characteristic
tics with confined action spectra may be isolated by
using a combination of test organisms such as Bacil-
The production of some products does not give lus subtilis and Streptomyces viridochromogenes or Clos-
the producing organism any selective advantage tridium pasteurianum—such a test may identify anti-
which may be exploited in an isolation procedure. biotics with a low activity against B. subtilis and a
Therefore, organisms are often isolated on a random high activity against the other test organism. Anti-
basis and techniques developed to rapidly screen biotics selected by this means may be new because
the isolated organisms for the desired characteristic. they would not have been detected by the more
Examples include the production of antibiotics, common tests using B. subtilis alone. Zahner cites
growth promoters and the synthesis of large concen- ‘the kirromycin group ofantibiotics which have been
trations of amino acids and nucleotides. detected by methods ofthis type.
It is desirable that the medium, on which the The detection of novel antimicrobials may be
isolates are grown prior to assay, allows maximum assisted by the use of very specific screening
expression of their genetic material. Nisbet (1982) techniques. For example, Brown et al. of Beechams
put forward some guidelines for the design of media Pharmaceuticals (1976) discovered the B-lactamase
which may allow good product formation and these inhibitor, clavulanic acid, using a screen which
are summarized in Table 3.2. detected a synergy between ampicillin and the test
sample in their inhibition of aB-lactamase produc-
TABLE 3.2. Guidelines
for ‘overproduction media’ (Nisbet, 1982) ing Klebsiella. Glaxo workers have utilized an in vitro
GUIDELINES enzyme screen designed to detect inhibitors of car-
1. Prepare a range of media in which different types ofnutrients boxypeptidase which is involved in the cross-linking
become growth-limiting (i.e. C, N, P, O). of peptidoglycan of bacterial cell walls. This work
2. For each type of nutrient depletion use different forms ofthe
has led to the detection of several novel antibiotics,
growth-sufficient nutrient.
3. Use a polymeric or complexed form of the growth-limiting currently under investigation (Nisbet, 1982).
nutrient.
4. Avoid the use ofreadily assimilated forms of carbon (glucose)
or nitrogen (NH,,) that may cause catabolite repression. SCREENING FOR PHARMACOLOGICALLY
5. Ensure that known cofactors are present (Co** Mg?*, Mn?*, ACTIVE COMPOUNDS
Fe?*), Umezawa (1972) pioneered the concept of screen-
6. Buffer to minimize pH changes.
ing for microbial products capable ofinhibiting key
enzymes of human metabolism. Umezawa’s
SCREENING FOR THE PRODUCTION OF approach was based on the logic that, ifa compound
ANTIBIOTICS inhibits a key human enzyme in vitro, it may havea
The detection of antimicrobial action may be pharmacological action in vivo. Active compounds
achieved by growing the potential producer on an detected in an in vitro screen may then be used in
agar plate in the presence of an organism (or animal-test systems. Hamill (1982) has reviewed
organisms) against which antimicrobial action is this area and the success of the techniques is indi-
required. Production ofthe inhibitor is indicated by cated in Table 3.3, constructed by Nisbet (1982).
30
The Isolation, Preservation and Improvement of Industrial Micro-organisms
Tasve 3.3. Pharmacological screens based on the enzyme-largel approach. Examples are restricted
to enzymes which detected new inhibitors (Nisbet, 1982)
SCREENING FOR THE PRODUCTION OF producing organism may then be isolated from the
GROWTH FACTORS original plate which was replicated. This procedure
The production of growth factors, such as amino is followed by quantitative assays where the
acids and nucleotides, is a difficult characteristic to organism is grown in liquid culture and the cell-free
employ as a selective force in an isolation procedure; broth estimated for the amino acid concentration.
thus, organisms have been isolated randomly and
producers detected by subsequent screening tests. SCREENING FOR THE PRODUCTION
The detection of growth-factor producers depends OF POLYSACCHARIDES
on the stimulation of the growth of auxotrophic Micro-organisms producing exopolysaccharides
bacteria by the isolate. Daoust (1976) described a have been isolated from a range of environments
screening procedure for the detection of amino acid (Lawson and Sutherland, 1978) but, again, the
producers. The majority of amino acid producers property of polysaccharide production is difficult to
belong to the genera Arthrobacter, Microbacterium, Bre- apply as a selective factor in the isolation. Lawson
vibacterium, Micrococcus and Corynebacterium so Daoust and Sutherland (1978) suggested attempting to iso-
suggested the exclusion of fungi from the search by late from materials, such as carbohydrate industrial
the incorporation of antifungal compounds, such as waste, which may be expected to be rich in the
cycloheximide, in the isolation media. The initial desired organisms. Isolates from such environments
screen for the production of amino acids is a qualita- may be grown on suitable media and producers
tive one, normally carried out on solid media. Isola- recognized by the mucilaginous appearance of the
tion plates containing between thirty and fifty col- colonies—the suspected isolates would then have to
onies are replica plated on to solid medium, capable be grown in liquid culture and the properties of the
polysaccharides assessed.
of supporting amino acid production, and grown for
2 to 3 days. The developed colonies are then killed
with ultraviolet light and are overlaid with agar THE PRESERVATION OF INDUSTRIALLY
containing a suspension of an organism auxotrophic IMPORTANT MICRO-ORGANISMS
for the desired product. After 16 hours incubation at
37° the assay organism should have grown as a halo The isolation ofa suitable organism for a commer-
1round any producing colony on the plate and the cial process may be a long and very expensive
31
Principles of Fermentation Technology
procedure and it is therefore essential that the isolate is suffered during the freezing of the culture but
retains the desirable characteristics that led to its there is very little loss during the storage period so
selection. Also, the culture used to initiate an indus- that the viability may be predictable even after a
trial fermentation must be viable and free from period of many months. The main disadvantages of
contamination. Thus, industrial cultures must be the system are the expense of maintenance at such
stored in such a way as to eliminate genetic change, low temperatures and the danger of failure of the
protect against contamination and retain viability. apparatus.
An organism may be kept viable by repeated sub-
culture into fresh medium, but, at each cell division,
there is a small probability of mutations occurring Storage in a dehydrated form
and because repeated sub-culture involves very
SOIL CULTURE
many such divisions, there is a high probability that
Moist, sterile soil may be inoculated with the
strain degeneration would occur. Also, repeated
culture and incubated for several days for some
sub-culture carries with it the risk of contamination.
growth to occur and then allowed to dry at room
Thus, preservation techniques have been developed
temperature for approximately 2 weeks. The dry
to maintain cultures in a state of ‘suspended anima-
soil may be stored in a dry atmosphere or, prefera-
tion’ by storing either at reduced temperature or in
bly, in a refrigerator. The technique has been widely
a dehydrated form.
used for the storage of fungi and actinomycetes and
Pridham et al. (1973) observed that of 1800
actinomycetes dried on soil about 50% were viable
Storage at reduced temperature
after 20 years storage.
ensure their quality and such a procedure has been ers and this has been observed frequently in the
outlined by Lincoln (1960): a single colony of the early stages in the development of a new product.
culture to be preserved was inoculated into a shake An explanation of this phenomenon for mycelial
flask and the growth of the culture observed to organisms may be that most new isolates are proba-
ensure a typical growth pattern; after a further bly heterokaryons (contain more than one type of
shake-flask sub-culture the broth was used to pre- nucleus) and the selection of the progeny of uninuc-
pare a large number of lyophilized ampoules. At leate spores results in the production of homo-
least 3% of the ampoules were reconstituted and the karyons (contain only one type of nucleus) which
cultures assessed for purity, viability and productiv- may be superior producers. However, the phenom-
ity. If the samples failed any one of these tests the enon is also observed with unicellular isolates which
entire batch was destroyed. Thus, by the use of such are certainly not heterokaryons. Therefore, it is
a quality-control system stock cultures may be worth while to periodically plate out the producing
retained, and used, with confidence. culture and screen a proportion of the progeny for
productivity; this practice has the added advantage
that the operator tends to become familiar with
THE IMPROVEMENT OF INDUSTRIAL
morphological characteristics associated with high
MICRO-ORGANISMS
productivity and, by selecting ‘typical’ colonies, a
strain subject to yield degeneration may still be used
Natural isolates usually produce commercially with consistent results.
important products in very low concentrations and Therefore, selection of natural variants may result
therefore every attempt is made to increase the in increased yields but it is not possible to rely on
productivity of the chosen organism. Increased such improvements, and techniques must be
yields may be achieved by optimizing the culture employed to increase the chances of improving the
medium and growth conditions, but this approach culture. Dulaney and Dulaney (1967) compared the
will be limited by the organism’s maximum ability spread in productivity of chlortetracycline ofnatural
to synthesize the product. The potential productiv- variants of Streptomyces viridifaciens with the spread in
ity of the organism is controlled by its genome and, productivity of the survivors of an ultraviolet treat-
therefore, the genome must be modified to increase ment. The results of their comparison are shown in
the potential yield. The cultural requirements ofthe Figs. 3.3 and 3.4, from which it may be seen that
modified organism would then be examined to pro- although there are more inferior producers amongst
vide conditions that would fully exploit the the survivors of the ultraviolet treatment there are
increased potential of the culture, while further also strains producing more than twice the parental
attempts are made to beneficially change the level, far greater than the best of the natural vari-
genome of the already improved strain. Thus, the ants. The use ofultraviolet light is only one ofa large
process of strain improvement involves the con- number of physical or chemical agents which
tinual genetic modification of the culture, followed increase the mutation-rate—such agents are termed
by reappraisals ofits cultural requirements. Genetic mutagens. The reader is referred to Sermonti (1969)
modification may be achieved by selecting natural and Elander and Espenshade (1976) for accounts of
variants, by selecting induced mutants and by the modes of action of mutagens. The vast majority
selecting recombinants. There is a small probability of induced mutations are deleterious to the yield of
ofagenetic change occurring each time a cell divides the desired product, but, as shown in Fig. 3.4, a
and when it is considered that a microbial culture minority are more productive than the parent. The
will undergo a vast number ofsuch divisions it is not problem ofobtaining the high-yielding mutants may
surprising that the culture will become more be approached from two theoretical standpoints;
heterogeneous. The heterogeneity of some cultures the number of desirable mutants may be increased
can present serious problems of yield degneration by ‘directed mutation’, i.e. the use of a technique
because the variants are usually inferior producers which will preferentially produce particular mut-
compared with the original culture. However, vari- ants at a high rate; or techniques may be developed
ants have been isolated which.are superior produc- to improve the separation of the few desirable types
33
Principles of Fermentation Technology
25 25
20 20
NS 15
10 of
% 10
cultures
cultures
of
%
Fic. 3.3. The spread in productivity of chlortetracycline of natu- Fic. 3.4. The spread in chlortetracycline productivity of the
ral variants of Streptomyces viridifaciens (Dulaney and Dulaney, survivors of aUV-treated population of Streptomyces viridifaciens
1967). (Dulaney and Dulaney, 1967).
from the large number of mediocre producers. The eukaryotic chromosome, even in synchronized cul-
choice of mutagen has been demonstrated to affect tures. However, the concept of ‘directed
the success of mutation programmes—for example, mutagenesis’ has come to fruition using in vitro DNA
Hostalek (1964) claimed that ultraviolet radiation manipulation techniques (Shortle ef al., 1981). In
was the most effective mutagen for increasing the these systems a cloned DNA molecule may be sub-
yield oftetracycline by strains ofStreptomyces aureofa- jected to in vitro enzymatic cleavage and manipula-
ciens. However, despite such evidence, the concept of tion such that the mutation is present at a particular
directed mutation is difficult to apply and, although site. Thus, reintroduction of the modified DNA into
it is obviously useful to take regard of the literature the cell should result in the production of a cell
in choosing a mutagen, the ‘progeny’ of even a mutated at a particular site. The incorporation ofa
favourable mutagen will still contain only a very selectable marker into the cloned DNA would assist
small number ofdesirable mutants necessitating the in the recovery of cells incorporating the modified
use ofa separation technique. Indeed, Alikhanian gene(s). These systems may not only be applicable
(1973) summarized the situation thus: ‘... the to organisms whose genetics are well documented
potentialities of specificity and choice of “specific” but also to organisms whose genetics are relatively
mutagens are so limited, that it is hardly possible to unknown. According to Malik (1982) DNA from
speak about the reality of the idea about a set of such organisms may be isolated, mutagenized and
mutagens which could ensure the obtaining of incorporated into the chromosome and_ the
phenotypically determined mutations’. phenotype of the transformed cell characterized.
The induction of mutations by N-methyl-N’- Malik also claims that such techniques may allow
nitro-N-nitrosoguanidine (NTG) appears to be the immediate application of genetic manipulation in
closest system to in vivo ‘directed mutation’. NTG commercially important antibiotic producing
treatment causes clusters of mutations around the organisms.
replicating fork of the bacterial chromosome. Thus, However, despite the potential of in vitro directed
provided the bacterial chromosome is well mapped, mutagenesis, the most widely used technique is to
it may be possible to predict the region of the attempt to separate the desirable mutants from the
chromosome susceptible to change, in a mutation other survivors of a mutation treatment, and this
treatment of asynchronous culture. The application situation is similar to the isolation of desirable
of this technique to eukaryotes presents many prob- organisms from nature. Where possible, the mutant
lems, not least of which is the possibility of mutations isolation procedure should use the improved charac-
occurring at the many replicating sites of the teristic of the desired mutant as a selective factor.
34
The Isolation, Preservation and Improvement of Industrial Micro-organisms
DN ide
genome in such a way as to prevent the transcription
of the gene into-messenger RNA, thus resulting in —-—-* Feedback control
the prevention of enzyme synthesis. Fic. 3.6. The control ofa biosynthetic pathway by the concerted
Feedback inhibition and repression frequently effects of products D and F on the first enzyme of the pathway.
PFT-D 35
Principles of Fermentation Technology
Ji Pi /
products. In Fig. 3.8 both D and F independently AoE Pres
a sO
reduce the activity of the first enzyme by 50% \ \
36
The Isolation, Preservation and Improvement of Industrial Micro-organisms
metabolite may enable the construction of a ‘blue- the accumulation of a large concentration ofa-keto-
print’ of the most useful industrial organism, i.e. one glutarate which the organism converts to glutamic
where the production of the metabolite is not acid. Oxaloacetate is regenerated in glutamate pro-
restricted by the organism’s control system. Such ducers by the activity of the glyoxylate pathway as
postulated mutants may be modified in three ways: shown in Fig. 3.11.
When C. glutamicum was grown in a medium
1. The organism may be modified such that the
containing a high concentration of biotin, the
end products which control the key enzymes of
organism synthesized glutamate at a level of 25-36
the pathway are lost from the cell due to some
ug mg! dry weight of cells, further production
~abnormality in the permeability of the cell
being assumed to be prevented by some form of
membrane.
feedback control by glutamate of its own synthesis
2. The organism may be modified such that it
(Demain and Birnbaum, 1968). Under conditions
does not produce the end products which con-
ofbiotin limitation, glutamate was released from the
trol the key enzymes ofthe pathway.
cells and accumulated to a level of up to 50 gdm’.
3. The organism may be modified such that it
The increased permeability of the biotin-limited
does not recognize the presence of inhibiting or
cells to glutamate corresponds with a change in the
repressing levels of the normal control metabo-
fatty acid and phospholipid content of the cell
lites.
envelopes. The crucial factor in the control of per-
meability appears to be the synthesis of membranes
MODIFICATION OF THE PERMEABILITY deficient in phospholipids (Nakao et al., 1973).
The best example of the modification of the per- The role of biotin in the glutamate fermentation
meability of a micro-organism is provided by the obviously necessitates the inclusion oflimiting levels
glutamic acid fermentation. Kinoshita et al. (1957a) of biotin in the medium. Thus, the use of crude
isolated a biotin requiring, glutamate-producing carbon sources such as cane molasses, which are
organism, subsequently named Corynebacterium rich in biotin, presented difficulties which have been
glutamicum, the permeability of which could be mod- overcome by the modification of the permeability by
ified by the level of biotin. Provided that the level of means other than biotin limitation. The inclusion of
biotin in the production medium was below 5 pug penicillin or fatty acid derivatives such as
dm ° then the organism would excrete glutamate, polyoxyethylene sorbitan mono-oleate (Tween 80)
but at concentrations of biotin optimum for growth during logarithmic growth in biotin-rich medium
the organism produced lactate (Kinoshita and caused an aberration ofthe cell envelope permeabil-
Nakayama, 1978). Thus, the permeability of C. ity resulting in the release of glutamate ((Udagawa
glutamicum may be controlled by the composition of et al., 1962).
the culture medium and, as such, is not an example The production of glutamate from hydrocarbons
of strain improvement in the sense that the term has (as carbon sources) has also presented difficulties in
been used in this chapter. However, the isolation of the control of permeability, in that the assimilation
C. glutamicum was a major advance in microbial of hydrocarbons results in the production of fatty
production of amino acids and the demonstration of acids which effectively bypasses the site of biotin
the role of permeability in the production of gluta- control (Nakao et al., 1972). The permeability of the
mate suggested the possibility of the genetic modifi- producing organism may be controlled by the addi-
cation of permeability to achieve high levels of tion of penicillin (Wang et al., 1979) but a genetic
productivity. Thus, it is relevant at this stage to solution to this problem has also been found. Nakao
consider the physiology of glutamate production by et al. (1970) isolated a glycerol requiring auxotrophic
biotin-limited culture. mutant of Corynebacterium alkanolyticum in which
Kinoshita’s isolate was not only deficient in its phospholipid synthesis was controlled by the supply
ability to produce biotin but also in the level of of glycerol. The mutant produced about 40 g dm~°
a-ketoglutarate dehydrogenase which normally glutamate when grown on n-paraffins in the pre-
converts a-ketoglutarate to succinate in the tricar- sence of 0.01% glycerol (Nakao et al., 1972). Thus,
boxylic acid cycle. Thus, a metabolic block results in an understanding of the mode of action of biotin
37
Principles of Fermentation Technology
® -----4h---~
Aap Be C+D E.
Feed limiting
concentration
of E
A A @ A
@ rae cs Shia pm =e
x \ x * a: i x
&
I Cc —_ | C ap Ee C \
SAAD
oe eee | AGNES
ee
Sau
Me De rai aeee
Nf WwW S “ Nf *, ye
E G E G E G
Feed limiting Feed limiting Feed limiting
concentration concentration concentration
of E of EandG of EandG
© 74
‘
vena te
/
= i Feed limiting
\ a Cee concentration
NES f|
iver
< G \ H S
ee
Fic. 3.12. Overproduction of primary metabolites by decreasing the concentration ofa repressing or inhibiting end product. ---* , Site
of auxotrophic mutation; -——-, feedback regulation; — overproduced product (Demain, 1972).
such, so that the best approach is to design a proce- auxotrophs. Several techniques have been
dure to select relevant auxotrophs from the survivors developed using different antimicrobials suitable for
of a mutation and subsequently screen the selected use with a range of micro-organisms.
auxotrophs for productivity. Productive strains Davis (1949) developed an enrichment technique
amongst the auxotrophs may be detected by over- utilizing penicillin as the inhibitory agent. The sur-
layering colonies of the mutants with agar suspen- vivors of amutation treatment were first cultured in
sions of bacteria auxotrophic for the required pro- complete medium, harvested by centrifugation,
duct. The high-producing mutants may be iden- washed and resuspended in minimal medium plus
tified by the growth of the overlay around the pro- penicillin. Only the growing prototrophic cells were
ducer. The most commonly used methods for the susceptible to the penicillin and the non-growing
recovery of auxotrophic mutants are the use of some auxotrophs survived. The cells were harvested by
form of enrichment culture or the use of atechnique centrifugation, washed (to remove the penicillin
to visually identify the mutants. and products released from lysed cells) and resus-
The enrichment processes employed are based on pended in complete medium to allow the growth of
the provision of conditions which adversely affect the auxotrophs, which could then be purified on
the prototrophic cells but do not damage the auxo- solid medium. The nature of the auxotrophs isolated
trophs. Such conditions may be achieved by expos- may be determined by the design of the so-called
ing the population, in minimal medium, to an anti- complete medium; if only one addition is made to
microbial agent which only affects dividing cells the minimal medium then mutants auxotrophic for
which should result in the death of the growing the additive should be isolated.
prototrophs but the survival of the non-growing Abe (1972) described the use of Davis’ technique
39
Principles of Fermentation Technology
40
The Isolation, Preservation and Improvement of Industrial Micro-organisms
Aspartate Glutamate
Asparty! semi-aldehyde
al N-acetylglutamy! phosphate
Dihydrodipicolinate Homoserine
/ N-acetylglutamic semialdehyde
| / Homoserine- (P)
N-acetyl ornithine
os
ee
eee
ee
ee
Threonine ——!
Ornithine
Methionine
/\
vf
| jf a-Ketobutyrate Citrulline
f
/
| )/
// \
Argino succinate
|
/
/
4 |
NON = = SSS
SS J
, \
X i
\ :
Ne | / -—---® Feedback inhibition
\ / ——> Biosynthetic route
Mi
\ Fic. 3.15. The control of the biosynthesis of arginine in C.
Lysine i‘ \ glutamicum.
—--
EE
EE
ee
TH
EE
SE
peeeeeeweeee
Ee a Le oi/
4]
Principles of Fermentation Technology
Adenylosuccinate ,
Seec
fee
as
fe
faa
ec
SS
ere
ee
ae
ee
ee
a
Sa
ee Mutants which are auxotrophic for AMP or
\ 5/ doubly auxotrophic for AMP and GMP have been
AMP ~ _ isolated which will excrete inosine at levels of up to
15 gdm * (Demain, 1978). AMP auxotrophs, lack-
ing adenylosuccinate synthase activity, require the
addition of small quantities of adenosine but will
—— Biosynthetic route
—---© Feedback inhibition accumulate inosine due to the removal of the inhibi-
—— Feedback repression tion and co-operative repression of PRPP amidino-
transferase, as shown in Fig. 3.17. AMP and GMP
Fic. 3.16. Control of the biosynthesis of the purine nucleotides in
Bacillus subtilis. 1. Reaction catalysed by PRPP amidino- double auxotrophs will produce inosine due to the
transferase. 2. Reaction catalysed by IMP dehydrogenase. 3. removal of the controls normally imposed by the two
Reaction catalysed by adenylosuccinate synthase. AMP = end products, as illustrated in Fig. 3.18. Such double
Adenosine monophosphate. IMP = Inosine monophosphate.
XMP = Xanthine monophosphate. GMP = Guanosine auxotrophs require the addition of both adenosine
monophosphate. and guanosine, in small concentrations.
PRPP
of the pathway is the feedback inhibition of the
second enzyme in the sequence, acetylglutamic acid
phosphorylating enzyme, although the first enzyme
may also be subject to regulation.
Kinoshita et al. (1957b) isolated a citrulline
requiring auxotroph of C. glutamicum which would
accumulate ornithine at a molar yield of 36% from
glucose, in the presence of limiting arginine and
* Site of the defective enzyme
excess biotin. The mutant lacked the enzyme con-
—— Biosynthetic route
verting ornithine to citrulline which resulted in the ----—-e Feedback inhibition
cessation of arginine synthesis and, therefore, the —— Feedback repression
removal of the control of the pathway. --<--e@ Control defective
Inosine monophosphate has been shown to Fic. 3.17. The control of the synthesis of purine nucleotides in a
demonstrate flavour-enhancing qualities and is pro- mutant with a defective adenylosuccinate synthase.
42
The Isolation, Preservation and Improvement of Industrial Micro-organisms
43
Principles of Fermentation Technology
44
The Isolation, Preservation and Improvement of Industrial Micro-organisms
Fic. 3.21. The genealogy of L-arginine-producing mutants of B. flavum TA, thiazolealanine; NG, N-methyl-N’-nitro-N-nitroso-
guanidine (Kubota et a/., 1973).
flavum is probably similar to that shown for C. Shiio and Sano (1969) investigated the use of
glutamicum in Fig. 3.15. The major control appears to prototrophic revertants of B. flavum for the produc-
be the feedback inhibition of the second enzyme of tion of lysine. These workers isolated prototrophic
the pathway by arginine. Kubota ef al. selected revertants from a homoserine dehydrogenase-defec-
mutants resistant to the arginine analogue, 2- tive mutant. The revertants were obtained as small-
thiazolealanine, and the genealogy of the mutants is colony forming strains and produced up to 23 g
‘shown in Fig. 3.21. Strain number 352 produced dm~* lysine. The overproduction of lysine was
25.3 g dm °* arginine. Presumably, the mutants shown to be due to the very low level of homoserine
were altered in the susceptibility of the second dehydrogenase in the revertants which, presuma-
enzyme in the pathway to inhibition by arginine. bly, resulted in the synthesis of threonine and
The second technique used for the isolation of methionine in quantities sufficient for some growth,
mutants altered in the recognition ofcontrol factors but insufficient to cause inhibition or repression.
is the isolation of revertant mutants. Auxotrophic Programmes for the improvement of strains pro-
mutants may revert to the phenotype of the mutant ducing primary metabolites rarely rely on the use of
‘parent’. Consider the hypothetical pathway illus- only one selection technique. Most projects employ
trated in Fig. 3.19 where P controls its own produc- a number of methods including the selection of
tion by feedback inhibiting the first enzyme (a) of natural variants and the selection of induced mut-
the pathway. A mutant does not produce the ants by a variety of means. The selection of bacteria
enzyme, a, and is, therefore, auxotrophic for P. overproducing threonine provides a good example
However, a revertant of the mutant produces large of the use of a variety of selection techniques.
concentrations of P. The explanation of the Attempts to isolate auxotrophic mutants of C.
behaviour of the revertant is that, with two muta- glutamicum producing threonine were unsuccessful
tions having occurred at loci concerned with the despite the fact that productive auxotrophic strains
production of enzyme a, the enzyme ofthe revertant of Escherichia coli had been isolated. Huang (1961)
is different from the enzyme of the original proto- demonstrated threonine production at a level of 2—4
trophic strain and is not susceptible to the control by g dm * by a diaminopimelate and methionine dou-
P. Revertants may occur spontaneously or ble auxotroph of E. coli. Kase et al. (1971) isolated a
mutagenic agents may be used to increase the fre- triple auxotroph mutant of E. coli which required
quency of occurrence, but the recognition of the diaminopimelate, methionine and isoleucine and
revertants would be achieved by plating millions of produced between 15 and 20g dm~* threonine. The
cells on medium which would allow the growth of control of the production of the aspartate family of
only the revertants, i.e. in the above example, on amino acids in EF. coli is shown in Fig. 3.22 and that
medium lacking P. in C. glutamicum in Fig. 3.14. The mechanism of
45
Principles of Fermentation Technology
‘Tete wae!
Homoserine 4 .
Aspartate ———» /spartyl Asparty| $8
Homoserine——© Threonine
——s phosphate semialdehyde —.» phosphate —+>
Ie
— =
ea
a ‘ =H
|
|
Dihydrodi-
|
picolinate a-Ketobutyrate
|
|
|
|
{
Methionine {
{
|
|
t
|
|
L Isoleucine
Fic. 3.22. Control of the aspartate family of amino acids in Escherichia coli.
|
KY 10484 (met~, AHV resistant)
|
KY 10230 (met~, AHV resistant)
Produced 2.3 g dm~-$ threonine Produced 3.7 g dm~$ threonine
|
KY 10440 (met™, AHV resistant,
|
KY 10251 (met~, AHV resistant,
AEC resistant) AEC resistant)
Produced 9.5 g dm-$ threonine Produced 9.0 g dm-3 threonine
and 5.5 g dm-3 lysine
Fic. 3.23. The genealogy of mutants of C. glutamicum producing t-threonine or L-threonine plus L-lysine. AHV, a-amino-B-hydroxy
valeric acid; AEC, S-(8-aminoethyl)-1-cysteine (Kase and Nakayama, 1972).
47
Principles of Fermentation Technology
careful manipulation of the environment ofthe pro- Clarke (1974) isolated mutants of
ducing culture, i.e. the medium should contain the Pseudomonas aeruginosa constitutive for alipha-
inducer of the enzyme but should not contain repres- tic amidase by culturing on a medium con-
sors or rapidly utilized carbon sources. However, it taining succinate as the carbon source and
may also be possible to isolate mutants which are formamide as the sole nitrogen source, for-
altered in the control of the synthesis of the enzymes mamide being a good substrate but a poor
and a number of techniques have been developed inducer of aliphatic amidase. Sikyta and
which may be applied to the selection of mutants Kyslik (1981) isolated a mutant of E. colt
which produce enzyme in the absence of inducers constitutive for penicillin amidase by cultur-
and in the presence of repressors. ing the organism in a chemostat containing
Mutants which produce normally inducible excess sodium lactate and _ limiting
enzymes in the absence of the inducer are termed phenylacetylglutamic acid (a poor inducer
constitutive mutants and may be isolated by the of enzyme) as the sole nitrogen source.
provision of cultural conditions which favour the (iv) Selection of constitutive mutants may be
mutant more than the wild type. Demain (1971) achieved by culturing a population in the
cited the following techniques which may be used presence of the enzyme inducer and an
for the isolation of constitutive mutants: inhibitor of induction.
(v) Avery useful technique for the recognition of
(i) Constitutive mutants may be isolated by constitutive mutants on solid medium is the
growing the population in a chemostat with use of chromogenic substrates which do not
the inducer as the limiting substrate. In act as inducers. An example is provided by
continuous culture, the limiting substrate is Demain (1971) in detecting the presence of
normally present at very low concentrations, mutants constitutive for B-galactosidase by
often below the threshold level for induction. spreading the progeny of a mutation treat-
Thus, the constitutive mutants present in ment on to solid medium containing glycerol
the population would have a selective advan-
as the carbon source and spraying the col-
tage over the inducible types. Novick and
onies produced with o-nitrophenyl-B-p-
Horiuchi (1969) selected mutants of E. coli
galactoside. Constitutive mutants would
constitutive for B-galactosidase by this
degrade the substrate with the release of
method, using lactose as the carbon source.
yellow o-nitrophenol resulting in the staining
Sikyta and Kyslik (1981) also isolated mut-
of the constitutive colonies.
ants ofE. coli constitutive for B-galactosidase
by culturing the organism in a lactose- Mutants may also be isolated which will produce
limited chemostat. catabolic enzymes in the presence ofrepressors. The
(ii) The cycling of a population between media two types of repression to which catabolic enzymes
with and without the inducer should enrich are susceptible are a form of feedback repression by
the culture with constitutive types. The con- the products (or related products) of the enzyme-
stitutive mutant would be at an advantage catalysed reaction and catabolite repression, caused
because the inducible type would display a by rapid growth on readily utilized carbon sources.
lag time when introduced into the medium It appears to be difficult to select for resistance to the
containing the inducer (Cohen-Bazaire and effects of feedback-type repression of catabolic
Jolit, 1953). enzymes although it is possible that the screening of
(ii1) The use ofa substrate of the enzyme which is revertants of mutants lacking the enzyme might
a poor inducer as an essential component of yield satisfactory results. A technique specific to the
the medium should result in the enrichment production of protease by species of Bacillus has led
of any constitutive mutants 1n a population. to the release of repression by amino acids; both
Jacob and Monod (1961) isolated mutants sporulation and protease formation is repressed by
constitutive for B-galactosidase by the use of amino acids and it is therefore possible to isolate
phenyl-B8-galactoside as the carbon source. mutants resistant to the repressive effects of amino
48
The Isolation, Preservation and Improvement of Industrial Micro-organisms
acids on protease production by isolating strains of the enzyme and a catabolite repressor
capable of sporulating in the presence of an other- present in the medium. An organism capable
wise inhibitory mixture of amino acids. Levisohn of utilizing both substrates should be at a
and Aronson (1967) isolated mutants of Bacillus competitive advantage in the chemostat and
cereus which produced about 10 times as much pro- would be selected out.
tease as the parent.
The above accounts of techniques for the isolation of
Several techniques have been used for the isola-
mutants changed in their response to inducers and
tion of mutants resistant to catabolite repression of
repressors includes examples taken from academic
certain enzymes:
research but with more knowledge of the control of
(i) The screening of revertants of mutants lack- production ofindustrially important enzymes these
ing the enzyme has resulted in the isolation techniques may find wider application.
of mutants altered in their susceptibility to Enzyme production by a bacterium may be
catabolite repression. Mutants insensitive to increased by increasing the number of copies of the
catabolite repression of B-galactosidase were gene coding for the enzyme. This may be achieved
isolated as revertants from a group oflactose- by genetic manipulation techniques (discussed in a
negative mutants. Smyth and Clark (1972) later section of this chapter) where the gene is
isolated mutants of P. aeruginosa resistant to introduced into a plasmid which may be present 1n
the catabolite repression of amidase from the the cell at a high copy number. An example is
revertants of amidase-negative mutants. provided by the work of Colson ef al. (1981) who
(1) Mutants resistant to the catabolite repres- isolated a Bacillus coagulans gene coding for a ther-
sion of enzymes degrading nitrogen contain- mostable a-amylase. The gene was introduced into
ing substrates have been isolated by the a plasmid which was then incorporated into an E.
culture of a population on solid medium coli where it was replicated and maintained at a high
containing a known catabolite repressor, as copy number. A very high level of gene expression
the carbon source, and the substrate of the (a-amylase production) was achieved in the &. coli
enzyme as the sole nitrogen source. Sikyta et strain.
al. (1982) isolated a mutant of£. coli resistant
to the catabolite repression of serine
deaminase by culturing the organism in a
The isolation of induced mutants
chemostat containing D-serine, as the limit-
producing improved yields of secondary
ing nitrogen source, and excess glucose.
metabolites where directed selection
Neidhart (1960) isolated mutants resistant
is difficult to apply
to catabolite repression of histidase by cul-
turing a population on medium with glucose
as the carbon source and histidine as the sole The discussion so far has considered the isolation
nitrogen source. The isolated mutants were of mutants producing products whose biosynthesis
also resistant to the catabolite repression of and control have been sufficiently understood to
urocanase and galactosidase which indicates prepare a ‘blueprint’ of the desirable mutant which
that this technique may be useful for the may then enable the construction of suitable selec-
selection of mutants with several enzymes tion procedures. The biosynthesis of secondary
released from repression, and may, there- metabolites is far less well understood than is that of
fore, be applied to the isolation of mutants primary metabolites and relatively little is known
overproducing enzymes of industrial about the mechanisms which limit the productivity
interest. of cultures synthesizing secondary metabolites.
(iii) The selective pressure of achemostat may Therefore, it has been extremely difficult to adopt
be applied to the selection of catabolite rep- the approach ofisolating specific regulatory mutants
ression resistant mutants by culturing a and it has been necessary to simply screen the
population in a chemostat with the substrate survivors of amutation treatment for superior pro-
49
Principles of Fermentation Technology
Taste 3.4 Improvement of antibiotic yields over a period ofyears testing will be determined by the practical limiting
(Alikhanian, 1962 and Riviere, 1977) factors of personnel, incubator and shaker space
Initial yield
and time. However, Davies (1964) demonstrated
attimeof Improvedyields Improved yields that a statistical approach could give valuable
discovery obtainedinthe in France, 1972 guidelines for the efficient utilization of physical
Antibiotic (unitscm~°) USSR (unitsem~*) (units cm”)
resources in strain-development programmes.
Penicillin 20 (1943) 8,000 (1955) 12,000—15,000 Davies based a computer simulation of amutation
Streptomycin 50 (1945) 5,000 (1955) 12,000-15,000 and screening programme on the availability of 200
Chlortetra-
cycline 200 (1948) 4,000 (1959) 12,000—15,000 shaker spaces and practical results of error variance,
Oxytetra- and the distribution ofyield likely to occur amongst
cycline 400 (1950) 6,000 (1959) — the mutants. He assumed that the majority of the
Erythromycin 100 (1955) 2,000 (1961) 3,000
progeny of a mutation would give a small, rather
than a large, increase in productivity and it would,
ductivity without the expedience of a selection therefore, be more feasible to screen a small number
technique wnich gives the higher producer an in the hope of obtaining a small increase rather than
advantage over the less desirable variants. screen a large number in the hope of obtaining a
Nevertheless, the empirical techniques of random large increase. The major difficulty inherent in this
screening for the selection of improved strains has approach is the error involved in determining small
proved to be extremely effective in increasing the increases which implies the replication of screening
yields of antibiotics, as illustrated in Table 3.4 tests, and, therefore, the use of more facilities.
(Riviere, 1977). However, it appears that attempts Davies used the computer simulation to investi-
are being made to adopt a more rational approach to gate the merits of replication in screening program-
selection techniques and to reduce the empirical mes and finally proposed the use of a two-stage
nature of strain-improvement programmes. These scheme where mutants were tested singly in the first
attempts include the streamlining of the empirical stage and then the better producers were tested in
techniques and the use of more directed selection quadruplicate in the second stage. Davies concluded
methods. Before discussing the attempts at directed that such two-stage schemes were adequate over a
selection the empirical approach will be considered, wide range of conditions, although a one-stage
along with the attempts made to improve this screen could be used if the testing error were small
approach, including miniaturized programmes. and the frequency of occurrence offavourable types
The empirical approach to strain improvement high, and a three-stage screen if the testing error
involves subjecting a population of the micro- were high and the frequency of occurrence of favour-
organism to a mutation treatment and then screen- able types low. A screening programme based on
ing a proportion of the survivors ofthe treatment for Davies’ proposals is shown in Fig. 3.24.
improved productivity. The assessment of the cho- The answer to Fantini’s second question (which
sen survivors is usually carried out in shake flasks colonies should be isolated?) is extremely difficult in
resulting in the procedure being costly, both in the field of secondary metabolites and in Davies’
terms of time and personnel. According to Fantini scheme the colonies were chosen at random. The
(1966) the two questions which arise in the design of selection of colonies on the basis of changed mor-
programmes for the isolation of improved strains phology has been considered by a number of workers
producing antibiotics are: and it appears that this is an undesirable technique.
(1) How many colonies from the survivors of a Elander (1966) demonstrated that it was preferable
mutation treatment should be isolated for to isolate normal morphological types as, although a
testing? morphological variant may be a superior producer,
(ii) Which colonies should be isolated? it might require considerable fermentation develop-
ment to materialize the increased production. Also,
In attempting to answer the first question Fantini Alikhanian (1962) claimed that the vast majority of
dismissed statistical approaches as impractical and morphological mutants of the antibiotic producing
claimed that the number of colonies isolated for actinomycetes tested were inferior producers. The
50
The Isolation, Preservation and Improvement of Industrial Micro-organisms
Spore suspension
Mutation treatment
|
200 colonies removed and each
suspended in sterile water
|
2 Son
UU U
3 stock slopes
Spore suspension of
each isolate used
of each isolate to inoculate an
inoculum
development flask
|
Best 4 conserved and re-mutated
Fic. 3.24. A strain-improvement programme for a secondary metabolite producing culture (Davies, 1964).
most common type of programme quoted is similar tic is assessed by the degree of inhibition of the
to that of Davies where the choice of colonies is overlayed indicator. The system is simple to apply
random and the screening for production is done in to strains producing low levels of antibiotic but must
a shake flask. However, several attempts have been be modified to allow the screening of high producers
made to miniaturize the procedure for the improve- where very large zones of inhibition would be
ment of antibiotic-producing strains. Such obtained. Also, some systems should be used to free
miniaturized systems are designed to enable the the superior producers from contaminating indi-
productivity ofall (or a significant proportion) of the cator organisms. Dulaney and Dulaney (1967) used
survivors to be assessed which should eliminate (or overlay techniques in the isolation of mutants pro-
reduce) the problem of choice of colonies for assess- ducing chlortetracycline: mutated spores were cul-
ment. The basis of the miniaturized techniques is tured on an agar medium in petri dishes for 6 days
the overlayering of mutant colonies, growing on and then covered with pieces of sterile cellophane.
solidified medium, with an indicator organism sensi- An overlay of agar containing the indicator
tive to the produced antibiotic. The level of antibio- organism was then added and the plates incubated
51
PFT-E
Principles of Fermentation Technology
overnight. The mutant colonies were kept free from on solid medium which would not allow the full
contamination by the cellophane and the size of the production potential of the culture to be detected.
inhibition zone could be controlled by the depth of However, nutrient-limiting conditions were
the base layer, the age of the colonies when over- achieved with the solidified medium by omitting the
layed, the depth of the overlay and the temperature main carbon source and reducing the corn steep-
of incubation. The system was calibrated initially liquor content. Mutated spores were plated over the
such that a single colony of a non-mutated strain surface of petri dishes of the nutrient limited
would not produce an inhibition zone but that two medium such that ten to twenty colonies developed
such adjacent colonies would. In practice, the size of per plate. The time of incubation of the plates was
the inhibition zone was controlled by the depth of not quoted but ‘when the colonies were of a size
the overlay. Dulaney and Dulaney obtained a far suitable for accurate measurement’ a pasteurized
greater enrichment in the number of desired spore suspension of B. subtilis containing 0.16 units
phenotypes by the overlay technique than by ran- cm? of penicillinase (to limit the size of the inhibi-
dom selection and testing in liquid medium. tion zones) was dispersed over the surface of the
Ichikawa et al. (1971) screened for the increased dish. The plates were then incubated for 18 to 24
production ofthe antibiotic kasugamycin by a Strep- hours and the inhibition zones examined. Suitable
tomyces sp., using a miniaturized technique termed colonies were freed of contaminating B. subtilis by
the agar piece method. In order to prevent interfer- culturing on nutrient agar containing sodium
ence between colonies, mutated spores were grown penicillin and streptomycin sulphate. The use of
on plugs of agar which were then placed on assay this technique enabled three operators to scan
plates containing agar seeded with the indicator 15,000 survivors from ultraviolet irradiated spore
organism, levels of the antibiotic being determined _ populations in 3 months.
by the size of inhibition zones. By combining this The major disadvantages of the miniaturized
technique with a medium improvement programme technique approach is that productivity expressed
the authors obtained a tenfold increase in productiv- on the solid medium may not be expressed in sub-
ity. Ichikawa’s method was modified by Ditchburn sequent liquid culture and conversely, and more
et al. (1974) for the isolation of Aspergillus nidulans important, colonies not showing activity on solid
mutants synthesizing improved levels of penicillin. media may be highly productive in liquid medium.
These workers obtained promising results using the Despite these limitations the above workers have
technique and claimed that its potential for the demonstrated that the approach has considerable
recovery of higher-yielding mutants, for a given merit and Ball (1978) has claimed that the increase
expenditure of labour time, was greater than the in throughput may be as much as 20 times that ofa
shake-flask method. The level of production of conventional shake-flask programme.
penicillin by A. nzdulans is very small compared with In several cases the examination of randomly
P. chrysogenum and the technique would have to be selected high-yielding mutants has indicated that
modified considerably to be used commercially. their superiority may have been due to modifications
Ball and McGonagle (1978) developed a of their control systems. Goulden and Chataway
miniaturized technique suitable for the assay of (1969) demonstrated that a mutant ofP. chrysogenum,
penicillin production by ‘high-yielding’ strains ofP. producing high levels of penicillin, was less sensitive
chrysogenum (producing up to 6000 units cm~’). to the control of acetohydroxyacid synthase by val-
These workers highlighted the design of the solid ine. Valine is one of the precursors of penicillin and
medium as a critical factor in the optimization ofthe its synthesis is controlled by its feedback inhibition
method. They claimed that the growth ofa colony of acetohydroxyacid synthase. Thus, removal of the
on agar-solidified medium would be unlikely to control ofvaline synthesis may result in the produc-
modify the medium to the same extent as the growth tion of higher valine levels and, hence, greater pro-
of the organism on the same medium in submerged duction of penicillin. Pruess and Johnson (1967)
liquid culture. Thus, the nutrient-limiting condi- demonstrated that higher-yielding strains of P.
tions that favour the onset of antibiotic production chrysogenum also contained higher levels of acyl-
might not be achieved by the growth of the culture transferase, the enzyme which catalyses the addition
(
52
The Isolation, Preservation and Improvement of Industrial Micro-organisms
ey |
THE ISOLATION OF AUXOTROPHIC ~ Lysine — Biosynthetic route
MUTANTS ——-—* Feedback inhibition
Although mutation of secondary metabolite pro-
Fic. 3.25. Biosynthesis of benzyl penicillin and lysine in Penicil-
ducers to auxotrophy has resulted frequently in lium chrysogenum.
their producing lower yields, cases of improved
productivity have also been demonstrated. For (isoleucine, valine or threonine) were poor produc-
example, Alikhanian et al. (1959) investigated the ers of the antibiotic.
tetracycline producing abilities of fifty-three auxo- Many secondary metabolites may be considered
trophic mutants, all of which produced significantly as end products of branched pathways which also
less tetracycline than the parent strain in normal give rise to primary metabolites. Thus, a mutation
production medium. Supplementation of the to auxotrophy for the primary end product may also
medium with the growth requirement resulted in influence the production of the secondary product.
the expression of productivity superior to the parent In P. chrysogenum lysine and penicillin share the same
by one mutant. common biosynthetic route to a-aminoadipic acid,
It is difficult to explain the precise reason for the as shown in Fig. 3.25. This biosynthetic route may
effect of mutation to auxotrophy on the production explain Bonner’s (1947) observation that 25% of
of secondary metabolites but in the majority of cases the lysine auxotrophs of P. chrysogenum he isolated
it has been demonstrated to be an effect on the could not produce pencillin. The role of lysine in the
secondary metabolic system rather than, simply, an pencillin fermentation has also led to the investiga-
effect on the growth of the organism. The simplest tion of lysine auxotrophs as potential superior
explanation of the deleterious effect on secondary penicillin producers. Demain (1967) demonstrated
metabolite yield is that the auxotroph is blocked in that lysine was inhibitory to the biosynthesis of
the biosynthesis of a precursor of the end product, pencillin. The explanation of this phenomenon is
for example, Polsinelli e¢ al. (1965) demonstrated considered to be the inhibition of homocitrate synth-
that auxotrophs of Streptomyces antibioticus which ase by lysine resulting in the depletion of a-
required any of the precursors of actinomycin aminoadipic acid required for penicillin synthesis
53
Principles of Fermentation Technology
(Demain and Masurekar, 1974). It may be post- pathway regulation in primary metabolism where
ulated that lysine auxotrophs blocked immediately the activity of one pathway is affected by the product
after a-aminoadipic acid would produce higher of an apparently unrelated sequence. The alterna-
levels of penicillin due to the diversion of the inter- tive explanation is that the effect on secondary
mediate towards penicillin synthesis and_ the metabolism is not due to the auxotrophy but to a
removal of any control of homocitrate synthase by second mutation accompanying the auxotrophy, i.e.
endogenous lysine. O’Sullivan and Pirt (1973) a double mutation. Demain cited two attempts to
investigated the production of penicillin by lysine determine whether the effects of auxotrophy on
auxotrophs of P. chrysogenum in continuous culture secondary metabolism were due to double muta-
but were unable to demonstrate improved produc- tions or to the auxotrophy. MacDonald et al. (1963)
tivity in a range of lysine-supplemented media. reverted a low-producing thiosulphate-requiring
Luengo et al. (1979) examined the lysine regulation mutant of P. chrysogenum to thiosulphate indepen-
of penicillin biosynthesis in low-producing and dence and examined penicillin productivity by the
industrial strains of P. chrysogenum. The industrial revertants. Approximately half of the revertants
strain was capable of producing up to 12,000 units reacquired their ‘grandparents’’ production level,
cm’ penicillin. These workers demonstrated that whereas the other half retained their poor productiv-
although the onset of penicillin synthesis in the ity. Polsinelli e¢ al. (1965) reverted five isoleucine—
industrial strain was less sensitive to lysine than in valine auxotrophs of S$. antibioticus (which also pro-
the low-producing strain the empirical industrial duced low levels of actinomycin compared with the
selection procedures had not completely removed parent strain) to prototrophy and discovered that
the mechanisms of lysine regulation of penicillin some were returned to normal production and
biosynthesis. Luengo et al. expressed the possibility _ others to higher production levels than the ‘grand-
that overproducers ofpenicillin may be obtained by parent’. Thus, in the case ofPolsinelli ef a/.’s mutants
the selection of mutants to lysine regulation. How- it is unlikely that the effect on secondary metabolism
ever, it should be noted that the industrial strain was due to a double mutation, but it is possible that
employed was a representative of only one series of this was the case for some of Macdonald et al.’s
penicillin producers and that other series may have strains. However, it should be remembered that
already been modified with respect to the regulatory both these groups of auxotrophs were poor secon-
effects of lysine. Nevertheless, this study does indi- dary metabolite producers blocked in routes directly
cate the possibility that selection of strains resistant involved with the secondary biosynthetic pathway.
to lysine control may be overproducers ofpenicillin. It may be more relevant to examine the nature of
It is far more difficult to explain the effect of auxotrophic strains blocked in apparently unrelated
auxotrophy for factors not associated with the pathways and produce improved levels of the secon-
biosynthesis of the secondary metabolite. Polsinelli dary metabolite. It is not possible to say whether
et al. (1965) demonstrated that seven out of twenty- any of the auxotrophic mutants previously discussed
seven auxotrophs of S. griseus produced more produced superior levels of the secondary metabolite
actinomycin than did the prototrophic parent. as a result of double mutations but it may be possible
Dulaney and Dulaney (1967) demonstrated to exploit this possibility in the future.
increased chlortetracycline yields in eight out of The treatment of bacterial cells with nit-
eleven auxotrophs of S. viridifaciens when grown in rosoguanidine (NTG) has been demonstrated to
supplemented media. In neither of these cases were result in clusters of mutations around the replicating
the auxotrophic requirements directly involved in fork of the chromosome (Guerola et al. 1971). Thus,
the biosynthesis of the secondary metabolites. if one of the mutations were selectable (for example,
Demain (1973) put forward two possible explana- auxotrophy) it may be possible to isolate a strain
tions to attempt to account for the behaviour of such containing the selectable mutation along with other
auxotrophs. The first explanation is that the auxo- non-selectable mutations which map close by. The
trophic factors were involved in ‘cross-pathway’ efficient application of this technique would depend
regulation with the secondary metabolite or its pre- on the accurate mapping of the gene involved in
cursors. Demain quoted several examples of cross- producing the secondary metabolite so that neigh-
54
The Isolation, Preservation and Improvement of Industrial Micro-organisms
bouring mutations may be selected. This technique Addition of tryptophan to the improved strain
may be valuable for the selection of mutants of the would not result in higher pyrrolnitrin synthesis,
bacilli and streptomycetes producing high levels of indicating that tryptophan supply was no longer the
antibiotics where mutations affecting synthesis may limiting factor and the organism was producing
be mapped for each biosynthetic step. Co-mutation sufficient endogenous tryptophan for antibiotic syn-
by. NTG may then be followed by selection for thesis.
changes to genes adjacent to those loci known to be The role of tryptophan in the control of the can-
involved in production of the particular secondary dicidin fermentation has been investigated by Mar-
metabolite. tin and co-workers (Martin, 1978). Candicidin is a
polyene macrolide antibiotic containing an aroma-
tic p-aminoacetophenone moiety derived from
THE ISOLATION OF RESISTANT MUTANTS
chorismic acid, precursor of the aromatic amino
The isolation of analogue-resistant mutants has
acids, so that tryptophan and candicidin may be
already been discussed in the field of primary
considered to be end products of a branched
metabolism, where the rationale was that a mutant
biosynthetic pathway. Tryptophan has _ been
resistant to the toxic effects of an analogue, which
demonstrated to inhibit the biosynthesis ofthe anti-
mimics the control characteristics of the natural
biotic which led Martin et al. (1979) to isolate
metabolite, may overproduce the natural metabo-
mutants resistant to tryptophan analogues. Such
lite. This approach has been adopted, or may be
mutants produced more candicidin than the parent
adopted, in a number ofguises in the field of secon-
strain, apparently due to the removal of tryptophan
dary metabolism.
control.
(1) Mutants may be isolated which are resistant Godfrey (1973) isolated mutants of Streptomyces
to the analogues of primary metabolic pre- lipmanii resistant to the valine analogue,
cursors of the secondary metabolite. trifluoroleucine. These mutants produced higher
(ii) Mutants may be isolated which are resistant levels of cephamycin than the parent strain, and
to the feedback effects of the secondary appeared to be deregulated for the isoleucine,
metabolite. leucine, valine biosynthetic pathway, indicating
(iii) Mutants may be selected which are resistant that valine may be a rate-limiting step in cephamy-
to the toxic effects of the secondary metabo- cin synthesis.
lite when added to the trophophase of the Methionine has been demonstrated to stimulate
producing organism. the biosynthesis of cephalosporin by Acremonium
(iv) Mutants may be isolated which are resistant chrysogenum and superior producers have been iso-
to the toxic effects of a precursor of the lated in the form of methionine analogue-resistant
secondary metabolite. mutants (Chang and Elander, 1979).
(i) Mutants resistant to the analogues of primary metabolic (11) Mutants resistant to the feedback effects of the secondary
precursors of the secondary metabolite metabolite
This approach has been adopted by Elander et al. There are many cases in the literature of a secon-
(1971) in the isolation of mutants of Pseudomonas dary metabolite preventing its own synthesis. Mar-
aureofaciens overproducing the antibiotic pyrrolnit- tin (1978) cites the following examples: chloram-
rin. Tryptophan is a precursor of pyrrolnitrin and phenicol, aurodox, cycloheximide, staphylomycin,
although it is stimulatory to production it is ristomycin, puromycin, fungicidin, candihexin,
uneconomic to use as an additive in an industrial mycophenolic acid and penicillin. The precise
process. Thus Elander et al. (1971) isolated mutants mechanisms of these controls is not clear but they
resistant to tryptophan analogues using the gradient appear to be specific against the synthesis of the
plate technique described previously. A strain was secondary metabolite and not against the general
eventually isolated which produced two to three metabolism of the cells.
times more antibiotic than the parent and was The mechanism of the control of its own synthesis
resistant to feedback inhibition by tryptophan. by chloramphenicol appears to be the repression of
55
Principles of Fermentation Technology
arylamine synthetase (the first enzyme in the path- only sensitive to levels of the antibiotic in excess of
way from chorismic acid to chloramphenicol) by that originally incorporated into the medium. In
chloramphenicol. Jones and Vining (1976) cases where it has been demonstrated that feedback
demonstrated that arylamine synthetase was fully control by the end product plays an important role
repressed by 100 mg dm * chloramphenicol, a level in limiting productivity it would probably be worth
of antibiotic which neither affected cell growth nor while to design such procedures.
the activities of the other enzymes of the chloram-
phenicol pathway. (iii) The isolation of mutants resistant to the toxic effects of
Martin (1978) quoted several examples of correla- the secondary metabolite in the trophophase
tions between the level of secondary metabolite It has been demonstrated for many secondary
accumulation and the level which causes ‘feedback’ metabolite producing organisms that the secondary
control, which may imply that the limiting factor to metabolite is toxic to the producing cell when it is
the increased yields of some secondary metabolites | present in the trophophase (growth phase) (De-
is the feedback inhibition by the end product. The main, 1974). Thus, it appears that a ‘switch’ in the
selection of mutants resistant to feedback inhibition metabolism ofthe organism in the idiophase enables
by a secondary metabolite is a far more difficult task it to produce an otherwise ‘autotoxic’ product. Fur-
than the isolation of strains resistant to primary thermore, it has been demonstrated that, in some
metabolic control. It is extremely unlikely that a cases, the higher the resistance to the secondary
toxic analogue of the secondary metabolite could be metabolite in the growth phase the higher is produc-
found where the toxicity lay in the mimicing of the tivity in the production phase. Dolezilova et al.
feedback control of the secondary metabolite, a (1965) demonstrated that the level of production of
compound which would not be necessary for growth. ‘nystatin by various strains of S. noursei was related to
However, the detection of mutants resistant to feed- the resistance of the strain to the antibiotic in the
back inhibition by antibiotics may be achieved by growth phase; a non-producing mutant was inhi-
the use of solid media screening techniques, similar bited by 20 units cm~*, the parent strain produced
to the miniaturized screening techniques previously 6000 units cm~* and was inhibited by 2000 units
described. The technique would involve culturing cm ° in the growth phase and a mutant producing
the survivors of a mutation treatment on solid 15,000 units cm~* was found to be resistant to
medium containing hitherto repressing levels of the 20,000 units cm~°.
antibiotic and detecting improved producers by The possible relationship between antibiotic
overlaying the colonies with an indicator organism. resistance and productivity may be used to advan-
The difficulty inherent in this technique is that the tage in the selection of high-producing mutants by
incorporated antibiotic, itself, will inhibit the culturing the survivors of a mutation treatment in
development of the indicator organism. This prob- the presence of a high level of the end product.
lem may be overcome by adjusting the depth ofthe Those strains capable of growth in the presence of a
overlay or the concentration of the indicator such high level of the antibiotic may also be capable of
that an inhibition zone could only be produced by a high productivity in the idiophase. This approach
level of antibiotic greater than that incorporated in has been used successfully for antifungal agesterols
the original medium. However, it is unlikely that (Bu’Lock, 1980), streptomycin (Woodruff, 1966)
this would be a satisfactory solution for the selection and ristomycin (Trenina and Trutneva, 1966) but
of a high-producing commercial strain. Another without success for novobiocin (Hoeksema and
approach would be to utilize an analogue of the Smith, 1961).
antibiotic which mimiced the feedback control by
the natural product but which did not have antimic- (iv) The isolation of mutants resistant to the presence of
robial properties. Inhibition of antibiotic synthesis toxic precursors in the growth phase.
by analogues has been demonstrated in the cases of A secondary metabolite precursor which is added
aurodox (Liu et al., 1972) and penicillin (Gordee to the fermentation medium may prove to be toxic to
and Day, 1972). An alternative may be to use a the producing organism in the growth phase. An
mutant indicator bacterium for the overlay which is example of this phenomenon is the toxic effect of
56
The Isolation, Preservation and Improvement of Industrial Micro-orga
nisms
phenylacetic acid in P. chrysogenum when used as a associated with the auxotrophy. Similarly, revertant
precursor in the penicillin fermentation. It appears mutants may affect secondary metabolism in a
that strains capable of withstanding higher concen- number of ways—direct effects on the pathway,
trations of the precursor may be able to synthesize cross-pathway effects and effects due to mutations
higher levels of the end product, suggesting that the other than the detected reversion.
production of the end product results in the detoxifi- Dulaney and Dulaney (1967) investigated the
cation of the precursor. Polya and Nyiri (1966) tetracycline productivities of apopulation of proto-
applied this hypothesis in selecting phenylacetic troph revertants of S$. viridifaciens derived from each
acid-resistant mutants of P. chrysogenum and of five auxotrophs. These workers predicted that
demonstrated that 7% of the isolates showed some revertants may be productive due to direct
enhanced penicillin production. influence of the mutations on tetracycline biosyn-
Ball (1978) stressed that the major difficulty in thesis but that others may be so because they con-
the selection of toxic precursor resistant mutants is tained other lesions. Superior producers were
that the site of resistance of amutant may not result obtained in all the prototroph-revertant populations
in the organism overproducing the end product—for apart from those derived from a homocysteine auxo-
example, the resistance may be due to an alteration troph. However, the frequency of the occurrence of
in the permeability of the mutant or due to the the superior producers was similar to that obtained
mutants’ ability to degrade the precursor to a harm- by the random selection of the survivors of amuta-
less metabolite unrelated to the desired end product. tion treatment in all but one of the populations. The
exceptional population was the revertants of a
methionine auxotroph, 98% of which produced
THE ISOLATION OF REVERTANT
between 1.2 and 3.2 times as much tetracycline as
MUTANTS
the original prototrophic culture. A possible expla-
As discussed in the section on primary metabo-
nation ofthe very favourable titres of the population
lites, a,mutant may revert to the phenotype of its
may be the role of methionine as the methyl] donor in
‘parent’, but the genotype of the revertant may not,
tetracycline biosynthesis and that methionine avail-
necessarily, be the same as the original ‘parent’.
ability limited the production of the secondary
Some revertant auxotrophs have been demonstrated
metabolite in the original prototroph. However,
to accumulate primary metabolites (Shiio and Sano,
addition of exogenous methionine to the prototroph
1969) and attempts have been made to apply the
did not result in superior productivity.
technique to the isolation of mutants overproducing
Polsinelli et al. (1965) also demonstrated that
secondary metabolites. Two approaches have been
reversion offive mutants of an actinomycin produc-
used in the field of secondary metabolites with
ing strain ofS. antibioticus blocked in the isoleucine—
respect to the isolation ofrevertants:
valine pathway resulted in the isolation of some
(i) The isolation of revertants of mutants auxo- superior mutants. Godfrey (1973) reported that the
trophic for primary metabolites which may reversion of a cysteine auxotrophic mutant of S.
influence the production of a secondary lipmanit resulted in improved production of
metabolite. cephamycin.
(ii) The reversion of mutants which have lost the These studies provide promising evidence that
ability to produce the secondary metabolite. the selection of revertants of auxotrophs of primary
metabolites involved in secondary metabolism may
(i) The isolation of revertants of mutants auxotrophic for yield a high number of productive mutants.
primary metabolites which may influence the production
of a secondary metabolite (ii) The isolation of revertants of mutants which have lost
As previously discussed, it is difficult to account the ability to produce the secondary metabolite
for some of the effects of auxotrophy on secondary The reversion of non-producing strains may result
metabolism and it appears that some may be due to in the detection of a high-producing mutant as that
as yet unresolved cross-pathway phenomena and mutant would have undergone at least two muta-
others due to the expression of other mutations tions associated with the production of the secon-
57
Principles of Fermentation Technology
dary metabolite. Dulaney and Dulaney (1967) tems for both strain improvement and for obtaining
plated the progeny ofa mutation ofa non-producing information on the genetics of industrial organisms.
strain of §. viridifaciens on to solid production
medium and screened for superior tetracycline pro- THE APPLICATION OF THE PARASEXUAL
ducers by an overlay technique. A mutant was CYCLE
isolated which produced 9 times the tetra- Many industrially important fungi do not possess
cycline yield of the original ‘parent’. The major a sexual stage and thereore it would appear difficult
difficulty inherent in this technique is that non- to achieve recombination in these organisms. How-
producing mutants of high-yielding strains may ever, Pontecorvo et al. (1953) demonstrated that
be incapable of being reverted due to extreme nuclear fusion and gene segregation could take place
deficiencies in their metabolism. Nevertheless, outside, or in the absence of, the sexual organs. The
reversion techniques may be extremely worth while process was termed the parasexual cycle and has
where an accurate ‘overlay’ screening system may been demonstrated in the imperfect fungi, A. niger
be employed. and P. chrysogenum, as well as the sexual fungus A.
nidulans. In order for parasexual recombination to
occur in an imperfect fungus, nuclear fusion must
occur between unlike nuclei in the vegetative hyphae
The use of recombination systems
of the organism. Thus, recombination may only be
for the improvement of industrial
achieved in an organism in which at least two
micro-organisms
different types of nuclei coexist, i.e. a heterokaryon.
The heterozygous diploid nucleus resulting from the
Hopwood (1979) defined recombination, in its -fusion of the two different haploid nuclei may give
broadest sense, as ‘any process which helps to gener- rise to a diploid clone and, in rare cases, a diploid
ate new combinations of genes that were originally nucleus in the clone may undergo an abnormal
present in different individuals’. The use of recombi- mitosis resulting in mitotic segregation and the
nation mechanisms for the improvement of indus- development of recombinant clones which may be
trial strains has been fairly limited compared with either diploid or haploid.
the use of mutation techniques. This may be Recombinant clones may be detected by their
attributable to the lack of basic knowledge of the display of recessive characteristics not expressed in
genetics of industrial micro-organisms and the rela- the heterokaryon. Analysis of the recombinants nor-
tive ease of applying the technique of mutation and mally demonstrates them to be segregant for only
selection to these organisms compared with the one, or a few linked, marker and culture of the
application of recombination techniques. The segregants results in the development of clones dis-
remarkable success of mutation programmes may playing more recessive characters than the initial
have further relegated the adoption of recombina- segregant. The process of recombination during the
tion systems to a situation where an apparent growth of the heterozygous diploid may occur in two
‘plateau’ in yield improvement had been reached in ways: mitotic crossing over, which results in diploid
the mutation programme. However, the demonstra- recombinants, and haploidization, which results in
tion by Pontecorvo and Sermonti (1954) of the haploid recombinants.
parasexual cycle in P. chrysogenum stimulated the Mitotic crossing over is the result of an abnormal
application of recombination to the improvement of mitosis. The normal mitosis of aheterozygous dip-
penicillin yields and major advances in the genetics loid cell is shown in Fig. 3.26. During mitosis, each
of the streptomycetes has resulted in the develop- pair of homologous chromosomes replicate to pro-
ment of recombination techniques for this indust- duce two pairs of chromatids and a chromatid ofone
rially important group. Also, techniques have been pair migrates to a pole of the cell with a chromatid of
developed which may enable the transfer of genetic the other pair. Division of the cell at the equator
information between different microbial taxa, over- results in the production of two cells, both of which
coming some recombination barriers. Thus, there is are heterozygous for all the genes on the chromo-
increasing interest in the use of recombination sys- some. Mitotic crossing over involves the exchange of
58
The Isolation, Preservation and Improvement of Industrial Micro-organisms
Diploid cell
where 2n = 2
Chromatid
replication
as)
Spindle
formation
Separation of
chromatids
Se)@
Diploid cell
where 2n = 2
Chromatid
replication
Spindle
formation
Abnormal
segregation
Soe
<)
Daughter cells
2n+ 1 and 2n—1
59
Principles of Fermentation Technology
rially important fungi has been hindered by a chrysogenum and demonstrated that there were five
number of problems (Elander, 1980). A major diff- different genes which deleteriously affected penicil-
culty is the influence of the auxotrophic markers lin production, four of which would appear to be on
used for the selection of the heterokaryon on the one chromosome. Work on the genetically more
synthesis of the desired product. As discussed ear- familiar A. nidulans has resulted in the allocation of
lier, auxotrophic mutations have been observed to three genes coding for raised penicillin yield to their
have quite unpredictable results on the production respective chromosomes and significant progress
of some secondary metabolites. Even conidial colour has been made in mapping these genes using a
markers have been shown to have deleterious effects sexual cross (A. nidulans possessing a sexual stage).
on product synthesis (Elander, 1980). Even when Another important piece of information gained by
suitable markers are available, the induction of the use ofthe parasexual cycle is that at least some of
heterokaryons in some industrial fungi has been the mutations involved in superior penicillin pro-
demonstrated to be a difficult process and special- duction are recessive in P. chrysogenum (Elander,
ized techniques have had to be developed to increase 1967) although Macdonald and Holt (1976) showed
the probabilities of heterokaryon formation. Mac- A. nidulans to possess some dominant mutations
donald and Holt (1976) have discussed the giving superior penicillin productivity. It is impor-
techniques used. A method which appears to remove tant to remember that the full benefit of such basic
some of the difficulties of heterokaryon formation is genetic investigations can only be achieved if the
protoplast fusion and this technique and its applica- biochemical expression of the studied mutations is
tions are discussed later in the chapter. discovered so that the genetic work must be carried
Despite the difficulties of inducing the parasexual out in association with detailed biochemical investi-
cycle in some industrial fungi it has been used in two - gations.
ways to study these organisms. The cycle may be The application of the parasexual cycle to the
used to investigate the genetics of the producing breeding of superior strains has met with some
strain as well as to isolate recombinants which may success in the production ofpenicillin. According to
be superior producers. Information which may be Queener and Swartz (1979) the following three
obtained on the basic genetics of the producing factors have made the breeding of superior penicillin
fungus using the parasexual cycle include the producing strains a more attractive proposition than
number of chromosomes (or linkage groups), the was previously thought:
allocation of genes, important in product synthesis,
to particular chromosomes and the mapping of (i) The United States patent on the use of the
important genes on a chromosome. Sermonti (1969) parasexual cycle for the breeding of micro-
and Macdonald and Holt (1976) have described the organisms to higher yields (Pontecorvo and
techniques which may be used to achieve these Roper, 1958) expired in January 1975.
objectives. (ii Ee, In recent years it has become more difficult
Investigation ofthe parasexual cycle of organisms to increase the yield of penicillin by mutation
producing penicillin has provided some information and selection techniques.
on the genetics of penicillin production. The two (iii) Companies such as Panlabs Inc. have been
producing organisms which have been used are P. producing P. chrysogenum strains for clients
chrysogenum (the commercial producer of penicillin) to employ for penicillin production. Thus,
and A. nidulans which is a low producer of the strains different from those previously
antibiotic but has been well studied genetically and employed by the producing firm are
has been used as a model system for the investigation available for breeding studies.
of the biosynthesis of penicillin. In studies on P. Queener and Swartz claimed that a number of the
chrysogenum, Macdonald et al. (1965) concluded that major penicillin-producing companies are involved
the fungus had three, or possibly four, chromosomes in programmes to increase yields by breeding
and Ball (1973) demonstrated that there were at superior strains. Both haploid and diploid recom-
least three linkage groups in the organism. Holt e¢ al. binants have been investigated and three basic types
(1976) investigated non-producing mutants of P. of cross have been attempted. These crosses have
60
The Isolation, Preservation and Improvement of Industrial Micro-organisms
been described by Ball (1978) as sister, ancestral haploid mutants and mutation and selection of the
and divergent. A sister cross is between strains diploid gave rise to a diploid strain producing higher
differing only in the markers used to induce the levels of penicillin than the parents. Ball (1978)
heterckaryon or to recognize the segregants. An claimed that the usefulness of diploids may only be
ancestral cross is one between strains, one developed short term, presumably implying problems of
from the other, differing in the possession of muta- degeneration, and may best be used as stepping
tions which increase productivity as well as in the stones to a recombinant haploid. It may be possible
presence of markers. The progeny of an ancestral to control the degeneration of a diploid by incor-
cross may then combine the high yield of the recent porating non-homologous recessive lethal muta-
mutant with, perhaps, some ofthe desirable proper- tions on separate chromosomes of an homologous
ties of the ancestor such as good sporulation or a pair (Ball, 1973). The deleterious effects of the
high growth rate which may have been lost in the mutation would be repressed by the dominant
selection programme. A divergent cross is one alleles in the diploid but would be expressed in a
between two strains derived from different mutant haploid derived from the diploid, resulting in any
‘lines’ which contain, presumably, different yield- haploids being non-viable.
improving mutations as well as different markers. The advantages to be gained from the industrial
The progeny ofa divergent cross may contain new use of parasexual recombinants may not be confined
combinations ofyield improving mutations. to the amalgamation of different yield improving
Early studies on haploid strains of P. chrysogenum mutations. Equally advantageous would be the
derived from parasexual crosses demonstrated that breeding out of characteristics which are deleterious
most of the progeny exhibited the genotype ofone of to high productivity. Also, the interspersion of
the parents, i.e. no recombination had occurred recombinant techniques in a mutation and selection
(Macdonald, 1968). Macdonald suggested that one programme may result in the recovery of strains
of the reasons for this lack of success may have been which are more susceptible to subsequent mutation
the differences in gross chromosomal morphology treatments. Ball (1978) has claimed that a recombi-
between the parents caused by certain mutagens nation programme, under certain circumstances,
used in the development of the strains. Thus, one may be as successful as orthodox mutation program-
precaution to be adopted in the development of mes. As more information is gained ofthe genetics of
strains which may be used in a parasexual cross is the important industrial fungi then parasexual
the avoidance of mutagens which may cause gross recombination should prove to be a very powerful
changes in chromosomal morphology. Sub- tool in strain manipulation.
sequently Ball (1978) has suggested that the careful
choice of markers and the use of strains giving THE APPLICATION OF RECOMBINATION
similar titres and being ‘not too divergent’ may SYSTEMS IN THE ACTINOMYCETES
result in achieving more recombinants. The actinomycetes are responsible for the produc-
Although diploids produced by the parasexual tion of a large range of antibiotics and, in the past,
cycle are frequently unstable, stable diploids have the predominant exercise in strain improvement has
been reported to have been used for the industrial been the induction and selection of mutants. How-
production of penicillin. Elander (1967) isolated a ever, as in the case of the imperfect fungi, recombina-
diploid from a sister cross of P. chrysogenum which tion techniques should enable the breeding of
was shown to be a better penicillin producer than superior industrial strains. Hopwood (1976) has
the parents and was morphologically more stable. considered the case for the use of conjugation in the
One explanation of the superior performance ofthe breeding of industrial strains of actinomycetes and
diploid may have been its resistance to strain degen- referred to the successful use of the technique by
eration caused by deleterious recessive mutations some industrial companies who were not free to
because such mutations would only have been publish their findings.
expressed in the diploid if both alleles had been Conjugation involves the transfer of chromosomal
modified. Calam et al. (1976) demonstrated that a material from one bacterial cell to another via a
diploid strain of P. chrysogenum was more stable than connection between the participating cells. The
61
Principles of Fermentation Technology
zygote which results from conjugation contains a some of the structural genes for the synthesis of
complete circular chromosome from one parent and methylenomycin A in S. coelicolor are plasmid borne.
a segment of the chromosome of the other parent. The involvement of plasmids in the synthesis ofthe
The segment may be incorporated into the circular other antibiotics quoted may be in the control of
chromosome by a crossing-over procedure which their synthesis or in the resistance of the producer
results in the generation of a recombinant chromo- cells to their own antibiotics. Chater (1979) quoted
some, i.e. a chromosome containing information several examples where loss of plasmids conferring
which has arisen from two different cells. Conjuga- resistance to endogenous antibiotics also resulted in
tion is a rare occurrence and, therefore, to isolate the the loss of the ability to synthesis the antibiotic.
recombinants it is necessary to construct a selection The involvement of plasmids in actinomycete
procedure which places the recombinants at an antibiotic production has led to speculations of ways
advantage over the parent cells. This advantage of improving yields by plasmid manipulation. Hop-
may be achieved by introducing different auxo-. wood (1978) discussed the possibilities of transfer of
trophic markers into the parents and culturing the synthesis of an antibiotic to a strain which is more
progeny ofthe association on minimal medium. The suitable for commercial production than the original
recombinants isolated in this manner would not be producer. Such transfers may involve plasmids
completely representative of all the recombinants involved in antibiotic synthesis as well as plasmid
produced because those recombinants which did vectors to transfer otherwise chromosomal genes.
not include both the wild-type alleles would not be The enhancement ofantibiotic synthesis by increas-
selected. Thus, it is desirable to use different combi- ing the number of copies of the gene coding for the
nations of markers in different crosses to obtain a synthesis of the antibiotic is a difficult task because
more representative sample of recombinants. This “increased productivity would probably only be
approach may only be adopted ifthe positions of the achieved if all the genes in the pathway were dupli-
markers on the chromosome are known and Hop- cated. The roles of protoplast fusion and genetic
wood (1976) has claimed that this is not an insuper- manipulation techniques in the actinomycetes are
able task and could be achieved relatively easily for considered in the following sections.
most industrial actinomycetes.
Recombinants isolated by the above techniques THE APPLICATION OF PROTOPLAST
may then be screened for increased productivity. FUSION TECHNIQUES
Crosses between ancestral strains should give rise to Protoplasts are cells devoid of their cell walls and
new combinations of alleles and may also result in may be prepared by subjecting cells to the action of
the removal of deleterious mutations or render a wall degrading enzymes in isotonic solutions. Pro-
strain more susceptible to subsequent mutation. toplasts may regenerate their cell walls and are then
More ambitious crosses, such as between divergent capable of growth as normal cells. Cell fusion, fol-
strains, and attempts to combine specific genes for lowed by nuclear fusion, may occur between proto-
improved productivity requires more basic informa- plasts of strains which would otherwise not fuse and
tion on the genetics of industrial strains. However, the resulting fused protoplast may regenerate a cell
the empirical use of conjugation between sister and wall and grow as a normal cell. Thus, protoplasts
ancestral strains and associated mutation and selec- may be used to overcome some recombination bar-
tion should provide promising results in the short riers. Protoplast fusion has been demonstrated in
term. Streptomyces sp. (Hopwood et al., 1977), Bacillus sp.
Plasmids (extrachromosomal DNA) have been (Fodor and Alfoldi, 1976), filamentous fungi
demonstrated to be involved in the synthesis of (Ferenczy et al., 1974) and yeasts (Sipiczki and
several actinomycete antibiotics. Okanashi (1979) Ferenczy, 1977).
quoted that plasmid involvement had been studied The frequency of recombination in the filament-
in aureothricin, cephamycin, chloramphen icol, ous fungi by protoplast fusion is such that it is
holomycin, kanamycin, kasugamycin, methyl eno- necessary to introduce auxotrophic markers to assist
mycin A, oxytetracycline, puromycin, strepto mycin in the recovery of fused protoplasts. Fusion of fungal
and turimycin. There is strong evidence that at least protoplasts appears to be an excellent technique to
62
The Isolation, Preservation and Improvement of Industrial Micro-organisms
obtain heterokaryons between strains where con- genome in recombination. Also, Hopwood (1979)
ventional techniques have failed, thus allowing the has developed techniques which have resulted in the
use of the parasexual cycle for breeding purposes in recovery ofa very high proportion of recombinants
situations where it had not been previously possible. from the fusion products of Streptomyces coelicolor
This situation is illustrated by the work of Peberdy protoplasts. By subjecting protoplasts to an expo-
et al. (1977) who succeeded in obtaining heteroka- sure ofultraviolet light, sufficient to kill 99% of them
ryons between P. chrysogenum and P. cyaneo-fulvum prior to fusion, Hopwood has claimed a tenfold
and demonstrated the formation of diploids which increase in recombinant detection for strains nor-
gave rise to recombinants after treatment with p mally giving a low yield of recombinants (1%) and
fluorophenylalanine or benomyl. Although it has a doubling of the recombination frequency for a
been claimed that P. chrysogenum and P. cyaneo-fulvum cross normally yielding 20% recombinants. If such
are not different species of Penicillium (Samson et al., yields of recombinants are possible for industrial
1977), Peberdy e¢ al. still demonstrated that proto- strains, then the technique may be extremely valu-
plast fusion could be successful where conventional able. The recombinants would be detectable by
techniques had failed. simply screening a random proportion of the prog-
A demonstration ofthe use of protoplast fusion for eny ofa protoplast fusion and the use of auxotrophic
an industrial fungus is provided by the work of markers to ‘force’ out the recombinants would not
Hamlyn and Ball (1979) on the cephalosporin pro- be necessary as it is in the detection of recombinants
ducer, C. acremonium. These workers compared the from conjugation experiments. Wesseling (1982)
effectiveness of conventional techniques of obtaining has reveiwed the potential of this technique for the
nuclear fusion between strains of C. acremonium with manipulation ofthe industrial actinomycetes.
the protoplast fusion technique. The results from
conventional techniques suggested that nuclear THE APPLICATION OF GENETIC
fusion was difficult to achieve. Electron microscopic MANIPULATION TECHNIQUES
examination of fused protoplasts indicated that up In recent years, the transfer of DNA between
to 1% underwent immediate nuclear fusion. Recom- different species of bacteria has been achieved using
binants were obtained in both sister and divergent both in vivo and in vitro techniques (Atherton et al.,
crosses. A cross between an asporulating, slow- 1979). Thus, genetic material derived from one
growing strain with a sporulating fast-growing species may be incorporated into another where it
strain which only produced one-third of the may be expressed.
cephalosporin level of the first strain eventually In vivo techniques make use of phage particles
resulted in the isolation of a recombinant which which will pick up genetic information from the
combined the desirable properties of both strains, chromosome of one bacterial species, infect another
i.e. a strain which demonstrated good sporulation, a bacterial species and in so doing introduce the
high growth rate and produced 40% more antibiotic genetic information from the first host. The informa-
than the higher-yielding parent. Chang ef al. (1982) tion from the first host may then be expressed in the
utilized protoplast fusion to combine the desirable second host. Such techniques depend on the exis-
qualities of two strains of Penicillium chrysogenum. tence of so-called promiscuous sex factors (Hop-
Protoplasts from two strains, differing in colony wood, 1979) which will infect and transfer genetic
morphology and the abilities to produce penicillin V information between a range of different bacteria
(the desired product) and OH-V penicillin (an and are not host specific. Plasmids which will transfer
undesirable product), were fused, followed by plat- between a wide range of gram negative organisms
ing on a non-selective medium. Out of 100 stable are available but no such factors are yet available for
colonies which were scored two possessed the desira- the gram positive bacteria. As discussed previously,
ble morphology, high penicillin V and low OH-V protoplast fusion has been achieved in two genera of
penicillin productivities. gram positive bacteria (Streptomyces and Bacillus)
Protoplast fusion has particular advantages over and this may serve as an alternative to the use of the
conjugation in the industrial actinomycetes, in that promiscuous phage of the gram negative type.
the technique involves the participation ofthe entire Whereas, the in vivo techniques depend on vectors
63
Principles of Fermentation Technology
64
The Isolation, Preservation and Improvement of Industrial Micro-organisms
65
00
Principles of Fermentation Technology
The major difficulty to be overcome in the use of products and the morphological form of the
genetically manipulated cultures is their potential organism.
instability in large-scale culture, especially continu-
ous processes. The genetically manipulated strain of THE SELECTION OF STABLE STRAINS
Methylomonas methylotrophus appears to be stable in The ability of the producing strain to maintain its
continuous culture, but this is not surprising since high productivity during culture maintenance and a
the modification renders the cell more efficient and, fermentation is a very important quality. Yield
thus, selection force in the chemostat would tend to decay during culture storage may be avoided by the
work to its advantage. However, strains which have use of maintenance techniques such as those dis-
been manipulated to produce products which give cussed earlier in this chapter, but loss of productivity
the cell no selective advantage would be selected during the fermentation is far more difficult to con-
against in long-term culture, especially in chemostat trol. A decrease in the productivity of acommercial
systems. Thus, the development of techniques to strain is normally due to the occurrence of lower-
stabilize manipulated cultures and the incorpora- yielding, spontaneous revertant mutants which fre-
tion of selectable marker genes consistent with the quently havea higher growth rate than the high-pro-
use of cheap fermentation media are critical to the ducing parent, so that yield decay is especially
successful fulfilment of much of the promise of gene- problematical in long-term fermentations such as
tic-manipulation techniques. fed-batch or continuous culture where the faster-
growing, lower producer may predominate, or even
replace, the high-producing original strain. This
The improvement of industrial strains
situation is illustrated by the commercial, amino
by modifying properties other than the
acid-producing organisms, many of which are
yield of product
insufhciently stable to be used in continuous-culture
processes. Bu’ Lock (1979) claimed that the future of
The previous sections of this chapter have consi- some of the amino acid fermentations was in doubt
dered ways of increasing the yields of metabolites because of their inability to compete with new chem-
produced by industrial micro-organisms. However, ical and immobilized enzyme processes unless con-
the design and economics of a commercial process tinuous fermentations could be used, which would
are influenced by properties of the organism other require the development of very stable strains. How-
than its productivity. For example, although a strain ever, several workers have attempted to control the
may produce a very high level of a metabolite it stability of amino acid-producing strains. As may be
would be unsuitable for a commercial process ifits seen from the previous section on the isolation of
productivity were extremely unstable, or if the mutants overproducing primary metabolites the
organism’s oxygen demand were such that it could amino acid producers tend to be auxotrophs,
not be satisfied in the industrial fermenter available analogue-resistant mutants or revertants; such
for the process. Therefore, characteristics of the mutations removing the normal mechanisms con-
producing organism which affect the process may be trolling the production of amino acids. The intro-
critical to its commercial success. Thus, it may be duction of more than one mutation giving the same
desirable to modify such characteristics of the pro- phenotype may give a more stable strain since all the
ducing organism which may be achieved by select- mutations would have to revert for the strain to lose
ing natural and induced variants and recombinants. its productivity.
Naturally, it is crucial that the modified strain Woodruff and Johnson (1970) selected a double
retains its desirable productivity so that the screen- auxotrophic mutant of Micrococcus glutamicus requir-
ing involved in these procedures should include ing both homoserine and threonine and compared
assay of the yield as well as the characteristics being its lysine-producing properties with those of a
selected. Some examples of the characteristics which homoserine auxotroph. The authors claimed that
may be importantin this context are: strain stability, the double mutant had a twofold advantage in that
resistance to phage infection, response to dissolved it produced higher levels of lysine compared with
oxygen, tolerance of medium components, the pro- the single auxotroph and was also less susceptible to
duction of foam, the production of undesirable by- reversion to low productivity, the stability of the
66
The Isolation, Preservation and Improvement of Industrial Micro-organisms
double auxotroph being such that it was a suitable phage infections because new host range phages
organism for the production oflysine by continuous may be introduced on to the plant or phage mutants
culture. appear. Plant hygiene is essential to minimize risk of
Nakayama (1972) discussed the problem ofstrain contamination and it is also possible to utilize chem-
reversion in the lysine fermentation and cited an ical agents in the fermentation which selectively
example of a fermentation where 87% of the cells inhibit phage replication (Hongo e¢ al., 1972).
from a 60—88-hour culture were revertants. This Infection of some fermentations with ‘wild’ micro-
very high level of revertants was probably due to the organisms may be made more easily controlled by
faster growth rate of the revertant compared with selecting commercial strains which are resistant to
the single auxotroph in homoserine-limited culture. various antibiotics. The antibiotics to which the
This situation was controlled, to a certain extent, by commercial strain is resistant may then be used to
the incorporation of antibiotics, such as erythromy- control the level of contaminants. The danger inher-
cin, which are more inhibitory to the rapidly grow- ent in this technique is that resistant contaminants
ing revertants than the homoserine limited auxo- will also tend to be selected but reasonable protec-
trophs. However, the use of mutants which have tion should be given provided that stringent sterili-
multiple markers appears to be a better solution, as zation of a contaminated fermenter is carried out
this considerably reduces the probability of rever- before the vessel is used again so that any antibiotic-
sion to the wild type as reversion of all the markers resistant contaminants are removed.
must occur, which should be an extremely rare
event. Sano and Shiio (1970) cited the use ofa C.
glutamicum mutant for the industrial production of THE SELECTION OF NON-FOAMING
lysine which was auxotrophic for homoserine and STRAINS
leucine and was resistant to S-(2-aminoethyl)-L-cys- Foaming during a fermentation may result in the
teine and produced 39 g dm° lysine. loss of broth, cells and product via the air outlet as
The stability of fungal diploids used in commer- well as putting the fermentation at risk from con-
cial fermentations may be controlled by a technique tamination (see Chapter 9). Thus, foaming is nor-
discussed by Ball (1973). Ball claimed that it may be mally controlled by either chemical or mechanical
possible to control the degeneration of a diploid into means (Hall et al., 1973), but this task may be made
haploids by incorporating non-homologous reces- easier if a non-foaming strain of the commercial
sive lethal mutations on separate chromosomes of organism can be developed. Foaming which occurs
an homologous pair in the diploid. The deleterious early in the fermentation is usually due to a compo-
effects of the mutations would be repressed by the nent in the medium whereas foaming late in the
dominant alleles in the diploid but would be expres- fermentation is normally a property of the growing
sed in a haploid derived from the diploid, resulting organism and, therefore, it is only this latter type of
in any haploids being non-viable. foam which may be controlled by strain selection.
An excellent example of selection of a non-foaming
strain is provided by the work of Mindlin e¢ al.
THE SELECTION OF STRAINS RESISTANT (1961) on oxytetracycline production by S$. rimosus.
TO INFECTION This work utilized recombination rather than
Bacterial fermentations may be affected very seri- mutagenesis and involved a cross between two
ously by phage infections, which may result in the strains ofS. rimosus which resulted in the isolation of
lysis of the bacteria. A possible method for reducing a hybrid which did not foam in rich media. However,
the risk of failure due to phage contamination is to it should be remembered that both the original
select bacterial strains which are resistant to the strains foamed and this is not an example of combin-
phages isolated in the fermentation plant (Hongo et ing together different characteristics previously
al., 1972). It is important that the apparent resistant expressed in different organisms. Nevertheless,
strains isolated are not lysogenic as the carrying ofa recombination probably has more potential in this
population of phages in the fermentation is a source field than mutagenesis and may prove to be a valu-
of potential lytic phage mutants. The use of phage- able tool in recombining characteristics from differ-
resistant mutants does not ensure immunity from ent strains in a strain-development programme.
PFT-F
67
Principles of Fermentation Technology
wort (and hence the conversion of the wort to THE ISOLATION OF MUTANTS PRODUCING
alcohol) and the ease of clarification of the beer. NEW FERMENTATION PRODUCTS
Although mutant selection is rarely applied, the The isolation of organisms from the natural envi-
selection of natural variants is a very common prac- ronment synthesizing commercially useful metabo-
tice in many breweries. As discussed in Chapter 6, lites is an expensive and laborious process. There-
the yeast produced from one beer fermentation may fore, other means of producing novel compounds
be used to inoculate a new fermentation, a practice which may be of some industrial significance have
which is very rarely employed in the rest of the been attempted. Probably the most successful alter-
fermentation industry. By the careful selection of native approach has been the semi-synthetic one
the yeast produced during a fermentation, strains where microbial products have been chemically
showing the desired degree of flocculence may be modified, for example the semi-synthetic penicillins.
selected. The first cells to rise to the top of a top Precursor feeding has also met with some success in
fermentation, or drop to the bottom of a bottom this context; by incorporating a precursor of a natu-
fermentation, are the most flocculent cells whereas ral product into the fermentation medium the level
the last to rise or drop (depending on the type of of the end product may be increased, for example,
fermentation) are the least flocculent cells. Thus, by the use of phenylacetic acid in the production of
selecting those cells that flocculate at an inter- penicillin G. The feeding of an analogue of the
mediate time the brewer isolates the natural variants normal precursor of a natural product frequently
which have the most desirable morphological prop- results in the production of an analogue of the
erties for the inoculation of the next fermentation. natural product. Hamill et al. (1970) demonstrated
that if 5, 6 or 7 substituted tryptophan replaced
THE SELECTION OF STRAINS WHICH ARE tryptophan, the normal precursor of the antifungal
TOLERANT OF LOW OXYGEN TENSION agent, pyrrolnitrin, in cultures of Pseudomonas
The provision of oxygen is frequently the timiting aureofaciens, then a series of substituted pyrrolnitrins
factor of many fermentations and it would, there- were obtained, some of which had improved antifun-
fore, be desirable to select an organism which was gal activity. However, the disadvantages of the
capable of producing the product at a lower oxygen analogue-precursor technique are that the end pro-
tension than normal. Mindlin and Zaitzeva (1966) duct tends to be very similar to the natural product
isolated a lysine-producing strain which maintained and the new product will be contaminated with the
its productivity under aeration conditions which normal product which the organism may still syn-
decreased the parental strain productivity by almost thesize from its self-produced natural precursors.
a half. Birch (1963) suggested that the problem of mixed
products may be overcome by isolating mutants
THE ELIMINATION OF UNDESIRABLE which would not produce the normal precursor but
PRODUCTS FROM A PRODUCTION STRAIN were capable of converting it into the end product.
Athough an industrial micro-organism may pro- Thus, the analogue precursor could be converted
duce large quantities of a desirable metabolite it into the novel end product without competition
may also produce a large amount of a metabolite from the normal endogenous precursor. Shier et al.
which is not required, is toxic or may interfere with (1973) succeeded in applying Birch’s idea in the
the extraction process. An example in the penicillin- study of neomycin production by Streptomyces fradiae
producing strains is the elimination of the produc- from the precursor, deoxystreptamine. By using a
tion of the yellow pigment, chrysogenein, by the replica plating technique, the survivors of a muta-
selection of non-pigmented mutants which made tion treatment were screened for the ability to inhibit
the extraction of the antibiotic much simpler the growth ofthe test organism, Bacillus subtilis, only
(Backus and Stauffer, 1955). in the presence of deoxystreptamine. By feeding the
Dolezilova et al. (1965) considered the production mutant isolated by this method, different antibiotics
of fungicidin (nystatin) and cyclohexamide by mut- were isolated.
ants of §. noursei. These workers demonstrated that Nagaoka and Demain (1975) have described this
mutants could be isolated which produced increased technique for the isolation of new products as ‘muta-
levels of fungicidin but produced no cycloheximide. tional biosynthesis’ and the mutants isolated as
69
Principles of Fermentation Technology
‘idiotrophs’. Besides the aminoglycoside—amino- Biss, M., Scuorre, J. L. and Consn, S. N. (1980) A DNA
cloning system for interspecies gene transfer in antibiotic
cyclitol antibiotics (of which neomycin is an exam- producing Streptomyces. Nature (London), 284, 526-531.
ple) mutational biosynthesis has been applied to the Bircu, A. J. (1963) The biosynthesis of antibiotics. Pure Appl.
macrolide antibiotics, the novobiocin antibiotics Chem. 7, 527-537.
Bonner, D. (1947) Studies on the biosynthesis ofpenicillin. Arch.
and the B-lactam antibiotics (Daum and Lemke,
Biochem. 13, 1.
1979). However, although many new products have Brown, A. G., BurreRwortu, D., Cote, M., Hanscomsg, G.,
been identified, up until 1979 none had been com- Hoop, J. D. and Reaprno, C. (1976) Naturally occurring
B-lactamase inhibitors with anti-bacterial activity.J.Anti-
mercially exploited, due mainly to poor production
biot. 29, 668-669.
yields (Daum and Lenke, 1979). Nevertheless, many Butt, A. T., Ertwoop, D. C. and Ratrence, C. (1979) The
of the mutants isolated in such programmes have changing scene in microbial technology. Society General Micro-
biology Symposium, 29, 1-28.
proved very useful in the elucidation of biosynthetic
Bu’Lock, J. D. (1979) Process needs and the scope for genetic
pathways. methods. In Genetics of Industrial Micro-organisms (Editors
Sebek, O. K. and Laskin, A. I.), pp. 105-111. American
Society for Microbiology, Washington.
Bu’ Lock, J. D. (1980) Resistance ofa fungus to its own antifungal
REFERENCES metabolites and the effectiveness of resistance selection in
screening for higher yielding mutants. Biotechnol. Lett. 3,
285-290.
Ass, S. (1972) Mutants and their isolation. In The Microbial Catam, C. T., Dacuisu, L. B. and McCann, E. P. (1976)
Production ofAmino Acids (Editors Yamada, K., Kinoshita, S., Penicillin: tactics in strain improvement. In Second Interna-
Tsunoda, T..and Aida, K.), pp. 39-66. Halstead Press, New tional Symposium on the Genetics of Industrial Micro-organisms
York. (Editor MacDonald, K. D.), pp. 273-287. Academic Press,
Apter, H. I. (1971) Techniques for the development of novel London.
micro-organisms. In Radiation and Radioisotopes for Industrial CatcuHEsIDE, D. G. (1954) Isolation of nutritional mutants of
Micro-organisms, pp. 241-250. International Atomic Energy Neurospora crassa by filtration enrichment.J. Gen. Micro. 11,
Agency, Vienna. 34-36.
AjtnomotTo (1983) European patent application 71023. Cuance, L. T. and ELanper, R. P. (1979) Rational selection for
ALIKHANIAN, S. I. (1962) Induced mutagenesis in the selection of improved cephalosporin C production in strains of
micro-organisms. Adv. Appl. Micro. 4, 1-5. Acremonium chrysogenum. Dev. Ind. Micro. 20, 367-380.
ALIKHANIAN, S. I. (1973) Principle results and unsolved problems Cuanc, L. T., Terasaxa, D. T. and ELanper, R. P. (1982)
in microbial selection. In Genetics of Industrial Micro-organisms, Protoplast fusion in industrial fungi. Dev. Ind. Micro. 23,
Vol. 2, pp. 9-18 (Editors Vanek, Z., Hostalek, A. and 21-29.
Cudlin, A.). Elsevier, Amsterdam. Cuater, K. F. (1979) Some recent developments in Strep-
ALIKHANIAN, S. I., Minpuin, S. Z., Gotpat, S. V. and VLa- tomyces genetics. In Genetics of Industrial Micro-organisms
pimizov, A. V. (1959) Genetics of organisms producing (Editors Sebek, O. K. and Laskin, A. I.), pp. 123-133.
tetracyclines. Ann. N.Y. Acad. Sci. 81, 914. American Soc. Microbiology, Washington.
ATHERTON, K. T., Byrom, D. and Dart, E. C. (1979) Genetic Crarkg, P. H. (1974) The evolution of enzymes for the utilisation
manipulation for industrial processes. Soc. Gen. Micro. Symp. of novel substrates. Society
forGeneral Microbiology Symposium,
29, 379-406. 24, 183-218.
AunstrRup, K., Outrrrup, H., ANDRESEN, O. and Dammann, C. CoueEn-Bazire, G. and Jouit, M. (1953) Isolation by selection
(1972) Proteases from alkalophilic Bacillus sp. Fermentation of E. coli mutants which spontaneously synthesise amylo-
Technology Today: Proc. 4th Int. Fermentation Symposium, maltose and B-galactosidase. Ann. Inst. Pasteur, 84, 937—
pp. 299-305. 945.
Backus. M. P. and Sraurfer,J. F. (1955) The production and Corson, C., Cornetius, P., DicNerrr, C., Warton, R. and
selection of a family of strains of Penicillium chrysogenum. Waton, C. (1981) Genetically engineered micro-organisms
Mycologia, 47, 429-463. for massive production of amylolytic enzymes and process
BALL, C. (1973) The genetics of Penicillium chrysogenum. Prog. Ind. for preparing same. European Patent Office, Patent Appli-
Micro. 12, 47-72. cation no. 81 3005782.
Batt, C. (1978) Genetics in the development of the penicillin Daun, S. J. and Lemxe,J. R. (1979) Mutational biosynthesis of
process. In Antibiotics and Other Secondary Metabolites Biosyn- new antibiotics. Ann. Rev. Micro. 33, 241-266.
thesis and Production (Editors Hutter, R., Leisinger, T., Daousr, D. R. (1976) Microbial synthesis of amino acids. In
Nuesch, J. and Wenr.in, W.), pp. 163-176. Academic Industrial Microbiology (Editors Miller, B. M. and Litsky,
Press, London. W.), pp. 106-127. McGraw-Hill, New York.
Batt, C. and McGonacte, M. P. (1978) Development and Davis, O. L. (1964) Screening for improved mutants in antibio-
evaluation ofa potency index screen for detecting mutants of tic research. Biometrics, 20, 576-591.
P. chrysogenum having increased penicillin yields. J. Appl. Davis, B. D. (1949) Nutritionally deficient bacterial mutants
Bact. 45, 67—74. isolated by penicillin. Proc. Natn. Acad. Sci., U.S.A. 35, 1-10.
BARTHOLOMEW, W. H., SHEEHAN, B. T., Suu, P. and Squires, R. DeEBABov, V. G. (1982) Gene engineering and microbiological
W. (1977) Abstracts of the 174th National Meeting of the American industry. In Overproduction of Microbial Products (Editors
Chemical Soc., Chicago, Illinois. Division of Microbial and Krumphanzl, V., Sikyta, B. and Vanek, Z.), pp. 345-352.
Biochemical Technology, Abstract number 37. Academic Press, London.
70
The Isolation, Preservation and Improvement of Industrial Micro-organisms
Demat, A. L. (1957) Inhibition ofpenicillin formation by lysine. produetion in Streptomyces lipmanii. Antimicrob. Agents and
Archives of Biochemistry and Biophysics, 67, 244-245. Chemother. 4, 73-79.
Dematn, A. L. (1971) Overproduction of microbial metabolites Gorper, E. R. and Day, L. E. (1972) Effect of exogenous
and enzymes due to alteration of regulation. Adv. Biochem. penicillin on penicillin synthesis. Antimicrob. Agents and
Eng. 1, 113-142. Chemother. 1, 315-322.
Demat, A. L. (1972) Cellular and environmental factors affect- Goutpen, S. A. and Cuataway, F. W. (1969) End product
ing the synthesis and excretion of metabolites.J.Appl. Chem. control of acetohydroxy acid synthetase by valine in P.
Biotech. 22, 346-362. chrysogenum Q176 and a high yielding mutant.J. Gen. Micro.
Demam, A. L. (1973) Mutation and production of secondary 59, 111-118.
metabolites. Adv. Appl. Micro. 16, 177-202. Guerota, N., INcrAnam,J. L. and Cerpa-Otmepo, E. (1971)
Demain, A. L. (1974) How do antibiotic producing organisms Introduction of closely linked multiple mutations by nit-
avoid suicide? Ann. N.Y. Acad. Sci. 235, 601-612. rosoguanidine. Nature New Biology, London, 230, 122-125.
Demarn, A. L. (1978) Production of nucleotides by micro- Hatt, M. J., Dickinson, S. D., PrircHarp, R. and Evans,J. I.
organisms. In Primary Products of Metabolism. Economic Micro- (1973) Foam and foam control in fermentation processes.
biology, 2, 187-209 (Editor Rose, A. H.). Academic Press, Prog. Ind. Micro. 12, 169-234.
London. Haiti, R. L. (1982) Screens for pharmacologically active fer-
Demat, A. L. and Birnspaum,J. (1968) Alteration of permeabil- mentation products. In Bioactive Microbial Products: Search and
ity for the release of metabolites from the microbial cell. Discovery (Editors Bu’ Lock, J. D., Nisbet, L. J. and Winstan-
Current Topics in Microbiology and Immunology, 46, 1-25. ley, D. J.), pp. 71-76. Academic Press, London.
Demany, A. L. and Masurekar, P. S. (1974) Lysine inhibition of Hamitt, R. L., Evanper, R. P., Mase,J. A. and Gorman, M.
in vive homocitrate synthesis in Penicillium chrysogenum.J. Gen. (1970) Metabolism of tryptophan by Pseudomonas aureofaciens.
Micro. 82, 143-151. Appl. Micro. 19, 721-725.
Dircusurn, P., Grppincs, B. and Macponatp, K. D. (1974) Hamtyn, P. F. and Batt, C. (1979) Recombination studies with
Rapid screening for the isolation of mutants in A. nidulans Cephalosporium acremonium. In Genetics of Industrial Micro-
with increased penicillin yields.J.Appl. Bact. 37, 515-523. organisms (Editors Sebek, O. K. and Laskin, A. I.),
DoteziLova, L., Sp1zEk,J.,VONDRACEK, M., PALEcKova, F. and pp. 185-191. American Soc. for Microbiology, Washington.
Vanek, Z. (1965) Cycloheximidine producing and fungici- Harrison, D. E. F. (1978) Mixed cultures in industrial fermenta-
din producing mutants of Streptomyces noursei. J. Gen. Micro. tion processes. Adv. in Appl. Micro. 24, 129-162.
39, 305-310. Harrison, D. E. F., Toprwaa, H. H. and Hamer, G. (1972)
Dutvaney, E. L. (1954) Induced mutation and strain selection in Yield and productivity in SCP production from methane
some industrially important micro-organisms. Ann. N.Y. and methanol. Fermentation Technology Today, Proc. 4th Int.
Acad. Sci. 60, 155-167. Fermentation Symposium, pp. 491-495.
Duvaney, E. L. and Duraney, D. D. (1967) Mutant popula- Harrison, D. E. F., Witkinson, T. G., Wren, S. J. and Har-
tions ofStreptomyces viridifaciens. Trans. N.Y. Acad. Sci. 29, 782— woop, J. H. (1976) Mixed bacterial cultures as a basis for
799. continuous production of SCP from C, compounds. In
ELAnpER, R. P. (1966) Two decades of strain development in Continuous Culture 6: Applications and New Fields (Editors
antibiotic-producing micro-organisms. Dev. Ind. Micro. 7, Dean, A. C. R., Ellwood, D. C., Evans, C. G. T. and
61-73. MELLING, J.), pp. 122-134. Ellis Horwood, Chichester.
ELanper, R. P. (1967) Enhanced penicillin synthesis in mutant Hecxty, R. J. (1978) Preservation of micro-organisms. Ado. in
and recombinant strains of P. chrysogenum. In Induced Muta- Appl. Micro. 24, 1-53.
tions and their Utilisation (Editor Stubble, H.), pp. 403-423. Hoexsema, H. and Smiru, C. G. (1961) Novobiocin. Prog. Ind.
Academic Verlag, Berlin. Micro. 3, 93-139.
Evanper, R. P. (1980) New genetic approaches to industrially Hott, G. F., Epwarps, St.L. and Macponatp, K. D. (1976)
important fungi. Biotech. Bioeng. 22, Supplement 1, 49-61. The genetics of mutants impaired in the biosynthesis of
Evanper, R. P. and EspensHave, M. A. (1976) The role of penicillin. In Second International Symposium on the Genetics of
microbial genetics in industrial microbiology. In Industrial Industrial Micro-organisms (Editor Macdonald, K. D.),
Microbiology (Editors Miller, B. M. and Litsky, Vii odbeh) pp. 199-212. Academic Press, London.
pp. 192-256. McGraw-Hill, New York. Honco, M., Ox1, T. and Ocata, S. (1972) Phage contamination
Evanper, R. P., Mase, J. A., Hamity, R. L. and Gorman, M.
and control. In The Microbial Production of Amino Acids
(1971) Biosynthesis of pyrrolnitrin by analogue resistant (Editors Yamada, K., Kinoshita, S., Tsunoda, T. and Aida,
mutants of Pseudomonas fluorescens. Fol. Microbiol. 16, 157-165. K), pp. 67-90, Halsted Press, New York.
Hopwoop, D. A. (1976) Genetics of antibiotic production in
Fantini, A. A. (1966) Experimental approaches to strain
improvement. Dev. Ind. Micro. 7, 79-87. Streptomycetes. in Microbiology—1976 (Editor Schlessinger,
D.), pp. 558-562. American Soc. for Microbiology,
Fantini, A. A. (1975) Strain development. Methods in Enzymology,
43, 2441. Washington.
Hopwoop, D. A.(1978) Extrachromosomally determined anti-
Ferenczy, L. Kevet, F. and Zsoxt, J. (1974) Fusion of fungal
protoplasts. Nature (London), 248, 793-794. biotic production. Ann. Rev. Micro. 32, 373-392.
Hopwoop, D. A. (1979) The many faces of recombination. In
Fopor, K. and Avro.p1, L. (1976) Fusion of protoplasts of
Genetics of Industrial Micro-organisms (Editors Sebek, O. K.
Bacillus megatherium. Proc. Nat. Acad. Sci., U.S.A. 73, 2147-
and Laskin, A. I.), pp. 1-9. American Soc. for Microbiology,
2150.
P. L. and Ivencar, M. R. S. (1968) An enrichment Washington.
Ganyu,
Hopwoop, D. A., Biss, M. J., Bruton, C. J., Cuarer, K. F.,
technique for isolation of auxotrophic micro-organisms. Hin-
FeITELSON,J. S. aand Gir,J.A. (1983) Cloning Streptomyces
dustan. Antibiot. Bull. 11, 12-21.
genes for antibiotic production. Trends in Biotechnology, 1(2),
Goprrey, O. W. (1973) Isolation of regulatory mutants of the
42-49.
aspartic and pyruvic families and their effect on antibiotic
71
Principles of Fermentation Technology
Hopwoop, D. A., Wricut, H. M., Biss, M. J. and Couen, S.N. Liu, C. M., McDaniex, L. E. and ScHaFrrner, C. P. (1972)
(1977) Genetic recombination through protoplast fusion in Candicidin biogenesis.J.Antibiot. 25, 116-121.
Streptomycetes. Nature (London), 268, 171-174. Luenco,J.M., Reviza, G., VitLanueva,J. R. and Martin, J.
Hostavek, Z. (1964) Relation between the carbohydrate F. (1979) Lysine regulation of penicillin biosynthesis in
metabolism ofStreptomyces aureofaciens and the biosynthesis of low-producing and industrial strains of Penicillium
chlortetracycline—The efect of interrupted aeration, inor- chrysogenum.J.Gen. Micro. 115, 207-211.
ganic phosphate and benzylthiocyanate on chlortetracycline McCormick,J.R. D., Sjocanper, N. O., Hirscu, U., JENSEN, E.
biosynthesis. Fol. Microbiol. 9, 78-88. R. and Dorrscuuk, A. P. (1957) A new family ofantibiotics,
Hs1en, D. P.and Munneckg, D. M. (1972) Accelerated microbial the demethyl-tetracyclines.J.Amer. Chem. Soc. 79, 4561.
degradation of concentrated parathion. Fermentation Technol- MacDona tp, K. D. (1968) The persistence of parental genome
ogy Today, Proc. 4th Int. Fermentation Symposium, pp. 551-554. segregation in P. chrysogenum after nitrogen mustard treat-
Huanc, H. T. (1961) Production of t-threonine by auxotrophic ment. Mutat. Res. 5, 302-305.
mutants ofE. coli. Appl. Micro. 9, 419-424. MacDona.p, K. D. and Hott, G. (1976) Genetics of biosyn-
Icu1kawa, T., Date, M., IsHrkura, T. and Ozaki, A. (1971) thesis and overproduction of penicillin. Sc. Prog. 63, 547—
Improvement of kasugamycin producing strains by the agar Dio
piece method and prototroph method. Folia Microbiologica, MacDona
tp, K. D., Hurcuinson, J. M. and Gitiett, W. A.
16, 218-224. (1963) Formation and segregation of heterozygous diploids
IrakurA, K., Hirose, T., CREA, R., Rices, A. D., HEYNEKAR, H. between a wild type strain and derivative of high yield in
L., Botivar, F. and Boyer, H. W. (1977) Expression in E. Penicillium chrysogenum.J.Gen. Micro. 33, 385-394.
coli of a chemically synthesised gene for the hormone MacDona tp, K. D., Hutcuinson, J. M. and Gitiett, W. A.
somatostatin. Science, 198, 1056-1063. (1965) Heterozygous diploids of Penicillium chrysogenum and
Jacoz, F. and Monop, J. (1961) On the regulation of gene their segregation patterns. Genetica, 36, 378-397.
activity. Cold Spring Harbor Symp. Quant. Biol. 26, 193-211. McGinnis, M. R., Papuye, A. A. and AjeLLo, L. (1974) Storage
Jounson, M. J. (1972) Techniques for selection and evaluation of of stock cultures of filamentous fungi, yeasts and some
cultures for biomass production. Fermentation Technology aerobic actinomycetes in sterile distilled water. Appl. Micro.
Today, Proc. 4th Int. Fermentation Symposium, pp. 473-478. 28, 218-222.
Jones, A. and Vinine, L. C. (1976) Biosynthesis of chloram- Matix, V. S. (1982) Genetics and biochemistry of secondary
phenicol in Streptomyces: Identification of p-amino-L- metabolism. Adv. Appl. Micro. 28, 28-116.
phenylalanine as a product from the action of arylamine Martin, J. F. (1978) Manipulation of gene expression in the
synthetase on chorismic acid. Can.J.Micro. 22, 237-244. development ofantibiotic production. In Antibiotics and Other
Kase, H. and Nakayama, K. (1972) Production of L-threonine Secondary Metabolites. Biosynthesis and Production (Editors Hut-
by analogue resistant mutants. Agric. Biol. Chem. 36, 1611— ter, R., Leisinger, T., Nuesch, J. and Wehrlin, W.).
1621. Academic Press, London.
Kase, H., Tanaka, H and Nakayama, K. (1971) Studies on Martin, S. M. and Skerman, B. D. (1972) World Directory of
L-threonine fermentation: Production of L-threonine by Micro-organisms. Wiley-Interscience, New York.
auxotrophic mutants. Agri. Biol. Chem. 35, 2089-2096. Martin, J. F., Giri., J. A., NAHARRO, G., Lrras, P. and ViL-
Karaairt, I. (1954) Study on chlortetracycline. Improvement of LANUEVA, J. R. (1979) Industrial microorganisms tailor-
chlortetracycline producing strain by several kinds of made by removal of regulatory mechanisms. In Genetics of
methods.J. Antibiot. 7, 45-52. Industrial Micro-organisms (Editors Sebek, O. K. and Laskin,
Krnosuita, S. and Nakayama, K. (1978) Amino acids. In Primary A. I.), pp. 205-209. American Soc. Microbiology,
Products of Metabolism. Economic Microbiology, 2, 210-262. Washington.
Academic Press, London. Minpbiin, S. Z., ALIKHANIAN, S. I., VLApImMIRov, A. V. and
KinosuirTa, S., Upaka, S. and Suimono, M. (1957a) Studies on Mixuainova, G. R. (1961) A new hybrid strain of an
the amino acid fermentations, part 1. Production of L-Glu. oxytetracycline producing organism, Streptomyces rimosus.
acid by various microorganisms. J. Gen. Appl. Micro. 3, Appl. Micro. 9, 349-353.
193-205. Minpin, S. Z. and Zartseva, Z. M. (1966) Effect of threonine
KrnosuirTa, S., NAKAYAMA, K. and Upaka, S. (1957b) Fermenta- and isoleucine on biosynthesis of L-lysine by a homoserine
tive production of L-Orn.J. Gen. Appl. Micro. 3, 276-277. deficient strain of Micrococcus glutamicus. Prikl. Biokhim. i
Kusota, K., Onpa, T., Kamiyo, H., Yosuinaca, F. and OKAM- Mikrobiol. 2(2), 108-174.
ura, S. (1973) Microbial production of L-arginine. I. Pro- Nacaoka, K and Demam, A. L. (1975) Mutational biosynthesis
duction of L-arginine by mutants of glutamic acid-producing of a new antibiotic, streptomutin A, by an idiotroph of
bacteria.J.Gen. Appl. Micro. 19, 339-352. Streptomyces griseus.J.Antibiot. 28, 627-647.
Kyowa Hakko Kocyo (1983) European patent application Nagata, S., Tarra, H., Hatt., A., Jounsrup, L., STREULI, M.,
73062. Ecson1, J., Bott, W., Canrett, K. and Weissmann, C.
Lawson, C. J. and SurHERLAND, I. W. (1978) Polysaccharides. Synthesis in E. coli of a polypeptide with human leukocyte
in Primary Products of Metabolism. Economic Microbiology, 2, interferon activity. Nature (London), 284, 316-320.
328-392 (Editor Rose, A. H.). Academic Press, London. Nakao, Y., Ixrucui, M., Suzuki, M. and Dot, M. (1970) Micro-
LEDERBERG, J. and LepEeRBERG, E. M. (1952) Replica plating bial production of L-Glu acid from n-paraffin by glycerol
and indirect selection of bacterial mutants. J. Bact. 63, auxotrophs. Agr. Biol. Chem. 34, 1875.
399-406. Nakao, Y., Kanamaru, T., Kirxucui, M. and YAMATODANI, S.
LeEvisoun, S. and Aronson, A. I. (1967) Regulation of extracellu- (1973) Extracellular accumulation of phospholipids, UDP-
lar protease production in B. cereus. J. Bact. 93, 1023-1030. N-acetylhexosamine derivatives and L-glutamic acid by
LincoLy, R. E. (1960) Control of stock culture preservation and penicillin treated Corynebacterium alkanolyticum. Agr. Biol.
inoculum build-up in bacterial fermentations. J. Biochem. Chem. 37, 2399-2404.
Microbiol. Tec. Eng. 2, 481-500. Nakao, Y., Kixucui, M., Suzuki, M. and Dor, M. (1972)
72
The Isolation, Preservation and Improvement of Industrial Micro-organisms
Microbial production of L-glutamic acid by glycerol auxo- taxonomic study of the Penicillium chrysogenum series. Antonie
trophs and production of L-glutamic acid from n-paraffins. van Leuwenhoek, 43, 169-175.
Agri. Biol. Chem. 36, 490-496. Sano, K. and Suno, I. (1970) Microbial production of L-lysine
Nakayama, K.(1972) Lysine and diaminopimelic acid. In The I1J—Production by mutants resistant to S-(L-aminoethyl)-
Microbial Production of Amino Acids (Editors Yamada, K., L-cysteine.J. Gen. Appl. Micro. 16, 373-391.
Kinoshita, S., Tsunoda, T. and Aida, K.), pp. 369-397. Sermonti, G. (1969) Genetics of Antibiotic Producing Organisms.
Halsted Press, New York. Wiley-Interscience, London.
Nakayama, K., Kirupa, S. and Kinosuira, S. (1961) Induction Surpal, H., Ener, H. and Hirose, Y. (1978) Purine nucleoside
of nutritional mutants of glutamic acid bacteria and their fermentations. Process Biochem. 13(11), 6-8.
amino acid accumulation.J.Gen. Appl. Micro. '7(1), 41-51. Suier, W. T., Ocawa, S., Hichens, M. and Rinewarrt, K. L.Jr.
Nemuart, F. C. (1960) Mutant of Aerobacter aerogenes lacking (1973) Chemistry and biochemistry of the neomycins.
glucose repression.J.Bact. 80, 536-543. XVII. Bioconversion of amino-cyclitols to aminocyclictol
Nisset, L. J. (1982) Current strategies in the search for bioactive antibiotics.J.Antibiot. 26, 551-561.
products.J. Chem. Tech. Biotech. 32, 251-270. Suuno, I. and Sano, K. (1969) Microbial production of L-lysine
Novick, A. and Horrucut, T. (1961) Hyperproduction of B- II—Production by mutants sensitive to threonine or
galactosidase by E. coli. Cold Spring Harbor Symp. Quant. Biol. methionine.J. Gen. Appl. Micro. 15, 267-287.
26, 239-245. SHortie, D., DiMaro, D. and Natuans, D. (1981) Directed
Oxanasut, M. (1979) Plasmids and antibiotic synthesis in Strep- mutagenesis. Ann. Rev. Gen. 15, 265-294.
tomycetes. In Genetics of Industrial Microorganisms (Editors Sixyta, B. and Kysuix, P. (1981) Selection of Escherichia coli with
Sebek, O. K. and Laskin, A. I.), pp. 134-140. American Soc. an increased endoenzyme synthesis in continuous culture.
Microbiology, Washington. Adv. Biotech. 1, 215-220 (Editors Moo-Young, M., Robinson,
O Lp, R. W. and Primrose, S. B. (1981) Principles ofGene Manipula- C. W. and Vezina, C.). Pergamon Press, Toronto.
tion. An introduction to Genetic Engineering, 2nd edition. Stxyta, B., Kysiix, P., Vouesky, B., Paviasova, E. and
Blackwell, Oxford. STEJSKALOVA, E. (1982) Over-production of endoenzymes
O’SuLtivan, C. Y. and Pirt, S. J. (1973) Penicillin production by in Escherichia coli—Selection of hyperproducing strain in a
lysine auxotrophs of P. chrysogenum.J.Gen. Micro. '76, 65-75. chemostat. In Overproduction of Microbial Products (Editors
PeBerpy, J. F., Eyssen, H. and Anne, J. (1977) Interspecific Krumphanzl, V., Sikyta, B. and Vanek, Z.), pp. 593-600.
hydridisation between Penicillium chrysogenum and Penicillium Academic Press, London.
cyaneo-fulvuum following protoplast fusion. Mol. Gen. Genet. Srpiczk1, M. and Ferenazy, L. (1977) Protoplast fusion of
157, 281-284. Schizosaccharomyces pombe auxotrophic mutants of identical
PERLMAN, D. and Kixucut, M. (1977) Culture maintenance. mating type. Molecular and General Genetics, 157, 77-83.
Annual Reports on Fermentation Processes, 1, 41-48 (Editor Smytu, P. F. and CiarkeE, P. H. (1972) Catabolite repression of
Perlman, D.). Academic Press, New York. Pseudomonas aeruginosa amidase.J.Gen. Micro. 73, ix.
PousinELui, M., ALBERTINI, A., CaAssani, G. and CIFErRRI, O. SzyBatsk1, W. (1952) Microbial selection. Part 1. Gradient plate
(1965) Relation of biochemical mutations to actinomycin technique for study of bacterial resistance. Science, 116,
synthesis in Streptomyces antibioticus. J. Gen. Micro. 39, 239- 46-48.
246. Trenina, G. A. and Trutneva, E. M. (1966) Use of ristomycin
Potya, K. and Nyrrt, L. (1966) Adstr. 9th Int. Congr. Microbiol., in selection of active variants of Proactinomyces fructiferi var.
PZ ristomycini. Antibiotiki. 11, 770-774.
Ee ns G. P. A. and Roper, J. A. (1958) United States Upacawa, K., Ape, S. and Kinosuita, S. (1962) J. Ferment.
Patent 2,820,742. Technol. (Osaka) (Japanese), 40, 614.
PontEecorvo, G., Roper, J. A., Hemmons, L. M., MAcponaLp, U.tricu, A., Sune, J., Coircwin, J., Pictet, R., Tiscuer, E.,
K. D. and Burton, A. W. J. (1953) The genetics of A. Rutter, W. J. and Goopman, H. M. (1977) Rat insulin
nidulans. Adv. Genetics, 5, 141-238. genes: Construction of plasmids containing the codon
PontEecorvo, G. and Sermontt, G. (1954) Parasexual recombi- sequence. Science, 196, 1313-1319.
nation in Penicillium chrysogenum.J.Gen. Micro. 11, 94-110. Umezawa, H. (1972) Enzyme Inhibitors ofMicrobial Origin, Univer-
Pripuam, T. G., Lyons, A. J. and PHrompatima, B. (1973) sity of Tokyo Press.
Viability of actinomycetes stored in soil. Appl. Micro. 26, Wang, D. I. C., Cooney, C. L., Demain, A. L., DunnixL, P.,
441-442. Humpurey, A. E. and Litty, M. D. (1979) Fermentation and
Pruegss, D. L. and Jounnson, M.J. (1967) Penicillin acyl trans- Enzyme Technology, Wiley-Interscience, New York.
ferases in P. chrysogenum.J.Bact. 94, 1502-1509. WESSELING, A. C. (1982) Protoplast fusion among the Actinomy-
Queener, S. and Swartz, R. (1979) Penicillins: biosynthetic cetes and its industrial applications. Dev. Ind. Micro. 23,
and semisynthetic. In Economic Microbiology 3: Secondary Pro- 31-40.
ducts of Metabolism (Editor Rose, A. H.), pp. 35-123. Wooprurr, H. B. (1966) The physiology of antibiotic produc-
Academic Press, London. tion: role of producing organisms. Symp. Soc. Gen. Micro. 16,
Riviere, J. (1977) Industrial Applications of Microbiology (Trans- 22-46.
lators Moss, M. O. and Smith, J. E.). Surrey University Wooprurr, H. B. and Jounson, M. (1970) U.S. Patent
Press, London. 3,524,797.
Row ey, B. I. and Butt, A. T. (1977) Isolation ofa yeast lysing ZauNER, H. (1978) The search for new secondary metabolites. In
Arthrobacter species and the production oflytic enzyme com- Antibiotics and Other Secondary Metabolites. Biosynthesis and Pro-
plex in batch and continuous culture. Biotech. Bioeng. 19, duction, FEMS Symp., 5, 1-18 (Editors Hutter, R., Leisinger,
879-900. : T., Nuesch,J.and Wehrli, W.). Academic Press, London.
Samson, R. A., Haprock, R. and Srork, A. C. (1977) A
73
CHAPTER 4
Itaconic acid (Nubel and Ratajak, 1962) Gibberellic acid (Calam and Nixon, 1960)
Cane molasses (as sugar) 150gdm~° Glucose monohydrate 20gdm_ 3
Amylase (Underkofler, 1966) Glutamic acid (Gore, Reisman and Gardner, 1968)
Ground soybean meal 1.85% Dextrose 270¢gdm~ 3
Autolysed Brewers yeast 1.50% NH,H,PO, 2g¢dm_ 3
fractions (NH,).HPO, 2¢dm_ 3
Distillers dried solubles 0.76% K,SO, 2g¢dm— 3
NZ-amine (enzymatic casein 0.65% MgsO,:7H,O 0.5¢dm— 3
hydrolysate) MnSO,:H,O 0.04 ¢dm~ 3
Lactose 4.75% FeSO,-7H,O 0.02 ¢dm * 3
MgSO,:7H,O 0.04% Polyglycol 2000 0.3g¢gdm~ 3
Hodag KG-1 antifoam 0.05% Biotin i2ugdm~ 3
Penicillin llugdm~ 3
Note. The choice of constituents in the six media is not a haphazard one. The rationale for medium design will be detailed in the
remainder ofthe chapter.
75
Principles of Fermentation Technology
TaBLe 4.2. Elemental composition of bacteria, yeasts and fungi (% by dry weight)
A ee ia ee Te a eS
Bacteria
(Luria, 1960; Yeasts Fungi
Herbert, 1976; (Aiba etal., 1973; (Lilly, 1965;
Element Aibaetal., 1973) Herbert, 1976) Aiba etal., 1973)
possible to calculate substrate concentrations neces- thesis and energy generation. The carbon require-
sary to produce required product yields. There may ment under aerobic conditions may be estimated
be medium components which are needed for pro- from the cellular yield coefficient (Y) which is
duct formation which are not required for biomass defined as:
production. Unfortunately it is not always easy to
Quantity of cell dry matter produced
quantify all the factors very precisely.
A knowledge of the elemental composition of a Quantity of carbon substrate utilized
named micro-organism, which should include the Some values are given in Table 4.3. Thus for bac-
content of C, H, O, N, S, P, Mg and K, is required teria with a Y for glucose of 0.5, which is 0.5 g cells
in the solution of the elemental balance equation. 1.0 g glucose', the quantity of glucose needed to
This information may not be available so that data obtain 30 g dm ® cells will be 30/0.5 = 60 g dm™>
which is given in Table 4.2 will serve as a guide to glucose. This medium would also need to contain
the absolute minimum quantities of N,S, P, Mg and approximately 3.0 gdm °N, 1.0gdm~* Pand0.3g
K to include in an initial medium recipe. Trace dm’°S.
elements may also be needed in smaller quantities An adequate supply of the carbon source is also
(Fe, Zn, Cu, Mn, Co, Mo, B). An analysis of relative essential for a product forming fermentation pro-
concentrations of individual elements in bacterial cess. In a critical study, analyses are made to deter-
cells and commonly used cultivation media quoted mine how the observed conversion of the carbon
by Cooney (1981) showed that some nutrients are source to the product compares with the theoretical
frequently added in substantial excess of that maximum yield. This may be difficult because of
required, e.g. P, K; however, others are often near limited knowledge of the biosynthetic pathways.
limiting values, e.g. Zn, Cu. The concentration of P
is deliberately raised in many media to increase the
TasLe 4.3. Cellular yield coefficients (Y) of bacteria on different carbon
buffering capacity. These points emphasize the need
substrates (data from Abbott and Clamen, 1973)
for considerable attention to be given to medium
design. Cellular yield
Substrate
Some micro-organisms cannot synthesize specific Se ee ie ere
coefficient
eee
nutrients, e.g. amino acids, vitamins or nucleotides. Methane 0.62
Once a specific growth factor has been identified it n-Alkanes 1.03
Methanol
can be incorporated into a medium in adequate 0.40
Ethanol 0.68
amounts as a pure compound or as a component of Acetate 0.34
a complex mixture. Maleate 0.36
Glucose (molasses)
The carbon substrate has a dual role in biosyn- eee
Ori
76
Media for Industrial Fermentations
Cooney (1979) has calculated theoretical yields for in brewing, and most critical in the mashing process,
penicillin G biosynthesis on the basis of material and historically influenced the siting of breweries
and energy balances using a biosynthetic pathway and the types of beer produced. Hard waters con-
based on reaction stoichiometry. The stoichiometry taining high CaSO, concentrations are better for the
for the overall synthesis is: English Burton bitter beers and Pilsen type lagers,
while waters with a high carbonate content are
ay9CgH).0¢ 4F b.NH3 55 C909 Ste dy H SO, ar éyPAA —>
better for the darker beers such as stouts. Nowadays,
noPen G ot poCO, = ggH,O
the water may be treated by deionization or other
where dy, by, Cy, do, €, Mo, fo and gy are the techniques and salts added, or the pH adjusted, to
stoichiometric coefficients and PAA is phenylacetic favour different beers so that breweries are not so
acid. Solution of this equation yields: dependent on the local water source. Detailed infor-
mation is given by Hough, Briggs and Stevens
‘"/¢CgHj20, + 2NH, + '/,0, + H,SO, + (1974);
C,H,O, — Cig¢H),04N>S ae 2CO, ete 9H,O In large continuous-culture SCP plants it is com-
mon practice to recycle the water streams. It may
In this instance it was calculated that the theoretical
even be possible to reuse all the water in the fer-
yield was 1.1 g penicillin G g' glucose.
menter feed with appropriate adjustment of nutrient
Using a simple model for a batch-culture penicil-
levels (Topiwala and Khosrovi, 1978; Hamer, 1979;
lin fermentation it was estimated that 28, 61 and
Levi, Shennan and Ebbon, 1979).
11% of the glucose consumed was used for cell mass,
maintenance and penicillin respectively. When
experimental results of a fed-batch penicillin fer- ENERGY SOURCES
mentation were analysed, 26% of the glucose has
been used for growth, 70% for maintenance and 6%
for penicillin. The maximum conversion yield for Energy for growth comes from either the oxida-
penicillin was calculated to be 0.053 g g ! glucose. tion of medium components or from light. Most
Thus the theoretical conversion value is many times industrial micro-organisms are chemo-organo-
higher than the experimental value. trophs, therefore the commonest source of energy
The other major nutrient which will be required is will be the carbon source such as carbohydrates,
oxygen which is provided by aerating the culture, lipids and proteins. Some micro-organisms can also
and this aspect is considered in detail in Chapter 9. use methane or methanol as carbon and energy
The design of a medium will influence the oxygen sources (Kosaric and Zajic, 1974).
demand ofa culture in that the more reduced carbon
sources will result in a higher oxygen demand. The
amount of oxygen required may be determined CARBON SOURCES
stoichiometrically, and this aspect is also considered
in Chapter 9. Examples of commonly used carbon sources
It is common practice to use carbohydrates as the
carbon source in microbial fermentation processes.
WATER
The most widely available carbohydrate is starch
obtained from maize grain. It is also obtained from
Water is the major component ofall fermentation other cereals, potatoes and cassava. Maize and
media, and is needed in many of the ancillary other cereals may also be used directly in a partially
services. Clean water of consistent composition 1s ground state, e.g. maize chips. Starch may also be
therefore required in large quantities from reliable readily hydrolysed by dilute acids and enzymes to
permanent sources. Some of the factors which need give a variety of glucose preparations (solids and
to be considered include pH, dissolved solids and syrups). Those produced by acid hydrolysis may
effluent contamination. also contain toxic products which may make them
The mineral content of the water is very important unsuitable for particular processes.
ad
Principles of Fermentation Technology
TaB_e 4.4. Carbohydrate composition of barley malt (Harris, 1962) Tasce 4.5. Analysis of beet and cane molasses (Rhodes and Fletcher,
(expressed as % dry weight oftotal) 1966) (expressed as % of total w/v)
Barley grains may be partially germinated and bohydrates. These compounds can be obtained ina
heat treated to give the material known as malt, pure form, which will simplify subsequent recovery
which contains a variety of sugars besides starch and purification processes. Methane, methanol and
(Table 4.4). Malt is the main substrate for brewing n-alkanes have been utilized as substrates for
beer in many countries. Malt extracts may also be biomass production (Hamer, 1979; Levietal., 1979).
prepared from malted grain. Methanol is used as a substrate by Methylophilus
Sucrose is obtained from sugar cane and sugar
beet. It is commonly used in fermentation media in TaBLe 4.6. Partial analysis of corn-steep liquor (Belik et al., 1957;
a very impure form as beet or cane molasses (Table Misecka and Zelinka, 1959; Rhodes and Fletcher, 1966)
4.5), which are the residues left after crystallization Total solids 51% w/v
of sugar solutions in sugar refining. Acidity as lactic acid 15% w/v
The use of lactose and crude lactose (milk whey Free reducing sugars 5.6% w/v
~ Free reducing sugars after hydrolysis 6.8% w/v
powder) in media formulations is now extremely Total nitrogen 4% w/v
limited since the introduction of continuous-feeding Aminoacids as % ofnitrogen
processes discussed in a later section. However, Alanine 25
Arginine 8
whey is still used as a substrate for biomass produc- Glutamic acid 8
tion (Meyrath and Bayer, 1979). Leucine 6
Commercial vegetable oils (olive, maize, cotton Proline a
Isoleucine 3.5
seed, linseed, soya beans, etc.) may be included in
Threonine 3)
media for two possible reasons. They may be there Valine 3:5
purely as a source of carbon, mainly as oleic, linoleic Phenylalanine 2.0
Methionine 1.0
and linolenic acid, or else as an antifoam in associa-
Cystine 1.0
tion with a surface active agent (see the later section
on antifoams for more details). Ash 1.25% w/v
Potassium 90%
Corn steep liquor (Table 4.6) is a by-product after Phosphorus 1-5 %
starch extraction from maize. Although primarily Sodium 0.3-1%
used as a nitrogen source, it does contain lactic acid, Magnesium 0.003-0.3%
Iron 0.01-0.3%
small amounts of reducing sugars and complex
Copper
polysaccharides. Certain other materials of plant Calcium 0.01-0.03%
origin, usually included as nitrogen sources, such as Zinc 0.003-0.08%
Lead
soyabean meal and Pharmamedia contain small but
Silver 0.001—0.003%
significant amounts of carbohydrate. In media for- Chromium
mulations used in the production of vaccines, pro-
B Vitamins
teins such as bovine albumin, beef meat and NZ-
Aneurine +149 ue o~
case may be used as carbon sources (Van Hemert, Biotin 0.34-0.38 ue g~
1974). Calcium pantothenate 14.5-21.5pug 2
Folicacid
There is a growing interest in other carbon con- 0.26-0.6ug g |
Nicotinamide 30-40 ug g~
taining materials such as alcohols, simple organic Riboflavine 3.9-4+.7 ugg!
acids and alkanes, which may be initially more
expensive than equivalent quantities of crude car- Also niacin and pyridoxine
SSS
ee ee
78
Media for Industrial Fermentations
methylotrophus for ICI Ltd.’s single-cell protein pro- will partially inhibit the growth of many micro-
cess. N-alkanes have been utilized for production of organisms. Starch suffers from the handicap that.
organic acids, amino acids, vitamins and co-factors, when heated in the sterilization process it
carbohydrates, nucleic acids, antibiotics, enzymes gelatinizes, giving rise to very viscous liquids, so
and proteins (Fukui and Tanaka, 1980). that only concentrations of up to 2% can be used
without modification (Solomons, 1969).
The main product of a fermentation process will It is now recognized that the rate at which the
often determine the choice of carbon source, particu- carbon source is metabolized can often influence the
larly if the product results from the direct dissimila- formation of biomass or production of primary or
tion of it. In fermentation such as ethanol or single- secondary metabolites. Fast growth due to high
cell protein production where raw materials are 60 concentrations of rapidly metabolized sugars is often
to 77% ofthe production cost, the selling price ofthe associated with low productivity of secondary
product will be largely determined by the cost of the metabolites. This has been demonstrated for the
carbon source (Whitaker, 1973; Moo-Young, 1977). following examples: griseofulvin (Rhodes, 1963),
In certain processes, impurities must be removed penicillin (Pirt and Righelato, 1967), siomycin
from the carbohydrate source, e.g. carbohydrates (Kimura, 1967), bacitracin (Weinberg, 1967),
other than refined sugar must be partially purified actinomycin (Marshall et al., 1968), streptomycin
from metallic ions before use in some citric acid (Inamine et al., 1969). At one time the problem was
“processes (Karrow and Waksman, 1947; Wood- overcome by using the less readily metabolized
ward, Snell and Nicholls, 1949; Smith et al., 1974). sugars such as lactose (Johnson, 1952), but many
The choice of substrate may also be influenced by processes now use semi-continuous or continuous
government legislation. Within the European feed of glucose or sucrose, discussed in Chapter 2.
Economic Community, the use of beet sugar and
molasses is encouraged, and the minimum price
NITROGEN SOURCES
controlled. The quantity of imported cane sugar
and molasses is carefully monitored and _ their
Examples of commonly used nitrogen
imported prices set so that they will not be competi-
sources
tive with sugar beet.
Local laws may also dictate the substrates which Most industrially used micro-organisms can
may be used to make a number of beverages. In the utilize inorganic or organic sources of nitrogen.
Isle of Man, the Manx Brewers Act (1874) forbids Inorganic nitrogen may be supplied as ammonia
the use of ingredients other than malt, sugar and gas, ammonium salts or nitrates. Ammonium salts
hops in the brewing of beer. There are similar laws such as ammonium sulphate will usually produce
applying to beer production in the German Federal acid conditions as the ammonium ion is utilized and
Republic. Scotch malt whisky may only be made the free acid will be liberated. On the other hand,
from barley malt, water and yeast. Within France, ammonia gas and nitrates will normally cause an
many wines may only be called by a certain name if alkaline drift as they are metabolized. Ammonium
the producing vineyard is within a limited geog- nitrates will first cause an acid drift as the ammonia
raphical locality. is utilized, and nitrate assimilation is repressed.
The method of media preparation, particularly When the ammonia has been exhausted, there is an
sterilization, may affect the suitability of carbohyd- alkaline drift as the nitrate is used as an alternative
rates for individual fermentation processes. It is nitrogen source (Morton and MacMillan, 1954).
often best to sterilize sugars separately because they One exception to this pattern is the metabolism of
may react with ammonium ions and amino acids to Gibberella fujikurot (Borrow et al., 1961, 1964). In the
form black nitrogen containing compounds which presence of nitrate the assimilation of ammonia is
79
Principles of Fermentation Technology
inhibited at pH 2.8-3.0. Nitrate assimilation con- 1974). For this reason ammonia or ammonium ion is
tinues until the pH has increased enough to allow the preferred nitrogen source. In fungi that have
the ammonia assimilation mechanism to restart. been investigated, ammonium ion represses uptake
Organic nitrogen may be supplied as amino acid, of amino acids by general and specific amino acid
protein or urea. In many instances growth will be permeases (Whitaker, 1976). In Aspergillus nidulans,
faster with a supply of organic nitrogen, and a few ammonia also regulates the production of alkaline
micro-organisms have an absolute requirement for and neutral proteases (Cohen, 1973). Therefore, in
amino acids. It might be thought that the main mixtures of nitrogen sources, individual nitrogen
industrial need for pure amino acids would be in the components may influence metabolic regulation so
deliberate addition to amino acid requiring mutants that there is preferential assimilation of one compo-
used in amino acid production. However, in lysine nent until its concentration is diminished.
production, methionine and threonine are obtained The nitrogen source has been shown to influence
from soybean hydrolysate since the pure amino the fermentation pattern. Antibiotic production
acids would be too expensive to be used (Nakayama, may be inhibited by a rapidly utilized nitrogen
1972a). Chemically defined amino acid media source. For example, in the production of polyene
devoid of protein are necessary in the production of antibiotics soy-bean meal is considered a good nitro-
certain vaccines when they are intended for human gen source because ofthe balance of nutrients. The
use (Rhodes and Fletcher, 1966; Van Hemert, main factors being the protein, the low phosphorus
1974). content and slow hydrolysis. They are thought to
Amino acids are more commonly added as com- help create physiological conditions in_ the
plex organic nitrogen sources which are non- trophophase which favour antibiotic production in
homogeneous and readily available. Nitrogen _the idiophase (Martin and McDaniel, 1977). In
sources, which are often part protein, which are gibberellin production the nitrogen source in the
commonly used include corn steep liquor, soya medium has been shown to have an influence on
meal, soy beans, pea-nut meal, cotton-seed meal directing the production of different gibberellins
(Pharmamedia and Proflo), Distillers’ solubles, and the relative proportions of each type (Jefferys,
casein hydrolysate, slaughter house wastes, fish 1970). Other predetermined aspects of the process
meal and yeast extract. can influence the choice of the nitrogen source.
Rhodes (1963) has shown that the optimum concen-
tration of available nitrogen for griseofulvin produc-
Factors influencing the choice of tion showed some variation depending on the form
nitrogen source of inoculum and the type of fermenter being used.
Obviously these factors must be borne in mind in
Control mechanisms exist by which nitrate reduc- the interpretation of results in media-development
tase, an enzyme involved in the conversion ofnitrate programmes. Some examples of the best nitrogen
to ammonium ion, is repressed in the presence of sources selected during development programmes
ammonia (Brown, MacDonald-Brown and Meers, are given in Table 4.7.
80
Media for Industrial Fermentations
81
Principles of Fermentation Technology
Taser 4.10. Trace elements influencing primary and secondary metabolite production
ee
Se
Product Trace element(s) Reference
Weinberg (1970) has reviewed the nine trace there is a vitamin deficiency it is often possible, by
elements of biological interest (Atomic numbers careful blending of materials, to eliminate the defi-
23-30, 42). Managanese, iron and zinc are most ciency (Rhodes and Fletcher, 1966). It is important
critical in secondary metabolism. In every secon- to remember that if only one vitamin is required it
dary metabolic system in which sufficient data has may be sometimes more economical to add the pure
been reported, the yield of the product varies vitamin, instead of using a larger bulk of a cheaper
linearly with the logarithmic concentration of the multiple vitamin source. Calcium pantothenate has
‘key’ metal. The linear relationship does not apply been used in one medium formulation for vinegar
at concentrations of the metal which are either production (Beaman, 1967). In processes used for
insufficient or toxic to cell growth. Some of the the production of glutamic acid, biotin must be
primary and secondary microbial products whose present in the medium (see Chapter 3). Some pro-
yields are affected by concentrations of trace metals duction strains may also require thiamine
greater than those required for maximum growth (Kinoshita e¢ al., 1972).
are given in Table 4.10.
Chlorine does not appear to play a nutritional role
in the metabolism of fungi (Foster, 1949). It is, NUTRIENT RECYCLE
however, required by some ofthe halophilic bacteria
(Larsen, 1962). Obviously, in those fermentations Large-scale continuous-culture fermenters for
where a chlorine-containing metabolite is to be SCP are designed to operate with recycle of the
produced the synthesis will have to be directed to liquid feed (Topiwala and Khosvori, 1978; Hamer,
ensure that the non-chloro-derivative is not formed. 1979; Levi et al., 1979). Obviously there will be a
The most important compounds are chlortetracy- need for appropriate adjustment of nutrient levels.
cline and griseofulvin. In griseofulvin production, For this reason, phosphoric acid may be used as a
adequate available chloride is provided by the inclu- reagent for flocculating bacteria (SCP) when har-
sion of at least 0.1% KCl (Rhodes et al., 1955), as vesting since it will be an essential nutrient for the
well as the chloride provided by the complex organic micro-organism. Recycling is obviously efficient and
materials included as nitrogen sources. Birkinshaw leads to economies. However, this technique may
(1963) also cites caldriomycin, nornidulin, mollisin lead to the gradual building up of unwanted metabo-
as chlorine-containing metabolites. lites in the feed-stock.
Many of the natural carbon and nitrogen sources The control of pH may be extremely important if
contain all or some of the required vitamins. When optimal productivity is to be achieved. A compound
82
Media for Industrial Fermentations
may be added to the medium to serve specifically as duct. Probably the earliest example is that of impro-
a buffer, or may also be used as a nutrient source. ving penicillin yields (Moyer and Coghill, 1946,
Many media are buffered at about pH 7.0 by the 1947). A range of different side chains can be incor-
incorporation of calcium carbonate (as chalk). If the porated into the penicillin molecule. The signifi-
pH decreases the carbonate is decomposed, if it cance of the different side chains was first
increases there may be some correction to neutrality appreciated when it was noted that the addition of
by microbial acid production. Obviously, phos- corn-steep liquor increased the yield of penicillin
phates which are part of many media also play an from 20 units cm~’ to 100 units cm~*. Corn steep
important role in buffering. liquor was found to contain phenylethylamine
The balanced use of the carbon and nitrogen which was preferentially incorporated into the
sources will also form a basis for pH control as penicillin molecule to yield benzyl penicillin
buffering capacity can be provided by the proteins, (Penicillin G). Having established that the activity
peptides and amino acids, such as in corn-steep of penicillin lay in the side chain, and that the
liquor. The pH may also be externally controlled by limiting factor was the synthesis of the side chain, it
addition of ammonia or sodium hydroxide and sul- became standard practice to add side-chain precur-
phuric acid (Chapter 8). sors to the medium, in particular phenylacetic acid.
Smith and Bide (1948) showed that addition of
phenylacetic acid and its derivatives to the medium
THE ADDITION OF PRECURSORS AND were capable of both increasing penicillin produc-
METABOLIC REGULATORS TO MEDIA tion threefold and to directing biosynthesis towards
increasing the proportion ofbenzyl penicillin from 0
to 93% at the expense ofother penicillins Phenylace-
Some components ofafermentation medium help
tic acid is still the most widely used precursor in
to regulate the production of the product rather than
penicillin production. Some important examples of
support the growth of the micro-organism. Such
precursors are given in Table 4.11.
additives include precursors, inhibitors and induc-
ers, all of which may be used to manipulate the
progress of the fermentation. Inhibitors
Phenylacetic acid and PenicillinG Penicillium chrysogenum Moyer and Coghill (1947)
related compounds
Phenoxy acetic acid Penicillin V Penicillium chrysogenum Soper etal. (1948)
Chloride Chlortetracycline Streptomyces aureofaciens Van Dyck and de Somer (1952)
Chloride Griseofulvin Penicillium griseofulvin Rhodes etal. (1955)
*Proprionate Riboflavin Lactobacillus bulgaricus Smiley and Stone (1955)
Cyanides Vitamin By, Proprionobacterium, Streptomyces spp. Mervyn and Smith (1964)
P-lononones Carotenoids Phycomyces blakesleeanus Reyes etal. (1964)
a-Amino butyric acid L-isoleucine Bacillus subtilis Nakayama (1972b)
D-threonine L-isoleucine Serratia marcescens
Anthranilic acid L-tryptophan Hansenula anomala
Glycine L-serine Corynebacterium glycinophilum
PFT-G
83
Principles of Fermentation Technology
~~ SO; Na*
the various microbial enzyme media. An extensive
review has been produced by Davies (1963). Unfor-
tunately enzyme manufacturers have not published
many details of the inducers used, so that the infor-
Since acetaldehyde is no longer available for re- mation can only be stated in general terms. A few
oxidation of NADHg, its place as hydrogen acceptor examples are given in Table 4.13.
is taken by dihydroacetone phosphate, produced The correct medium formulation is obviously
during glycolysis. The product of this reaction is important for obtaining the optimum yield.
glycerol-3-phosphate, which is converted to Aunstrup (1974) reported that the protein should be
glycerol. present at a high concentration and the carbohyd-
The application of general and specific inhibitors rate should not be in excess for protease production.
are illustrated in Table 4.12. In most cases the The carbohydrate often causes catabolite repression
inhibitor is effective in increasing the yield of the (Davies, 1963; Inamine, Lago and Demain, 1969;
desired product and reducing the yield of undesired Reese, 1972). Optimum conditions have been
related products. A number of studies have been achieved in production of a variety of types of
made with potential chlorination inhibitors, e.g. enzyme by adding the carbohydrate during the
bromide, to minimize chlortetracycline production fermentation so that it remains at a low concentra-
during a tetracycline fermentation (Gourevitch et tion throughout the fermentation (Guntelberg,
al., 1956; Lepetit S.p.A., 1957; Goodman et al., 1959; 1954; Davies, 1963; Kalabokias, 1972; Reese, 1972):
Lein et al., 1959; Szumski, 1959). One unusual application of an inducer is the use
Inhibitors have also been used to affect cell-wall of yeast mannan in streptomycin production (In-
structure and increase the permeability for release amine, Lago and Demain, 1969). During the fer-
of metabolites. The best example is the use of penicil- mentation varying amounts of streptomycin and
lin and surfactants in glutamic acid production mannosidostreptomycin are produced. Since
(Phillips and Somerson, 1960). mannosidostreptomycin has only 20% of the biolog-
84
Media for Industrial Fermentations
ical activity of streptomycin, the former is an unde- oxygen transfer rate. This topic will be consi-
sirable product. The production organisms Strep- dered in a later section of this chapter.
tomyces griseus can be induced by yeast mannan to
produce a-mannosidase which will convert manno-
sidostreptomycin to streptomycin. Fast metabolism
media. Polymers in solution, particularly starch and preprogrammed rather than based on the require-
other polysaccharides, may contribute to the ments of the culture at the time the feed was added.
rheological behaviour of the fermentation broth The availability of reliable steam sterilizable oxygen
(Tuffile and Pinho, 1970). As the polysaccharide is and pH electrodes (Chapter 8) has enabled the
degraded, the effects on rheological properties will ‘direct monitoring ofa fermentation and led to the
change. Allowances may also have to be made for development of techniques for controlled additions.
polysaccharides being produced by the micro- Hockenhull (1959) was able to develop an improved
organism (Banks, Mantle and Syczyrbak, 1974; process for griseofulvin production by controlling
Leduy, Marsan and Coupal, 1974). This aspect is the pH with additions of carbohydrate. The supply
considered in more detail in Chapter 9. of carbohydrate was adjusted so that the pH of the
medium conformed to an experimentally deter-
mined ‘ideal’ pH curve. Similar techniques have
RESTRICTED NUTRIENT LEVELS been developed for controlled additions of glucose to
penicillin fermentations by monitoring the pH (Pan
et al., 1972) or the oxygen level (Squires, 1972).
In the early 1900s it was recognized that unless
the concentration of molasses was carefully control-
led in media used for the production ofyeast cells, an ANTIFOAMS
excess of molasses would cause too much growth
leading to anaerobic growth developing because of In most microbiological processes, foaming is a
lack of oxygen. This would lead to ethanol produc- problem. The commonest cause is due to proteins in
tion instead of yeast biomass production. Incremen- the medium, which may denature at the air—broth
tal feeding of nutrients to an initially dilute medium interface and form a skin which does not rupture
was introduced between 1915 and 1920 (Reed and readily. The foaming can cause removal ofcells from
Peppler, 1973). This concept of batch feed or con- the medium which will lead to autolysis and the
tinuous feed of single components or multiple com- further release of microbial cell proteins will proba-
ponents has now been used in many fermentation bly increase the stability of the foam. Ifuncontrolled
processes resulting in better control of the fermenta- then the air-filter exits of the fermenter become wet
tion process and more efficient utilization of media and there is danger of microbial infection and the
components (Table 4.14). possibility of siphoning leading to loss of product.
In the past, if nutrient feeds were needed they Hall et al. (1973) have recognized five patterns of
were based on limited analytical data and were foaming in fermentations:
86
Media for Industrial Fermentations
1. Foaming remains at a constant level through- these requirements have been found to be most
out the fermentation. suitable in different fermentation processes (Sol-
2. A steady fall in foaming during the early part omons, 1969):
of the fermentation, after which it remains
constant. 1. Alcohols; stearyl and octyl decanol.
2u ksters.
3. The foaming falls slightly in the early stages of
3. Fatty acids and derivatives, particularly
the fermentation then rises.
glycerides, which include cottonseed oil, lin-
4. The fermentation has a low initial foaming
seed oil, soy-bean oil, olive oil and castor oil.
capacity which rises.
4. Silicones.
5. A more complex foaming pattern during the
On Sulphonates.
fermentation which may be a combination of
6. Miscellaneous; Alkaterge C, oxazaline, poly-
two or more of the previously described pat-
propylene glycol.
terns.
These antifoams are generally added when foaming
If excessive foaming is encountered there are two
occurs during the fermentation. Because many anti-
ways of approaching the problem (Hall e¢ al., 1973):
foams are of low solubility they need a carrier such
1. To attempt a partial purification of some of the as lard oil, liquid paraffin or castor oil, which may be
complex nutrients and a modification of some metabolized and affect the fermentation process
of the physical parameters (pH, temperature, (Solomons, 1967).
aeration and agitation). This assumes that the Unfortunately, the concentrations of many anti-
foam is due to a component in the medium and foams which are necessary to control fermentations
not a metabolite. There has only been limited will reduce the oxygen-transfer rate by as much as
work in this field. 50%, therefore antifoam additions must be kept to
2. The foam is unavoidable and antifoam should an absolute minimum (Solomons and Perkins, 1958;
be used. This is the more standard approach. Phillips et al., 1960; Bull and Kempe, 1971). If the
oxygen-transfer rate is severely affected by antifoam
Antifoams are surface active agents, reducing the
addition then mechanical foam breakers may have
surface tension in the foams. They appear to act
to be considered as a possible alternative.
competitively by replacing the compounds which
are causing the foaming. The antifoams themselves
are unable to produce stable foams (Solomons,
1967). REFERENCES
An ideal antifoam ought to have the following
properties: AssorTt, B. J. and Ciamen, A. (1973) The relationship of sub-
strate, growth rate and maintenance coefficient to single cell
1. Should disperse readily and have fast action protein production. Biotech. Bioeng. 15, 117-127.
on an existing foam. Apa, S., Humpureey, A. E. and Mituts, N. F. (1973) Scale-up. In
Biochemical Engineering (2nd edition), pp. 195-217. Academic
~ Should be active at low concentrations. Press, New York.
3. Should be long acting in preventing new foam Arima, K., IMANAKA, H., Kousaka, M., Fukapa, A. and Tam-
urRA, G. (1965) Studies on pyrrolnitrin, a new antibiotic. 1.
formation.
Isolation of pyrrolnitrin.J.Antibiotics, Ser A, 18, 201-204.
4, Should be non-toxic to the micro-organism. Arxinson, B. and Maviruna, F. (1983) Stoichiometric aspects of
5. Should be non-toxic to humans and animals. microbial metabolism. In Biochemical Engineering and
Biotechnology Handbook, pp. 114-203. Macmillan, London.
@ Should not cause any problems in the extrac- AunstruP, K. (1974) Industrial production of proteolytic
tion of the product. enzymes. In Industrial Aspects of Biochemistry, Part A,
7. Should not cause any handling hazards. pp. 23-46 (Editor Spencer, B.). North Holland, Amsterdam.
Avanzini, F. (1963) Preparation of tetracycline antibiotics.
8. Should be cheap. Chlortetracycline from Streptomyces aureofaciens. British
9. Should have no effect on oxygen transfer. Patent 939,476.
l 0. Should be heat sterilizable. Banks, G. T. (1977) Aeration of mould and streptomycete
culture fluids. Topics in Enzyme and Fermentation Biotechnology,
The following compounds which meet most of 1, 72-110.
87
Principles of Fermentation Technology
Banks, G. T., Mantis, P. G. and Syezyrpar, GC, A. (1974) Large Dexporrrer, F. Ho and West,J.M_ (1960) Rheological proper-
scale production of clavine alkaloids by Claviceps fastformis._). ties offermentation broths. Ad. Apel. Mierediel. 2, 265-273.
Gen, Microdiel. 82, 345-361. Dewan, A. L. (1968) Regulatory mechanisms and the industrial
Barri, M. R. (1967) Process for producing citric acid. U.S. production ofmicrobial metabolites, Llendia, 31 395-418.
Patent 3,335,067. Bors, J. R., Lanner, W. V. and Beyer, G. F. (1919) Production
Braman, R. G. (1967) Vinegar fermentation, In Micredieal Technel- of glycerine from sugar by fermentation. nd. Exg. Chom. 11,
egy, pp. 344-359 (Editor Peppler, H. J.). Reinhold, New 842-845.
Tork, Finn, R. K. (1954) Agitation-aeration in the laboratory and in
Breousr, B., BerNBAvUER, K. and Winann, G, (1961) Process for industry. Beet. Ree 18, 254-274.
the conversion of benzimimidazole containing vitamin By Fogarty, W. M. and Warp, OQ, P. (1974) Pectinases and pectic
factors, particularly Factor III to vitamin Byy, U.S, Patent polysaccharides. Preg. Industr. Micrediel. 13, 39-119.
2,976, 220. Foster, J. W. (1949) Chemical Activities of the Fangi, Academic
Beutx, E., Heroip, M. and Dosxoen, J. (1957) The determina- Press, New York.
tion of B complex vitamins in corn-steep extracts by micro- Foster, J. W., Woonrvurr, H. B. and McDanuzz, L. E. (ISS)
biological tests. Chem. Zeesti, LL, 51-56. (Chem. Ads. 51 Microbiological aspects of penicillin. 3. Preduction of
11508c). penicillin in subsurface cultures of Penicilliam netatem.].Bact.
Brreinsnaw, J. H. (1963) Miscellaneous products of fungal 46, 421-482.
metabolism, In Bieckemisiry ofIndustrial Micre-erganisms, pp. Furur, S. and Tanaxa, A, (1980) Production of useful com-
452-488 (Editors Rainbow, C. and Rose, A. H.). Academic pounds from alkane media in Japan. Ade. Bieckem. Exe. 17,
Press, London. 1-35.
Borrow, A., Brown, S., Jerrerys, E. G.. Kesser, R. H. J. Gavucerro, V. and Gorrus, D. (1958) The biesynthesis of
Luoyp, E. C., Luoyrn, P. D. , Rorawetr, A., Rotawes
rt, B. chloramphenicol. IIT. Effects ofmicronutrients on synthesis,
and Swart, J. CG. (1964) The kinetics of metabolism of Myecalagia, 50, 490-500.
Gibberella_fajikerot in stirred culture. Car. ]. Micrediol. 10 Garnsr, H. R., Fauwy, M., Pamurs, R. L., Rormuer, H.,
407-444. Terrautt, P, A. and Bowonas, N_ (1950) Chemical changes
Borrow, A., Jerrerys, E.G. Ressser, R. H. J. Leon, E. C., in the submerged growth of Sirepiompees grisens. Bact. Prec,
Luoyp, P. B. and Nixon, I. S. (1961) Metabolism of Gié- pp. 189-140.
berella_fujikuret in stirred culture. Can.J.Micrebiel. 7, 227- _ Goonmay, J.J..Matrisaiy, M.. Youne, RW. and McCormrex,
276, J. R. D. (1959) Inhibition of the incorporation of chlorine
Borrow, A., Jrrrerys, E. G. and Nixox, I. 8. (1960) Gibberellic — into the tetracycline molecule._j. Beet. 78, 492-499,
acid. British Patent 838,033. Gore, J. H., Rewwax, H. B. and Garpnsr, C. H. (1968)
Brae, A. (1947) Antibacterial compounds 1. The isolation of L-ghatamic acid by continuous fermentation. U.S. Patent
gentisyl alcohol in addition to patulin from the filtrate of a 3,402,102.
penicillin culture. Some derivatives of gentisyl alcohol. Helv. Goursvirer, A. Mistsx, M. and Lary, J. (1956) Gompetitive
Chim. Acta, 30, 1-8. imhibition by bromide of incorporation of chloride into the
Brown, C. M., MacDonatp, D. S. and Messrs, J. F. (1974) tetracycline molecule, Anitéietics end Chemotherapy, 5, 448-—
Physiological aspects of microbial inorganic nitrogen 452.
metabolism. Adv. in Micredial Phys. 11, 1-52. Grove,J.F. (1967) Griseofalvin, In driiéietics, H, pp. 123-188
Butt, D. N. and Kener, L. L. (1971) Influence ofsurface active (Editors Gottlieb, D. and Shaw, P. D.). Springer-Verlag,
agents on oxygen absorption at the free interface in a stirred Bertin.
fermentor. Bieieck. Bioeng. 13, 529-547. Gunteserc, A. V. (1954) Method for the production of plakal-
Caram, CG. T. and Nixon, I. S. (1960) Gibberellic aad. Briush bumin-forming proteinase from Bacillus sebtilis, Conght. rend.
Patent 839,652. - trav. lab. Carlsberg Ser. Chem. 29, 27-35.
Carrincton, T. R. (1971) The development ef commercial Haut, M. J., Drexisson, S. D., Prrrenarp, R. and Evans, J. I.
processes for the production of 6-aminopenicillanic acid (1973) Foams and foam control in fermentation processes.
(6-APA). Proc. Rey. Soc, Lendox B, 179, 321-338. Prog. Industry. Micrediel. 12, 169-254.
Crarincs, C. A., Bus, J.A.. Deruria, M. D. and Price, K. EL Haut, M. J. and Hassarz, GC. H. (1970) Production of the
(1974) Fermentation and mutation studies with a butirosin monomycins, novel depsipeptide antibiotics, Apel. Micediel.
producing strain of Bacillus ctreaans. Dev. Indusir, Micrediel. 19, 108-112.
15, 101-113. Hamer G, (1979) Biomass from natural gas, In Eraremic Micre-
Crarr, D. S., Iro, K. and Horrrsu, H. (1965) Effect of man- bielegy, Vol. 4, pp. 315-360 (Editor Rose, A. H.). Academie
ganese and other heavy metals on submerged citric acid Press, London. - .
fermentation of molasses. Bioieck. Bioeng. 8, 463-471, Harnep, R. L. (1969) Riboflavine production. U.S. Patent
Conen, B. L. (1973) The neutral and alkaline proteases of $3,475,274,
Aspergillus nidulans._]. Gen. Microbiel. 77, 5321-528. Harrs, G_ (1962) The structural chemistry of barley and malt.
Cooney, GC. L, (1979) Conversion yields in penicillin production: In Barley and Malt, Biolegy, Bivckemistry and Tecknolegy,
theory vs, practice. Process Biochem. 14(5), 31-33. pp. 431-382 (Editor Cook, A. H.). Academic Press, London.
Cooney, C. L. (1981) Growth of micro-organisms. In Bieiecknel- Harrison, D. EB. F_, Torrnwara, H. and Hamer, G. (1972) Yield
egy, Vol. 1, pp. 73-112 (Editors Rehm, H. J. and Reed, G.). and productivity in single cell protein production from
Verlag Chemie, Weinheim. methane and methanol. In Fermentation Teotrelegy Taden,
Daviss, R. (19638) Microbial extracellular enzymes, their uses pp. 491-495 (Editor Terai, G.). Society of Fermentation
and some factors affecting their formation. In Bieckemistz> of Technology, Japan. ‘
Industrial Micro-erganisms, pp. 68-150 (Editors Rainbow, C. Harrison,J.S. (1971) Yeast production. Prag. Industr, Micrebiel.
and Rose, A. H.). Academic Press, London. 1Q, 129-177,
88
Media for Industrial Fermentations
Hersert, D. (1976) Stoichiometric aspects of microbial growth. Litty, V. G. (1965) The chemical environment for growth. 1.
In Continuous Culture 6: Applications and New Fields, pp. 1-30 Media, macro- and micronutrients. In The Fungi, Vol. 1,
(Editors Dean, A. C. R., Ellwood, D. C., Evans, C.G.T. and pp. 465-478 (Editors Ainsworth, G. C. and Sussman,
Melling, J.). Ellis Horwood, Chichester. A.S.). Academic Press, New York.
Hickey, R. J. (1945) The inactivation ofiron by 2,2”-bipyridine Liu, C. M., McDantet, L. E. and Scuarrner, C. P. (1975)
and its effect on riboflavine synthesis by Clostridium Factors affecting the production of candicidin. Antimicrob.
acetobutylicum. Arch. Biochem. 8, 439-447. Agents. Chemother. '7, 196-202.
Hockennutt, D. J. D. (1959) Improvements in or relating to Luria, S. E. (1960) The bacterial protoplasm: composition and
antibiotics. British Patent 868, 958. organisation. In The Bacteria, Vol. 1, pp. 1-34 (Editors
Horxksema, H. and Smiru, C. G. (1961) Novobiocin. Prog. Industr. Gunsalus, I. C. and Stanier, R. Y.). Academic Press, New
Microbiol. 3, 91-139. York.
Hos-er, P. and Jounson, M. J. (1953) Penicillin from chemically MacLennan, D. G., Gow, J. S. and Srrincrer, D. A. (1973)
defined media. Ind. Eng. Chem. 45, 871-874. Methanol-bacterium process for SCP. Process Biochem. 8(6),
Houcn,J. S., Briccs, D. E. and Stevens, R. (1971) Malting and 22-24.
Brewing Science, Chapter 7. Chapman and Hall, London. Mayjumpar, M. K. and Mayumpar, S. K. (1965) Effects of
Inaming, E., Laco, B. D. and Deman, A. L. (1969) Regulation of minerals on neomycin production by Streptomyces fradiae.
mannosidase, an enzyme of streptomycin biosynthesis. In Appl. Microbiol. 13, 190-193.
Fermentation Advances, pp. 199-221 (Editor Perlman, D.). Mauzann, R. C., Puitups, R. F. and Hanson, A. M. (1959)
Academic Press, New York. Riboflavin. U.S. Patent 2,876,169.
InsKEeEP, G. C., Bennett, R. E., Dubey,J.F. and SHeparp, M. MarsHALL, R., RepFietD, B., Katz, E. and Weisspacx, H.
W. (1951) Bacitracin, product of biochemical engineering. (1968) Changes in phenoxazinone synthetase activity during
Ind. Eng. Chem. 43, 1488-1498. the growth cycle of Streptomyces antibioticus. Arch. Biochem.
Jerrerys, E. G. (1970) The gibberellin fermentation. Adv. Appl. Biophys. 123, 317-323.
Microbiol. 13, 283-316. Martin,J. F. and McDaniet. L. E. (1975) Kinetics of biosyn-
Jounson, M. J. (1946) Metabolism of penicillin producing thesis of polyene macrolide antibiotics in batch cultures: cell
moulds. Ann. New York Acad. Sci. 48, 57-66. maturation time. Biotech. Bioeng. 17, 925-938.
Joxunson, M.J. (1952) Recent advances in penicillin fermenta- Martin,J.F.and McDanizt, L. E. (1977) Production ofpolyene
tions. Bull. World Health Org. 6, 99-121. macrolide antibiotics. Adv. Appl. Microbiol. 21, 1-52.
- Karasoxtas, G. (1972) Preparation of microbial alkaline pro- Mervyn, L. and Smiru, E. L. (1964) The biochemistry of vitamin
tease by fermentation with B. subtilis var. licheniformis. U.S. Bj, fermentation. Prog. Industr. Microbiol. 5, 150-201.
Patent 3,623,956. MeyratuH, J. and Bayer, K. (1979) Biomass from whey. In
Karrow, E. O. and Waxsman, S. A. (1947) Production ofcitric Economic Microbiology, Vol. 4, pp. 207-269 (Editor Rose, A.
acid in submerged culture. Jnd. Eng. Chem. 39, 821-825. H.). Academic Press, London.
Katz, E., Prenra, P. and Stvak, A. (1958) The role of nutrition MiseckA, J. and ZELINKA,J. (1959) The effect of some elements
in the synthesis of actinomycin. Appl. Microbiol. 6, 236-241. on the quality of corn steep with reference to the production
Keay, L. (1971) Microbial proteases. Process Biochem. 6(8), 17— of penicillin. Biologia (Bratislava), 14, 591-596. Chem. Abs.
21. 54, 23171c.
Kimura, A. (1967) Biochemical studies on Streptomyces sioyaensis. Mrvyairt, N., Mryosut, T., Aoki, H., Kousaxa, M., IkusHima,
II. Mechanism ofthe inhibitory effect of glucose on siomycin H., Kunueira, K., Saxart, H. and Imanaka, H. (1970)
formation. Agr. Biol. Chem. (Tokyo), 31, 845-852. Studies on thiopeptin antibiotics. 1. Characteristics of
Kinosuita, S. and Tanaka, K. (1972) Glutamic acid. in The thiopeptin B. J. Antibiotics, 23, 113-119.
Microbial Production of Amino Acids, pp. 263-324 (Editors Mryosut, N., Mryarrei, H., Aox1, H., Kousaxa, M., Saxal, H.
Yamada, K., Kinoshita, S., Tsunoda, T. and Aida, K.). and Imanaka, H. (1972) Bicyclomycin, a new antibiotic. 1.
Halsted Press—Wiley, New York. Taxonomy, isolation and characterization.J.Antibiotics, 25,
Korrter, H., Knicut, S. G. and Frazier, W. C. (1947) The 269-275.
effect of certain mineral elements on the production of Mizusawa, K., IcHisHawa, E. and Yosuipa, F. (1966) Proteoly-
penicillin in shake flasks.J.Bact. 53, 115-123. tic enzymes of the thermophilic Streptomyces. 11. Identifica-
‘Kosaric, N. and Zayjic, J. E. (1974) Microbial oxidation of tion of the organism and some conditions ofprotease forma-
methane and methanol. Adv. Biochem. Eng. 3, 89-125. tion. Agric. Biol. Chem. (Tokyo), 30, 35-41.
Lane, G. and Wuite, R. J. (1973) Rifamycin fermentation Moo-Younc, M. (1977) Economics of SCP production. Process
studies. Process Biochem. 8(7), 14-16. Biochem. 12(4), 6-10.
Larsen, H. (1962) Halophilism. In The Bacteria, Vol. 4, Morton, A. G. and MacMitian, A. (1954) The assimilation of
pp. 297-342 (Editors Gunsalus, I. C. and Stanier, Re Yo) nitrogen from ammonium salts and nitrate by fungi.J.Expil.
Academic Press, New York. Botany (London), 5, 232-252.
Lepuy,J.,Marsan, A. A. and Coupat, B. (1974) A study of the Moss, A. R. and Kein, R. (1946) Improvements in or relating to
rheological properties of a non-newtonian fermentation the manufacture ofriboflavin. British Patent 615,847.
broth. Biotech. Bioeng. 16, 61-76. Moyer, A. J. and Cocuit, R. D. (1946) Penicillin IX. The
Len, J., Sawmivcer, L. F. and Cueney, L. C. (1959) Chlorina- laboratory scale production of penicillin in submerged cul-
tion inhibitors affecting the biosynthesis of tetracycline. ture by Penicillium notatum Westling (NRRL 832).J.Bact. 51,
Appl. Microbiol. 7, 149-157. 79-93.
Leretit, S.p.A. (1957) Bromotetracycline. British Patent Moyer, A. J. and Cocuttt, R. D. (1947) Penicillin X. The effect
TT 2NAS: ; of phenylacetic acid on penicillin production. J. Bact. 53,
Levi,J.D., SHENNAN, J. L. and Esson, G. P. (1979) Biomass 329-341.
from liquid n-alkanes. In Economic Microbiology, Vol. 4, Nakayama, K. (1972a) Lysine and diaminopimelic acid. In The
pp. 361-419 (Editor Rose, A. H.). Academic Press, London. Microbial Production of Amino Acids, pp. 369-397 (Editors
89
Principles of Fermentation Technology
Yamada, K., Kinoshita, $., Tsunoda, T. and Aiba, K.). Sotomons, G. L. (1967) Antifoams. Process Biochem. 2(10), 47-48.
Halsted Press—Wiley, New York. Sotomons, G. L. (1969) Materials and Methods in Fermentation.
Nakayama, K. (1972b) Micro-organisms in amino acid fermen- Academic Press, London.
tation. In Fermentation Technology Today, pp. 433-438 (Editor Sotomons, G. L. and Perxmys, M. (1958) The measurement and
Terui, G.). Society of Fermentation Technology, Japan. mechanism ofoxygen transfer in submerged culture._/.Appl.
Nuset, R. D. and Rarayak, E. J. (1962) Process for producing Chem. 8, 251-259.
itaconic acid. U.S. Patent 3,044,941. Soper, Q. F., WHITEHEAD, C. W., BEHRENS, O. K., Corse,
J. J.
Pan, C. H., Herter, L. and ELanper, R. P. (1972) Control ofpH and Jones, R. G. (1948) Biosynthesis of penicillins. VII.
and carbohydrate addition in the penicillin fermentation, Oxy- and mercaptoacetic acids. J. Amer. Chem. Soc. 70,
Dev. Industr. Microbiol. 13, 103-112. 2849-2855.
Pan, S. C., Bonanno, S. and Wacman, G. H. (1959) Efficient Sgurres, R. W. (1972) Regulation of the penicillin fermentation
utilization offatty oils as energy source in penicillin fermen- by means of a submerged oxygen-sensitive electrode. Dev.
tation. Appl. Microbiol. 7, 176-180. Industr. Microbiol. 13, 128-135.
Pertman, D. (1970) The evolution of penicillin manufacturing Szumsxr, S. A. (1959) Chlortetracycline fermentation. U.S.
processes. In The History of Penicillin Production, pp. 25-30 Patent 2,871,167.
(Editor Elder, A. L.). Chem. Eng. Prog. Symp. Series, 60(100). Tanner, F. W., Voynovicx, C. and Van Lanen, J. M. (1945)
Puiturs, K. L., Spencer, J. F. T., Sartans, H. R. and Rox-, Riboflavin production by Candida species. Science (New York),
BuRGH, J. H. (1960) Effect of antifoam agents on oxygen 101, 180-181.
transfer in deep tank fermentations. J. Biochem. Microbiol. Toprwaca, H. H. and Kuosrovi, B. (1978) Water recycle in
Technol. Eng. 2, 81-91. biomass production processes. Biotech. Bioeng. 20, 73-85.
Purturrs, T and Somerson, N. L. (1960). Glutamic acid. U.S. Trivut, A., MicHetint, V., Mantovani, V. and Pirt, S. J. (1977)
Patent 3,080,297 Estimation of productivities in repeated fed batch cephalos-
. Pirt, S. J. and RicHeraro, R. C. (1967) Effect of growth rate on porin fermentation./.Appl. Chem. Biotechnol. 27, 219-224.
the synthesis of penicillin by Penicillium chrysogenum in batch TurrFiLe, C. M. and Prnuo, F. (1970) Determination of oxygen-
and chemostat cultures. Appl. Microbiol. 15, 1284-1290. transfer coefficients in viscous streptomycete fermentations.
Reep, G. and Peppter, H. J. (1973) Yeast Technology, Chapter 5. Biotech. Bioeng. 12, 849-871.
Avi, Westport. Uemura, T., Sucisakr, Z. and Takamura, Y. (1972) Valine. In
Reese, E. T. (1972). Enzyme production from insoluble sub- The Microbial Production of Amino Acids, pp. 339-368 (Editors
strates. Biotechnol. and Bioeng. Symp. 3, 43-62. Yamada, K., Kinoshita, S., Tsunoda, T. and Aida, K.).
Reyes, P., CuicHEsTER, C. O. and Nakayama, T. O. M. (1964) Halsted Press—Wiley, New York.
Mechanism ofB-ionone stimulation of carotenoid and ergos- UNDERKOFLER, L. A. (1966) Production of commercial enzymes.
terol biosynthesis in Phycomyces blakesleeanus. Biochim. Biophys. In Enzymes in Food Processing, Chapter 10 (Editor Reed, G.).
Acta, 90, 578-592. Academic Press, New York.
Ruopes, A. (1963) Griseofulvin: production and biosynthesis. Van Dyck, P. and bE Somer, P. (1952) Production and extraction
Prog. Industr. Microbiol. 4, 165-187. methods of aureomycin. Antibiotics and Chemotherapy 2,
Ruopes, A., Crosse, R., FERGuson, T. P. and FLercHer, D. L. 184-198.
(1955) Improvements in or relating to the production of the Van Hemert, P. (1974) Vaccine production as a unit process.
antibiotic griseofulvin. British Patent 784,618. Prog. Industr. Microbiol. 13, 151-271.
Ruopes, A. and Fiercuer, D..L. (1966) Principles of Industrial WALLENFELS, K., BENDER, H. and Racnep, J. R. (1966) Pul-
Microbiology. Pergamon Press, Oxford. lulanase from Aerobacter aerogenes. Biochem. Biophys. Res. Com-
Roets,J.A., VAN DEN Bere,J.and VoncKkEn, R. M. (1974) The mun. 22, 254-261.
rheology of mycelial broths. Biotech. Bioeng. 16, 181-208. Wernser, E. D. (1967) Bacitracin, gramicidin and tyrocidin, In
Sensi, P. and Turemann,J. E. (1967) Production of rifamycins. Antibiotics, Vol. 2, pp. 240-253 (Editors Gottlieb, D. and
Prog. Industr. Microbiol. 6, 21-60. SHaw, P. D.). Springer-Verlag, Berlin.
SHEPHERD, M. W. (1963) Citric acid. U.S. Patent 3,083,144. Werner, E. D. (1970) Biosynthesis of secondary metabolites:
Suu, P. and Jonnson, M. J. (1948) Interdependence of medium roles of trace metals. Adv. Microbial Phys. 4, 144.
constituents in citric acid production by submerged fermen- WeinserG, E. D. (1974) Secondary metabolism: control by
tation.J.Bact. 56, 577-585. temperature and inorganic phosphate. Dev. Jndustr. Micro-
SincH, A., BRuze.ius, E. and Heprne, H. (1976) Streptomycin, biol. 15, 70-81.
a fermentation study. EuropeanJ.Appl. Microbiol. 3, 97-101. Werserg, E. D. and Tonnis, S. M. (1966) Action of chloram-
Saitey, K. L. and Strong, L. (1955) Production of riboflavine by phenicol and its isomers on secondary biosynthesis processes
Ashbya gossypii. U.S. Patent 2,702,265. of Bacillus. Appl. Microbiol. 14, 850-856.
Situ, C. G. (1956) Fermentation studies with Streptomyces niveus. Wuiraker, A. (1973) Fermentation economics. Process Biochem.
Appl. Microbiol. 4, 232-236. 8(9), 23-26.
Situ, E. L. and Big, A. E. (1948) Penicillin salts. Biochem.].42, Wauirraker, A. (1976) Amino acid transport into fungi: an essay.
XV1I-XVill. Trans. British Mycol. Soc. 67, 365-376. ;
SirH, J. E., Nowaxowska-Waszczuk, K. and ANDERSON,J. G. Winoisy, W. W. and Muartre, N.S. (1965) Microbial amylases.
(1974) Organic acid production by mycelial fungi. In Jndus- Adv. Appl. Microbiol. 7, 273-304. :
trial Aspects of Biochemistry, Part A, pp. 297-317 (Editor Woopwarp, J. G., Snect, R. L. and Nicnotts, R. S. (1949)
Spencer, B.). North Holland, Amsterdam. Conditioning molasses and the like for the production of
Soc. RicHarD, ALLENET ET Cre (1921) Acetone and buty] alcohol. citric acid. U.S. Patent 2,492,673.
British Patent 176,284.
90
al bad belEloh Nod)
Sterilization
INTRODUCTION Avoidance of contamination may be achieved by:
(i) Using a pure inoculum to start the fermenta-
A fermentation product is produced by the culture tion, as discussed in Chapter 6.
of a certain organism, or organisms, in a nutrient ) Sterilizing the medium to be employed.
medium. Ifthe fermentation is invaded by a foreign (iii) Sterilizing the fermenter vessel.
micro-organism then the following consequences (iv) Sterilizing all materials to be added to the
may occur: fermentation during the process.
(v) Maintaining aseptic conditions during the
fermentation.
(i) The medium would have to support the
growth ofboth the production organism and The extent to which these procedures are adopted is
the contaminant, resulting in a loss of pro- determined by the likely probability of contamina-
ductivity. tion and the nature ofits consequences. Some fer-
(ii) If the fermentation is a continuous one then mentations are described as ‘protected’—that is, the
the contaminant may ‘outgrow’ the produc- medium may be utilized by only a very limited range
tion organism and displace it from the fer- of micro-organisms, or the growth of the process
mentation. organism may result in the development ofselective
(iii) The foreign organism may contaminate the growth conditions, such as a reduction in pH. The
final product, e.g. single-cell protein where brewing of beer falls into this category; hop resins
the cells, separated from the broth, consti- tend to inhibit the growth of many micro-organisms
tute the product. and the growth of brewing yeasts tends to decrease
(iv) The contaminant may produce compounds the pH of the medium. Thus, brewing worts are
which make subsequent extraction of the boiled, but not necessarily sterilized, and the fer-
final product difficult. menters are thoroughly cleaned with disinfectant
(v) The contaminant may degrade the desired solution but are not necessarily sterile. Also, the
product; this is common in bacterial con- precautions used in the development of inoculum
tamination of antibiotic fermentations where for brewing are far less stringent than, for example,
the contaminant would have to be resistant in an antibiotic fermentation. However, the vast
to the normal inhibitory effects of the anti- majority of fermentations are not ‘protected’ and, if
biotic and degradation of the antibiotic is a contaminated, would suffer some of the conse-
common resistance mechanism, e.g. the quences previously listed. The approaches adopted
B-lactam antibiotics by B-lac-
degradation of to avoid contamination will now be discussed in
tamase-producing bacteria. more detail, apart from the development of aseptic
(vi) Contamination of a bacterial fermentation inocula which is covered in Chapter 6 and the
with phage could result in the lysis of the aseptic operation of fermentation vessels which is
culture. — discussed in Chapters 6 and 7.
91
Principles of Fermentation Technology
MEDIUM STERILIZATION
92
Sterilization
InN
Time
Time
Fic. 5.2a. Initial population increase resulting from the heat 7 as Whole culture
activation of spores in the early stages of a sterilization process Sensitive organism
(Richards, 1968).
—w— Resistant organism
In —
InN
Time
93
Principles of Fermentation Technology
ne : (5.4)
dT se he
125 120 115 110 105 100
where £& is the activation energy, Temperature (°C)
R is the gas constant,
Fic. 5.4. The effect of sterilization time and temperature on the
and_ YT is the absolute temperature. Del factor achieved in the process. The figures on the graph
On integration equation (5.4) gives: indicate the Del factors for each straight line (modified after
Richards, 1966).
k= Ae WRT (5.5)
Therefore:
where A is the Arrhenius constant.
On taking natural logarithms, equation (5.5) V=lIn No
becomes: N,
E but In Be af kt
Ink =InA RT (5.6)
N,
From equation (5.6) it may be seen that a plot ofIn k and ki= Atego!)
against the reciprocal of the absolute temperature thus, V =4ienc® (5.8)
will give a straight line. Such a plot is termed an
Arrhenius plot and enables the calculation of the On rearranging, equation (5.8) becomes:
activation energy and the prediction of the reaction E V
rate for any temperature. By combining together Inn t = —
yas+lIn {[—}-
(5) (5.9)
equations (5.3) and (5.5), the following expression
may be derived for the heat sterilization of a pure Thus a plot of the natural logarithm of the time
culture at a constant temperature: required to achieve a certain V value against the
reciprocal of the absolute temperature will yield a
In hy = A.t.e ART (5.7) straight line, the slope of which is dependent on the
Nt activation energy, as shown in Fig. 5.4. From 5.4 it
is clear that the same degree ofsterilization (V) may
Deindoerfer and Humphrey (1959) used the term be obtained over a wide range of time and tempera-
In N,/N, as a design criterion for sterilization, which ture regimes; that is, the same degree ofsterilization
has been variously called the Del factor, Nabla may result from treatment at a high temperature for
factor and sterilization criterion represented by the a short time as from a low temperature for a long
term V. Thus, the Del factor is a measure of the time.
fractional reduction in viable organism count pro- This kinetic description ofbacterial death enables
duced by a certain heat and time regime. the design of procedures (giving certain V factors)
94
Sterilization
95
Principles of Fermentation Technology a
TaBe 5.1. The effect ofsterilization conditions on the subsequent yields ofriboflavin and vitamin Bi» fermentations (Wang
et al., 1S79)
Corn Animal
steep steep
Glucose liquor liquor Riboflavin
(%) (%) (%) (y cm7?)
Extracted
Distiller soybean Vitamin 6
solubles meal By
(%) (%) (ycm™*)
4+ — Batch: 250°F—l 20 min: pH = 4.8 0.1
4 — Continuous: 325°F—13 min: pH = 4.8 12
2 2 Batch: 250°F—90 min: pH = 5.7 13
2 2 Continuous: 325°F—13 min: pH = 5.7 2.0
not constant and, therefore, the calculation of V Fic. 5.7. The graphical integration method applied to the
would require the integration of equation (5.8) for increase in temperature over a time period. t), ty, etc., represent
equal time intervals (Richards, 1968).
the time-temperature regime observed. Deindoerfer
and Humphrey (1959) produced integrated forms of
the equation for a variety of temperature-time Calculation of the holding time at
profiles including linear, exponential and hyper- constant temperature (121°C)
bolic. However, the regime observed in practice is
From the previous calculations the overall Del
frequently difficult to classify making the applica-
‘factor as well as the Del factors of the heating and
tion of these complex equations problematical.
cooling parts of the cycle have been determined.
Richards (1968) demonstrated the use ofagraphical
Therefore, the Del factor to be achieved during the
method ofintegration and this is illustrated in Fig.
holding time may be calculated by difference;
5.7. The time axis is divided into a number ofequal
increments, ¢;, fy, tg, etc., Richards suggesting 30 as 1.€. Viverding = Vsgecan a Visas rar cooling*
a reasonable number.
The temperature of the holding period is normally
For each increment, the temperature correspond-
121°C so the holding time, at that temperature, to
ing to the mid-point time is recorded. It may now be
approximated that the total effect of the heating-up
achieve the desired Del factor should be calculated.
Using the example where the overall Del factor was
period is equivalent to the sum of the effects of the
mid-point temperatures for each time increment. shown to be 32.2, if it is taken that the heating Del
factor was 9.8 and the cooling Del factor 10.1, the
The value of the specific death rate of B. stearother-
holding Del factor may be calculated:
mophilus spores at each mid-point temperature may
be deducted from the Arrhenius equation using the Viciang = 80 Din Out AO al
thermal death characteristic published by Dein- A ehines = 1s
doerfer and Humphrey (1959). The value of the Del
factor corresponding to each time increment may But V = kt, and from the data of Deindoerfer and
then be calculated from the equation: Humphrey (1961) the specific death rate of B.
stearothermophelus spores at 121°C is 2.54 min !.
Vi = hy
Vo = hot, Therefore t= ae or t = = = 4.84 min.
2.54
V3 = ket,
etc. If the contribution made by the heating and cooling
parts of the cycle were ignored then the holding time
The sum ofthe Del factors for all the increments will
would be given by the equation:
then equal the Del factor for the heating-up period.
The Del factor for the cooling-down period may be t= Veen = 32.2
= = 12.68 min.
calculation in a similar fashion.
98
Sterilization
Thus by considering the contribution made to the Taste 5.2. Del values for B. stearothermophilus spores for the
sterilization process by the heating and cooling heating-up period over a temperature range of 100 to 130°, assuming a rate
of temperature change of \° min ' and negligible spore destruction at
parts of the cycle a considerable reduction in expo- temperatures below 100° (Richards, 1968)
sure time is achieved.
TEE k min”! Vv
99
PFT-H
Principles of Fermentation Technology
resulting in an increase in the Del factor. For exam- alleviated by aeration as it would be during
ple, if a pilot sterilization were carried out in a the fermentation proper. Thus, if the
1000-dm? vessel with a medium containing 10° requirement for agitation during in situ
organisms cm° requiring a probability of contami- sterilization were removed, the fermenter
nation of| in 1000 the Del factor would be: could be equipped with a less powerful
motor. Obviously, the sterilization kettle
—
10° x 10°. x 10°
— = |I
10"
—e would have to be equipped with a powerful
V In 1072 n (2
motor but this would provide sterile medium
=In 10° = 345 for several fermenters.
(iv) The fermenter would be spared the corrosion
If the same probability of contamination were which may occur with medium at high tem-
required in a 10,000-dm?* vessel using the same perature.
medium the Del factor would be:
The major disadvantages of a separate medium
10° x 10° x 10* 10° sterilization vessel may be summarized as:
In |————.——_ } = In
es 103 10-3
(i) The cost of constructing a batch medium
= In 10'° = 36.8. sterilizer is much the same as that for the
Thus, the Del factor increases with an increase in fermenter.
the size of the fermenter volume. The holding time (ii) Ifa cooker serves a large number of fermen-
in the large vessel may be calculated by the graphical ters complex pipework would be necessary
integration method or the rapid method of Richards to transport the sterile medium with the
(1968), as discussed earlier, based on the tempera- inherent dangers of contamination.
ture time profile of the sterilization cycle in the large (iii) Mechanical failure in a cooker supplying
vessel. medium to several fermenters would render
all the fermenters temporarily redundant.
Methods of batch sterilization The provision of contingency equipment
would be prohibitively costly.
The batch sterilization of the medium for a fer- The decision on which batch sterilization system to
mentation may be achieved either in the fermenta- adopt would depend on the scale of the operation
tion vessel or a separate mash cooker. Richards and the characteristics of the medium. However, it
(1966) considered the relative merits of in situ is probably more common that the medium is
medium sterilization and the use of a special vessel. sterilized in the fermenter rather than in a separate
The major advantages of aseparate medium sterili- vessel.
zation vessel may be summarized as:
(1) One cooker may be used to serve several
THE DESIGN OF CONTINUOUS
fermenters and the medium may be
STERILIZATION PROCESSES
sterilized as the fermenters are being cleaned
and prepared for the next fermentation, thus
saving time between fermentations. The design of continuous sterilization cycles may
(ii) The medium may be sterilized in a cooker in be approached in exactly the same way as for batch
a more concentrated form than would be sterilization systems. The continuous system
used in the fermentation and then diluted in includes a time period during which the medium is
the fermenter with sterile water prior to heated to the sterilization temperature, a holding
inoculation. This would allow the construc- time at the desired temperature and a cooling period
tion of smaller cookers. to restore the medium to the fermentation tempera-
(iii) In some fermentations, the medium is at its ture. The major advantage ofthe continuous process
most viscous during sterilization and the is that temperature may be used as a process vari-
power requirement for agitation is not able, thus reducing the holding time and reducing
100
Sterilization
Holding section
Fic. 5.8. Flow diagram ofa typical continuous plate heat exchange sterilizer.
the degree of nutrient degradation. The required maintained for the holding period and then cooled
Del factor may be achieved by the combination of to the fermentation temperature in another plate
temperature and holding time which gives an heat exchanger. The continuous injector flash cooler
acceptably small degree of nutrient decay. is shown in Fig. 5.9 and consists of asteam injector
There are two types of continuous sterilizer which which injects steam directly into the unsterile
may be used for the treatment of fermentation medium, a holding section to maintain the medium
media: the continuous-plate heat exchanger and the at the required temperature and a flash cooler where
continuous injector-flash cooler. The continuous- the hot medium is passed through an expansion
plate heat exchanger is illustrated in Fig. 5.8 and valve into a vacuum chamber resulting in its virtu-
consists of a series of steam-heated plates against ally instantaneous cooling. Temperature profiles of
which the medium is circulated to raise the tempera- the two systems are shown in Fig. 5.10a and 5.10b
ture to the desired level, at which the medium is from which it may be seen that the use of the injector
Steam in
Cooling Cooling Unsterile
water in water out medium in
Steam
out to
vacuum
Cooling heat Pre-heat heat
Expansion
exchanger exchanger
valve
Sterile
medium out
Fic. 5.9. Flow diagram ofa typical continuous injector-flash cooler sterilizer.
101
Principles of Fermentation Technology
2~3 min and cooling periods are so rapid that they may be
ignored in calculations of temperature time regimes,
140 °C whereas these would have to be taken into account
Flash cooling when considering the plate-exchange system, albeit
2 the contribution being considerably smaller than in
=
igo} the batch system. Richards (1968) quoted the fol-
©
Q: lowing example to illustrate the range of tempera-
‘S
oO
Ke ture-time regimes which may be employed to
achieve the same probability of obtaining sterility.
The Del factor for a sterilization was 45.7 and the
Time
following temperature time regimes were calculated
to give the same Del factor:
20 20
seC 2~3 min sec
Temperature Holding time
144 °C
130° 2.44 min
2 135° OE ae
=)
140° 18.9 sec
6
© 150° DE eSEG
Q
=
o
=
A=In22 (5.11)
type results in much shorter heating and cooling x,
periods. where x, . : : :
Banks (1979) summarized the advantages and is the concentration of essential heat labile
disadvantages of the steam injection system: nutrient in the original medium,
x, is the concentration of essential heat labile
Advantages: nutrient in the medium after a sterilization
(i) It may be used for media containing sus- time, ¢.
pended solids. As considered earlier, the destruction ofa nutrient
Low capital cost. may be considered a first-order reaction and its rate
Easy cleaning and maintenance. would be affected by temperature as described by
Shorter heating and cooling periods. the Arrhenius equation:
(v) Higher steam utilization efficiency.
xy
———S= Se
—kt
Disadvantages: Xo
(i) Foaming may occur during both heating where £ is the reaction rate constant
and cooling.
(ii) Direct contact of the medium with steam OG al
all Cot
dense dilution and requires ‘clean’ steam, Thus, taking natural logarithms,
free from anticorrosion additives.
therefore NSO
but fe A~ ART),
103
Principles of Fermentation Technology
complete and a positive pressure is maintained; there is always the possibility ofan organism passing
otherwise a vacuum may develop and unsterile air through the filter, regardless of the filter’s depth.
may be drawn into the vessel. Despite the apparent superiority of the absolute
filters, the fermentation industry has tended to rely
upon the fibrous-type filter because it tends to be
more robust, cheaper, and has a lower pressure drop
STERILIZATION OF THE FEEDS than the absolute filter. However, more robust abso-
lute filters are now available which have been
specifically designed for the fermentation industry.
A variety of additives may be administered to a
For example, Pall Process Filtration Ltd.
fermentation during the process and it is essential
(Portsmouth, England) have produced a polytetra-
that these materials are sterile. The sterilization
fluorethylene cartridge filter incorporating a uni-
method depends on the nature of the additive and
form pore size of 0.2 wm. The filter, which is steam
the volume and feed rate at which it is administered.
sterilizable, exhibits a high voids volume (80%)
If the additive is fed in large quantities then continu- —
which provides free space for air flow resulting in a
ous sterilization may be desirable, for example,
low pressure drop. Smith (1981) cited the use of
Aunstrup et al. (1979) cited the use of continuous
absolute filtration in the sterilization of air for the
heat sterilization for the feed medium of microbial
ICI biomass continuous fermenter.
enzyme fermentations. Aunstrup ¢¢ al. also referred
to the use of filtration for the sterilization of certain
feeds. Batch sterilization of feed liquids normally
involves steam injection into the material held in The theory of fibrous filters
storage vessels. Whatever the sterilization system
employed it is essential that all ancillary equipment
According to Aiba et al. (1973) the removal of
and feed pipework associated with the additions are
micro-organisms from an aerosol by a fibrous filter
sterilizable.
may occur by one, or a combination, ofthe following
mechanisms:
Chapter 9). Although it is possible to sterilize air by Aiba et al. have given detailed quantitative analysis
heat treatment, the most commonly used steriliza- of these mechanisms but this account will be limited
tion process is filtration. Filters for the removal of to a description of the overall efficiency of operation
micro-organisms from an environment may be of fibrous filters. Several workers (Ranz and Wong,
divided into two large groups—those in which the 1952; Chen, 1975) have put forward equations relat-
pores in the filter are smaller than the particles ing the collection efficiency of a filter bed to various
which are to be removed and those in which the characteristics of the filter and its components.
pores are larger than the particles which are to be However, a simpler description cited by Richards
removed. The former type may be regarded as an (1967) may be used to illustrate the basic principles
absolute filter, so that filters of this type (provided of filter design.
they are not physically damaged) are claimed to be Ifit is assumed that if a particle touches a fibre it
100% efficient in removing micro-organisms. Filters remains attached to it, and that there is a uniform
of the latter type are made of a fibrous material such concentration of particles at any given depth in the
as cotton, glass, slag or steel wool and the gaps filter, then each layer of a unit thickness of the filter
between the fibres are usually in the range 0.5 to 15 should reduce the population entering it by the same
fm. Thus, in theory, the removal of micro- proportion; which may be expressed mathemati-
organisms by a fibrous filter cannot be absolute as cally as:
104
Sterilization
dN
x
—=-—KN (1) or 2.303 log io -.= —K Xo
1
eee KY Fic. 5.12. The effect of increasing linear air velocity on K and Xyy
~tO ~ ofa filtration system (Richards, 1967).
105
Principles of Fermentation Technology
TABLE 5.3. Xo) values for the removal of a range of micro-organisms by a variety of filtration materials
(Humphrey, 1960 and Riviere, 1977)
Diameter
fibres Air speed Xo
Filter material (wm) Micro-organism (cm/s) (cm)
107
CHAPTER 6
TaBLe 6.1. Jnoculum development and production media for a range ofprocesses
a ee ee
Inoculum development
Process medium Production medium References
The sub-master culture is used to inoculate a shake involved the use of one sub-master culture to
flask (250 or 500 cm*) which, in turn, is used as develop a bulk inoculum which was subdivided,
inoculum for a larger flask, or a laboratory fer- stored in a frozen state and used as inocula for
menter, which may then be used to inoculate a several months. A single colony, derived from a
pilot-scale fermenter. Culture purity checks are car- sub-master culture, was inoculated into liquid
ried out at each stage to detect contamination as medium and grown to maximum log phase. This
early as possible. Although the results of these tests culture was then transferred into nineteen times its
may not be available before the culture has reached volume of medium and incubated again to the
the production plant at least it is known at which maximum log phase, at which point it was dispensed
stage in the procedure contamination occurred. For in 20-cm® volumes, plug frozen and stored at below
a sporulating organism the process may be modified —20°. At least 3% of the samples were tested for
to facilitate the use of aspore suspension as inoculum purity and productivity in subsequent fermentation
and this will be discussed in more detail later in this and, provided these were suitable, the remaining
chapter. samples could be used as initial inocula for sub-
Lincoln (1960) suggested a more elaborate proce- sequent fermentations. One of the thawed samples
dure for the development of inoculum for bacterial was used as a 5% inoculum for a seed culture which,
fermentations which, with minor modifications, is in turn, was used as a 5% inoculum for the next
applicable to any type of culture. The procedure stage in the programme. This procedure ensured
109
Principles of Fermentation Technology
that a proven inoculum was used for the penultimate Despite these precautions yeasts are rarely used
stage in inoculum development. for more than five to ten consecutive fermentations
The .yeasts, bacteria, fungi and streptomycetes (Thorne, 1970; Reed and Peppler, 1973) necessitat-
have different requirements for inoculum develop- ing the periodical production of a pure inoculum.
ment and these are dealt with separately. This would involve developing sufficient biomass
from a single colony to pitch a fermentation at a level
of approximately 2 grams ofpressed yeast per litre.
Hansen (1896) pioneered the use of pure inocula
THE DEVELOPMENT OF INOCULA FOR and devised a yeast propagation scheme utilizing a
YEAST PROCESSES 10% inoculum volume at each stage in the pro-
gramme and employing conditions similar to those
The most important industrial fermentations used during brewing. However, modern propaga-
utilizing yeasts are the brewing of beer and the tion schemes use inoculum volumes of 1% or even
production of biomass, of which the longest estab- lower and may use conditions different from those
lished is the production of bakers’ yeast. used during brewing. Therefore, continuous aera-
tion may be used during the propagation stage
which seems to have little effect on the beer produced
Brewing in the subsequent fermentation (Curtis and Clark,
1957). However, Hoggan (1977) stressed that the
It is common practice in the British brewing performance of fermentations in conical vessels was
industry to use the yeast from the previous fermenta- particularly affected by the growth conditions
tion to inoculate (pitch, in brewing terms) a fresh _employed in the propagator. Hoggan considered it
batch of wort. The dangers inherent in this practice necessary to use similar temperatures in the prop-
are the introduction of contaminants and the degen- agator and the fermentation and that the duration of
eration of the strain, the most common degenera- the propagation cycle be the same as the fermenta-
tions being a change in the degree offlocculence and tion time in the production vessel. Furthermore,
attenuating abilities of the yeast. In breweries should the yeast be held in the propagator for more
employing top fermentations these dangers are than a few hours after maximum growth, floccula-
minimized by collecting yeast to be used for future tion would occur which would result in a prolonged
pitching from ‘middle skimmings’. During the fer- lag phase in the subsequent fermentation.
mentation the yeast cells flocculate and float to the A number of yeast propagators (which are basic-
surface, the first cells to do this being the most ally closed, aerated vessels) have been described in
flocculent and the last cells the least flocculent. As the literature. The simplest type of propagator is a
the head of yeast develops, the surface layer (the single-stage system resembling an unstirred, aer-
most flocculent and highly contaminated yeasts) is ated fermenter which is inoculated with a shake-
removed and discarded and the underlying cells flask culture developed from a single colony (Gilli-
(the ‘middle skimmings’) are harvested and used for land, 1971). Curtis and Clark (1957) and Thorne
subsequent pitching. Therefore, the ‘middle skim- (1970) described two-stage systems which could be
mings’ will contain cells which have the desired operated semi-continuously. Thorne’s propagator
flocculence and which have been protected from consisted of two linked vessels, 1.5 and 150 dm?
contamination by the surface layer of the yeast head. respectively. The smaller vessel was filled with wort,
The pitching yeast may be treated to reduce the sterilized, cooled, aerated and inoculated with a
level of contaminating bacteria and remove protein flask-grown culture. After growth for 3 to 4 days the
and dead yeast cells by such treatments as reducing culture was forced by air pressure into the second
the pH of the slurry to 2.5 to 3, washing with water, vessel which had been filled with sterilized, cooled
washing with ammonium persulphate and treat- wort and aerated. An aliquot of 1.5 dm? was forced
ment with antibiotics such as polymixin, penicillin back into the first vessel after mixing. In a further 3
and neomycin (Mand et al., 1964; Strandskov, 1964; to 4 days the larger vessel contained sufficient
Roessler, 1968). biomass to pitch a 1000-dm* fermenter and the first
110
The Development of Inocula for Industrial Fermentations
111
Principles of Fermentation Technology
TABLE 6.2. The inoculum development programme for the production of given in Table 6.3. The reconstitution of the stock
bacitracin by Bacillus subtilis (Inskeep et al., 1951) culture involved a heat-shock treatment to stimulate
Stage Cultural conditions Incubation time
spore germination and to eliminate the weaker
spores.
Its 4-dm? shake flask inocu- 18 to 24 hours Lurie (1975) described the South African acetone-
lated with a stock
culture butanol process and stated that only 10 dm® of
2. Stage | culture inoculated 6 hours inoculum was used for a 100,000-dm* fermenter.
into 750-dm? fermenter
The use of such a small inoculum necessitated the
3 750-dm* culture inocu- Grown to the point
lated into 6000-dm? of greatest achievement ofas near perfect conditions as possible
fermenter production of to prevent contamination and to avoid an abnor-
cells
mally long lag phase.
4. 6000-dm? culture
inoculated into
120,000-dm? production
fermenter
THE DEVELOPMENT OF INOCULUM
FOR FUNGAL PROCESSES
vinegar. The acetic acid bacteria used in the vinegar
process are highly aerobic and are extremely sensi-
The preparation of inocula for fungal fermenta-
tive to oxygen starvation. Therefore, to avoid dis-
tions is probably more involved than that for bacte-
turbing the system, the cells at the end ofafermenta-
rial and yeast processes (Hockenhull, 1980). The
tion are used as inoculum for the next batch by
majority of industrially important fungi are capable
removing approximately 60% of the culture and
of asexual sporulation so it is common practice to
restoring the original level with fresh medium (Con-
use a spore suspension as seed during an inoculum
ner and Allgeier, 1976). The advantage of a highly
development programme. Three basic techniques
active inoculum apparently outweighs the disad-
have been developed to produce a high concentra-
vantages of possible strain degeneration and conta-
tion of spores for use as an inoculum.
minant accumulation.
An example of the development of inoculum for
an anaerobic bacterial process is provided by the
clostridial acetone-butanol fermentation. Although Sporulation on solidified media
this process is now only in operation in South Africa
it was a very important development in the fermen- Most fungi will sporulate on suitable agar media
tation industry during, and soon after, the First but a large surface area must be employed to pro-
World War and may even be an economic process in duce sufficient spores. Parker (1950) described a
the future (Bu’ Lock, 1975). The inoculum develop- ‘roll-bottle’ technique for the production ofspores of
ment programme, as described by Beech (1952), is Penicillium chrysogenum. 300 cm? quantities of medium
TaBLe 6.3. The inoculum development programme for the clostridial acetone-butanol fermentation (Beech, 1952)
112
The Development of Inocula for Industrial Fermentations
containing 3% agar were sterilized in 1-dm? cylin- Tasie 6.4. Medium for the submerged sporulation of P. patulum
drical bottles, which were then cooled to 45° and (Rhodes ef al. 1957)
rotated on a roller mill so that the agar set as a
cylindrical shell inside the bottle. The bottles were Whey powder, to give irefees gins
y ae Nitrogen 0.05%
inoculated with a spore suspension from a sub-mas- KH,PO, 0.4%
KC] 0.05%
ter slope and incubated at 24° for 6 to 7 days. Parker Corn-steep liquor solids to give approx. 0.04% N 0.38%
claimed that although the use of the ‘roll-bottle’
involved some sacrifice in ease of visual examina-
tion, it provided a large surface area for cultivation tation is provided by the griseofulvin process:
of spores in a vessel of aconvenient size for handling Rhodes et al. (1957) described the necessary condi-
in the laboratory. tions for the submerged sporulation of the griseoful-
vin-producing fungus, Penicillium patulum, and the
medium utilized is given in Table 6.4. These authors
Sporulation on solid media
found that for prolific sporulation the nitrogen level
had to be limited to between 0.05 and 0.1% w/v and
that good aeration had to be maintained. Also, an
Many fungi will sporulate profusely on the surface
interaction was demonstrated between the nitrogen
of cereal grains from which the spores may be
level and aeration in that the lower the degree of
harvested. Substrates such as barley, hard wheat
aeration the lower the concentration of nitrogen
bran and ground maize are all suitable for the
needed to induce sporulation. Submerged sporula-
sporulation ofa wide range of fungi. The sporulation
tion was induced by inoculating 600 cm? of the
of a given fungus is particularly affected by the
above medium, in a 2-dm’® shake flask, with spores
amount of water added to the cereal before steriliza-
from a well-sporulated Czapek-Dox agar culture
tion and the relative humidity of the atmosphere,
and incubating at 25° for 7 days. The resulting
which should be as high as possible, during sporula-
suspension of spores was then used as a 10%
tion (Vezina and Singh, 1975). Singh e¢ al. (1968)
inoculum for a vegetative seed stage in a stirred
have described a system for the sporulation ofAsper-
fermenter, the seed culture subsequently providing
gillus ochraceus in which a 2.8 dm? Fernbach flask
a 10% inoculum for the production fermentation.
containing 200 grams of‘pot’ barley or 100 grams of
The vegetative seed and production media are given
moistened wheat bran produced 5 X 10!! conidia
in Table 6.1.
after six days at 28° and 98% relative humidity. This
was 5 times the number obtainable from a Roux
bottle batched with Sabouraud agar and 50 times
The use of the spore inoculum
the number obtainable from such a vessel batched
with Difco Nutrient Agar, incubated for the same
time period. Vezina et al. (1968) have published a The stage in an inoculum development prog-
list of Phycomycetes, Ascomycetes, Deuteromycetes ramme at which a large-scale spore inoculum is used
and Basidiomycetes which are capable of sporulat- varies according to the process but it appears to be
ing heavily on cereal grains. common practice that it is the penultimate stage
Sansing and Cieglem (1973) have described the that is so inoculated. The inoculum development
mass production of spores of several Aspergillus and programme for the penicillin fermentation
Penicillium species on whole loaves of white bread. described by Parker (1950) is given in Table 6.5.
From Table 6.5 it may be seen that the penulti-
mate stage was inoculated with a spore suspension
Sporulation in submerged culture (from a ‘roll-bottle’) and that this stage may have
produced either a vegetative or a submerged spore
Many fungi will sporulate in submerged culture inoculum for the final fermentation. For the
provided a suitable medium is employed (Vezina et griseofulvin process, Rhodes et al. (1957) stated that
al., 1965). An example ofthe use ofthis technique for the spore suspension obtained from the submerged
the production of inoculum for an industrial fermen- sporulation stage could either be used for direct
113
Principles of Fermentation Technology
TaBLe 6.5. The development of inoculum of Penicillium chrysogenum (Borrow et al., 1961). Hansen (1967) described an
Sor the production ofpenicillin (Parker, 1950) inoculum development programme for the gibberel-
Iststage Master culture—mould spores in sterile
lin fermentation. Cultures were grown on long slants
sand (25 X 150 mm test tubes) of potato dextrose agar for
2nd stage Working slope culture—spore culture on agar 1 week at 24°. Growth from three slants was scraped
medium in test tube
3rd stage ‘Roll-bottle’ culture—spore culture on agar
off and transferred to a 9-dm? carboy containing 4
mediumin |-dm* bottles dm? of a liquid medium composed of 2% glucose,
4th stage Plant inoculum culture—spores or mycelium 0.3% MgSO,:7H,O, 0.3% NH,Cl and 0.3%
grown in first stage of plant fermentation
units
KH, PO,. The medium was aerated for 75 hours at
Sth stage Penicillin production culture—mycelium 28° before transfer to a 100-dm* seed fermenter
grown in the 2nd stage of plant containing the same medium.
fermentation units
The major problem in using vegetative mycelium
as initial seed is the difficulty of obtaining a uniform,
standard inoculum. The procedure may be
inoculation of the production fermentation or it
improved by fragmenting the mycelium in an
could be germinated in an inoculum development
homogenizer, such as a Waring blender, prior to use
medium to yield a vegetative inoculum for the final
as inoculum. This method provides a large number
fermentation. The latter course was preferred and
of mycelial particles and therefore a large number of
an inoculum volume of 7—10% was used.
growing points. Worgan (1968) has given a detailed
When considering the production of gluconic acid
account of the use of this technique in the prepara-
by Aspergillus niger, Lockwood (1975) discussed the
tion of inocula for the submerged culture of the
merits of inoculating the final fermentation directly
higher fungi.
with a spore suspension as compared with germinat-
ing the spores in a seed tank to give a vegetative
inoculum. Direct spore inoculation would avoid the
cost of installation and operation of the seed tanks The effect of the inoculum on the
whereas the use of germinated spores would reduce morphology of fungi in submerged culture
the fermentation time of the final stage, thus allow-
ing a greater number of fermentations to be carried
When filamentous fungi are grown in submerged
out per year. However, labour costs for the produc-
culture the type of growth varies from the ‘pellet’
tion of the vegetative inoculum could be almost as
form, consisting of compact discrete masses of
high as for the final fermentation although some of
hyphae, to the filamentous form in which the hyphae
these costs may be recovered, in that gluconic acid
form a homogeneous suspension dispersed through
produced in the penultimate stage would be recover-
the medium (Whitaker and Long, 1973).
able from the final fermentation broth and would
The morphology ofa fungus in submerged culture
contribute to the buffering capacity throughout the
has been shown to be critical in many industrial
fermentation. Thus, Lockwood claimed that the
fermentations and two of the factors determining
choice of inoculum for the production stage depends
fungal form is the medium composition and the
on the length ofthe cycle of the fermentation process,
concentration of spores in a spore inoculum. Usu-
plant size and capacity and the availability and cost
ally, a high spore inoculum will tend to produce a
of labour.
disperse form of growth whilst a low one will favour
pellet formation (Foster, 1949). The effect of the
concentration ofa spore inoculum on the morphol-
Inoculum development for vegetative fungi
ogy ofPenicillium chrysogenum is given in Table 6.6.
The citric acid, penicillin, submerged mushroom
Some fungi will not produce asexual spores and culture and fungal protein processes have been
therefore an inoculum ofvegetative mycelium must shown to be particularly affected by the morphology
be used. Gibberella fujikuroi is such a fungus and is of the producing fungus and this is summarized in
used for the commercial production of gibberellin Table 6.7. Therefore, in the commercial production
114
The Development of Inocula for Industrial Fermentations
TABLE 6.6 The effect of spore concentration and medium on the morphology
THE DEVELOPMENT OF INOCULUM FOR
of P. chrysogenum in liquid culture (Camici et al., 1952)
a em STREPTOMYCETE PROCESSES
Spore concentration in
Medium the medium Morphology
The vast majority of industrially important
Corn-steep More than 10 x 10°dm~* Filamentous streptomycetes will produce asexual spores. There-
dextrin Less than 10 X 10°dm-? Pellets
fore, as is the case for the fungi, itis common practice
Czapek-dox More than 3.0 X 10° dm? Filamentous to use a spore suspension as seed during an
Less than 3.0 X 10° dm~° Pellets inoculum-development programme. Sporulation on
Glucose- More than 2.0 X 10° dm™?
agar media appears to be the most commonly used
Filamentous
lactose- Less than 2.0 X 10°dm ° Pellets technique for the production ofa high concentration
ammonium of spores. However, Stark and Smith (1961)
lactate
———
eee observed that the agar used to solidify the medium
may have an inhibitory effect on the growth and/or
sporulation of the organism and that this detrimen-
of these products, it is critical to grow the fungus in tal action may be reduced by washing the agar to
the desired morphological form which necessitates remove toxic substances. Some representative
the use of an inoculum which will achieve this end. media for the sporulation of some streptomycetes
If the production fermentation is to be inoculated are given in Table 6.8.
with a spore suspension then the spore concentra- Very little information is available on the mor-
tion must be such as to produce the production phology of streptomycetes grown in submerged cul-
culture in the desired morphological form; if a veg- ture (Williams et al., 1974) but it appears that
etative inoculum is to be used for the production pelleting does not play as critical a role in streptomy-
fermentation then, again, the concentration of its cete fermentations as it does in fungal ones. Williams
spore inoculum must be such as to produce the et al. reported the formation of‘colonies’ (pellets) by
vegetative inoculum in the desired morphological antibiotic-producing submerged cultures of strep-
form. Furthermore, where a spore inoculum is ger- tomycetes growing in commercial media. However,
minated in a seed fermenter the medium employed these authors used only one level of inoculum and
must be such that the vegetative inoculum is in the therefore the formation of colonies could not be
correct form. Whitaker and Long (1973) have sum- related to the spore concentration. Lawton et al.
marized the effects of medium composition on the (1984) have demonstrated that a wide range of
morphology of fungi in submerged culture. pellet morphological types are found amongst the
The responses ofthe fermentations listed in Table streptomycetes and that the morphological form of
6.7 to the morphology of the producing organism some species is influenced by the concentration of
appears to be due to aeration effects and these will spores in the inoculum, the medium composition
be dealt with in more detail in Chapter 9. and the shear forces operating during culture. How-
TaBse 6.7. The effect offungal morphology on the performance of some industrial fermentations
Optimum
morphological
Fermentation Organism form References
115
) (PEICE
Principles of Fermentation Technology
ever, little information appears to be currently avail- tion of a plant fermenter from a laboratory vessel.
able on the effect of morphological form on product The apparatus is shown in Fig. 6.2. The connecting
formation by streptomycetes in submerged culture. point A is normally covered with the blank plug a
and prior to inoculation this plug is slightly
loosened, valve E closed and valve F opened to allow
THE ASEPTIC INOCULATION OF steam to exit at A. Valve F is then closed and E
PLANT FERMENTERS opened so that when a is removed sterile air will be
released from the vessel. After removal, plug a is
placed in strong disinfectant. Blank plug b is then
The inoculation of plant scale fermenters may
involve the transfer of culture from a laboratory removed and a coupling made between B and A.
fermenter, or spore suspension vessel, to a plant
fermenter, or the transfer from one plant fermenter
Steam supply
to another.
116
The Development of Inocula for Industrial Fermentations
Sterile air
Plant fermenter
117
Principles of Fermentation Technology
from n-paraffins by glycerol auxotroph GL-21. Agr. Biol. Sincu, K., Sencar, S$. N. and Vezina, C. (1968) Large scale
Chem. 36, 1141-1146. transformation of steroids by fungal spores. Appl. Microbiol.
Lawton, P., Wuiraker, A., Opett, D. and Stowe tt, J. D. 16, 393-400.
(1984) A study on gross morphology of actinomycetes in Smitu, G. M. and Catam, C. T. (1980) Variations in inocula and
shake fiasks. Presented at the Society for General Micro- their influence on the productivity of antibiotic fermenta-
biology meeting, Reading University, January 1984. tions. Biotechnol. Letters, 2(6), 261-266.
Lincoin, R. E. (1960) Control of stock culture preservation and SNELL, R. L. and Scuweicer, L. B. (1949) Citric acid by
inoculum build-up in bacterial fermentation. J. Biochem. fermentation. U.S. Patent 2,492,667.
Microbiol. Tech. Eng. 2, 481-500. Sotomons, G. L. (1975) Submerged culture production of myce-
Locxwoop, L. B. (1975) Organic acid production. In The lial biomass. In The Filamentous Fungi, Vol. 1, pp. 249-264
Filamentous Fungi, Vol. 1, pp. 140-157. (Editors Smith,J. E. (Editors Smith,J.E. and Berry, D. R.). Arnold, London.
and Berry, D. P.). Arnold, London. Stark, W. M. and Smirn, R. L. (1961) The erythromycin
Lurie, J. (1975) In Large Scale Fermentations for Organic Solvents. fermentation. Prog. Ind. Micro. 3, 211-230.
Octagon Papers, Vol. 2, pp. 70-73 (Editors Powell, A. J.and STRANDSKOV, F. B. (1964) Yeast handling in a brewery. Amer. Soc.
Bu’Lock,J. D.). Dept. of Extra Mural Studies, The Univer- Brewing Chemists Proc., pp. 76-79.
sity, Manchester. Szuecs, J. (1958) Growing mushroom mycelium. U.S. Patent
Manp1, B., GRuNEWALD,J. and VorRKE.Ius, G. A. (1964) The 2,850,841.
application of time-saving brewing methods. Brauwelt, Tuorne, R. S. W. (1970) Pure yeast cultures in brewing. Process
104(80), 1541-1543. Biochem. 5(4), 15-22.
Meyrat, J.and SucHAnek, G. (1972) Inoculation techniques— UnvDERKOFLER, L. A. (1976) Microbial enzymes. In Industrial
effects due to quality and quantity ofinoculum. In Methods in Microbiology, pp. 128-164 (Editors Miller, B. M. and Litsky,
Microbiology, Vol. 7B, pp. 159-209 (Editors Norris,J.R. and W.). McGraw-Hill, New York.
Ribbons, D. W.). Academic Press, London. Vezina, C., SEHGAL, S. N. and Sineu, K. (1968) Transformation
Parker, A. (1950) Aseptic technique in industrial scale fermen- of organic compounds by fungal spores. Adv. Appl. Microbiol.
tations. In Recent Advances in the Fermentation Industries. Royal 10, 221-268.
Institute of Chemistry, London. Vezina, C. and Sineu, K. (1975) Transformation of organic
Puitures, D. H. (1966) Oxygen transfer in mycelial pellets. compounds by fungal spores. In The Filamentous Fungi, Vol.
Biotech. Bioeng. 8, 456-460. 1, pp. 158-192 (Editors Smith, J. E. and Berry, D. R.).
Reep, G. and Pepper, H. J. (1973) In Yeast Technology, pp. Arnold, London.
664-668. Avi, Westport. Vezina, C., Sincu, K. and Seue@at, S. N. (1965) Sporulation of
Ruopes, A., Crosse, R., Fercuson, T. P. and FLercuer, D. L. filamentous fungi in submerged culture. Mycologia, 57, 722—
(1957) Improvements in or relating to the production of the 736.
antibiotic griseofulvin. British Patent 784,618. Wuitaker, A. and Lone, P. A. (1973) Fungal pelleting. Process
Roessier, J. G. (1968) Yeast management in the brewery, Part Biochem. 8(11), 27-31.
II. Yeast handling and treatment. Brew. Digest, 43(9), 94, Wi.uiams, S. T., EnrwistLe, S. and Kurytowicz, W. (1974)
96, 98, 102, 115. The morphology of streptomycetes growing in media used
Rosen, K. (1977) Production of bakers’ yeast. Process Biochem. for commercial production of antibiotics. Microbios. 11, 47—
12(3), 10=12: 60.
Sansinc, G. A. and Creciem, A. (1973) Mass propagation of Woraan, J. T. (1968) Culture of the higher fungi. Prog. Ind.
conidia from several Aspergillus and Penicillium species. Appl. Micro. 8, 73-139.
Microbiol. 26, 830-831.
119
CHAPTER 7
Design of a Fermenter
INTRODUCTION perforated pipes. In later modifications, mechanical
impellers were used to increase the rate of mixing
A research team led by Chaim Weizmann in and to break up and disperse the air bubbles. This
Great Britain during the First World War (1914— procedure led to the compressed-air requirements
1918) developed a process for the production of being reduced by a factor of5. Baffles on the walls of
acetone by a deep liquid fermentation using Clos- the vessels prevented a vortex forming in the liquid.
tridium acetobutyliticum which led to the eventual use Even at this time it was recognized that the cost of
of the first truly large-scale aseptic fermentation energy necessary to compress air could be 10 to 20%
vessels (Hastings, 1978). Contamination, particu- ‘of the total production cost. As early as 1934,
larly with bacteriophages, was often a serious prob- Strauch and Schmidt patented a system in which
lem, especially during the early part ofa large-scale the aeration tubes were provided with water and
production stage. Initially, no suitable vessels were steam for cleaning and sterilizing.
available and attempts with alcohol fermenters Prior to 1940, the other important fermentation
fitted with lids were not satisfactory as steam sterili- products besides bakers’ yeast were ethanol,
zation could not be achieved at atmospheric pres- glycerol, acetic acid, citric acid, other organic acids,
sure. Large mild-steel cylindrical vessels with enzymes and sorbose (Johnson, 1971). These pro-
hemispherical tops and bottoms were constructed cesses used highly selective environments such as
that were capable of being sterilized with steam acidic or anaerobic conditions or the use of an
under pressure. Since the problems of aseptic addi- unusual substrate, resulting in contamination being
tions of media or inocula had been recognized, steps a relatively minor problem compared with the
were taken to design and construct piping, joints acetone fermentation or the subsequent aerobic
and valves in which sterile conditions could be antibiotic fermentations.
achieved and maintained when required. Although The decision to use submerged culture techniques
the smaller seed vessels were mechanically stirred, for penicillin production, where aseptic conditions,
the large production vessels were never mechani- good aeration and agitation were essential, was
cally stirred, and the large volumes ofgas produced probably a very important factor in forcing the
during the fermentation continually agitated the development of carefully designed and purpose-
vessel contents. This meant that considerable exper- built fermentation vessels. In 1943, when the British
tise was built up in the construction and operation of government decided that surface culture production
this aseptic anaerobic process for production of was inadequate, none of the fermentation plants
acetone-butanol. were immediately suitable for deep fermentation,
The first true large-scale aerobic fermenters were although the Distillers Company solvent plant at
used in Central Europe in the 1930s for the produc- Bromborough only needed aeration equipment to
tion of compressed yeast (de Becze and Liebmann, - make it suitable for penicillin production (Hastings,
1944). The fermenter consisted ofa large cylindrical 1971).
tank with air introduced at the base via networks of Construction work on the first large-scale plant to
120
Design of a Fermenter
produce penicillin by deep fermentation was started TABLE 7.1. Service provisions
for a fermentation plant
on 15th September 1943, at Terre Haute in the
Compressed air
United States of America, building steel fermenters Sterile compressed air (at 1.5 to 3.0 atmospheres)
with working volumes of 54,000 dm* (12,000 gal- Chilled water (12 to 15°)
lons) (Callaham, 1944). The plant was operational Cold water (4°)
Hot water
on 30th January 1944. Unfortunately no other con- Steam (high pressure)
struction details were quoted for the fermenters. Steam condensate
Initial agitation studies in baffled stirred tanks to Electricity
Stand-by generator
identify variables were also reported at this time by Drainage ofeffluents
Cooper, Fernstrom and Miller (1944), Foust, Mack Motors
and Rushton (1944) and Miller and Rushton (1944). Storage facilities for media components
Control and monitoring equipment for fermenters
The geometry of the stirred vessel (Cooper e¢ al., Maintenance facilities
1944) has been used in the design of later conven- Extraction and recovery equipment
tional fermenters. Accessibility for delivery of materials
121
Principles of Fermentation Technology
Stirrer
shaft
Aseptic
inoculation
Aseptic
pipe inoculation
Stirrer pipe
shaft
seal
Working
Working level level
Baffle Impeller
Baffle
Sampling
point enDe a
Impeller Sampling
point aa
Sterile air age |j~— |<
Air
Fic. 7.1. Diagram ofa fermenter with one multi-bladed impeller.
sparger
Taste 7.2. Details of geometrical ratios offermenters with single multi-bladed impellers
(see Fig. 7.1)
TaB_e 7.3. Details of geometrical ratios of fermenters with three multi- (1 to 30 dm*) it is possible to use glass and/or
bladed impellers (see Fig. 7.2) stainless steel, Glass is useful because it gives smooth
surfaces, is non-toxic, corrosion proof and it is usu-
Jackson Aibaetal. Paca etal.
Dimension (1958) (1973) (1976) ally easy to examine the interior of the vessel. Two
basic types of fermenter are used:
Operating volume — 100,000 dm? 170 dm?
(total) 1. A glass vessel with a round or flat bottom and
Liquid height (L) — — 150cm
a top flanged carrying plate (Fig. 7.3). The
L/D (tank diameter) — — Ihodl
Impeller diameter 0.34-0.5 0.4 0.33 large glass containers originally used were
Baffle width/D 0.08-0.1 0.095 0.098 borosilicate battery jars (Brown and Peterson,
Impeller height/D 0.5 0.24 0.37 1950). All vessels of this type have to be
P/V 0.5-1.0 os 0.74
P/W 0.5-1.0 0.85 0.77 sterilized by autoclaving.
ay 0.5-1.0 0.85 0.77 A glass cylinder with stainless-steel top and
P/Z Jel 0.91 bottom plates (Fig. 7.4). These fermenters
H/D 1.0-1.6 2 2.95
may be sterilized in situ, but 30 cm diameter is
122
Design of a Fermenter
the upper size limit to safely withstand work- Fic. 7.4. Glass fermenter with top and bottom plates (L.H.
Engineering, Stoke Poges, England).
ing pressures (Solomons, 1969).
123
Principles of Fermentation Technology
_- Stirrer
Top
bearing : shaft
as
Fermenter
top plate Bottom
VA
\s
bearing @
Set screw
oO
©
Bearing housing
Drive pin
Rotating seal
Stationary seal KH ) |
NAS
aN; AY BWH-O
Oo
er
oY
eo
oo
aa
RSs
Stationary seal-insert
MAGNETIC DRIVES
The problems of providing a satisfactory seal
when the impeller shaft passes through the top or
bottom plate of the fermenter may be solved by the
use of amagnetic drive in which the impeller shaft
does not pierce the vessel (Cameron and Godfrey,
1969). A magnetic drive (Fig. 7.10) consists ofa
driving and a driven magnet. The driving magnet is
held in bearings in a housing on the outside ofthe
head plate and connected to a drive shaft. The Fic. 7.10. A line drawing ofa magnetic coupling with driving and
internal driven magnet is placed on one end of the driven magnets integral with lid of vessel: (1) driving magnet; (2)
housing for driving magnet assembly; (3) lid of vessel; (4)
impeller shaft and held in bearings in a suitable bearing; (5) adjustable collar with locking screw for locating
housing. When multiple ceramic magnets have been bearing; (6) driving shaft terminated for cable drive (Cameron
used, it has been possible to transmit power across a and Godfrey, 1969).
127
Principles of Fermentation Technology
Seal
gap of 16 mm. Using this drive it has been possible Top end cap (a)
(1) Ball bearing
- to agitate water in baffled vessels of up to 300 dm*
capacity at speeds of 300 to 2000 rpm. It would be Bearing
housing bod
® ASS
an. 2
SY (5) Grub screw
Air
outlet
Support
grid
BRESKK KR KI
pooneeeeeeee
ns] —
:
Stainless
pon ent s] steel
4
SOS RG
SS
ORM SR
MA AS] CaINi
Meeecechechcectceeeren g
PS
SSSR O HY eeeenenen Air flow
Seetececec cececececococecerere neres
Petctetetcecececenenererere
poceecsnssesesa neceel control
Eee Fibrous
KK
SOKO
ececececenens filter
;
packing Condensate
drain and
steam bleed
ee Support
grid
heat, filtration through fibrous material and filtra- Fic. 7.13. An arrangement of packed air filter and fermenter
tion through granular material. Heat is generally (Richards, 1968).
too costly for full-scale operation.
The chosen filter must remove micro-organisms sterilizing mineral slag-wool filters because channel-
to a high level of efficiency, to be relatively cheap, ling occurred on prolonged or repeated sterilization.
robust and have a low pressure drop. The simplest They recommended some form of dry heating in
air filters (Fig. 7.12) consist ofa steel casing with an which the filter should be held at 160° to 180° for 2
air inlet in the base and an outlet at the top. The hours.
packing material (5 to 15 wm diameter), usually An alternative approach is to use a steam-jacketed
glass wool, glass fibre or mineral slag wool is sup- air filter (Fig. 7.14). At the beginning of a steriliza-
ported on a grid or perforated plate and held to the Air
required packing density by a further grid or plate.
A fermenter air filter will be sterilized in association
with the fermenter. Two procedures are commonly
followed depending on the construction of the filter
unit.
Figure 7.13 is the simple unit described by
Richards (1968). During sterilization the main non-
sterile air inlet valve A is shut, and initially the
sterile air valve B is closed. Steam is applied at valve Porous
C and air is purged downwards through the filter to filter
a bleed valve at the base. When the steam is issuing
Steam
freely through the bleed valve, the valve B is opened jacket Condensate
to allow steam to pass into the fermenter as well as
the filter. It is essential to adjust the bleed valve to
ensure that the correct sterilization pressure is main- Sterile air
tained in the fermenter and filter for the remainder Condensate
of the sterilization cycle. Terjesen and Cherry (1947)
showed that this procedure was unsuitable for Fic. 7.14. Design for a simple steam-jacketed air-filter.
130
Design of a Fermenter
Aeration and agitation stream of steam and condensate out of the sampling
port. Valve A is then opened slightly to cool the
piping. The broth is discarded. Valve C is then
Equipment designs for the achievement ofaseptic
closed and a sample is collected. Valve A is then
aeration and agitation already have been described
closed and the piping is resterilized and left in the
~ earlier in this chapter.
out of use arrangement.
An alternative arrangement for a sampling port is
illustrated in Fig. 7.16. In normal operation valves
The addition of inoculum, nutrients
and B are closed and valves C and D are opened
and other supplements
until the piping has been sterilized. Valve C is then
partially closed, valve D completely closed and
To prevent contamination during an addition it is valve B partially opened to allow a slow stream of
essential that both the addition vessel and the fer- steam and condensate out of the sampling port. Just
menter be maintained under a positive pressure and prior to sampling, valve A is opened briefly to cool
the addition port be equipped with a steam supply.
The aseptic inoculation of laboratory, pilot and
plant fermenters is described in detail in Chapter 6.
Vessel
Sampling
131
/PFT-J
Principles of Fermentation Technology
. }
the pipe and the broth is discarded. Valve C is then
closed and a sample is collected. Valve A is then
closed’ and the piping is resterilized. In between == Rasty
77} Handwheel
collecting samples valves C and D are left partially
rar
ww cE
open. ASA Bridge
Lax
222222TD
os
AS
Pillar
Foam control
Stem
132
Design of a Fermenter
Handwheel
Stainless steel
i piston
Disc spring
washer
Stem
Upper valve
: ring
Bonnet gasket
Fic. 7.19. Piston valve (Kemplay, 1980).
Lubricant screw
Globe valves
133
Principles of Fermentation Technology
S
protected seals, a minimal pressure drop and a Handle adaptor
<1 TN
LES
trol.
Stem seals +
Needle valves
Butterfly valves
134
Design of a Fermenter
Check valves
135
Principles of Fermentation Technology
most commonly used in fermenter designs. The applications and have been developed for specific
functional part is a hinged disc which closes against purposes or closely related processes. Some are
a seat ring when the intended direction of flow is historical developments, such as the earliest forms
accidentally reversed. of packed tower, others were being developed in
parallel with the standard mechanically stirred fer-
menter during the 1940s while other approaches are
Pressure-control valves more recent. Aspects of this topic have been
reviewed by Prokop and Votruba (1976), Katinger
When planning the design of a plant for a specific (1977), Hamer (1979), Levi, Shennan and Ebbon
process, the water, steam and air should be at (1979), Solomons (1980) and Sittig (1982).
different, but specified pressures and flow rates in
different parts of the equipment. For this reason it is
essential to control pressures precisely and this can The packed tower
be done using reduction or retaining valves.
The packed tower is now primarily of historical
PRESSURE-REDUCTION VALVES
interest in the fermentation industry except for
Pressure-reduction valves are incorporated into
effluent treatment. A vertical cylindrical column
pipelines when it is necessary to reduce from a
was packed with pieces of some relatively inert
higher to a lower pressure, and be able to maintain
material, e.g. wood shavings, twigs, coke, an aggre-
the lower pressure in the downstream side within
gate or polyethylene. Initially both medium and
defined limits irrespective of changes in the inlet
cells were fed into the top of the packed bed. Once
pressure or changes in demand for gas, steam or
the cells had adhered to the support and were
water.
growing well as a thin film, fresh medium was added
at the top of the column and the fermented medium
PRESSURE-RETAINING VALVES
was removed from the bottom of the column. The
A pressure-retaining valve will maintain pressure
best known example is the vinegar generator, in
in the pipeline upstream ofitself, and the valve is
which ethanol was oxidized to acetic acid by strains
designed to open with a rising upstream pressure. It
of Acetobacter supported on beech shavings; the first
is constructed with a reverse action of the pressure-
recorded use was in 1670 (Mitchell, 1926). More
reducing valve.
recently packed towers have been used for sewage
and effluent treatment (Noble e¢ al., 1964). In treat-
ment ofgas liquor, a column was packed to a height
Safety valves
of 7.9 m with ‘Dowpac’, a polystyrene derivative.
The main advantages compared with other methods
Safety valves must be incorporated into every air
of effluent treatment being its simplicity of operation
or steam line and vessel which is subject to pressure
and a saving in land because of the increased surface
to ensure that the pressure will never exceed the safe
areas within the column.
upper limit recommended by a manufacturer or a
code of practice. In the simplest valves, a spindle is
lifted from its seating against the pressure of gas,
The tower fermenter
steam or liquid. Once the pressure falls below the
value set by the tensioned spring, the spindle should
return to its original position. Itis difficult to formulate a single definition which
encompasses all the types of tower fermenter. Their
main common feature appears to be their
OTHER FERMENTATION VESSELS height: diameter ratio or aspect ratio. Such a defini-
tion has been given by Greenshields e¢ al. (1971) who
Some ofthe other forms of fermentation vessel will described a tower fermenter as an elongated non-
now be considered. These vessels have more limited mechanically stirred fermenter with an aspect ratio
136
Design of a Fermenter
of at least 6:1 for the tubular section or 10:1 overall, (1977). The fermenters used by all these workers
through which there is a unidirectional flow ofgases. were basically similar. Each consisted of acolumn
Several different types of tower fermenter exist and forming the body of the vessel and a number of
these will be examined in broad groups based on perforated plates which were positioned across the
their design. fermenter, dividing it into compartments. Approxi-
The simplest types of fermenter are those that mately 10% of the horizontal plate area was perfo-
consist ofatube into which substrate is pumped and rated. The possibility of introducing media into
air sparged at the base and the product removed individual stages independently was discussed by
from the top. These could also be operated as batch Lee et al. (1971). Besik (1973) decribed a down-flow
fermenters. This type of fermenter was first tower in which substrate was fed in at the top and
described for citric acid production (Snell and overflowed through down spouts to the next section
Schweiger, 1949). This batch fermenter was in the while air was supplied from the base.
form of a glass column having a height: diameter
ratio of 16:1 with a volume of 3dm*. Humid sterile
air was supplied through a sinter at the base. Steel, The Waldhof-type fermenter
Lentz and Martin (1955) reported an increase in
scale to 36 dm’ for a fermenter of this type.
The investigations on yeast growth in sulphite
In 1965 the brewing industry began to use tower
waste liquor in Germany, Japan and the United
fermenters which were more complex in design.
States of America led to the development of the
Hall and Howard (1965) described small-scale fer-
Waldhof-type fermenter (Inskeep et al., 1951;
menters that consisted of water-jacketed tubes of
Watanabe, 1976). Inskeep et al. (1951) have givena
. various dimensions which were inclined at angles of
description of a production vessel based on a mod-
9 to 90° to the horizontal. Air and mash were passed
ified design of the original design of Zellstofffabrik
in at the base and effluent beer was removed at the
Waldof. The fermenter was of carbon steel, clad in
top.
stainless steel, 7.9 min diameter and 4.3 m high with
A vertical tower fermenter design (Chapter 2)
a centre draught tube 1.2 m in diameter. The
was patented by Shore ef al. (1964). Perforated
draught tube was held by tie rods attached to the
plates were positioned at intervals in the tower to fermenter walls. The operating volume was 225,000
maintain maximum yeast production. The settling
dm? of emulsion (broth and air) or 100,000 dm? of
zone, which could be of various designs, was to
broth without air. Non-sterile air was introduced
provide a zone free of rising gas so that the cells
into the fermenter through a rotating pin-wheel type
could settle and return to the main body ofthe tower of aerator, composed of open-ended tubes rotating
and the clear beer could be removed. This design at 300 rpm (Fig. 7.27). The broth passed down the
must be considered as an intermediate between tube from the outer compartment and reduced the
single- and multistage systems. Towers of up to foaming. Fresh broth was fed in at the top of the
20,000 dm? capacity and capable of producing up to vessel and excess was harvested by the overflow
90,000 dm? day_' have been installed. Greenshields pipe.
and Smith (1971) commented that it was difficult to
predict the upper operating limits for these fermen-
ters. Experiments with particular yeast strains in Acetators and cavitators
pilot-size towers were essential to establish optimum
full-scale operating conditions.
The next group of fermenters are those which are Fundamental studies by Hromatka and Ebner
multistage systems, first described by Owen (1948) (1949) on vinegar production showed that if
and Victorero (1948) for brewing beer although Acetobacter cells were to remain active in a stirred
these systems were not used on an industrial scale. aerated fermenter, the distribution of air had to be
Later workers have included Kitai, Tone and Ozaki almost perfect within the entire contents of the
(1969), Prokop et al. (1969), Lee, Erikson and Fan vessel. They solved the full-scale problem by the use
(1971), Besik (1973) and Schugerl, Lucke and Oels of a self-aspirating rotor (Ebner, Pohl and Enenkel,
137
Principles of Fermentation Technology
Section ‘A—A’
Vinegar —— I id =i
out
Stage 1
139
Principles of Fermentation Technology
Direction of
Pipe for CO, Heat flow
entry and exchanger
pressure
cleaning
--~---}———— Air/ammonia
Sterile Sparge pipes
medium =
inlet
Vessel cleaning Fig. 7.33a. Air-lift fermenter with outer loop (Taylor and Senior,
and pressure 1978).
Conical jacket delivery pipe
outlet
CO, washing Exhaust gas
Thermometer antares
CO, injection } Outlet
cock cock Conical jacket inlet Gas
Yeast cock disengagement
with sightglass space
In the ICI Ltd continuous process, air and gase- Air outlet
ous ammonia are introduced at the base of the Suction |
he (one-third)
142
Design of a Fermenter
P. (1951) Food yeast from sulphite liquor. /nd. Eng. Chem. 43, Oxpsnus,J.Y. (1966) Fermentation mixing scale-up techniques.
1702-1711. Biotech. Bioeng. 8, 3-24.
Irvinc, G. M. (1968) Construction materials for breweries. Owen, W. L. (1948) Continuous fermentation. Sugar, 43, 36-38.
Chem. Eng. (New York), 75(14), 100, 102-104. Paca, J., Errcer, P. and Grecr, V. (1976) Hydrodynamic
Jackson, T. (1958) Development of aerobic fermentation proces- behaviour and oxygen transfer rate in a pilot plant fer-
ses: penicillin. In Biochemical Engineering, pp. 185-221 (Editor menter.J.Appl. Chem. Biotechnol. 26, 310-317.
Steel, R.) Heywood, London. Pavapino, S. and Ucouin1, F. (1954) A ‘compensated’ stuffing
Jounson, M. J. (1971) Fermentation—yesterday and tomorrow. box and bearing unit for fermenters of semi-industrial and
Chem. Tech. 1, 338-341. industrial capacity. Rend. Inst. Sup. Sanita. Roma (English
Kang,J. H., Fintay, A. C. and Annan, P. F. (1945) Production Edition, 17, 121-144.
of itaconic acid. U.S. Patent 2,385,283. Parker, A. (1950) Aseptic technique in industrial scale fermen-
Karincer, H. W. D. (1977) New fermenter configuration. In tations. In Recent Advances in the Fermentation Industries. Royal
Biotechnology and Fungal Differentiation, FEMS Symp. 4, 137- Institute of Chemistry, London.
155 (Editors Meyrath, J and Bu’Lock, J. D.). Academic PikuLik, A. (1976) Selecting and specifying valves for new plants.
Press, London. Chem. Eng. (New York), 83, 168-190.
Kemptay,J. (1980) Valve Users Manual. Mechanical Engineering Prokop, A., Errkson, L. E., FERNANEZ,J.and Humpnurey, A. E.
Publications, London. (1969) Design and physical characteristics of a multistage
Krrat, A., Tone, H. and Ozaxt, A. (1969) The performance of a continuous tower fermenter. Biotech. Bioeng. 11, 945-966.
perforated plate column as a multistage continuous fer- Prokop, A. and Vorrusa, J. (1976) Bioengineering problems
menter. 1. Washout and growth phase differentiation in the connected with the use of conventional and unconventional
column.J.Ferm. Tech. 47, 333-339. raw materials in fermentation. Folia Microbiol. 21, 58-69.
Ktopprer, W. J., Roperts, R. H., Royston, M. G. and Autt, R. RicHarps,J.W. (1968) Design and operation ofaseptic fermen-
G. (1965) Continuous fermentation in a tower fermenter. ters. In Introduction to Industrial Sterilization, pp. 107-122.
European Brewing Convention Proceedings, 10th Congress (Stock- Academic Press, London.
holm), pp. 238-259. Rivett, R. W., JoHnson, M. J. and Pererson, W. H. (1950)
Kro it, C. L., ForMANEK, S., Covert, A. S., Cutter, L. A., Laboratory fermentor for aerobic fermentations. Ind. Eng.
West,J.M. and Brown, W. E. (1956) Equipment for small Chem. 42, 188-190.
scale fermentations. Ind. Eng. Chem. 48, 2190-2193. Roxgoroucu, J. M., Spencer, J. F. T. and Savians, H. R.
Kuraisu1, M., Terao, I., Onkoucni, H., Marsupa, N. and (1956) Description of5 liter stainless steel fermenters. Can.J.
Naaat, I. (1977) SCP—process development with methanol Technology, 34, 389-396.
as substrate. DECHEMA Monograph (1978), 83 (1704— Royston, M. G. (1966) Tower fermentation of beer. Process
1723), 111-124. Biochem. 1(4), 215-221.
Leg, S. S., Errxson, L. E. and Fan, L. T. (1971) Modelling and ScHOLLER, M. and Seer, M. (1940) Yeast production and
optimization ofa tower type activated sludge system. Biotech. fermentation. U.S. Patent 2,188,192.
Bioeng. Symp. 2, 141-173. ScHreierR, K. (1975) High-efficiency fermenters with deep-jet
Levi, J. D., SHENNAN,J. L. and Epson, G. P. (1979) Biomass aerators. Chemiker Zeitung. 99, 328-331.
from liquid n-alkanes. In Economic Microbiology, Vol. 4, pp. ScHUGERL, K., LuckeE, J. and Oets, U. (1977) Bubble column
362-419 (Editor Rose, A. H.). Academic Press, London. bioreactors. Adv. Biochem. Eng. 7, 1-84.
LuNnDELL, R. and Larno, P. (1976) Engineering of fermentation SHore, D. T., Royston, M. G. and Watson, E. G. (1964)
plants. Part 1. Design aspects. Process Biochem. 11(3), 13-17. Improvements in or relating to the production of beer.
Lyons, J. L. and Askianp, C. L. (1975) Lyons’ Encyclopedia of British Patent 959,049.
Valves. Van Nostrand-Reinhold, New York. Sittic, W. (1982) The present state of fermentation reactors./.
Maxon, W. D. (1959) Aeration—agitation studies on the Chem. Tech. Biotechnol. 32, 47-58.
novobiocin fermentation.J. Biochem. Microbiol. Tech. Engng. Smitu, S. R. L. (1980) Single cell protein. Phil. Trans. Roy. Soc.
(London) B, 290, 341-354.
1, 311-324.
May, O. E., Herricx, H. T., Moyer, A. J. and WELLS, P. A. SNELL, R. L. and Scuweicer, L. B. (1949) Production of citric
acid by fermentation. U.S. Patent 2,492,667.
(1934) Production by submerged mold growths under
increased air pressure. Ind. Eng. Chem. 26, 575-578. Sotomons, G. L. (1969) Materials and Methods in Fermentation.
Mayer, E. (1961) Vinegar by oxidative fermentation of alcohol. Academic Press, London.
USS. Patent 2,997,424. Sotomons, G. L. (1980) Fermenter design and fungal growth. In
Miter, F. D. and Rusuton,J. H. (1944) A mass velocity theory Fungal Biotechnology, pp. 55-80 (Editors Smith,J. E., Berry,
for liquid agitation. Ind. Eng. Chem. 36, 499-503. D. R. and Kristiansen, B.). Academic Press, London.
Mircue tt, C. A. (1926) Vinegar: Its Manufacture and Examination Spivey, M.J. (1978) The acetone-butanol-ethanol fermentation.
(2nd edition). Griffin, London. Proc. Biochem. 13(11), 2-4, 25.
Mutter, R. and Kresricn, K. (1966) Technology of the micro- Sreet, R., Lentz, C. P. and Martin, S. M. (1955) Submerged
biological preparation of organic substances. Angwante Chem. citric acid production of sugar beet molasses: increase in
(International Edition), 5, 653-662. scale. Can.J.Microbiol. 1, 299-311.
Natuan, L. (1930) Improvements in the fermentation and mat- STEEL, R. and Maxon, W. D. (1961) Power requirements of a
uration ofbeers.J.Inst. Brewing, 36, 538-550. typical actinomycete fermentation. Ind. Eng. Chem. 53, 739-
Netson, H. A., Maxon, W. D. and Exrerpink, T. H. (1956) (a2:
Equipment for detailed fermentation studies. Jnd. Eng. Chem. STEEL, R. and Maxon, W. D. (1966a) Dissolved oxygen measure-
48, 2183-2189. ; ments in pilot and production scale novobiocin fermenta-
tions. Biotech. Bioeng. 8, 97-108.
Nos te, T. G., Jackman, M. I. and Bancer, E. M. H. (1964) The
STEEL, R. and Maxon, W. D. (1966b) Studies with a multiple-rod
biological treatment of gas liquor in aeration tanks and
packed towers. J. Inst. Sewage Purification, 5, 440-463. mixing impeller. Biotech. Bioeng. 8, 109-115.
143
Principles of Fermentation Technology
STEEL, R. and Miter, T. L. (1970) Fermenter design. Adv. Appl. Ucenser, G. H., Gerritson, H. and Huisman,J. (1972) Experi-
Microbiol. 12, 153-188. ences with a giant cylindro-conical tank. Master Brew. Assoc.
SrrRaucH and Scumipt (1932) German Patent 552,241. Cited by Amer. Tech. Quart. 9, 117-122.
de Becze and Liebmann (1944). VictToRERO, F. A. (1948) Apparatus for continuous fermentation.
Szucs,J. (1944) Method of producing citric acid by fermentation. U.S. Patent 2,450,218.
US. Patent 2,353,771. Waksma\, S. A. and Karow, E. O. (1946) Citric acid production
Tacucui, H. and Miyamoto, S. (1966) Power requirement in by fermentation. U.S. Patent 2,394,031.
non-newtonian fermentation broth. Biotech. Bioeng. 8, 43-54. Wacker, J. A. H. and Hotpswortn, H. (1958) Equipment
Taytor, I. J. and Senior, P. J. (1978) Single cell proteins: a new design. In Biochemical Engineering, pp. 223-273 (Editor Steel,
source. Endeavour (N.S.), 2, 31-34. R.). Heywood, London.
TERJESEN, S. G. and Cuerry, G. B. (1947) The removal of WATANABE, K. (1976) Production of RNA. In Microbial Production
micro-organisms from air by filtration. Trans. Inst. Chem. of Nucleic Acid related substances, pp. 55-65 (Editors Ogata, K.,
Engrs. 25, 88-96. Kinoshita, S., Tsunoda, T. and Aida, K.). Kodansha, Tokyo
Tuaysen, A. C. (1945) Production of food yeast. Food (May), pp. and Halsted, New York.
116-119. WecricH, R. H. and SHurTER, R. A. (1953) Development ofa
TuretscuH, H. (1967) Manufacture, fabrication and joining of typical aerobic fermentation. Ind. Eng. Chem. 45, 1153-1160.
commercial piping. In Piping Handbook (5th edition), pp.
7.1-7.300 (Editor King, R. C.). McGraw-Hill, New York.
144
CHAPTER 8
145
Principles of Fermentation Technology
Automatic control
Steam
When an automatic control loop is used, certain
Controller ——
| _ 5 Set-point modifications are necessary. The measuring ele-
ment must generate an output signal which can be
| |
| | " monitored by an instrument. In the case of tempera-
Control=—— | ture control, the thermometer is replaced by a
|
wie - Signal to j— Measured value thermocouple, which is connected to a controller
operate valve | which in turn will produce a signal which will
— Thermocouple operate the steam valve (Fig. 8.2).
Water Automatic control systems can be classified into
four main types:
Pipe
ye
1. Two-position controllers (On/Off).
Fic. 8.2. Simple automatic control loop for temperature control. 2. Proportional controllers.
3. Integral controllers.
4. Derivative controllers.
Temperature Positive
deviation {
Maximum
Set-point
|
Potential
Controlled deviation
variable
tecceccsesccccceres eocceee
Minimum Weg eh caer tteete eee
— == |ottset
Set-point
Negative
deviation
Time
Time —
Fic. 8.4, Oscillatory pattern of the temperature of a domestic
water tank (no water being drawn off) using On/Off control of the
heating element. Output without control
eeeeeee Proportional action
PFT-K
Principles of Fermentation Technology
a
(Change in input)
+20
(Change in output)
Temperature
controller
Thermocouple—
Then Al = K,A0.
K, may contain conversion units if there is an electri-
cal input and a pressure output or vice versa.
If the input to the controller is 1 unit of change
then:
(a) with a controller gain of 1, the output will be Pressure
1 unit, regulated
valve
(b) with a controller gain of 2, the output will be
2 units, etc. Fic. 8.6. A fermenter with a temperature-controlled heating
jacket.
On many controllers K, is graduated in terms of
proportional band instead of controller gain.
which will cause the control valve to go from fully
open to fully closed (28.75° to 31.25°). In this case
Now K,& i,
PB (proportional band)
the proportional band will be
Tabet 8.2, The effect ofcontroller gain (K,) on band width ofaproportional temperature controller
i
e ae y
a
Measured Pressure output (psig) at Pressure output (psig) at Pressure output (psig) at
temperature K,= | psig/1° K,= 2 psig/1° K,= 4psig/1°
BS Ys, 15 15
34° 14 15 15
367 13 15 15
$2:5% 125 15 15
Dae 12 14 15
31,20" 11.25 Nas 15
31° 1] 12 14
Set 30° 10 Prop. band 100% 10 ¢Prop, band 50% 10 ¢Prop. band 25%
point 29° 9 8 6
28.75° 8.75 To 5
28° 8 6 5
QisDe: 7:9 J 4,
Ai 7 5 5
26° 6 %) 5
25° 5 5 4)
for control of the chosen parameter, and the system on its own. The response curve has therefore been
takes longer to settle down. There is, however, no deliberately omitted from Fig. 8.5.
offset which is advantageous in many control pro- Figure 8.7 demonstrates the response ofderivative
cesses. control to sinusoidal error inputs. The output is
always in a direction to oppose changes in error,
DERIVATIVE CONTROL both away from and towards the set point as shown
When derivative control is applied, the controller in Fig. 8.7, which in this example results in a 90°
senses the rate of change of the error signal and phase shift. This opposition to a change has a fast
contributes a component of the output signal that is damping effect and this property is very useful to
proportional to a derivative of the error signal. Thus: combine with other modes of control which will be
discussed.
M=M,+ T, <
149
Principles of Fermentation Technology
Although thermocouples have many potential main disadvantage is the marked non-linear tem-
applications, they have not been used much for perature versus resistance curve.
temperature measurement in fermenters because
they are normally operated at ambient temperatures TEMPERATURE CONTROL
and unfortunately tend to be susceptible to cold- The use of water jackets or pipe coils within a
junction problems within 50° of this range (Solo- fermenter as a means of temperature control has
mons, 1969). MacLennan (1970) has described a been described in Chapter 7. In many small systems
temperature control for 3- and 12-dm°* fermenters there is a heating element, 300 to 400 W capacity is
which utilized a thermocouple in the fermenter and adequate for a 10-dm* fermenter, and a cooling
a thermocouple cold-point reference junction. water supply which are on or off depending on the
need for heating or cooling. The heating element
ELECTRICAL RESISTANCE THERMOMETERS should be as small as possible to reduce the size of
It is well known that the electrical resistance of the ‘heat sink’ and resulting overshoot when heating
metals changes with temperature variation. This is no longer required. In some cases it may be better
property has been utilized in the design of resistance to run the cooling water continuously at a steady
thermometers. The bulb of the instrument contains rate and only have the heating element connected to
the resistance element, a mica framework (for very the control unit. This can be an expensive mode of
accurate measurement) or a ceramic framework operation if the water flows directly to waste. For
(robust but for less accurate measurement) around small-scale use, Churchill Instruments Co. Ltd.
which the sensing element is wound. A platinum (Walingate Road, Perivale, Greenford, Middlesex,
wire of 100 Q resistance is normally used. Leads England) make a unit which will pump recirculating
emerging from the bulb are connected to the thermostatically heated water through fermenters
“measuring element. The reading is normally of up to 10 dm® capacity and give temperature
obtained by the use of aWheatstone bridge circuit control of £0.1°.
and is a measure of the average temperature of the In large fermenters, where the need for heating
sensing element. This type of thermometer does during the fermentation is not normally required, a
have a greater accuracy (+0.25%) than some ofthe regulatory valve at the cooling-water inlet may be
other measuring devices and is more sensitive to sufficient to control the temperature. There may be
small temperature changes. There is a fast response provision for circulation of refrigerated brine if
to detectable changes, and there is no restriction on excessive cooling is required. Steam inlets to the coil
distance between the very compact sensing point and jacket must be present if a fermenter is being
(30 X 5 mm) and the display point of reproducible used for batch sterilization of media.
readings. These thermometers are normally
enclosed in stainless-steel sheaths if they are to be
used in large vessels and ancillary equipment. Flow measurement and control
Serre
= AH =,
Thermister
ne
Thermister
A voltage signal can be obtained by this method of
measurement which can be utilized in data logging.
Meter to measure Control of gas flow is usually by needle valves.
power Often this method of control is not sufficient, and it
Fic. 8.8. Thermal mass flowmeter. is necessary to incorporate a self-acting flow-control
valve. At a small scale, such valves as the ‘flowstat’
can cater for a wide range offlow rates. The accuracy are available (G. A. Platon, Ltd., Wella Road,
depends on having the gas at a constant pressure, Basingstoke, Hampshire, England). Fluctuations in
but errors of up to +10% offull-scale deflection are pressure in a flow-measuring oriface cause a valve or
quoted (Howe et al., 1969). The errors are greatest at piston pressing against a spring to gradually open or
low flow rates. Ideally, rotameters should not be close so that the originally preselected flow rate is
sterilized and are therefore normally placed between restored. In a gas ‘flowstat’, the oriface should be
a gas inlet and a sterile filter. There is no provision _upstream when the gas supply is at a regulated
for on-line data logging with the simple rotameters. pressure and downstream when the supply pressure
Metal tubes can be used in situations where glass is fluctuates and the back pressure is constant. Valves
not satisfactory. In these cases the float position is operating by a similar mechanism are available for
determined by magnetic or electrical techniques, larger scale applications.
but this provision has not been normally utilized for
fermentation work. Rotameters can also be used to LIQUIDS
measure liquid flow rates, provided abrasive parti- The flow of non-sterile liquids can be monitored
cles or fibrous matter is not present. by a number of techniques (Howe et al., 1969) but
The use of oxygen and carbon dioxide gas measurement offlow rates of sterile liquids presents
analysers for effluent gas analysis requires the pro- a number of problems which have to be overcome.
vision of very accurate gas-flow measurement if the On a laboratory scale flow rates may be measured
analysers are to be used effectively. For this reason manually using a sterile burette connected to the
thermal mass flowmeters have been utilized for the feed pipe and timing the exit of ameasured volume.
range 0 to 500 dm? min“'.These instruments have a The possible use of rotameters has already been
+1% full-scale accuracy and work on the principle mentioned in the previous section. A more expensive
of measuring a temperature difference across a heat- method is to use an electrical flow transducer (Howe
ing device placed in the path of the gas flow (Fig. et al., 1969) which can cope with particulate matter
8.8). Temperature probes such as thermistors are in suspension and measure a range offlow rates from
placed upstream and downstream ofthe heat source, very low to high (50 cm? min™! to 500,000 dm?
which may be inside or outside the piping. min') with an accurate of +1%. In this flowmeter
The mass flowrate ofthe gas, Q, can be calculated (Fig. 8.9) there are two windings outside the tube,
from the specific heat equation: supplied with an alternating current to create a
magnetic field. The voltage induced in the field is
H = QC,(T) — T;) proportional to the relative velocity of the fluid and
where H = heat transferred, the magnetic field. The potential difference in the
Q = mass flow rate of the gas, fluid can be measured by a pair of electrodes, and is
C,, = specific heat of the gas, directly proportional to the velocity of the fluid.
152
Instrumentation and Control
BSSYepMCR
<iRe
Insulating
liner Piston
Electrode
assembly
In batch and fed-batch culture fermenters, a range by changing the stroke rate, the length of the
cheaper alternative is to measure flow rates in- piston stroke and by using a different piston size.
directly by load cells (see Weight section). The Sizes are available from cm* h~' to thousands of
fermenter and all ancillary reservoirs are attached dm’ h7! and all can be operated at relatively high
- to load cells which monitor the increases and working pressures. Unfortunately, they cannot be
decreases in weight of the various vessels at regular used to pump fibrous or particulate suspensions.
time intervals. Provided the specific gravities of the Piston pumps are more expensive than comparable
liquids are known it is possible to estimate flow rates sized peristaltic pumps but do not suffer from tube
fairly accurately in different feed pipes. This is failure.
another technique which may be used with particu- Leakage can occur via the shaft housing of a
late suspensions. piston pump. The problem can be prevented by the
Another indirect method of measuring flow rates use of adiaphragm pump. This pump uses a flexible
aseptically is to use a metering pump which pumps diaphragm to pump fluid through a housing (Fig.
liquid continuously at a predetermined and accurate 8.10) with ball valves to control the direction offlow.
rate. A variety of metering pumps are commercially The diaphragm may be teflon, neoprene, stainless
available including motorized syringes, peristaltic steel, etc., and is actuated by a piston. A range of
pumps, piston pumps and diaphragm pumps. sizes of pumps is available for flow rates up to
Motorized syringes are only used when very small thousands of dm*h!'.
quantities of liquid have to be added slowly to a
vessel. In a peristaltic pump, liquid is moved for-
wards gradually by squeezing a tubing held in a
Pressure measurement
semicircular housing. A variety of sizes of tubes can
be used in different pumps to produce different
known flow rates over a very wide range. Suspen- Pressure is one of the crucial measurements that
sions can be handled since the liquid has no direct must be made when operating many processes.
contact with moving parts. Pressure measurements may be needed for several
A piston pump contains an accurately machined reasons, the most important of which is safety.
ceramic or stainless-steel piston moving in a cylinder Industrial and laboratory equipment is designed to
normally fitted with double ball inlet and outlet withstand a specified working pressure plus a factor
valves. The piston is driven by a constant-speed of safety. It is therefore important to fit the equip-
motor. Flow rates can be varied within a defined ment with devices that will sense, indicate, record
153
Principles of Fermentation Technology
Process
connection
Pinion
—Connecting
\ link Fic, 8.12. Nested diaphragm-type pressure sensor (Liptak,
Movement /
sector halTravelling 1969).
angle
Safety valves
+
|
I
!
Timer Detector
oats |
|
|
I
Safety valves (Chapter 7) should be incorporated Probe ==
I
at various suitable places in all vessels and pipe I
Antifoam
layouts which are likely to be operated under pres- reservoir
sure. The valve should be set to release the pressure Antifoam
as soon as it increases markedly above a specified inlet
working pressure.
A variety of sensors can be used to measure the Fic. 8.13. Foam sensing and control unit.
power consumption ofa fermenter. On a large scale,
a watt meter attached to the agitator motor will give
a fairly good indication of power uptake. This tachometer will be determined by the type of signal
measuring technique becomes less accurate as there which is required for recording and/or process con-
is a decrease in scale to pilot scale and finally to trol for regulating the motor speed and other ancil-
laboratory fermenters, the main contributing factor lary equipment. Provision is often made on small
being friction in the stuffing box (Chapter 7). Tor- laboratory fermenters to vary the rate of stirring. In
~ sion dynamometers can be used in small-scale appli- most cases it is now standard practice to use an a.c.
cations. Since the dynamometer has to be placed on slip motor that has an acceptable torque curve that
the shaft outside the fermenter the measurement is coupled to a thyristor control. At pilot or full scale,
will once again include the friction in the bearings in the need to change rates of stirring is normally
the stuffing box. For this reason strain gauges reduced. When necessary it can be done using gear
mounted on the shaft within the fermenter are the boxes, modifying the sizes of wheels and drive belts,
most accurate method of measurement and over- or by changing the drive motor, the most expensive
come frictional problems (Aiba, Okamoto and alternative.
Satoh, 1965; Brodgesell, 1969). Aiba et al. (1965)
mounted four identical strain gauges at 45° to the
axis in a hollow shaft. Lead wires from the gauges Foam sensing and control
passed out ofthe shaft via an axial hole and electrical
signals were then picked up by an electrical slip-ring
The formation of foam is a difficulty in many
arrangement. Theoretical treatment of the strain
types of microbial fermentation which can create
gauge measurements has been covered by Aiba et al.
serious problems if not controlled. It is common
(1973). practice to add an antifoam to a fermenter when the
culture starts foaming above a certain predeter-
mined level. The methods used for foam sensing and
Rate of stirring
antifoam additions will depend on process and
economic considerations. The properties of anti-
In all fermenters it is important to monitor the foams have been discussed elsewhere (Chapter 4),
rate of rotation (rpm) of the stirrer shaft. The as has their influence on dissolved oxygen concen-
tachometer used for this purpose may employ elec- trations (Chapter 9).
tromagnetic induction, voltage generation, light A foam sensing and control unit is shown in Fig.
sensing or magnetic force as detection mechanisms 8.13. A probe is inserted through the top plate ofthe
(Brodgesell, 1969). Obviously the final choice of fermenter. Normally the probe is a stainless-steel
155
Principles of Fermentation Technology
rod, which is insulated except at the tip, and set ata solid or tubular steel cylinder, the compressive
defined level above the broth surface. When the strain of which under axial load may be measured.
foam rises and touches the probe tip, a current is The cell is calibrated by measuring compressive
passed through the circuit of the probe, with the strain over the appropriate range of loading. A
foam acting as an electrolyte and the vessel acting as convenient method of measuring the strain is by
an earth. The current actuates a pump or valve and means of electrical resistance strain gauges that are
antifoam is released into the fermenter for a few cemented to the surface of the cylinder. The changes
seconds. Process timers are routinely included in the of resistance with strain are proportional to load and
circuit to ensure that the antifoam has time to mix are determined by appropriate electrical apparatus.
into the medium and break down the foam before Most commonly an electrical resistance strain
the probe is programmed after a preset time interval gauge is a grid of metal foil, generally constantin
to sense the foam level again and possibly actuate alloy, deposited on a thin sheet ofa plastic material
the pump or valve. Alternatively antifoam may be. as in Fig. 8.14a. Gauge lengths of some 5 to 25 mm
added slowly at a predetermined rate by a small are generally used with resistances in the order of 50
pump so that foaming never occurs and there is to 500 ohms. The gauge is stuck by means of epoxy
therefore no need for a sensing system. or other cement to the surface ofthe steel, connecting
A number of mechanical antifoam devices have wires are soldered to the tags and the whole unit is
been described including discs, propellers, brushes covered with a waterproofing compound.
or hollow cones attached to the agitator shaft above The gauge measures strain, whether this be
the surface of the broth. The foam is broken down caused by loading or by temperature changes. In a
when it is thrown against the walls of the fermenter. typical cell, changes of temperature are compen-
Other devices which have been manufactured sated and other advantages obtained by arranging
include horizontal rotating shafts, centrifugal half the gauges axially on the cylinder and the other
separators and jets spraying on to deflector plates half circumferentially as in Figs. 8.14 b, cand d. The
(Hall et al., 1973; Viesturs, Kristapsons and Levi- gauges are connected to form a four arm bridge as in
tans, 1982). Unfortunately most of these devices Fig. 8.14e. This arrangement gives temperature
have to be used in conjunction with an antifoam. compensation, and by distributing the gauges
around the cylinder the effect of any eccentricity of
loading is eliminated. There is the further advantage
Weight that under compressive load while the axial gauges
are reduced in resistance, the circumferential gauges
A load cell offers a convenient method of finding are stretched and increased in resistance. Tempera-
the weight of an object. A cell designed to take ture compensation could be obtained by placing
tensile forces may be placed on a crane hook to gauges 2, 4, 6 and 8 (the ‘compensating gauges’; Fig.
measure hanging loads, while the weight of a 8.14d) on a piece of unstressed steel maintained at
stationary object can be found by means of compres- the cell temperature. Gauges 1, 3, 5 and 7 would
sion load cells placed in or at the foot of the vessel then be the ‘working gauges’. By placing the ‘com-
supports. It is the latter type ofload cell that will be pensating’ gauges on the cylinder in a position
described. where they receive a strain of the opposite sign to the
When designing the support system for a fer- ‘working’ gauges they contribute to the disturbance
menter or other vessel, the weight of which is to be of the electrical balance of the four arm bridge. The
measured by load cells, the principle of the three- leads B and D (Fig. 8.14e) are connected to an
legged stool should be remembered. Three feet will electrical power source and A and C to a galvano-
always rest in stable equilibrium even though the meter (or ammeter). Generally the resistances are
supporting surface is uneven. If more feet are pro- adjusted so that the meter reads zero when the cell is
vided, the additional feet must each be fitted with under no load. Alternatively the meter may be set at
means of adjustment or precision packing to ensure zero for the tare load and read a load increment.
bearing on all the feet. Commercially available meters are calibrated to
A load cell is essentially an elastic body, usually a read directly in kg or lb.
156
Instrumentation and Control
(SSSSSSSSSSSSSSSSS
2
SSS
Tag
Gauge length |
| 5 to 25 mm
F A2 Bo C, Dz
Fic. 8.14b. Position of gauges on cylinder. AyeB jC} Dy A3B3C3D3
PWdQO
Fic. 8.14c. Plan showing gauge location around cylinder. Fic. 8.14f. Gang switch for three load cells.
157
Principles of Fermentation Technology
< oO = oe = a ”
The external fittings to the load cell are designed f ——-——
Gl ass SH
HSSkos Aeon
node S — Electrolyte
St;
to suit the conditions of use. The steel cylinder is tubing Extras Anode (Ag-AgCl) =
usually capped with plattens, machined to give ERReH| (Pb Helix) =
ERS SRE =
parallel bearing surfaces and provided with bolt- ili
Silicone ERR
Eeerece —Electrol
ectrolyte =
2
tubing | RH =
i BX =
holes or a socket for attaching to the apparatus to be Bees
: mal Glass wool ‘St
Rat i
weighed. The upper surface may be domed to pro- <B 7 Coeds aves :NRLs
RG O-ring
-—ri
vide a ball seating or arranges with a seat for a steel Membrane (Ag spiral) t : Membrane
ball, this being to ensure axial load. The cylinder is
fitted with a guard cover and the wires A, B, C and (a) (b)
D brought to a connector on the cover (Fig. 8.14 g Fic. 8.15. Construction of dissolved-oxygen electrodes: (a) gal-
and h). vanic, (b) polarographic (Lee and Tsao, 1979).
158
Instrumentation and Control
employ potassium hydroxide, chloride, bicarbonate If it is found necessary to increase the dissolved
or acetate as an electrolyte. The sensing tip of the oxygen concentration in a medium this may be
electrode is a teflon, polyethylene or polystyrene achieved by increasing the air flow rate or the rpm of
membrane which allows passage of the gas phase so the impeller or a combination of both processes.
that an equilibrium is established between the gas Another way is to increase the rate of oxygen to
phases inside and outside the electrode. Because of nitrogen in the input gas using a variable propor-
the relatively slow movement of oxygen across the tionating valve while maintaining a constant gas
membrane, this type ofelectrode has a slow response flow rate (Siegall and Gaden, 1962). The cost ofthis
of the order of 60 seconds to achieve a 90% reading technique would normally restrict its use to
of true value (Johnson, Borkowski and Engblom, laboratories and pilot plants.
1964). Buhler and Ingold (1976) quote 50 seconds
for 98% response for a later version. These elec- Exit-gas analysis
trodes are therefore suitable for monitoring very
slow changes in oxygen concentration and are nor- The measurement and recording of the effluent
mally chosen because of the compact size and rela- gas composition is important in many fermentation
tively low cost. Unfortunately this type of electrode studies. By observing the concentrations of carbon
is very sensitive to temperature fluctuations, and dioxide and oxygen in the entry and exit gases in the
should be compensated for temperature using a fermenter and knowing the gas flow rate it is possible
thermistor circuit. The electrodes also have a limited to determine the oxygen uptake of the system, the
life because of corrosion of the anode. carbon dioxide evolution rate and the respiration
Polarographic electrodes, which are bulkier than rate of the microbial culture.
galvanic electrodes, are more commonly used in The oxygen concentration can be determined by
‘pilot and production fermenters, needing instru- a paramagnetic gas analyser. Oxygen has a strong
ment ports of 19 or 25 mm diameter. They have affinity for a magnetic field, a property which is
silver anodes which are negatively polarized with shared with only nitrous and nitric oxide. The
respect to reference cathodes of platinum or gold, analysers may be of a deflection or thermal type
using aqueous potassium chloride as the electrolyte. (Brown et al., 1969).
Response times of 0.05 to 15 seconds to achieve a In the deflection analyser, the magnetic force acts
95% reading have been reported (Lee and Tsao, on a dumb-bell test body that is free to rotate about
1979). The electrodes which can be very precise may an axis (Fig. 8.16). The magnetic force which is
be both pressure and temperature compensated.
Side view
Although a polarographic electrode may initially
cost 600% more than a galvanic equivalent, the
maintenance costs are considerably lower as only Quartz
suspension
the membrane should need replacing. Magnet
Dumbbell
Dissolved oxygen concentrations may also be test body
determined by a tubing method which has been Magnetic
described by Phillips and Johnson (1961) and field
Roberts and Shepherd (1968). The probe consists of
a coil of permeable teflon or propylene tubing within
the fermenter through which is passed a stream of
Light
dividing To
helium or nitrogen. The oxygen which diffuses from mirror Restabilizing
receiver
159
Principles of Fermentation Technology
Detector Amplifier
Chopper
Read out
160
Instrumentation and Control
161
Principles of Fermentation Technology
value in any system depending on the equilibrium and product formation ought to be continuously
of: monitored. This ideal situation has not yet been
achieved but a number of techniques are currently
Reduced form = Oxidized form + electron(s)
being developed.
(negative value) (positive value)
The measuring electrode consists of gold,
platinum or iridium which is welded to a copper Ion-specific sensors
lead. The interpretation of results presents difficul-
ties since the micro-organisms can be at a different
Ion-specific sensors have been developed by
redox potential from the broth. It is the ability to
Orion Research and Radiometer to measure NH3,
detect low concentrations of oxygen in media
Ca?*, Kt, Mg’*, PO}, SO7, etc. (Orion Research
(1 ppm) where redox electrodes may have a good
Inc, 11 Blackstone Street, Cambridge, Mass., USA;
application for determining oxygen availability in
Radiometer—Aagard Nielsen and _ Schroder,
anaerobic or micro-aerophilic processes operated
Emdrupvej 72, Copenhagen-NV, Denmark). How-
ona small scale (Kjaergaard and Joergensen, 1979).
ever, none of these probes are steam sterilizable.
The redox potential may be controlled by sparging
with oxygen or nitrogen or by adding cysteine,
ascorbic acid, sodium thioglycollate (Daniels eé¢ al.,
Enzyme electrodes
1965) or glucose (Kjaergaard and Joergensen,
1979):
Eventually enzyme electrodes might also be used
in some analyses (Updike and Hicks, 1967; Enfors
Carbon dioxide electrodes ~ and Molin, 1978). A suitable enzyme which pro-
duces a change in pH or forms oxygen in the enzyme
reaction is chosen and immobilized on a membrane
The measurement of dissolved carbon dioxide is
held in close contact to a pH or oxygen electrode.
possible with an electrode, since a pH or voltage
Unfortunately, the oxygen demand of an enzyme
change can be detected as the gas goes into solution.
may restrict the substrate concentration which
The first available electrode consisted of acombined
might be detected in a medium. Enfors (1981) has
pH electrode with a bicarbonate buffer (pH 5.0)
reported that it has been possible to overcome this
surrounding the bulb and ceramic plug, with the
problem for glucose determinations by co-
solution being retained by a PTFE membrane held
immobilizing glucose oxidase and catalase. Ideally,
by an O-ring. Unfortunately this electrode is not
steam sterilizable. This basic design has been a sterilized electrode that can be inserted into a
modified by immobilizing a thin layer of aqueous fermenter is required. Hewetson, Jong and Gray
bicarbonate over the glass sensing membrane of the (1979) prepared sterilized penicillinase electrodes
pH electrode, which in turn is covered by a steel- by assembling sterile components or by standing
mesh reinforced silicone membrane and an O-ring. assembled components in chloroform.
It is possible to calibrate the electrode in situ using
gas mixtures of known CO, partial pressure and
check it with buffer solutions. This new version can Microbial electrodes
be steam sterilized (Puhar et al., 1980).
163
PFT-L
Principles of Fermentation Technology
Mainframe
computer
Printout
Analogue
(hard copy)
to digital
convertor Visual
display unit
Interface Dedicated
Tele-type
mini
computer
Data store
Digital
to analogue
cl)
AE
convertor Graphic unit
Clock
Addition
reservoir
Alarms
Sensor S
Fic. 8.20. Simplified layout of computer-controlled fermenter with only one control loop shown.
It is possible to analyse data, compare it with Fig. 8.20. A sensor S in the fermenter produces a
model systems in a data store, and use control signal which may need to be amplified and con-
programs which will lead to process optimization. ditioned in the correct analogue form. At this stage
However, process optimization by this method is it is necessary to convert the signal to a digital form
not a widely used procedure in the fermentation which can be subsequently transmitted to the com-
industries at present. It is important to be aware of puter. An interface is placed in the circuit at this
these different applications, since this will influence point. This interface serves as the junction point for
the size and type of computer system which will be the inputs from the fermenter sensors to the com-
appropriate for the precise role which it is intended puter and the output signals from the computer to
to perform, whether in a laboratory, a pilot plant, or the fermenter controls such as a pump T. There is a
manufacturing plant, or a combination of these need for a digital to analogue conversion between
three. the interface and the pump T. The micro-computer
itself is dedicated solely to one or more fermenters.
This computer is coupled to a real-time clock, which
Components of a computer-linked system determines how frequently readings from the sen-
sor(s) should be read and possibly recorded. The
A simple outline of the main components of a other ancillary equipment linked directly to the
computer-linked fermentation system is given in computer might include a visual display unit, a data
164
Instrumentation and Control
store, a teletype, a graphic display unit, a print out TABLE 8.3. Gateway sensors (Aiba et al., 1973)
and alarms.
The micro-computer is often connected to a large Information that may be
determined from the sensor
main frame computer for random access, not on a Sensor signal
real-time scale, but for long-term data storage and
retrieval and for complex data analysis which will pH Acid product formation
Dissolved oxygen Oxygen-transfer rate
not be subsequently utilized in real-time control. Oxygen in exit gas
Oxygen-uptake rate
It is also possible to develop programs so that Gas-flow rate
Carbon dioxide in exit gas
on-line instruments can be regularly checked and Carbon dioxide evolution rate
Gas-flow rate
recalibrated when necessary. Swartz and Cooney Oxygen-uptake rate
Respiratory quotient
(1979) were able to routinely. recalibrate a Carbon dioxide evolution rate
paramagnetic oxygen analyser and an infrared car- Sugar-level and feed rate
Yield and cell density
Carbon dioxide evolution rate
bon dioxide analyser every 12 hours utilizing a
program which connected a gas of known composi-
tion to the analysers and subsequently monitored
the analyser outputs.
Data analysis
Data logging
Because a computer can undertake so many cal-
culations very rapidly, it is possible to design pro-
The simplest task for a computer is data logging. grams to analyse fermentation data in a number of
Parameters such as those listed in Table 8.1 can be ways. A linked main frame computer may be used
.measured by sensors which produce a signal which for part of this analysis as well as the dedicated
is compatible with the computer system. mini-computer.
Programs have been developed so that by refer- A number ofthe monitoring systems were described
ence to the real-time clock, the signals from the as ‘Gateway Sensors’ by Aiba et al. (1973) and are
appropriate sensors will be scanned sequentially in given in Table 8.3. Gateway sensors are so called
a predetermined pattern and logged in a data store. because the information they yield can be processed
Typically, this may be 2- to 60-second intervals, and to give further information about the fermentation.
the data is printed out on a visual display unit. In The respiratory quotient of a culture may be
preliminary scanning cycles the values are com- calculated from the metered gas-flow rates and
pared with predefined limit values, and deviations analyses for oxygen and carbon dioxide leaving a
from these values result in an error print out, or if known volume ofculture in the fermenter. This pro-
more extreme then an alarm may be activated. In cedure was used to monitor growth of Candida utilis
the final cycle of a sequence, say every 5 to 60 in a 250-dm® fermenter, to follow or forecast events
minutes, the program instructs that the sensor read- during operation (Nyiri, Toth and Charles, 1975).
ings are permanently recorded on a print out or ina If one defines the fraction of substrate which is
data store. converted to product then it is possible to write mass
At the same time as on-line data is being recorded balances for C, H, O and N with the measurement of
from sensors, analytical data for microbial growth, only a few quantities (Oy, GO,, NHs, etc.). All the
substrate utilization and product formation, which other quantities can be calculated, including
have to be determined separately may be logged into biomass and yield if the biomass elemental com-
the data store for specific known times. Thus, it is position is known (Chapter 2). This procedure was
now possible to continuously record data for a range used for the analysis of a bakers’ yeast fermentation
of parameters from a number of fermenters simul- (Cooney, Wang and Wang, 1977). Biomass produc-
taneously using minimal manpower provided the tion can be regarded as a stoichiometric relationship
capital outlay is made for fermenters with suitable in which substrate is converted, in the presence of
instrumentation coupled with adequate computer oxygen and ammonia to biomass, carbon dioxide
facilities. and water:
165
Principles of Fermentation Technology
Carbon source-energy + oxygen + ammonium puter failure, control can be returned to each control
— cells + water + carbon dioxide loop by manual override. If Direct Digital Control
(DDC) is to be used, control units will no longer be
Thus, the equation can be rewritten in the form:
required as the sensors will be interfaced directly
ae O, +bO, aig cNH, with the computer. The computer (DDC) again
> dC, HgO,Nz AP eH,O + fCO, determines the set-point values, but at the same
time gives better control since the control algorithms
where a, 6, c, d, e and
f are moles of the respective
are mathematically stored functions rather than
reactants and products. C,H,O, is the molecular
electrical functions. This procedure allows for
formula of the substrate where x, y and z are the
greater flexibility and more precise representation
specific carbon, hydrogen and oxygen atom num-
but computer failure will present problems unless
bers. Biomass is represented by C H O N where a,
there is some manual back-up facility. Armiger and
B, y and € are the corresponding atomic numbers of
Moran (1979) have discussed the relative merits of
each element in the cell. This technique was
DDC and DSC systems. Although DDC was
developed by Cooney, Wang and Wang (1977) to
cheaper in large-scale operations because of equip-
use with bakers’ yeast fermentations (Wang,
ment cost savings, DSC was preferred for pilot
Cooney and Wang, 1977).
plants.
When DDC or DSC are being used, values for a
Process control
control profile are introduced into the computer
through a teletype by the operator prior to starting
Arminger and Moran (1979) have recognized the fermentation. The operator may alter the control
three levels of process control that might be incor- values during the fermentation. Should the process
porated into a system. Each higher level involves - readings not fall within the prescribed limits a com-
more complex programs and needs a greater overall mand program may actuate a pump or valve via the
understanding of the process. The first level of interface and a digital to analogue converter. Should
control, which is already routinely used in the chem- there be evidence of excessive use of acid, alkali or
ical industries, involves sequencing operations, such antifoam, then an alarm program may be actuated.
as manipulating valves or starting or stopping Up to the present time computers have been most
pumps, instrument recalibration, on-line mainten- widely used with highly instrumented pilot-plant
ance and fail-safe shut-down procedures. In most of fermenters. Evidence for their use in large-scale
these operations the time base is at least in the order control applications is restricted, particularly for
of minutes, so that high-speed manipulations are established plants.
not vital. Two applications are sterilization cycles The most advanced level of control is concerned
and medium batching. with increased productivity and process optimiza-
The next level of computer control involves con- tion. Unfortunately this type of control is still in its
necting into the system individual control loops of infancy. At the present time the number of on-line
the type which already have been considered earlier sensors which are available and suitable for long-
in this chapter, e.g. temperature, pH, foam control, term use are limited. It is difficult or impossible to
etc. These control loops make it possible to maintain measure biomass directly, and very few products
environmental parameters at defined values (set- can be measured on-line by the electrodes or other
points). One approach is to utilize Digital Set-point sensors which are currently available. These
Control (DSC). When monitoring control loops, the parameters may be measured separately and the
computer is programmed to scan the outputs and readings added to the appropriate computer data
compare them with set-point values previously file.
defined in a program. At specified times in the Advances have been made in mathematical
program the set-points on the individual controllers modelling and potential applications using both
may be changed, but the modes of control are analogue and digital computers (Nyiri, 1972a;
limited to those on an analogue device, i.e. propor- Blanch and Dunn, 1974; Hampel, 1979; Kossen,
tional, integral and derivative. In the event of com- 1979). Equations have been derived to describe the
166
Instrumentation and Control
behaviour ofafermentation which makes it possible tions. Second European Congress of Biotechnology, Eastbourne,
for the computer system to identify the behaviour p. 141.Society for Chemical Industry, London.
Enrors, S. O. and Moun, N. (1978) Enzyme electrodes for
pattern and control a fermentation for optimum
fermentation control. Proc. Biochem. 13(2), 9-11, 24.
yield or some other parameter. Computer control at Fiynn, D.S. (1974) Computer control of fermentation processes.
pilot scale has been used to optimize bakers’ yeast Biotech. Bioeng. Symp. 4, 597-605.
Grayson, P. (1969) Computer control of batch fermentations.
production (Ramirez et al., 1981), to stop a fermen-
Proc. Biochem. 4(3), 43-44.
tation when enzyme production was uneconomical Hatt, M. J., Dickinson, S. D., PrircHarp, R. and Evans,J. I.
(Lundell e¢ a/., 1981) and for regulating the tempera- (1973) Foams and foam control in fermentation processes.
Prog. Ind. Microbiol. 12, 169-234.
ture in a continuous-culture yeast fermentation to
Hamre, W. (1979) Applications of microcomputers in the study
the highest value at which a steady state could be of microbial processes. Adv. Biochem. Eng. 13, 1-33.
sustained (Harrison e¢ al., 1981). Harrison, D. E. F., Asusy, R. E., DiamMonp, K. R., Hanna, F.
K., Hinton, O., Leunc, C. and Wyner, P. E. O (1981) A
microprocessor controlled fermentation system used for
REFERENCES automatic optimisation and strain selection. Abstracts of
Communications. Second European Congress of Biotechnology,
p. 144. Society of Chemical Industry, London.
Ara, S., Humpureey, A. E. and Mixus, N. F. (1973) Biochemical Harrison, D. E. F. and Cuance, B. (1970) Fluorimetric
Engineering (2nd edition), Chapter 12. Academic Press, New technique for monitoring changes in the level of reduced
York. nicotinamide nucleotides in continuous cultures of micro-
Ara, S., Oxamoro, R. and Sarton, K. (1965). Two sorts of organisms. Appl. Microbiol. 19, 446-450.
measurements with ajar type of fermenter—power require- Heinzie, E., Dunn, I. J and Bourne,J. R. (1981) Continuous
ments of agitations and capacity coefficient of mass transfer measurement of gases and volatiles during fermentations
in bubble aeration.J.Ferm. Tech. 43, 137-145. using mass spectrometry. In Abstracts of Communications.
ARMIGER, W. B. (1979) Computer Applications in Fermentation Second European Congress of Biotechnology, Eastbourne, p. 24.
Technology. Biotech. Bioeng. Symp. 9. Society of Chemical Industry, London.
ArRMIGER, W. B. and Humpnrey, A. E. (1979) Computer applica- Hewetson,J. W., Jonc, T. H. and Gray, P. P. (1979) Use of an
; tions in fermentation technology. In Microbial Technology immobilized penicillinase electrode in the monitoring of the
(2nd edition), Vol. 2, pp. 375-401 (Editors Peppler, H. J. penicillin fermentation. Biotech. Bioeng. Symp. 9, 125-135.
and Perlman, D.). Academic Press, New York. Howe, W.H., Kop,J.G., Sev, R. and Lirrak, B. G. (1969) Flow
ArmicEr, W. B. and Moran, D. M. (1979) Review ofalternatives measurement. In Instrument Engineer’s Handbook, Vol. 1,
and rationale for computer interfacing and system configu- pp. 411-567 (Editor Liptak, B. G.). Chilton, Philadelphia.
ration. Biotech. Bioeng. Symp. 9, 215-225. Jacos, H. E. (1970) Redox potential. In Methods in Microbiology,
BreyYe er, W., Ernsece, A. and Frecuter, A. (1981) Fluorometric Vol. 2, pp. 91-123 (Editors Norris, J. R. and Ribbons, D.
studies in bioreactors: method and applications. Abstracts of W.). Academic Press, London.
communications. Second European Congress of Biotechnology, Jounson, M. J., Borkowski,J.and ENcBiom, C. (1964) Steam
Eastbourne, p. 142. Society of Chemical Industry, London. sterilisable probes for dissolved oxygen measurement.
Briancu, W. H. and Dunn, I. J. (1974) Modelling and simulation Biotech. Bioeng. 6, 457-468.
in biochemical engineering. Adv. Biochem. Eng. 3, 127-165. KJAERGAARD, L. (1977) The redox potential: its use and control
Bravarb, J. P., CGorponnieR, M., KERNEVEZ, J. P. and in biotechnology. Adv. Biochem. Eng. 7, 131-150.
LeBeau tt, J.M. (1979) On-line identification of parameters KJAERGAARD, L. and JOERGENSEN, B. B. (1979) Redox potential
in a fermentation process. Biotech. Bioeng. 21, 1239-1249. as a state variable in fermentation systems. Biotech. Bioeng.
BropcEsELL, A. (1969) Miscellaneous sensors. In Jnstrument Symp. 9, 85-94.
Engineer’s Handbook, Vol. 1, pp. 919-943 (Editor Liptak, B. Koea, S., Bury, C. R. and Humpureey, A. E. (1967) Computer
G.). Chilton, Philadelphia. simulation of fermentation systems. Appl. Microbiol. 15,
Brown,J. E., Kaminski, R. K., Blake, A. C. and BRoDGESELL, 683-689.
A. (1969) Analytical measurements. In Instrument Engineer’s Kossen, N. W. F. (1979) Mathematical modelling of fermenta-
Handbook, Vol. 1, pp. 713-918 (Editor Liptak, B. G.), Chil- tion processes: scope and limitations. In Microbial Technology:
ton, Philadelphia. Current State, Future Prospects. Soc. Gen. Microbiol, Symp. 29,
Bunter, H. and Incotp, W. (1976) Measuring pH and oxygen pp. 327-357.
in fermenters. Proc. Biochem. 11(3), 19-22, 24. Ler, Y. H. and Tsao, G. T. (1979) Dissolved oxygen electrodes.
Cooney, C. L., Wane, H. Y. and Wane, D. I. C. (1977) Adv. Biochem. Eng. 13, 36-86.
Computer-aided balancing for prediction of fermentation Liptak, B. G. (1969) Pressure measurement. In Instrument
parameters. Biotech. Bioeng. 19, 55-67. Engineer’s Handbook, Vol. 1, pp. 171-263 (Editor Liptak, B.
Danis, W. F., ParKER, D. A., JoHNson, R. W. and SCHNEIDER, G.). Chilton, Philadelphia.
L. E. (1965) Controlled pH and oxidation-reduction with a Luioyp, A. K. (1973) Instrumentation in the brewing industry. In
new glass-tissue culture fermenter. Biotech. Bioeng. 7, 529- Instrumentation in the Process Industries, pp. 38-94 (Editor Lip-
503: tak, B. G.). Chilton, Philadelphia.
Eise.e, A., Ristropu, D. L. and Humpurey, A. E. (1978) LunpELL, R., Karvonen, E. and Oyama, H. (1981) Reduced
Mixing times and glucose uptake measured with a fluori- energy consumption in enzyme manufacturing by computer
meter. Biotech. Bioeng. 20, 1487-1492. control. In Abstracts of Communications. Second European Con-
Enrors, S. O. (1981) An enzyme electrode for control of glucose gress of Biotechnology, Eastbourne, p. 27. Society of Chemical
concentration in fermentation broths. Abstracts ofCommunica- Industry, London.
167
Principles of Fermentation Technology
MacLennan, D.C. (1970) Principles of automatic measurement Roserts, A. N. and SHEPHERD, A. S. (1968) Dissolved oxygen
and control of fermentation growth parameters. In Methods measurement in continuous aseptic fermentation. Proc.
in Microbiology, Vol. 2, pp. 1-22 (Editors Norris,J.R. and Biochem. 3(2), 23-24.
Ribbons, D. W. Academic Press, London. Suinskey, F. G. (1973) pH and plon Control in Process and Waste
MacLennan, D. G. and Pirt, S. J. (1966) Automatic control of Streams, Chapter 9. Wiley, New York.
dissolved oxygen concentration in stirred microbial cultures. SreGatL, S. O. and Gapen, E. L. (1962) Automatic control of
J. Gen. Microbiol. 45, 289-302. dissolved oxygen levels in fermentations. Biotech. Bioeng. 4,
Meiners, M. (1982) Computer applications in fermentation. In 345-356.
Overproduction of Microbial Products, pp. 637-649 (Editors Sotomons, G. L. (1969) Materials and Methods in Fermentation.
Krumphanzl, V., Sikyta, B. and Vanek, Z.). Academic Academic Press, London.
Press, London. Suzuk1, S. and KarusgE, I. (1979) Microbial electrode sensors for
Meskanen, A., LUNDELL, R. and Larno, P. (1976) Engineering of cephalosporins and glucose. In Immobilized Microbial Cells,
fermentation plants. Part I1I—Automation. Proc. Biochem. pp. 221-236 (Editor Venkatsubramanian, K.). American
11(5), 31-36. Chemical Society, Washington.
Moss, F. J. and Busu, F. (1967) Working design for a 5 liter Swartz, J. R. and Cooney, C. L. (1979) Indirect fermentation
controlled continuous culture apparatus. Biotech. Bioeng. 9, measurements as a basis for control. Biotech. Bioeng. Symp. 9,
585-602. 95-101.
Naaat, S. (1979) Mass and energy balance for microbial growth Uppixg, S. J. and Hicks, G. P. (1967) The enzyme electrode.
kinetics. Adv. Biochem. Eng. 11, 49-83. Nature, 214, 986-989.
Nyiri, L. K. (1972a) Applications of computers in biochemical Viesturs, U. E., Kristapsons, M. Z. and Levirans, E. S. (1982)
engineering. Adv. Biochem. Eng. 2, 49-95. Foam in microbiological processes. Adv. Biochem. Eng. 21,
Nyirt, L. K. (1972b) A philosophy of data acquisition, analysis 169-224.
and computer control of fermentation processes. Dev. Ind. Wane, H. Y., Cooney, C. L. and Wane, D. I. C. (1977)
Microbiol. 13, 136-145. Computer-aided baker’s yeast fermentations. Biotech. Bioeng.
Nyirei, L. K., JErrerts, R. P. and Humpurey, A. E. (1974) 19, 69-86.
Application of computers to the analysis and control of Wang, H. Y., Cooney, C. L. and Wang, D. I. C. (1979) On-line
microbiological processes. Biotech. Bioeng. Symp. 4, 613-628. gas analysis for material balances and control. Biotech.
Nyirl, L. K., Toru, G. M. and Cuartes, M. (1975) Measure- Bioeng. Symp. 9, 13-23.
ment of gas-exchange conditions in fermentation processes. _ Wuaarrte, P., ABorHEy, S., Hone, E. and Rocers, P. L. (1978)
Biotech. Bioeng. 17, 1663-1678. Microprocessor control of respiratory quotient. Biotech.
Puiturrs, D. H. and Jounson, M. J. (1961) Measurement of Bioeng. 20, 1459-1463.
dissolved oxygen in fermentations. J. Biochem. Microbiol. Yamasuira, S., Hisosut, H. and Inacak1, T. (1969) Automatic
Technol. Eng, 3, 261-275. control and optimization of fermentation processes—
Punar, E., Erseve, A., Bunter, H. and Incoip, W. (1980) glutamic acid. In Fermentation Advances, pp. 441-463 (Editor
Steam sterilizable pCO, electrode. Biotech. Bioeng. 22, 2411— Perlmian, D.) Academic Press, New York.
2416. Yamasuira, S. and Murao, C. (1978) Fermentation Process. J.
Ramirez, A., Duranp, A. and Biacuere, H. T. (1981) Optimal Soc. Instr. and Control Eng. 6(10), 735-740.
baker’s yeast production in extended fed-batch culture using ZasRISKIE, D. W. and Humpurey, A. E. (1978) Continuous
a computer coupled pilot-fermentor. In Abstracts of Communi- dialysis for the on-line analysis of diffusible components in
cations. Second European Congress of Biotechnology, Eastbourne, fermentation broths. Biotech. Bioeng. 20, 1295-1301.
p. 26. Society of Chemical Industry, London.
168
CHAPTER 9
169
Principles of Fermentation Technology
TABLE 9.1. Oxygen requirements ofa range of micro-organisms grown on a range of substrates (after
Mateles, 1979)
Oxygen requirement
Substrate Organism (g Op per gdry wt) Reference of growth data
170
Aeration and Agitation
of
productivity
Relative
acids
amino
oxygen in the broth. The effect of dissolved oxygen Fic. 9.2. The effect of dissolved oxygen on the production of
concentration on the specific oxygen uptake rate amino acids by Brevibacterium flavum (Hirose and Shibai, 1980).
171
Principles of Fermentation Technology
monitored and controlled, and facilitates the K;, may be considered as the reciprocal of the resis-
addition of supplements and the removal of tances to the transfer of oxygen from gas to liquid
samples. Some fermenters are so designed and (C* — C,) may be considered as the ‘driving
that adequate oxygen transfer is obtained force’ across the resistances. It is extremely difficult
without agitation and the design of these to measure both K, and ‘a’ in a fermentation and,
systems is also discussed in Chapter 7. therefore, the two terms are generally combined in
the term K;a, the volumetric transfer coefficient, the
Bartholomew et al. (1950) represented the transfer units of which are reciprocal time, normally h_!.
of oxygen from air to the cell, during a fermentation, The volumetric transfer coefficient is a measure of
as occurring in a number ofsteps: the aeration capacity of a fermenter, under the test
conditions, the larger the K,a the higher the aeration
(1) The transfer of oxygen from an air bubble
capacity of the system.
into solution.
The balance between the supply of dissolved
(ii) The transfer of the dissolved oxygen through
oxygen by the fermenter and the dissolved oxygen
the fermentation medium to the microbial
demand of the organism is reflected in the dissolved
cell.
oxygen concentration in the fermentation medium.
(ii) The uptake of the dissolved oxygen by the
If the K;a of the fermenter is such that the oxygen
cell.
demand of the organism cannot be met, the dis-
solved oxygen concentration will decrease below the
These workers demonstrated that the limiting step critical level (C,,;,). Ifthe K;ais such that the oxygen
in the transfer of oxygen from air to the cell in a demand of the organism can be easily met the
Streptomyces griseus fermentation was the transfer of dissolved oxygen concentration will be greater than
oxygen into solution. The resistance to oxygen trans- Coit and may be as high as 70 to 80% of the
fer imposed by the second step could be overcome by saturation level. Thus, the K;a of the fermenter must
increased agitation. These findings have been shown be such that the optimum oxygen concentration for
to be correct for non-viscous fermentations but it has product formation can be maintained in solution
been demonstrated that transfer may be limited by throughout the fermentation.
either of the other two stages in certain highly
viscous fermentations. The difficulties inherent in
such fermentations are discussed later in this
chapter. DETERMINATION OF K,;a VALUES
The rate of oxygen transfer from air bubble to the
liquid phase may be described by the equation:
The determination of the K,a of a fermenter is
dC,
di SA GONG
a 2G al 9.1
essential! in order to establish its aeration efficiency
and to quantify the effects of operating variables on
the provision of oxygen. This section considers the
where C, is the concentration of dissolved oxygen merits and limitations of the methods available for
in the fermentation broth, in mmoles the determination of K;a values.
dm’,
t is time in hours,
dc,. : : The sulphite oxidation technique
—— is the change in oxygen concentration
over a time period, i.e. the oxygen trans-
fer rate, in mmoles O, dm~*h"'," Cooper et al. (1944) were the first to describe the
K, is the mass transfer coefficient (cm ha). determination of oxygen-transfer rates in aerated
a is the gas/liquid interface area per liquid vessels by the oxidation of sodium sulphite solution.
volume (cm? cm7°), This technique does not require the measurement of
C* is the saturated dissolved oxygen con- dissolved oxygen concentrations but relies on the
centration, in mmoles dm~*. rate of conversion of a 0.5M solution of sodium
173
Principles of Fermentation Technology
dC,
di = K,a(C* <= C,)
carried out during a fermentation which should give
a more realistic assessment of the fermenter’s effh-
ciency. Because of the complex nature of fermenta-
and depicted in Fig. 9.4. Integration of equation tion broths the probe used to monitor the change in
(9.1) yields: dissolved oxygen concentration must be of the mem-
brane-covered type which may necessitate the use of
In (C* — C,) = —K,at. (9.3) the response correction factors referred to previ-
Thus, a plot ofIn (C* — C,) against time will yield ously. The procedure involves stopping the supply
a straight line of slope —K,a, as shown in Fig. 9.5. of air to the fermentation which results in a linear
This technique has the advantage over the sulphite decline in the dissolved oxygen concentration due to
oxidation method in that it is very rapid (normally the respiration of the culture, as shown in Fig. 9.6.
taking up to 15 minutes) and may utilize the fermen- The slope of the line AB in Fig. 9.6 is a measure of
tation medium, to which may be added dead cells or the respiration rate of the culture. At point B the
mycelium at a concentration equal to that produced aeration is resumed and the dissolved oxygen con-
during the fermentation. However, employing the centration increases until it reaches concentration
fermentation medium with, or without, killed X. Over the period, BC, the observed increase in
biomass necessitates the use of a membrane-type dissolved oxygen concentration is the difference
electrode, the response time of which may be between the transfer of oxygen into solution and the
inadequate to reflect the true change in the rate of uptake of oxygen by the respiring culture, as ex-
oxygenation over a short period of time. The probe pressed by the equation:
response time (7;,) is defined as the time needed to
record 63% of a stepwise change and this should be
Nees
ae =Kia(C* — C,) — xQO, (9.4)
much smaller than the mass transfer response time
of the system (1/K,a). According to Van’t Riet where x is the concentration of biomass and QO, is
175
Principles of Fermentation Technology
A
Xx t Cc
Dissolved Dissolved
oxygen
oxygen
concentration
concentration
Time
dC, + xQ0,
Fic. 9.6. Dynamic gassing out for the determination of K,a
dt
values. Aeration was terminated at point A and recommenced at
point B. Fic. 9.7. The dynamic method for the determination of K,a
values. The information is gleaned from Fig. 9.6 by taking
tangents to the curve, BC, at various values of C7.
the specific respiration rate (mmoles of oxygen g™!
biomass h_'). The term xQO, is given by the slope
of the line AB in Fig. 9.6. Equation (9.4) may be ing the fermentation so that the range of measure-
rearranged as: ments which could be used in the K;a determination
- would be very small. Thus, it may be difficult to
—] /dd
Cia
jg =| (>
di xQ Os) | c™. 9.5
(9:5) apply the technique during a fermentation which
has a maximum oxygen demand close to the supply
Thus, from equation (9.5), a plot of C;, versus capacity of the fermenter. It should also be remem-
dC,/dt + xQOpz will yield a straight line, the slope of bered that the technique relies upon measurements
which will equal —1/K,a, as shown in Fig. 9.7. taken at one point which may not be representative
The dynamic gassing-out method has the advan- of the bulk conditions in the vessel if the system is
tage over the previous methods of determining the inadequately mixed.
K,a during an actual fermentation and may be used
to determine K,a values at different stages in the
process. The technique is also rapid and only
requires the use ofa dissolved-oxygen probe, of the
membrane type. The major limitation in the opera-
tion of the technique is the range over which the
increase in dissolved oxygen concentration may be
measured. It is important not to allow the oxygen Dissolved
oxygen
concentration to drop below C,,;, during the concentration
deoxygenation step as the specific oxygen uptake
rate will then be limited and the term xQO, would
not be constant on resumption of aeration. The J corical
176
Aeration and Agitation
The oxygen-balance technique electrodes and use an average value. The value of C*
is frequently taken as that value which is in equilib-
The K;a ofa fermenter may be measured during a rium with the oxygen concentration ofthe gas outlet.
fermentation by the oxygen balance technique Wang et al. (1979) claimed that this approach was
which determines, directly, the amount of oxygen adequate for small-scale fermenters but on a large
transferred into solution in a set time interval. The scale there may be a considerable difference between
procedure involves measuring the following the dissolved oxygen concentration in equilibrium
parameters: with the inlet and outlet gases. Therefore, these
workers suggested that the behaviour of the gas in
(i) The volume of the broth contained in the
transit in the fermenter would approximate to plug
vessel in dm? (I).
flow conditions and a logarithmic mean value for the
(11) The volumetric air flow rates measured at
dissolved oxygen concentration should be used, that
the air inlet and outlet indm?* min | (Q; and
Q,, respectively).
is:
The total pressure measured at the fermenter OTR
ir giazcy aC NC a
Kea 9.7
air inlet and outlet, at atm. absolute (P; and
P., respectively). (Cr a Cr)
CERO
1 —————
(iv) The temperature of the gases at the inlet and
outlet in °K (7; and T,, respectively).
(v) The mole fraction of oxygen measured at the where C*,, and C* ,,are the dissolved oxygen concen-
inlet and outlet (»; and _y,, respectively). trations in equilibrium with gaseous oxygen at the
fermenter inlet and outlet, respectively, in mmoles
The oxygen-transfer rate may then be determined dm.
from the following equation (Wang et al., 1979): The oxygen-balance technique appears to be the
hd Le” simplest method for the assessment of K,a and has
OTR:= (O;P3/T; — QP, %/1,) (9-6) the advantage of measuring aeration efficiency dur-
cE
ing a fermentation. The sulphite oxidation and static
where 7.32 X 10° is the conversion factor equalling gassing-out techniques have the disadvantage of
(60 min h!') [mole/22.4 dm° being carried out using either a salt solution or
(STP)] (273°K/1 atm). an uninoculated, sterile fermentation medium.
These measurements require accurate flow Although, as Banks (1977) suggests, these
meters, pressure gauges and temperature-sensing techniques are adequate for the comparison of
devices as well as gaseous oxygen analysers (see equipment or operating variables it should not be
Chapter 8). The ideal gaseous oxygen analyser is a assumed that the values obtained are those actually
paramagnetic oxygen analyser which is sufficiently operating during a fermentation. This may be the
accurate to detect changes of | to 2%. case for bacterial or yeast fermentations where the
The K;a may be determined, provided that C;, rheology of the suspended cells in the broth is
and C* are known, from equation (9.1): similar to sterile medium or a salt solution, but it is
certainly not true for fungal and streptomycete pro-
OTK
OTR =
= K,a(C * — Cay Or Kya ae (C*—
ee G)en cesses where the rheology is quite different.
Tuffile and Pinho (1970) compared a number of
C, may be determined using a membrane-type dis- methods for the determination of K,a values in
solved-oxygen electrode and in this case the slow viscous streptomycete fermentations. The
response time is not an important factor because a techniques used were static gassing-out, dynamic
rate of change is not being measured, simply a gassing-out and the oxygen-balance method. Tuffile
steady-state oxygen level. However, it should be and Pinho did not make it clear whether non-respir-
remembered that an electrode simply measures the ing mycelium was present during their static
oxygen concentration at one point and it is, there- gassing-out procedure but from their results it would
fore, advisable to monitor the oxygen concentration appear that it was present in the vessel. Thus, the
at a number ofpoints in the vessel with a number of rheology of the fermenter contents would appear to
177
Principles of Fermentation Technology
Tasre 9.3. K,a values for a 300-dm’ fermenter containing a 90-hour cost of the monitoring equipment involved should
culture of S. aureofaciens (Tuffile and Pinho, 1970) be a worthwhile investment.
Method of K;a Measured oxygen uptake Kya
Wang et al. (1979) quoted a method for the deter-
determination (mmoles dm~*h_') (hy') mination of K;a based on the type of stoichiometric
determinations discussed earlier in the section on
Static gassing out =< 66.4
Dynamic gassing out 6.6 58.2 oxygen requirements. According to Wang et al., the
Oxygen balance 20.1 108.0 volumetric absorption rate of a culture may be
described as:
N, = px(K'/Yp) (9.8)
have been similar for the different determinations.
The K,;a values, determined by the different where N, is the oxygen absorption rate or demand
techniques, for a 300-dm?® fermenter containing a rate in mmoles of oxygenh |,
90-hour culture of Streptomyces aureofaciens are shown xis the cell concentration in grams dry
in Table 9.3. weight ofcells dm’,
From Table 9.3 it may be seen that the K;a values Yy is the yield coefficient of oxygen in grams
for the two gassing-out techniques were very similar of cells per gram of oxygen,
but there was a considerable difference between the K' is the conversion factor equalling 10°/32
oxygen uptake rates and the K;as determined by the or 31.3 mmoles oxygen g | oxygen.
dynamic method and the balance method. Tuffile The Y) value may be determined from the
and Pinho (1970) claimed that the low oxygen-
stoichiometry of the conversion of the carbon source
uptake rate determined by the dynamic method was and oxygen into cells, using equations such as those
due to air bubbles remaining in suspension in the
discussed earlier derived by Darlington (1964),
mash during the dynamic gassing-out period. Thus,
Johnson (1964) and Mateles (1971). However, these
the decline in oxygen concentration after the cessa-
equations assume that the only products are cells,
tion of aeration was not a measure of the oxygen-
carbon dioxide and water which makes them
uptake rate but the difference between oxygen
inapplicable to processes producing an extracellular
uptake and the transfer of oxygen from entrapped
product. Assuming that the stoichiometric equa-
bubbles. It was demonstrated that a large number of
tions are suitable, the value of the demand rate
bubbles remained suspended in the medium 15
calculated in equation (9.8) may be used as the
minutes after aeration had been stopped. The use of
oxygen-transfer rate. The K;a value may then be
the low oxygen-uptake rate in the calculation of the
determined from the equation:
K,a would result in an artificially low K,a being
determined. From these results it appears that the OTR
je
oxygen balance technique is the best method for Se
estimation of K;as in viscous fermentations. How-
C;, may be determined using a dissolved oxygen
ever, it is unfortunate that Tuffile and Pinho did not
calculate a K;a value from the aeration data of the
electrode.
Although this technique is probably the simplest
dynamic method using the oxygen-uptake rate
to use it should only be employed when the fermen-
determined from the balance method, especially
tation is accurately described by the stoichiometric
considering the similarity of the results they
obtained with the two gassing-out techniques, the model.
Before considering the factors which may affect
static one not relying on the determination of
the K;a of afermenter it is necessary to consider the
oxygen-uptake rates. Nevertheless, Heijnen et al.
behaviour offluids in agitated systems.
(1980) also observed anomalies in determining K,a
values in viscous systems due io the presence ofvery
small bubbles having a much longer residence time FLUID RHEOLOGY
than the more abundant large bubbles in the vessel.
Overall, it would appear that the balance method is Fluids may be described as Newtonian or non-
the most desirable technique to use and the extra Newtonian depending on whether their rheology
178
Aeration and Agitation
Bingham plastic
Shear
stress Casson body
(7)
Newtonian fluid
Dilatant
Shear
stress
(7)
Pseudoplastic
Shear rate (y)
179
PFT-M
Principles of Fermentation Technology
plastic viscosity. Thus, the flow ofa Bingham plastic lium chrysogenum broths behaved as pseudoplastic
is described by the equation: fluids (contradicting the earlier results of Dein-
doerfer and Gaden (1955)). Taguchi et al. (1968)
T=T7 + NY (9.10) demonstrated that an Endomyces culture broth dis-
where 77 is the coefficient ofrigidity and played pseudoplastic rheology.
Ty is the yield stress.
Deindoerfer and Gaden (1955) claimed that
Penicillium chrysogenum culture broths behaved as Dilatant rheology
Bingham plastics. Sato (1961) also related the
rheological properties of a kanamycin broth to The apparent viscosity of a dilatant liquid
Bingham plastic behaviour. increases with increasing shear rate. The flow ofa
dilatant liquid may also be described by equation
(9.12) butin this case the value ofthe flow-behaviour
Pseudoplastic rheology index is greater than 1, the greater the value the
greater the flow characteristics deviate from those of
The apparent viscosity of a pseudoplastic liquid a Newtonian fluid. Thus, the values of K and n may
decreases with increasing shear rate. Most polymer be obtained from a plot of log shear stress against log
solutions behave as pseudoplastics, the explanation shear rate.
of the decreased apparent viscosity at higher shear
rates being the fact that the long chain molecules
tend to align with each other resulting in easier flow. Casson body rheology
The flow ofapseudoplastic liquid may be described
by the equation:
Casson (1959) described a type of non-Newtonian
T= K(y)” (9.11) fluid, termed a Casson body, which behaved as a
pseudoplastic in that the apparent viscosity
where K is the consistency coefficient and
decreased with increasing shear rate but displayed a
n is the flow behaviour index or power law
yield stress and, therefore, also resembled a Bing-
index.
ham plastic. The flow characteristics of a Casson
K has the same units as viscosity and may be taken
body may be described by the following equation:
as the apparent viscosity when the shear rate is 1.
The flow-behaviour index is a dimensionless con- Vr= V+ K.Vy (9.13)
stant and is less than unity for a pseudoplastic
where K, is the Casson viscosity.
liquid, the smaller the value of n, the greater the flow
A plot of V7 against Vy will give a straight line, the
characteristics of the liquid deviate from those ofa
slope of which will equal the Casson viscosity and
Newtonian fluid. Equation (9.11) may be converted
the intercept of the V7 axis will equal V7.
to the logarithmic form as
Roels et al. (1974) claimed that the rheology of a
log
tT= log
K + nlog y. (9.12) penicillin broth could be best described in terms ofa
Casson body.
Thus, a plot of log shear stress against log shear rate Therefore, to determine the rheological nature of
will produce a straight line, the slope of which will a fluid it is necessary to construct a rheogram which
equal the flow-behaviour index and the intercept on requires the use of a viscometer which is accurate
the shear-stress axis will be equal to the logarithm of over a wide range of shear rates. Furthermore, the
the consistency coefficient. testing of mycelial suspensions may present special
Tuffile and Pinho (1970) demonstrated that the difficulties. These problems have been considered in
culture fluids of two Streptomyces sp. exhibited detail by Metz et al. (1979), whose paper should be
pseudoplastic behaviour late in the fermentation. consulted for methods of assessing the rheological
Deindoerfer and West (1960) claimed that Penicil- properties of mycelial fluids.
180
Aeration and Agitation
The air-flow rate has a relatively small effect on The degree of agitation has been demonstrated to
K,a values in conventional agitated systems, as have a profound effect on the oxygen-transfer efh-
illustrated in Fig. 9.11. The range of air-flow rates ciency of an agitated fermenter. Banks (1977)
employed rarely fall outside the range 0.5 to 1.5 claimed that agitation assisted oxygen transfer in
volumes of air per volume of medium per minute the following ways:
(volumetric air-flow rate). If very high flow rates are (i) Agitation increases the area available for
used then extremely low oxygen transfer rates may oxygen transfer by dispersing the air in the
be achieved due to the impeller becoming ‘flooded’. culture fluid in the form of small bubbles.
Flooding is the phenomenon where the air-flow rate (ii) Agitation delays the escape of air bubbles
is so high that the impeller is rotating in a gas phase from the liquid.
and therefore cannot assist in the transfer of gas into (iii) Agitation prevents coalescence of air bub-
solution. The volumetric air-flow rate is frequently bles.
kept constant on increasing the scale of a fermenta- (iv) Agitation decreases the thickness of the
tion. However, this procedure may result in extreme liquid film at the gas/liquid interface by
foaming of the broth due to the fact that the creating turbulence in the culture fluid.
volumetric air-flow rate increases by a cubic func-
tion whereas the escape of the introduced air is The degree of agitation may be measured by the
related to the surface area of the liquid at the top of amount of power consumed in stirring the vessel
the vessel and, therefore, increases by a square contents. The power consumption may be assessed
function. Some workers keep the superficial air by using a dynamometer, by using strain gauges
attached to the agitator shaft and by measuring the
velocity,
electrical power consumption of the agitator motor
volumetric air-flow rate (see Chapter 8). The assessment of electrical con-
cross-sectional area of the fermenter sumption is only suitable for use with large-scale
181]
Principles of Fermentation Technology
vessels. Several attempts have been made to corre- impeller, thus affecting the value of the exponent
late power consumption (and, hence, agitation) with term.
K,a values so that mathematical descriptions of the Richards (1961) stressed that it is extremely likely
relationship may be used for predicting power that any derived relationship between K,;a and
requirements in design and scale-up exercises. How- power consumption per unit volume is affected by a
ever, one of the major difficulties to arise in this large number of variables not included in the
context is that the correlations have normally been relationships. Such variables include impeller speed
carried out using Newtonian fluids, whereas many and size, culture rheology and the superficial air
of the fermentation broths are non-Newtonian in velocity. Richards derived the following relationship
nature. which includes some of the variables affecting K;a:
K,a (P,/V)°4V9°N°° (9.15)
THE RELATIONSHIP BETWEEN K,;a AND
POWER CONSUMPTION where Nis the impeller rotational speed.
Cooper et al. (1944) measured the Kyas of a Richards applied his semi-empirical relationship
number of agitated and aerated vessels containing to his own and Cooper ¢ al.’s results, all obtained
one impeller (up to a volume of 66 dm*), using the from the sulphite oxidation technique using a wide
sulphite oxidation technique, and derived the fol- range of vessel size and conditions, and although he
lowing expression: conceded the correlation was not good, a reasonable
P 0.95 linear fit was obtained. Taguchi et al. (1968)
Ka = (2) wel (9.14) examined the relationship between K,a and power
consumption in an Endomyces fermentation that dis-
where P, is the power absorption in an aerated sys- _ played pseudoplastic (non-Newtonian) characteris-
tem, tics, the study being carried out in fermenters rang-
V is the liquid volume in the vessel, ing in working volume from 20,000 to 30,000 dm?.
V, is the superficial air velocity which equals These workers derived the following relationship
the between K;a and power consumption:
volumetric air flow rate p33
their coalescence which involves the transmission of sumption (and, hence, agitation and, there-
more power than shear rate. The explanation was fore, K;as) in vessels ofdifferent size.
supported by a correlation between oxygen transfer
rate and impeller tip speed, which was independent
THE RELATIONSHIP BETWEEN POWER
of impeller diameter:
CONSUMPTION AND OPERATING VARIABLES
OTR =k'(S)!° (9.18) Rushton et al. (1950) investigated the relationship
between power consumption and operating vari-
where § is the impeller tip speed and equal to 7ND,
ables in baffled, agitated vessels using the technique
where JN is the impeller rotational speed
of dimensional analysis. They demonstrated that
and D is the impeller diameter, and
power absorption during agitation of non-gassed
k’ is a constant.
Newtonian liquids could be represented by a dimen-
Thus, although these workers demonstrated a sionless group termed the power number, defined
relationship between power consumption and K,;aa by the expression:
more widely applicable relationship was shown
between oxygen-transfer rate and impeller-tip ¢
N, p =—
pN°D° (
9.19 )
speed. Therefore, it should be remembered that
power consumption is not the only factor which may where N, is the power number,
indicate the degree of agitation in a system and it is P is the external power from the agitator,
important to consider the function of agitation when p is the liquid density,
correlating it with measurable parameters. The N is the impeller rotational speed,
correlation between oxygen-transfer rate and impel- D is the impeller diameter.
ler-tip speed suggests that the maintenance of a Thus, the power number is the ratio of external force
constant impeller-tip speed in non-Newtonian fer- exerted (P) to the inertial force imparted (oN*D”) to
mentations may provide equivalent aeration condi- the liquid. The motion of liquids in an agitated
tions. According to Wang et al.. (1979) many com- vessel may be described by another dimensionless
panies attempt to maintain constant impeller-tip number, known as the Reynolds number, which is a
speed, on scale up, as well as maintaining a constant ratio ofinertial to viscous forces:
K,a which may be achieved by varying the impeller
by pD°?N
diameter to vessel diameter ratio in the larger vessel. Ne e (9.20)
From this discussion it is evident that the K,a of be
an aerated, agitated vessel is affected significantly where Ne, is the Reynolds number and
by the consumption of power during stirring and, bw is the liquid viscosity.
hence, the degree of agitation. Although it is not Yet another dimensionless number, termed the
possible to derive a relationship between K,a and Froude number, relates inertial force to gravita-
power consumption which is applicable to all situa- tional force and is given by the term:
tions it is possible to derive a relationship between
the two which is operable within certain limits. If it x ND?
Ney (9:2)
is accepted that such relationships between power g
consumption and K,a exist it is of considerable where Ny, is the Froude number and
importance to relate power consumption to operat- g is the gravitational force.
ing variables which may affect it. Quantitative Rushton e¢ al. (1950) demonstrated that the power
relationships between power consumption and number was related to the Reynolds and Froude
operating variables may be useful in: numbers by the expression:
(i) Estimating the amount of power that an agi- IN = CONR oe NE) (9:22)
tation system will consume under certain
circumstances, which could assist in fer- where cis the constant dependent on_ vessel
menter design. geometry but independent ofvessel size,
(ii) In providing similar degrees of power con- x and y are exponents.
183
Principles of Fermentation Technology
(1) The laminar or viscous flow zone, where the where the subscripts sm and L refer to the small and
logarithm of the power number decreases large scales respectively. Maintaining the same
linearly with an increase in the logarithm of power input per unit volume:
the Reynolds number. The slope of the graph
Poy) Vege tepN nD
is equal to x, the exponent in equation (9.24),
Fy
=e V,
ese
cpN;, Dy
(9.28)
and is obviously equal to —1. The power
absorbed in this region is a function of the where Vis the volume.
viscosity of the liquid and the Reynolds Assuming the vessels to be geometrically similar
number is less than 10. then c will be the same regardless of scale and as the
184
Aeration and Agitation
same broth would be employed p would remain the where Q is the volumetric air-flow rate and
same for both systems: k isa constant, relating to fermenter
geometry.
Von N = D pa
This expression may be used to predict power con-
sumption in gassed systems where turbulent flow is
For geometrically similar vessels known to be operating but it should be remembered
D V. Ns that in non-mycelial fermentations the greatest
power demands often occur during agitation: when
Dy, oe. | the system is not gassed, that is during the steriliza-
tion of the medium in situ. Thus, in designing the
Therefore, substituting for D,,,/D,; in (9.29):
system care must be taken to ensure that the agitator
V 2/9
motor is sufficiently powerful to agitate the ungassed
N,, = Nm(2) . (9.30)
L system. Banks (1979) has suggested that the modifi-
cation of Rushton’s equation, in the light of Michel
If transient flow conditions occur in a fermentation and Miller’s work, would result in little practical
then it is necessary to construct a complete power advantage and that very similar results would be
curve for such predictions and this is discussed later obtained. '
in the chapter. From the foregoing account it may be seen that
The work of Rushton et al. (1950) was carried out reasonable techniques exist to relate operating vari-
using ungassed liquids whereas the vast majority of ables to power consumption and, hence, to the
fermentations are aerated. It is widely accepted that degree of agitation which may be shown to have a
aeration ofa liquid decreases the power consump- proportional effect on K,a. However, these
tion during agitation because an aerated liquid, techniques apply to Newtonian fluids and are not
containing suspended air bubbles, is less dense than directly applicable to the study of non-Newtonian
an unaerated one. Oyama and Endoh (1955) systems. Non-Newtonian fluids do not have con-
studied the effect of aeration on power consumption stant viscosities, which creates difficulties in utiliz-
in agitated systems and related the ratio of gassed to ing relationships which rely on being able to deter-
ungassed power consumption to a dimensionless mine the fluid viscosity. These difficulties may be
term, called the aeration number, N,. Oyama and avoided if the agitation system is capable of main-
Endoh presented the following equation: taining turbulent-flow conditions during the fer-
mentation because under such conditions power
Pp
~ = f(N 991 consumption is independent of the Reynolds
number and, hence, of viscosity. However, the high
where P and P, are the ungassed and gassed power viscosities of the majority of mycelial fermentation
consumptions, respectively, and fis a function. broths make fully turbulent flow conditions impos-
sible, or extremely difficult, to achieve. Such fermen-
N,= Van (9.32) tations tend to exhibit transient zone flow conditions
ND which necessitate the construction of complete
Michel and Miller (1962) demonstrated that the power curves to correlate power consumption with
aeration number was applicable when operating operating variables. The fact that the viscosity ofa
variables, such as impeller-tip speed and diameter, non-Newtonian liquid is affected by shear rate
were varied over a wide range. These workers means that the viscosity ofa non-Newtonian fermen-
demonstrated an empirical correlation for gassed tation broth will not be uniform throughout the
power consumption which was applicable over a fermenter because the shear rate will be higher near
wide range of operating conditions, in the turbulent the agitator than elsewhere in the vessel. Thus, the
flow region: determination of the impeller Reynolds number is
made difficult by not knowing the viscosity of the
P?ND? 0.45
P,=k (9:33) fermentation broth. Metzner and Otto (1957) pro-
5
posed a solution to this paradox by introducing the
185
Principles of Fermentation Technology
concept of average shear rate (y) related to the technique is not applicable to highly viscous non-
agitator shaft speed in the vessel, by the equation: Newtonian broths where turbulent-flow conditions
cannot be achieved.
y=kN (9.34)
where & is a proportionality constant.
Metzner and Otto determined the value of the The effect of medium and culture rheology
proportionality constant to be 13 for pseudoplastic on K,a
fluids in conventional, baffled reactors agitated by
single, flat-blade turbines. Several groups of workers
As can be seen from the previous section, the
have determined values of k under a wide range of
rheology of a fermentation broth has a marked
operating variables and the values range from
influence on the relationship between K,a and the
approximately 10 to 13. Metzner et al. (1961)
degree of agitation. The objective ofthis section is to
suggested that a compromise value of 11 could be
discuss the effects of medium and culture rheology
used for calculation purposes, with relatively little
on oxygen transfer during a fermentation. A fermen-
loss of accuracy, which would obviate the necessity
tation broth consists of the liquid medium, in which
to determine k for each circumstance. Therefore,
the organism grows, the microbial biomass and any
provided the rheological properties of a fermenta-
product which is produced by the organism. Thus,
tion broth are known, an apparent average viscosity
the rheology of the broth is affected by the com-
of the fluid may be calculated using the average
position ofthe original medium and its modification
shear rate which would enable the calculation ofthe
by the growing culture, the concentration and
impeller Reynolds number for each value of the
morphology of the biomass and the concentration of
impeller rotational speed, thus enabling a power
microbial products. Therefore, it should be appar-
curve to be constructed. Such a power curve may be
ent that there is wide variation in rheological
used to predict the power requirements ofa fermen-
properties between different fermentation broths
tation and to scale up a fermentation on the basis of
and significant change in broth rheology may occur
power consumption per unit volume. As pointed out
during a fermentation.
by Banks (1977) there is very little evidence in the
literature of the application of Metzner and Otto’s
technique to fermentation systems, despite the fact MEDIUM RHEOLOGY
that there is no obvious reason to suggest that it is Fermentation media frequently contain starch as
inapplicable to such systems. The most likely ex- a carbon source which may render the medium
planation of this anomaly is the lack of information non-Newtonian and relatively viscous. However, as
available on the rheology of non-Newtonian fermen- the organism grows it will degrade the starch and,
tation broths, and until this area is more thoroughly thus, modify the rheology of the medium and reduce
investigated the application ofsuch potentially use- its viscosity. Such a situation was described by
ful techniques will continue to be limited. Tuffile and Pinho (1970) in their study of the growth
The effect of gassing on power consumption in of Streptomyces aureofaciens on a starch-containing
non-Newtonian systems has not been studied to any medium. Before inoculation, the medium displayed
great extent. It is certainly true that gassing will Bingham plastic characteristics with a well-defined
decrease power consumption in a non-Newtonian yield stress and an apparent viscosity of approxi-
system but there are few published quantitative mately 18 pseudopoise; after 22 hours the
descriptions of the relationship. Taguchi and organism’s activity had decreased the medium vis-
Miyamoto (1966) considered the gassed power con- cosity to a few pseudopoise and modified its
sumption of a pseudoplastic Endomyces broth with behaviour to that of a Newtonian liquid; from 22
respect to Michel and Miller’s (1962) description hours onwards the apparent viscosity of the broth
of gassed power consumption. These workers gradually increased, due to the development of the
demonstrated that the relationship held true for the mycelium, up to a maximum of approximately 90
non-Newtonian system in the turbulent zone but pseudopoise and the rheology of the broth became
not in the viscous and transition zones. Thus, this increasingly pseudoplastic in nature. Thus, this
186
Aeration and Agitation
Time—> Time —+
---- Dissolved oxygen concentration
Culture oxygen uptake rate
Fie. 9.13. The effect of oxygen limitation on the culture oxygen uptake rate: (a) A typical bacterial fermentation. (b) A typical fungal
fermentation (Banks, 1977).
example suggests that the rheological problems pre- gen uptake remains constant in a unicellular system
sented by the medium are minor compared with whereas it decreases in a mycelial one. Banks
those presented by a high mycelial biomass, especi- claimed that the only possible explanation for such
ally when it is considered that the total oxygen a decrease is the increasing viscosity of the culture
demand is relatively low in the early stages of a caused by the increasing mycelial concentration.
fermentation. Several groups of workers have demonstrated the
detrimental effect of the presence of mycelium on
THE EFFECT OF MICROBIAL BIOMASS ON K;a oxygen transfer. Figure 9.14 represents some of the
The biomass concentration and its morphological data of Deindoerfer and Gaden (1955) illustrating
form, in a fermentation, has been shown to have a the effect of Penicillium chrysogenum mycelium on Ka.
profound effect on oxygen transfer. Most bacterial
and yeast fermentations tend to give rise to relatively
100
non-viscous, Newtonian broths in which conditions
of turbulent flow may be achieved. Such fermenta-
tions present relatively few oxygen-transfer prob-
lems. However, the non-Newtonian broths offungal
and streptomycete fermentations present major
difficulties in oxygen provision, the productivity of 50
many such fermentations being limited by oxygen
availability. Banks (1977) stressed the difference in (%
original)
Ka
of
the pattern of oxygen uptake between unicellular
and mycelial fermentations, as illustrated in Fig.
9.13. In both unicellular and mycelial fermentations
ON 0!25 0:40:65 0:85 lO) e214:
the pattern of total oxygen uptake is very similar
Mycelium concentration (%w/v)
during the period of exponential growth, up to the
point of oxygen limitation. However, during oxygen Fic. 9.14. The effect of Penicillium chrysogenum mycelium on K;a in
limitation, when arithmetic growth occurs, the oxy- a stirred fermenter (Deindoerfer and Gaden, 1955).
187
Principles of Fermentation Technology
to give rise to a highly viscous, non-Newtonian Righelato (1979) discussed the effects of
broth whereas the pellet form tends to produce an mycelium morphology on culture rheology and oxy-
essentially Newtonian system with a much lower gen transfer and came to the conclusion that the
viscosity. Metz et.al. (1979) demonstrated that pellet most desirable way for a mycelium to grow in
suspensions could be non-Newtonian but confirmed submerged culture is in the form of short, hyphal
that they gave rise to low viscosity broths. Thus, the fragments which would produce a broth less sus-
morphological form of a mycelial organism in sub- ceptible to diffusion limitation than a pelleted one,
merged culture has a major effect on the broth and less viscous than one containing long filaments.
rheology and may, therefore, be expected to influ- The degree of agitation has been shown to influence
ence aeration efficiency. the morphology of the culture and may be used as a
Carilli e¢ a/. (1961) compared the aeration efficien- means of controlling the broth rheology. Metz and
cies obtained in 3000-dm* fermenters containing Kossen (1977) claimed that strong agitation may
different morphological forms of Penicillium prevent the formation of pellets produced by the
chrysogenum and Aspergillus niger. Two strains of P. coagulation of spores. Also, these reviewers pointed
chrysogenum were employed, one which grew as short, out that strong agitation influences the development
highly branched hyphae and the other as long, and structure of pellets, once formed, and tends to
relatively unbranched hyphae. The short, branched give rise to smaller, more compact pellets. Ziegler et
hyphae gave rise to a relatively low viscosity broth in al. (1980) demonstrated that the high shear rate
which the oxygen transfer rate was approximately produced in a tubular loop reactor tended to reduce
twice that achieved with the more viscous broth of the clumping of mycelial organisms. Thus, the
the unbranched form. By manipulating the cultural degree of shear may be utilized to prevent, or encour-
conditions of A. niger, Carilli et al. were able to age, the formation of pellets and to control their
produce the fungus in either filamentous or pellet morphology once formed. However, attempts to
form and demonstrated that the pellet form gave rise encourage the formation of the desirable short
to a broth exhibiting half the viscosity of the hyphal fragment morphology (as compared with
filamentous broth. Also, oxygen limitation occurred the long filaments) by increasing the shear stress on
far earlier in the fermentation when the organism the mycelium has met with only limited success.
grew in the filamentous form. Dion et al. (1954) showed that the morphology of
Although the pellet type of growth tends to pro- Penicillium chrysogenum was influenced by the degree
duce a low viscosity Newtonian broth in which of agitation in that short, branched mycelium was
turbulent flow conditions may be achieved it, also, produced at high agitation rates compared with
may give rise to problems of oxygen availability if long hyphae produced at low agitation rates.
the pellets become too large. A large pellet may be so Katinger (1977) claimed that most of the attempts
compact that its centre may be unaffected by the to produce short hyphal fragments by increasing the
turbulent forces occurring in the bulk of the fermen- shear stress on the mycelium has resulted in the
tation broth so that the passage of oxygen within the mycelium tending to be ‘ground down’ rather than
pellet is dependent on simple diffusion which may ‘cut to pieces’. Van Suijdam and Metz (1981) inves-
result in the centre ofthe pellet being oxygen limited. tigated the effect of shear stress on the hyphal length
Thus, to maintain the intra-pellet oxygen concentra- of P. chrysogenum and demonstrated that an enorm-
tion at an adequate level it would be necessary to ous increase in energy input was necessary to obtain
maintain a high dissolved oxygen concentration to a substantial decrease in hyphal length, which
ensure an effective diffusion gradient. Kobayashi et suggests that it is oflittle practical value to decrease
al. (1973) demonstrated this phenomenon in pellets the viscosity of mycelial broths. Righelato (1979)
of A. niger where large pellets required a higher also claimed that it is unlikely that shear forces
dissolved oxygen concentration to maintain the could account for the break up of mycelia and that
same specific oxygen-uptake rate as smaller pellets. autolysis and lysis of some hyphal compartments
If oxygen limitation does occur within a pellet then may be more important controlling factors, perhaps
only the outer layer of the pellet would contribute to implying that the phenomenon may be more under
its growth and the centre may autolyse. genetic, rather than physical, control. Kuenzi
189
Principles of Fermentation Technology
(1978) compared the broth viscosities of anumber of THE EFFECT OF MICROBIAL PRODUCTS ON
Cephalosporium mutants grown under identical con- AERATION EFFICIENCY
ditions and showed that the viscosities differed by a Generally speaking, the product ofa fermentation
factor of times 7 to times 8; however, the morpholog- contributes relatively little to the viscosity of the
ical forms within the broths were not quoted so a culture broth. However, the exception to this state-
direct correlation may not be made between mor- ment is the production ofbacterial polysaccharides,
phological mutations and broth viscosity, in this where the broths tend to be highly viscous
case. The selection, and breeding, of morphologi- (30,000 cp, Sutherland and Ellwood, 1979) and
cally favourable strains is discussed in more detail in non-Newtonian. Charles (1978) demonstrated that
Chapter 3. the bacterial cells in a polysaccharide fermentation
Kuenzi (1978) reported that the viscosity of a made a minimal contribution to the high culture
Cephalosporium broth was considerably reduced by viscosity which was primarily due to the polysac-
growing the organism at 27° rather than 25°C. The charide product. Normally, microbial polysac-
higher temperature resulted in fragmentation of the charides tend to behave as pseudoplastic fluids,
hyphae whereas at the lower temperature the fungus although some have also been shown to exhibit a
grew in the form of long filaments which increased yield stress. Thus, bacterial polysaccharide fermen-
the culture viscosity to such an extent that it was tations present problems of oxygen transfer and
impossible to provide the culture with sufficient bulk mixing similar to those presented by mycelial
oxygen using the same equipment. fermentations. The pseudoplastic nature of the
Several workers have discussed the possible polysaccharide broth means that it becomes less
advantages of reducing the viscosity of a mycelial viscous with increasing shear stress so that adequate
fermentation, in its later stages, by diluting the _ aeration may only be achieved at high stirrer rates.
broth with either water or fresh medium. Sato However, viscosity would increase with increasing
(1961) increased the yield of akanamycin fermenta- distance from the agitator and air introduced into
tion, displaying Bingham plastic rheology, by 20% the fermenter would tend to rise through the vessel
by diluting the broth 5% by volume with sterile by the route of least resistance, that is, through the
water. Taguchi (1971) achieved a 50% reduction in well-stirred, less viscous, central zone. Thus, stag-
the viscosity of an Endomyces broth by diluting 10% nant areas, receiving little oxygen, may occur in the
with water or fresh medium. Blanch and Bhavaraju vessel. Wang and Fewkes’ (1977) discussion of the
(1976) put forward a scheme for the control of significance of the ratio of the impeller turbulent
viscosity and dissolved oxygen concentration in a shear stress to impeller pumping in mycelial fermen-
hypothetical fermentation. These workers proposed tations is also relevant to the development of stag-
that, as the critical dissolved oxygen concentration nant zones in the bacterial polysaccharide fermenta-
is approached, a set volume of broth could be tions. The agitator employed in these fermentations
removed from the fermenter and replaced with fresh should create the correct balance of turbulence (and,
medium. The process could be repeated in a step- hence, the transfer of oxygen into solution) and
wise manner as the system became oxygen limited pumping power to circulate the broth through the
which could be determined by dissolved oxygen region ofhigh shear.
concentration or viscosity measurements. Thus, by No detailed reports have yet been published of the
using such techniques the viscosity may be control- design ofagitators used in polysaccharide fermenta-
led and maintained below the level which may cause tions.
oxygen limitation. Kuenzi (1978) reported an
instance where the very slow feeding of medium toa
The effect of foam and antifoams on oxygen
Cephalosporium culture resulted in the organism
growing in the form of long filaments which pro- transfer
duced a highly viscous culture which could not be
adequately aerated. The design offed-batch proces- The high degree of aeration and agitation
ses such that efficient control may be achieved over required in a fermentation frequently gives rise to
the process is discussed in Chapter 2. the undesirable phenomenon offoam formation. In
190
Aeration and Agitation
extreme circumstances the foam may overflow from Bevt, G. H. and Gatto, M. (1971) Effect ofimpurities on oxygen
the fermenter via the air outlet or sample line result- transfer. Process Biochem. 6(4), 33-35.
ing in the loss of medium and product, as well as Biancu, H. W. and Buavarayu, S. M. (1976) Non-Newtonian
fermentation broths—Rheology and mass transfer. Biotech.
increasing the risk of contamination. The presence Bioeng. 18, 745-790.
of foam may also have an adverse effect on the Cariiut, A., Carn, E. B., Gaucanpi, G. and Morisi, G. (1961)
oxygen-transfer rate. Hall e al. (1973) pointed out Aeration studies. III. Continuous measurements of dissol-
ved oxygen during fermentations in large fermenters. Sci.
Waldhof and vortex-type fermenters (see Chapter Repts. Ist. Super Sanita, 1, 177-189.
7) were particularly affected due to the bubbles Casson, N. (1959) Rheology of Disperse Systems. Pergamon Press,
becoming entrapped in the continuously recirculat- Oxford.
Cuar-es, M. (1978) Technical aspects of the rheological proper-
ing foam resulting in high bubble residence times ties of microbial cultures. Adv. Biochem. Eng. 8, 1-62.
and, therefore, oxygen-depleted bubbles. The pres- Cooney, C. L. (1979) Conversion yields in penicillin production:
ence of foam in a conventional agitated, baffled Theory vs. practice. Process Biochem. 14(5), 31-33.
Cooper, C. M., Fernstrom, G. A. and Mirter, S. A. (1944)
fermenter may also increase the residence time of Performance of agitated gas-liquid contacters. Ind. Eng.
bubbles and therefore result in their being depleted Chem. 36, 504-509.
DaruincTon, W. A. (1964) Aerobic hydrocarbon fermentation—
of oxygen. Furthermore, the presence of foam in the
A practical evaluation. Biotech. Bioeng. 6(2), 241-242.
region of the impeller may prevent adequate mixing DeinDorrFeER, F. H. and Gapen, E. L. (1955) Effects of liquid
of the fermentation broth. Thus, it is desirable to physical properties on oxygen transfer in penicillin fermen-
tation. Appl. Micro. 3, 253-257.
break down a foam before it causes any process
DernporrRFER, F. H. and West, J. M. (1960) Rheological
difficulties and, as discussed in Chapter 7, this may examination of some fermentation broths. J. Microbiol.
be achieved by the use of mechanical foam breakers Biochem. Technol. Eng. 2, 165-175.
Dion, W. M., Carixu, A., SERMONTI, G. and Cuan, E. B. (1954)
or chemical antifoams. All antifoams are surfactants
The effect of mechanical agitation on the morphology of
and may, themselves, be expected to have some Penicillium chrysogenum Thom in stirred fermenters. Rend. Ist
effect on oxygen transfer. The predominant effect Super Sanita, 17, 187-205.
FEeREn, C. J. and Squires, R. W. (1969) The relationship between
observed by most workers is that antifoams tend to
critical oxygen level and antibiotic synthesis of capreomycin
decrease the oxygen-transfer rate, as discussed by and cephalosporin C. Biotech. Bioeng. 11, 583-592.
Aiba et al. (1973) and Hall et al. (1973). Thus; a Go pBeEr«, I., Rock,J. S., Ben-Bassat, A. and Marte es, R. f.
(1976) Bacterial yields on methanol, methylamine, formal-
balance must be struck between the necessity for dehyde and formate. Biotech. Bioeng. 18, 1657-1668.
foam control and the deleterious effects of the con- Hatt, M. J., Dickinson, S. D., PrircHarp, R. and Evans,J. I.
trolling agent. In fermentations which are limited (1973) Foams and foam control in fermentation processes.
Prog. Ind. Micro. 12, 171-234.
by the availability of oxygen the situation is especi- HeEyNEN, J.J., Rrier, K. W. and Wotruuts, A. J. (1980) Influence
ally difficult and the minimum level of antifoam of very small bubbles on the dynamic K;a measurement in
necessary to control the fermentation should be viscous gas-liquid systems. Biotech. Bioeng. 22, 1945-1956.
HEINEKEN, F. G. (1970) Use of fast-response dissolved oxygen
used. probes for oxygen transfer studies. Biotech. Bioeng. 12, 145—
154.
HEINEKEN, F. G. (1971) Oxygen mass transfer and oxygen
REFERENCES respiration rate measurements utilising fast response oxygen
electrodes. Biotech. Bioeng. 13, 599-618.
Hirose, Y. and Surat, H. (1980) Effect ofoxygen on amino acids
Ara, S., Humpurey, A. E. and Mixus, N. (1973) Biochemical fermentation. Adv. Biotechnology, 1, 329-333 (Editors, Moo-
Engineering. Academic Press, London. Young, M., Robinson, C. W. and Vezina, C.). Pergamon
Banks, G. T. (1977) Aeration of moulds and streptomycete Press, Toronto.
culture fluids. Topics in Enzyme and Fermentation Biotechnology, Jounson, M. J. (1964) Utilisation of hydrocarbons by micro-
1, 72-110 (Editor, Wiseman, A.), Ellis Horwood, Chiches- organisms. Chem. Ind. 36, 1532-1537.
ter. Jounson, M. J. (1967) Growth of microbial cells on hydro-
Banks, G. T. (1979) Scale-up of fermentation processes. Topics in carbons. Science, 155, 1515-1519.
Enzyme and Fermentation Biotechnology, 3, 170-267 (Editor, KartinceEr, H. W. D. (1977) New fermentation configurations. In
Wiseman, A.), Ellis Horwood, Chichester. Biotechnology and Fungal Differentiation, FEMS Symp. 4, 137-
BarTHOLOoMEw, W. H. (1960) Scale-up of submerged fermenta- 156 (Editors Meyrath,J.and Bu’Lock,J. D.).
tions. Adv. App. Micro. 2, 289-300. Kosayasui, T., VAN DepEem, G. and Moo-Youne, M. (1973)
BaRTHOLOMEW, W. H., Karrow, E. O., Srat, M. R. and Oxygen transfer into mycelial pellets. Biotech. Bioeng. 15,
WitHeELy, R. H. (1950) Oxygen transfer and agitation in 27-45.
submerged fermentations. Mass transfer of oxygen in sub- Kuenz1, M. T. (1978) Process design and control in antibiotic
merged fermentations of Streptomyces griseus. Ind. Eng. Chem. fermentation. In Antibiotics and Other Secondary Metabolites,
42(9), 1801-1809. Biosynthesis and Production, FEMS Symp. 5, 39-56 (Editors,
191
Principles of Fermentation Technology
192
CHAPTER 10
1. The intracellular or extracellular location of Fic. 10.1. Isolation stages in the recovery of a soluble extra-
the product. cellular product from a harvested fermentation broth.
193
Principles of Fermentation Technology
precise chromatographic techniques and crystalli- able antifoam. The ionic strength of the production
zation. Other products are isolated using modifica- medium may be too high, resulting in the harvested
tions of this flow-stream. supernatant needing dilution with demineralized
Attempts to simplify this outline extraction pro- water before it can be processed. Such a negative
cedure for antibiotic recovery using ‘whole broth’ procedure should be avoided if possible by unified
processing have met with limited success. The research and development programmes. Media
technique of ‘whole broth’ processing involves initial formulation is dominated by production require-
removal of large particles, which is then followed by ments, but the protein content of complex media
passage of the broth (including cells) through well- should be critically examined in view of subsequent
mixed ion-exchange columns or counter-current sol- enzyme recovery.
vent extraction units to extract the antibiotic This view is also shared by Topiwala and
directly. This topic will be discussed in more detail Khosrovi (1978), when considering water recycle in
in a later section of this chapter. biomass production. They stated that the inter-
It may be possible to modify the handling charac- action between the different unit operations in a
teristics of the broth so that it can be handled faster recycle process made it imperative that commercial
with simpler equipment making use of anumber of plant design and operation should be viewed in an
techniques: integrated fashion.
Flow sheets for recovery of penicillin, citric acid
1. Selection of amicro-organism which does not
and micrococcal nuclease are given in Figs. 10.2,
produce pigments or undesirable metabolites.
10.3 and 10.4, to illustrate the range of techniques
2. Modification of the fermentation conditions to
which are used in microbiological recovery pro-
reduce the production of undesirable meta-
cesses. A series of comprehensive flow sheets for
bolites.
citric acid, dextran, ethanol, intracellular and
Precise timing of harvesting.
extracellular enzymes, penicillin, vitamin B,, and
pH control after harvesting.
Temperature treatment after harvesting.
Addition offlocculating agents. Harvest broth from fermenter
194
The Recovery and Purification of Fermentation Products
{ COLLECT RESIN
pack into an adjustable column
Activated charcoal to decolourize
remove unadsorbed protein by passing 0.3 M ammonium acetate pH 6.0
t |
Cation and anion exchange resins ELUTION
{ DIALYSIS
overnight against demin. HO
Crystals of citric monohydrate separated in continuous centrifuges
|
| CONCENTRATION
CH, hollow fibre unit (Amicon)
Driers at 50—60°
mol. wt. cut off 10,000
tein (Hamer, 1979) and polysaccharides (Pace and Fic. 10.4. Purification of micrococcal nuclease (Darbyshire,
Righelato, 1980; Smith and Pace, 1982). 1981).
FOAM SEPARATION
REMOVAL OF MICROBIAL CELLS AND
OTHER SOLID MATTER
Foam separation depends on using methods
which exploit differences in surface activity of
Microbial cells and other insoluble materials are materials. The material may be whole cells, molecu-
normally separated from the harvested broth by lar such as a protein or colloidal, and is selectively
filtration or centrifugation. Because ofthe small size adsorbed or attached to the surface of gas bubbles
of many microbial cells it will be necessary to con- rising through a liquid, to be concentrated or sepa-
sider the use offilter aids to improve filtration rates rated and finally removed by skimming (Fig. 10.5).
while heat and flocculation treatments are employed It may be possible to make some materials surface
as techniques for increasing sedimentation rates in active by the application of surfactants such as fatty
centrifugation. Foam separation as a method for cell acids, amines and quaternary ammonium com-
recovery is not widely used. pounds. Materials made surface active and collected
195
PFI-N
Principles of Fermentation Technology
Fic. 10.5. Schematic flow diagram for foam fractionation (Wang FILTRATION
and Sinskey, 1970)
>?
Slurry
Se
Unfortunately s and & are not determined rapidly.
In most practical cases L is not readily measured
Filter cake but can be defined in terms of:
Filter cloth
V = volume offiltrate passed in time ¢
Support to
filter cloth and
Filtrate
v = volume of cake deposited by unit volume of
filtrate.
vV
Then L=—.
A
Fic. 10.6. Diagram ofa simple filtration apparatus.
Substituting in equation (10.1)
dV_ KA*Ap
as filtrate passes through the filter cloth. As the filter (10.2)
dt pvV
cake builds up in thickness the resistance to flow will
gradually increase. Thus, if the pressure applied to This is a general equation relating rate of filtration
the surface of the slurry is kept constant the rate of to pressure drop, cross-sectional area of the filter
flow will gradually diminish. Alternatively, if the and filtrate retained. Equation (10.2) can be inte-
-flow rate is to be kept constant the pressure will grated for filtration at constant pressure.
gradually have to be increased. The reduction in 2
flow rate may also be decreased by blocking of holes VENER (10.3)
pv
in the filter cloth and closure of voids between
particles, if the particles are soft and compressible. Integrating equation (10.3):
When particles are compressible it may not be 2
feasible to apply increased pressure. pe ens
BV
(10.4)
Flow through a uniform and constant depth por-
ous bed can be represented by the Darcy equation: Now in equation (10.4), Apis constant, wis generally
equal to 1, v can be determined by laboratory
A investigation and A” remains constant. Thus there is
Rate of flow = ge foals 10.1
dt BL ( ) a linear relationship between V? and ¢. By carrying
out small-scale filtration trials it is therefore possible
where = liquid viscosity,
to obtain a value for K. It is then possible to reapply
L = depth ofthe filter bed,
the equation for large-scale filtration calculations.
Ap = pressure differential across the filter
Although it is also possible to derive the equation
bed, for the pressure which is necessary to maintain a
A = areaof the filter exposed to the liquid,
constant filtration rate, it has little practical applica-
K = constant for the system. tion. The pressure is made up of two components.
K itself is a term which depends on the specific Firstly the pressure needed to pass the constant
surface area s (surface area/unit volume) of the volume through the filter resistance and secondly an
particles making up the filter bed and the voidage >» increasing pressure component which is propor-
when they are packed together. The voidage is the tional to the resistance from the increasing cake
amount of filter-bed area which is free for the filtrate depth. This filtration procedure would be complex
to pass through. It is normally 0.3 to 0.6 of the to perform practically, and other methods of filtra-
cross-sectional area of the filter bed. Thus K tion are used to achieve constant flow rates, e.g.
(Kozeny’s constant) can be expressed as vacuum drum filters.
197
Principles of Fermentation Technology
ee
widely used material. It has a voidage of approxi- Ans'
i
Geetha
SEU
mately 0.85, and improves the porosity ofa resulting ro
eek aio
filter cake leading to a faster flow rate if it is mixed ah
es 2
Ua
NES
te
‘
e e yp
199
Principles of Fermentation Technology
Direction
of ee é Stationary automatic
valve ring
Filtered cake
Vent to
atmosphere
Strings returning Comb
to drum
Cake discharge
Level of material Fic. 10.10. Cake discharge on a drum filter using strings (Talcott
to be filtered et al., 1980).
200
The Recovery and Purification of Fermentation Products
where v = rateofsedimentation(cms_'),
problem is overcome by precoating the drum witha d = particle diameter (cm),
layer of filter-aid 2 to 10 cm thick. The cake which w = angular velocity about the axis (radians
builds up on the drum during operation, is cut away sah
by the knife blade (Fig. 10.12) which mechanically r = distance of the particle from the axis of
rotation (cm),
advances towards the drum at a controlled slow
» — density ofparticle (g emi)
rate. Alternatively the blade may be operated manu-
P; = density ofliquid (g cm),
ally when there is an indication of‘blinding’ which
may be apparent from a reduction in the filtration = viscosity of liquid (g cm! second” ').
rate. In either case the cake is removed together with It is evident from this formula that the factors
a very thin layer of precoat. influencing the rate of sedimentation are the differ-
A study ofprecoat drum filtration has been made ence in density between the cells and the liquid, the
by Bell and Hutto (1958). The operating variables diameter of the cells and the viscosity of the liquid.
studied included drum speed, extent of drum sub- Ideally, the cells should have a large diameter, there
mergence, knife advance speed and applied vacuum. should be a large density difference between cell and
The work indicated that optimization for a new liquid and the liquid should have a low viscosity. In
process might require prolonged trials. practice, the cells are usually very small, of low
density and often suspended in viscous media.
THE CENTRIFUGE
Cell aggregation and flocculation
Micro-organisms and other similar sized particles
can be removed from a broth by using a centrifuge It is well known that aggregates of microbial cells,
when filtration is not a satisfactory separation although they have the same density as the indi-
method. Although a centrifuge may be expensive vidual cells, will sediment faster because of the
when compared with a filter it may be essential increased diameter of the particles (Stokes law).
when: This sedimentation process may be achieved natur-
ally with selected strains of brewing yeasts, particu-
1. Filtration is slow and difficult. larly if the wort is chilled at the end of fermentation,
2. The cells or other suspended matter must be and leads to a natural clearing of the beer. A number
obtained free offilter aids. of other factors besides temperature can influence
3. Continuous separation to a high standard of cell flocculation, which is due to the neutralization
hygiene is required. of anionic charges, primarily carboxyl and phos-
Non-continuous centrifuges are of extremely limited phate groups, on the surfaces of the microbial cells.
capacity and therefore not suitable for large-scale These include changes in the pH and the presence of
separation. The centrifuges used in harvesting fer- a range of compounds which alter the ionic environ-
201
Principles of Fermentation Technology
Feed
ment. Nakamura (1961) described the use of various
compounds for flocculating bacteria, yeasts and
algae, including alum, calcium and ferric salts.
Other agents which are now used include tannic
<——
Porous
acid, titanium tetrachloride and cationic agents | lining
pi
such as quaternary ammonium compounds, alkyl | Filter
amines and alkyl pyridinium salts. Gasner and = | cake
—_
Wang (1970) reported a many hundredfold increase
in the sedimentation rate of Candida intermedia when =
recoveries of over 99% were readily obtained. They |___— Perforated
—_=
HohlLU
continuous feed is used, and when the basket is filled
with the filter cake it is possible to wash the cake
before removing it. The bowl may suffer from blind-
ing with soft biological materials so that high cen- Fic. 10.14. L.S. ofa multichamber centrifuge.
202
The Recovery and Purification of Fermentation Products
Slurry in
Liquids
discharge
Poy a Solids
discharge
Fic. 10.15a. Diagram ofa solid-bow] scroll centrifuge (Pennwalt Ltd., Camberley, England).
Fic. 10.15b. Cutaway view ofa Sharples Super-D-Canter continuous solid-bowl centrifuge, Model P-5400 (Pennwalt Ltd., Camberley,
England).
chamber. The smaller particles collect in the outer the solids fraction, the size and total number of
chambers where they are subjected to greater gravi- vessels must be of the correct volume for the solids of
tational forces. Although the vessels can have a a batch run.
greater solids capacity than tubular bowls and there
is no loss of efficiency as the bow! fills with solids, SOLID-BOWL SCROLL CENTRIFUGE
their mechanical strength and design limits their (DECANTER CENTRIFUGE)
speed to a maximum of6500 rpm for a rotor of46cm This type of centrifuge is used for continuously
diameter with a holding capacity of up to 76 dm’. handling coarse materials such as sewage sludge
Because ofthe time needed to dismantle and recover (Fig. 10.15). The slurry is fed through the spindle of
203
Principles of Fermentation Technology
Feed
Effluent
r~
| G
, |
ow
T G
Nozzle for
discharge
’
:
|
r 4 yo eek
|
|
'
of solids Fic. 10.16b. L.S. of disc-bowl centrifuge with intermittent dis-
charge. (Solids discharged when rotor opens intermittently along
Fic. 10.16a. L.S. of disc-bowl centrifuge with nozzle discharge. the section C—C,.)
an archimedean screw within the horizontal rotating centrifuges is the facility to remove solids automati-
solids bowl. The solids settling on the walls of the cally through a series of nozzles in the circumference
bowl are scraped to the conical end of the bowl. The of the rotor or by opening up the rotor in the
slope of the cone helps to remove excess liquid from intermittent discharge design. Unfortunately, the
the filter cake before discharge. The liquid phase is solid fraction is not completely dewatered and the
discharged from the other end of the bowl. The arrangement of the discs makes this type of cen-
speed of this type of centrifuge is limited to 2000 to trifuge laborious to clean. Feed rates range from 45
4000 rpm because of the lack of balance within the to 1800 dm? min“.
bowl. The largest sizes may have a throughput for
200,000 dm? h7! ofliquid and process 40 tonnes h!
of solids. THE TUBULAR-BOWL CENTRIFUGE
This is a centrifuge to consider using for particle
size ranges of 0.1 to 200 wm and up to 10% solids in
DISC-BOWL CENTRIFUGE the in-going slurry. The main component of the
This centrifuge relies for its efficiency on the centrifuge is a cylindrical bowl (or rotor) A, which
presence ofdiscs in the rotor (Fig. 10.16). A central may be of a variable design depending on applica-
inlet pipe is surrounded by a stack of stainless-steel tion, suspended by a flexible shaft B driven by an
conical discs. Each disc has spacers so that a stack overhead motor or air turbine C. Figure 10.17a
can be built up. The broth to be separated flows shows an arrangement used in a Sharples Super
outwards from the central feed pipe, then upwards centrifuge. The inlet to the bowl is via a nozzle
and inwards at an angle of 45° to the axis of rotation. attached to the bottom bearing D. The feed which
The close packing ofthe cones assists rapid sedimen- consists of solids and light and heavy liquid phases
tation and the solids then slide to the edge of the is introduced by the nozzle E. During operation
bowl, provided there are no gums or fats in the solids sediment on the bowl wall while the liquids
slurry, and eventually accumulates on the inner separate into the heavy phase in the zone G and the
wall of the rotor. Ideally, the sediment should form light phase in the central zone H. The two liquid
a sludge which flows rather than a hard particulate phases are kept separate in their exit from the bowl
or lumpy sediment. The main advantage of these by an adjustable ring. The heavy phase flows over
204
The Recovery and Purification of Fermentation Products
f
Fic. 10.17a. Section of a Sharples Super-Centrifuge (Pennwalt
Ltd., Camberley, England).
205
Principles of Fermentation Technology
CELL DISRUPTION
206
The Recovery and Purification of Fermentation Products
®@® OO
SA lI
rl
ESS CLI Rae
NUTRALELRASALCARRLUETE
207
Principles of Fermentation Technology
concentration will cause disruption of anumber of of molecules. Polar liquids mix with each other and
cell types. However, the effect on microbial cell dissolve salts and other polar solids. The solvents for
types is normally minimal. It has proved to be a non-polar compounds are liquids of low or nil
successful technique for the extraction of luciferase polarity.
from Photobacterium fischeri (Hastings, Riley and The dielectric constant is a measure of the degree
Massa, 1965). A batch of 120 dm? of broth was of molar polarization of acompound. Ifthis value is
harvested and the cells collected as a cell paste in a known it is then possible to predict if a compound
Sharples centrifuge. Enzyme extraction was will be polar or non-polar. The dielectric constant D
achieved by osmotic lysis using a ratio of | g of cell of asubstance can be measured by determining the
paste to 4.cm? of cold distilled water with stirring for electrostatic capacity C of a condenser containing
15 to 30 minutes. A second extraction gave a small the substance between the plates. If Co is the value
yield of additional enzyme. Only low levels of for the same condenser when completely evacuated
soluble protein were released using this technique. then
Although the method is probably one of the most The value of D, can be calculated since C,; and C,
gentle which is available, unfortunately it is rela- can be measured and D, is known. The dielectric
tively expensive and the presence of the enzyme(s) constants for a number ofsolvents are given in Table
complicates a purification process. 10.1.
LIQUID-LIQUID EXTRACTION
TasLe 10.1. Dielectric constants of solvents at 25° (arranged
in order of increasing polarity)
The separation of a component from a liquid Solvent Dielectric constant
Ste 2 ee
mixture by treatment with a solvent in which the eee
208
The Recovery and Purification of Fermentation Products
Pee
Mixer Raffinate Mixer Raffinate Mixer Raffinate
separator separator separator
1 n
Solvent
Solvent Solvent
209
Principles of Fermentation Technology
Depleted
Raffinate Raffinate raffinate
Mixer Mixer
separator separator separator
1 2 n =— Solvent
Solvent enriched
with product
shaft. The two liquid streams are then discharged In acid conditions this ionization is suppressed and
via the shaft. Flow rates in excess of 100,000 dm*h7! the penicillin is more soluble in organic solvents. At
are possible in the largest models. Probably the most pH 2 to 3 the distribution ratio of total acid will be
useful property of this type of extractor is the low-
wt (RCOOH) org
hold up volume of liquid in the machine compared
with the throughput. (RCOOH)aq + (RCOO’)aq_
Penicillin G is an antibiotic which is recovered For penicillin this value may be as high as 40 ina
from fermentation broths by centrifugal counter- suitable solvent (Podbielniak et al., 1970).
current solvent extraction. In neutral pHs in water, The penicillin extraction process may involve the
penicillin is ionized: - four following stages:
Feed tubes
Mechanical
Large
ascotube
Mechanical
seals
Small
ascotube
Drain
Fic. 10.24. Diagram of the Podbielniak extractor (Queener and Swartz, 1979). HLI, LLI, HLO and LLO indicate heavy and light
liquid in and out.
210
The Recovery and Purification of Fermentation Products
3. Extraction from aqueous buffer into solvent. be recycled through the system. In some processes
4. Extraction of the solvent to obtain the penicil- the more difficult problem will be to remove all the
lin salt. solvent from the raffinate because ofthe value of the
solvent and problems which might arise from con-
At each extraction stage progressively smaller
tamination ofthe product.
volumes are used to achieve concentration of the
Distillation may be achieved in three stages:
penicillin (see also Fig. 10.2).
Unfortunately, penicillin G has a half-life of 15 1. Evaporation, the removal of solvent as a vap-
minutes at pH 2.0 at 20°. The harvested broth is our from a solution.
therefore initially cooled to 0° to 3°. The cooled 2. Vapour-—liquid separation in a column, to
broth is acidified to pH 2 to 3 with sulphuric or separate the lower boiling more volatile com-
phosphoric acid immediately before extraction. ponent from other less volatile components.
This acidified broth is quickly passed through a 3. Condensation of the vapour, to recover the
Podbielniak centrifugal counter-current extractor more volatile solvent fraction.
using about 20% by volume of the solvent in the
Evaporation is the removal of solvent from a
counter flow. Ideally the hold-up time should be
solution by the application of heat to the solution.
about 60 to 90 seconds. Enough aqueous NaOH or
There are a wide range of evaporators available.
KOH (about 20% by volume) is now added as the
Some are operated on a batch basis and others
penicillin-rich solvent passes through a second Pod-
continuously. Most industrial evaporators employ
bielniak so that the penicillin is removed to the
tubular heating surfaces. Circulation of the liquid
aqueous phase (pH 7.0 to 8.0) as the salt.
past the heating surfaces may be induced by boiling
RCOOH (organic phase) —> RCOO’Na* + H,O or by mechanical agitation.
In batch distillation (Fig. 10.25) the vapour from
These two stages may be sufficient to concentrate
the penicillin adequately from a broth with a high
titre. Cooling water
outlet
Penicillin will crystallize out of aqueous solution Vapour
||Condenser
at a concentration of approximately 1.5 X 10° units
UTE
cm°. If the initial harvested broth contains 60,000
!)
Cooling water
inlet
units cm°, and two five fold concentrations are b> D
Tray or Reflux
achieved in the two extraction stages then the
perforated
penicillin liquor should crystallize. If the initial plate
broth titre is lower than 60,000 units cm? or the
extractions are not so effective, the solvent and
buffer extractions will have to be repeated. Distillatio
At each stage the spent liquids should be checked column
1st 2nd nth
for residual penicillin and solvent usage carefully fraction fraction fraction
ent Mixture to
J be distilled
SOLVENT RECOVERY Heating
outlet
211
PFTI-O
Principles of Fermentation Technology
Chromatographic techniques are also used in the cationic or anionic exchange properties. Cationic
final stages
of purification of anumber ofproducts. ion-exchange resins normally contain a sulphonic
acid, carboxylic acid or phosphonic acid active
group. Anionic ion-exchange resins normally con-
Adsorption chromatography tain a secondary amine, quaternary amine or
quaternary ammonium active group. The appro-
priate resin for a particular purpose will depend on
Adsorption chromatography involves binding of
various factors such as bead size, pore size, diffusion
the solute to the solid phase primarily by weak Van
rate, resin capacity, range of reactive groups and the
de Waals forces. The materials used for this purpose
life of the resin before replacement. Weak-acid
to pack columns include inorganic adsorbants
cationic ion-exchange resins can be used in the
(active carbon, aluminium oxide, aluminium
isolation and purification of streptomycin, neomy-
hydroxide, magnesium oxide, silica gel) and organic
cin and similar antibiotics.
macro-porous resins.
In the recovery of streptomycin, the harvested
Di-hydro-streptomycin can be extracted from
filtrate is fed on to a column of a weak-acid cationic
filtrates using activated charcoal columns. It is then
resin such as Amberlite IRC 50 which is in the
eluted with methanolic hydrochloric acid and
sodium form. The streptomycin is adsorbed on to
purified in further stages (Nakazawa et al., 1960).
the column and the sodium ions are displaced.
Some other applications for small-scale antibiotic
purification are quoted by Weinstein and Wagman R—COO'Na’* + streptomycin
(1978). Active carbon may be used to remove pig- (resin)
ments to clarify broths. Penicillin-containing sol- — R—COO’ streptomycin* + NaOH
vents may be treated with 0.25 to 0.5% active (resin)
‘carbon to remove pigments and other impurities
Flow rates of between 10 and 30 bed volumes per
(Sylvester and Coghill, 1954).
hour have been used. The resin bed is then rinsed
Macro-porous adsorbants have also been tested.
with water and eluted with dilute hydrochloric acid
The first synthetic organic macro-porous adsor-
to release the bound streptomycin.
bants, the Amberlite XAD resins, were produced by
Rohm and Haas in 1965. These resins have surface R—COO’' streptomycin* + HCl
polarities which vary from non-polar to highly polar (resin)
and do not possess any ionic functional groups. — R—COOH + streptomycin*™ Cl’
Voser (1982) considers their most interesting appli- (resin)
cation to be in the isolation of hydrophilic fermenta-
A slow flow is used to ensure the highest recovery of
tion products. He stated that these resins would be
streptomycin using the smallest volume of eluant.
used at Ciba-Geigy in recovery of cephalosporin C
In one step the antibiotic has been both purified and
(acidic amino acid), cefotiam (basic amino acid),
may be concentrated more than 100-fold.
desferrioxamine B (basic hydroxamic acid) and
The resin column is regenerated to the sodium
paramethasone (neutral steroid).
form by passing an adequate volume of NaOH
slowly through the column and rinsing with distilled
water to remove excess sodium ions.
Ion exchange
R—COOH + NaOH > R—COO’ Na* + H,O
(resin) (resin)
Ion exchange can be defined as the reversible
exchange of ions between a liquid phase and a solid The resin can have a capacity of | g of streptomycin
phase which is not accompanied by any radical g 'resin. Commercially it is not economic to regen-
change in the solid structure. In the case of organic erate the resin completely, therefore the capacity
ion exchange resins, the main component is a di- will be reduced.
vinyl styrene co-polymer to which are attached the In practice the filtered broth is taken through two
reactive groups which give an individual resin its columns in series while a third is being eluted and
213
;
Principles of Fermentation Technology
Gel filtration
De
' ()
more rapidly than the larger ones, and penetrate the
pores of the gel to a greater degree. This means that Rotation Individual
solutes out
once elution is started, the larger molecules which
are still in the spaces in the gel will be eluted first. Fic. 10.27. The principle of continuous-partition chromato-
graphy. -——-, faster-moving component; © 0 00, slower-mov-
One early industrial application, although it is on
ing component (Fox, 1969).
a relatively small scale, is the purification of vaccines
(Latham et al., 1967). Tetanus and diphtheria
broths for batches of up to 100,000 human doses are are agarose, cellulose, dextrose and polyacrylamide.
passed through a 13-dm* column ofG 100 followed Agarose activated with cyanogen bromide is one of
by a 13-dm? column of G 200. This technique the most commonly used supports for the coupling
yielded a fairly pure fraction which was then concen- _of amino ligands.
trated ten-fold by pressure dialysis to remove the Purification may be several thousand-fold with
eluant buffer (NayHPO,). good recovery of active material. Affinity
chromatography was initially used in protein isola-
tion and purification, particularly enzymes. Since
Affinity chromatography then many other large-scale applications have been
developed for enzyme inhibitors, antibodies and
interferon (Janson and Hedman, 1982), and on a
Affinity chromatography is potentially a separa-
smaller scale for nucleic acids, cell organelles and
tion technique with many applications since it is
whole cells (Yang and Tsao, 1982).
possible to use it for separation and purification of
most biological molecules on the basis of their func-
tion or chemical structure. This technique depends
on the highly specific interactions between pairs of Continuous chromatography
biological materials such as enzyme-substrate,
enzyme-inhibitor, antigen—antibody, etc. The Although the concept of continuous enzyme isola-
molecule to be purified is specifically adsorbed by a tion is well established (Dunnill and Lilly, 1972),
binding substance (ligand) which is immobilized on the stage of least development is continuous
an insoluble support (matrix). The ligand is chromatography. Fox et al. (1969) developed a con-
attached to the matrix by physical absorption or tinuous-fed column for this purpose (Fig. 10.27). It
chemically by a covalent bond. The latter method is consisted of two concentric cylindrical sections
preferred whenever possible. Porath (1974) and clamped to a base plate. The space (1 cm wide)
Yang and Tsao (1982) have reviewed methods and between the two sections was packed with the
coupling procedures. appropriate resin or gel giving a total column capac-
Coupling procedures have been developed using ity of 2.58 dm’. A series oforifaces in the circumfer-
cyanogen bromide, bisoxiranes, disaziridines and ence of the base plate below the column space led to
periodates, for matrixes of gels and beads. Four collecting vessels. The column assembly was rotated
polymers which are often used for matrix materials on a slow-moving turntable (0.4-2.0 rpm). The mix-
214
The Recovery and Purification of Fermentation Products
ULTRA FILTRATION
brane of a very fine pore size. A range of membranes Fic. 10.28. Counter-current spray drier.
with different molecular weight cut-offs (500 to
500,000) are available which makes possible the
is often the last stage of a manufacturing process
separation of macro-molecules such as proteins,
(Coulson and Richardson, !968; Dombrowski and
enzymes, hormones and viruses.
Munday, 1968; Forrest, 1968). It involves the final
When considering the feasibility of ultra filtration
removal of water from a heat-sensitive material
it is important to remember that factors other than
ensuring that there is minimum loss in viability,
the molecular weight of the solute affect the passage
activity or nutritional value. Drying is undertaken
of molecules through the membranes (Melling and
because:
Westmacott, 1972). There may be concentration
polarization caused by accumulation ofsolute at the (a) The cost of transport can be reduced.
membrane surface, which can be reduced by good (b) The material is easier to handle and package.
agitation. Secondly, a slurry of protein may accumu- (c) The material can be more conveniently stored
late on a membrane surface which is not easy to in the dry state.
remove by agitation. Finally, equipment and energy
It is important that as much water as possible is
costs may be considerable because of the high pres-
initially removed by centrifugation or in a filter
sures which are necessary (70,000 to 700,000 kg
press to minimize heating costs in the drying
m ”), which also limits the life of the membranes.
process.
Nevertheless, ultra filtration is still a developing
A spray drier (Fig. 10.28) is most widely used for
technique. Details of some large-scale applications
drying of biological materials when the starting
are given by Lacey and Loeb (1972).
material is in the form ofa liquid or paste which can
be initially atomized into small droplets through a
DRYING nozzle or by contact with a rotating disc. The drop-
lets may fall into a spiral stream ofhot gas at 150° to
The drying of micro-organisms or their products 250°. The high surface area:volume ratio of the
215
Principles of Fermentation Technology
— Scraper
blade WHOLE BROTH PROCESSING
f] Steam-
| heated
drum The concept of recovering a metabolite directly
Product \ from an unfiltered fermentation broth is of consider-
able interest because ofits simplicity, the reduction
in process stages and the potential cost savings. It
Fic. 10.29. Cross-section of adrum drier.
may also be possible to remove the desired fermenta-
tion product continuously from a broth during fer-
droplets results in a rapid rate of evaporation and
mentation so that inhibitory effects due to product
complete drying in a few seconds. The evaporative
formation and product degradation can _ be
cooling effect prevents the material from becoming
minimized throughout the production phase.
overheated and damaged. The gas-flow rate must
Bartels et al. (1958) developed a process for
be carefully regulated so that the gas has the capac-
_ adsorption of streptomycin on to a series of cationic
ity to contain the required moisture content at the
ion-exchange resin columns directly from the fer-
cool-air exhaust temperature (75° to 100°).
mentation broth, which had only been screened to
When the material is more heat stable, a drum
remove large particles so that the columns would
dryer may be used (Fig. 10.29). A solution is run on
not become blocked. This procedure could only be
to a steam-heated drum which is slowly rotating,
used as a batch process. Belter et al. (1973)
evaporation takes place, and the dry solids can be
developed a similar process for the recovery of
continuously scraped off the drum with a knife
novobiocin. The harvested broth was first filtered
blade.
through a vibrating screen to remove large particles.
The broth was then fed into a continuous series of
CRYSTALLIZATION well-mixed resin columns fitted with screens to
retain the resin particles, plus the absorbed
novobiocin, but allow the streptomycete filaments
Crystallization is an established method used in plus other small particulate matter to pass through.
the initial recovery of organic acids and amino acids, The first resin column was removed from the extrac-
and more widely used for final purification of a tion line after a predetermined time and eluted with
diverse range of compounds. methanolic ammonium chloride to recover the
In citric acid production, the filtered broth is novobiocin.
treated with Ca(OH), so that the relatively insol- Karr et al. (1980) developed a reciprocating plate
uble calcium citrate crystals will be precipitated extraction column (Fig. 10.30) to use for whole
from solution. Checks are made to ensure that the broth processing ofabroth containing 1.4 g dm7? of
Ca(OH), has a low magnesium content since mag- a slightly soluble organic compound and 4% undis-
nesium Citrate is more soluble and would remain in solved solids provided chloroform or methylene
solution. The calcium citrate is filtered off and chloride were used for extraction. Methyl-iso-buty]
treated with sulphuric acid to precipitate the cal- ketone, di-ethyl ketone and iso-propyl acetate were
cium as the insoluble sulphate and release the citric shown to be more efficient solvents than chloroform
acid. After clarification with active carbon, the for extracting the active compound, but they pre-
aqueous citric acid is evaporated to the point of sented problems since they extracted impurities
216
The Recovery and Purification of Fermentation Products
1.49 kW drive
217
Principles of Fermentation Technology
Arxinson, B. and Maviruna, F. (1983) Biochemical Engineering whole fermentation broth with a Karr reciprocating plate
and Biotechnology Handbook, Chapters 12 and 13. Macmillan, extraction column. Can. Jour. Chem. Eng. 58, 249-252.
London. Kominek, L. A. (1975) Cycloheximide production by Streptomyces
Arxinson, B. and Sarnter, P. (1982) Development of down- griseus; alleviation of end-product inhibition by dialysis-
stream processing.J.Chem. Tech. Biotechnol. 32, 100-108. extraction fermentation. Antimicrob. Chemotherap. 7, 856-860.
Aunstrup, K. (1979) Production, isolation and economics of Kroner, K. H., Husrepr, S., Granpa, S. and Kura, M. R.
extracellular enzymes. In Applied Biochemistry and Bioengineer- (1978) Technical aspects of separation using aqueous two-
ing, Vol. 2, pp. 27-69 (Editors Wingard, L. B. and Katzir- phase systems in enzyme isolation processes. Biotech. Bioeng.
Katchalski, E.). Academic Press, New York. 20, 1967-1988.
Bartets, C. R., Kreman, G., Korzun,J. N. and Irisu, D. B. Lacey, R. E. and Logs, S. (1972) Industrial Processing with Mem-
(1958) A novel ion-exchange method for the isolation of branes. Wiley-Interscience, New York.
streptomycin. Chem. Eng. Prog. 54, 49-51. Latuam, W. C., Micuersen, C. B. and Epsatz, G. (1967)
BELL, G. B. and Hurvo, F. B. (1958) Analysis of rotary pre-coat Preparative procedure for the purification of toxoids by gel
filter operation. 1. New concepts. Chem. Eng. Prog. 54, 69-76. filtration. Appl. Microbiol. 15, 616-621.
Better, P. A., Cunnincuam, F. L. and Cuen, J. W. (1973) Locxwoop, L. B. and Irwin, W. E. (1964) Citric acid. In
Development of a recovery process for novobiocin. Biotech. Kirk-Othmer Encyclopedia of Chemical Technology (2nd edition),
Bioeng. 15, 533-549. Vol. 5, pp. 524-541. Wiley, New York.
Brooxman,J. S. G. (1974) Mechanism ofcell disintegration in a: Macnusson, K. E. and Epeso, L. (1976) Large-scale disintegra-
high pressure homogenizer. Biotech. Bioeng. 16, 371-383. tion of micro-organisms by freeze-pressing. Biotech. Bioeng.
Coutson,J.M. and Ricuaropson,J. F. (1968) Chemical Engineering 18, 975-986.
(2nd edition), Vol. 2. Pergamon Press, Oxford. Me tine, J. and Wesrmacortt, D. (1972) The influence of pH
DarsysuirE,J. (1981) Large scale enzyme extraction and recov- value and ionic strength on the ultrafiltration characteristics
ery. In Topics in Enzyme and Fermentation Biotechnology, Vol. 5, ofa penicillinase produced by Escherichia coli strain W3310.
pp. 147-186 (Editor Wiseman, A.). Ellis Horwood, Chiches- J. Appl. Chem. 22, 951-958.
ter. Mixier, S. A. et al. (1973) Liquid-solid systems. In Chemical
Domsrowski, N. and Munpay, G. (1968) Spray drying. In Engineers Handbook (5th edition), pp. 19-78 (Editors Perry,
Biochemical and Biological Engineering Science, Vol. 2, pp. 209— R. H. and Chilton, C. H.). McGraw-Hill Kogakusha,
320 (Editor Blakebrough, N.). Academic Press, London. Tokyo.
Dunnitt, P. and Litty, M. D. (1972) Continuous enzyme . Mocren, H., Linpsiom, M. and HEpEnskoe, G. (1974) Mechan-
isolation. Biotech. Bioeng. Symp. 3, 97-113. ical disintegration of micro-organisms in an industrial
Farry, W. T., Neuseck, C. E. and Reese, E. T. (1971) Produc- momogenizer. Biotech. Bioeng. 16, 261-274.
tion and applications of enzymes. Adv. Biochem. Eng. 1, Nakamura, H. (1961) Chemical separation methods for common
77-111. microbes.J.Biochem. Microbiol. Technol. Eng. 3, 395-403.
Forrest,J. C. (1968) Drying processes. In Biochemical and Biolog- Nakazawa, K., Surpata, M., Tanase, K. and Yamamoto, H.
ical Engineering Science, Vol. 2, pp. 97-135 (Editor Blake- (1960) Di-hydrostreptomycin. U.S. patent 2,931,756.
brough, N.). Academic Press, London. Nicuotas, R. A. and Fox, J. B. (1969) Continuous chromato-
Fox,J.B., Garnoun, R. C. and Eciinton, W.J. (1969) Continu- graphy apparatus. II. Application.J.Chromatog. 43, 61-65.
ous chromatography apparatus. I. Construction. /. Pace, G. W. and RicHeLato, R. H. (1980) Production of
Chromatog. 43, 48-54. extracellular polysaccharides. Adv. Biochem. Eng. 15, 41-70.
Gasn_er, C. C. and Wang, D. I. C. (1970) Microbial recovery Pace, G. W. and Smiru, I. H. (1981) Recovery of microbial
enhancement through flocculation. Biotech. Bioeng. 12, 873— polysaccharides. In Abstracts of Communications. Second Euro-
887. pean Congress of Biotechnology, Eastbourne, p. 36. Society of
Grirves, R. B. and Wana, S. L. (1966) Foam separation of Chemical Industry, London.
Escherichia coli with a cationic surfactant. Biotech. Bioeng. 8, PopsieLniaAk, W.J., Katser, H. R. and Z1eGenHoRN, G. J. (1970)
323-336. Centrifugal solvent extraction. In The History of Penicillin
Hamer, G. (1979) Biomass from natural gas. In Economic Micro- Production. Chem. Eng. Prog. Symp. 66 (100), 44-50.
biology, Vol. 4, pp. 315-360 (Editor Rose, A. H.). Academic Poratu, J. (1974) Affinity techniques—General methods and
Press, London. coupling procedures. In Methods in Enzymology, Vol. 34,
Hastines, J. W., Ritey, W. H. and Massa, J. (1965) The pp. 13-30 (Editors Jakoby, W. B. and Wilchek, M.).
purification, properties and chemiluminescent quantum Academic Press, New York.
yield ofbacterial luciferase.J.Biol. Chem. 240, 1473-1481. Purcuas, D. B. (1971) Industrial Filtration ofLiquids (2nd edition).
HETHERINGTON, P.J., FoLtows, M., Dunuixt, P. and Litty, M. Leonard Hill, London.
D. (1971) Release of protein from baker’s yeast (Saccharomyces QUEENER, S. and Swartz, R. W. (1979) Penicillins: biosynthetic
cerevisiae) by disruption in an industrial homogenizer. Trans. and semisynthetic. In Secondary Products of Metabolism,
Ind. Chem. Eng. 49, 142-148. Economic Microbiology, Vol. 3, pp. 35-122 (Editor Rose, A.
Honic, W. and Kuta, M. R. (1976) Selectivity of protein precipi- H.). Academic Press, London.
tation with polyethylene glycol fractions of various molecu- Rewacek, J. and Scuaerer, J. (1977) Disintegration of micro-
lar weights. Anal. Biochem. 72, 502-512. organisms in an industrial horizontal mill of novel design.
Hucues, D. E., Wimpenny,J.W. T. and Lioyp, D. (1971) The Biotech. Bioeng. 19, 1523-1534.
disintegration of micro-organisms. In Methods in Microbiology, Rusin, A. J., Gasser, E. A., HENDERSON, O., Jounson,J.D. and
Vol. 5B, pp. 1-54 (Editors Norris, J. R. and Ribbons, Lamp, J. C. (1966) Microflotation: new low gas-flow rate
D. W.). Academic Press, London. foam separation technique for bacteria and algae. Biotech.
Janson, J. C. and Hepman, P. (1982) Large-scale chromato- Bioeng. 8, 135-151.
graphy ofproteins. Adv. Biochem. Eng. 25, 43-99. Sameymma, H. (1972) Methods for extraction and purification. In
Karr, A. E., GeBert, W. and Wane, M. (1980) Extraction of The Microbial Production of Amino Acids, pp. 227-259 (Editors
218
The Recovery and Purification of Fermentation Products
Yamada, K., Kinoshita, S., Tsunoda, T. and Aida, K.). Toriwata, H. H. and Kuosrovi, B. (1978) Water recycle in
Kodanska, Tokyo. biomass production processes. Biolech. Bioeng. 20, 73-85.
Scawen, M. D., Arxinson, A. and DarBysuireE,J. (1980) Large VoseER, W. (1982) Isolation of hydrophylic fermentation pro-
scale enzyme purification. In Applied Protein Chemistry, ducts by adsorption chromatography. J. Chem. Tech.
pp. 281-324 (Editor Grant, R. A.). Applied Sciences Pub- Biotechnol. 32, 109-118.
lishers, London. Wane, D. I. C., Cooney, C. L., Demain, A. L., Dunnixt, P.,
Smiru, I. H. and Pace, G. W. (1982) Recovery of microbial Humpnurery, A. E. and Litiy, M. D. (1979) Enzyme isola-
polysaccharides.J.Chem. Tech. Biotechnol. 32, 119-129. tion. In Fermentation and Enzyme Technology, pp. 238-310.
SopeEck, G., Mont, J., Kominex,J. and SaLzBrunn, G. (1981) Wiley, New York.
Production ofcitric acid according to the submerged fermen- Wang, D. I. C. and Sinsxey, A. J. (1970) Collection of microbial
tation. Process. Biochem. 16(6), 9-11. cells. Adv. Appl. Microbiol. 12, 121-152.
Swartz, R. R. (1979) The use of economic analyses of penicillin Wane, H. Y., Kominex, L. A. and Jost, J. L. (1981) On-line
G manufacturing costs in establishing priorities for fermen- extraction fermentation processes. In Advances in Biotechnol-
tation process improvement. Ann. Rep. Ferm. Processes, 3, ogy, Vol. 1, pp. 601-607 (Editors Moo-Young, M., Robinson,
75-110. C. W. and Vezina, C.). Pergamon Press, Toronto.
SYLVESTER, J. C. and Cocuii1, R. D. (1954) The penicillin WemnsteIn, M.J.and Wacnan, G. H. (1978) Antibiotics, Isolation,
fermentation. In Industrial Fermentations, Vol. 2, pp. 219-263 Separation and Purification. Elsevier, Amsterdam.
(Editors Underkofler, L. A. and Hickey, R. J.). Chemical Wimprnny,J.W. T. (1967) Breakage of micro-organisms. Process.
Publishing Co., New York. Biochem. 2,(7), 41-44.
Tatcorr, R. M., Wittus, C. and Freeman, M. P. (1980) Filtra- Yano, C. M. and Tsao, G. T. (1982) Affinity chromatography.
tion. In Kirk-Othmer Encyclopedia of Chemical Technology (3rd Adv. Biochem. Eng. 15, 19-42.
edition), Vol. 10, pp. 284-337. Wiley, New York.
219
CHAPTER 11
Effluent Treatment
there are adequate levels of dissolved oxygen in Taste 11.1. Factors to investigate in a factory effluent survey
rivers, lakes, reservoirs, etc., if they are to be man-
aged satisfactorily. Ideally, the oxygen concentra- Daily flow rate (does it fluctuate during the day?)
Turbidity
tion should be at least 90% ofthe saturation concen- Suspended solids—their nature
tration at the ambient temperature and salinity. It is BOD
therefore important to know how effluents contain- COD
pH range
ing soluble and particulate organic matter can influ- Temperature range
ence the dissolved oxygen concentration. One Presence of toxic metals, free chlorine, sulphides, cyanides,
widely used method ofassessment is the ‘biochemi- phenols, ete.
Odours, tastes
cal oxygen demand’ (BOD), which is a measure of Colour
the quantity of oxygen required for the oxidation of Hardness
organic matter in water, by micro-organisms pre- Detergents
Radioactivity
sent, in a given time interval at a given temperature.
Details of the test procedure are given in books on
water testing by such bodies as the American Society
FACTORY SURVEYS
for Testing and Materials (1966). The oxygen con-
centration of the effluent, or a dilution of it, is
determined before and after incubation in the dark A complete survey of factory operations is essen-
at 20° for 5 days. The oxygen decrease can then be tial for any individual site before an economical
calculated and the results presented as mg of oxygen waste-treatment programme can be planned. It is
consumed per dm? of sample. Mineral nutrients and desirable to divide the factory into as many units as
a bacterial inoculum may be added to the initial possible, as knowledge of the various material
sample to ensure optimal growth conditions. This streams may show unexpected losses of finished
test is only an estimate of biodegradable material, product, solvent wastage, excessive use of water or
hence recalcitrant or inhibitory compounds might unnecessary contamination of water which might be
be overlooked. ? recycled in the factory. The factors, and concentra-
Because the BOD test takes 5 days it may be tions where appropriate, listed in Table 11.1 ought
necessary to resort to the ‘chemical oxygen demand’ to be known at each production rate under which
(COD), a chemical test which only takes a few hours each individual unit may operate in a representative
to complete. The test is based on treating the sample time period.
with a known amount of boiling acidic potassium The survey (Table 11.1) may indicate a need for
dichromate solution for 2'/, to 4 hours and then better control of water usage. It should be possible
titrating the excess dichromate with ferrous sulphate ~ to identify sources of uncontaminated and contami-
or ferrous ammonium sulphate. The oxidized nated water that might be reused in the factory.
organic matter is taken as being proportional to the Concentrated waste streams should be kept separate
potassium dichromate utilized. Most compounds if they contain materials that can be profitably
are oxidized virtually to completion in this test, recovered. It is also more economical to treat a
including those which are not biodegradable. In concentrate rather than a large volume of a diluter
circumstances where substances are toxic to micro- effluent because of the saving on pumps and settling-
organisms, the COD test may be the only suitable tank capacities.
method available for assessing the degree of treat- The various wastes may be tested in a laboratory
ment required. The BOD:COD ratios for sewage and on a pilot scale to assess the best potential
are normally between 0.2:1 and 0.5:1. The ratio methods of chemical and biological treatment. Once
values for domestic sewage may be fairly steady. the pHs of the effluents are known, samples may be
When industrial effluents of variable composition mixed to see ifa neutral pH is reached. A variety of
and loading are discharged, the ratio may fluctuate tests may be used to establish methods for reducing
considerably (Ballinger and Lishka, 1962; Davis, salt concentrations, coagulating suspended particles
19/1). and colloidal materials, and for breaking emulsions.
Papi.
Principles of Fermentation Technology
The commonly used biological tests include will result in potential high BODs. This is precisely
respirometry, aeration-flask tests (Otto e¢ al., 1962) what is being achieved in all large-scale fermenta-
and continuous-culture experiments. Small flask tion processes. An initial medium rich in organic
respirometers (Warburg or Gilson) are used initially matter is converted to biomass and primary and
to establish the conditions to use in bio-oxidation of secondary metabolites. Unfortunately, the product
the effluent, and test for the presence of toxic often represents a small proportion ofthe initial raw
materials. Large respirometers (Simpson and material even in an efficiently operated fermenta-
Anderson, 1967) are useful for predicting effluent tion. The spent wastes remaining after the distilla-
treatment rates and oxygen requirements. The tion of whisky may account for 90% ofthe initial raw
residues in the flasks can be analysed to see if there organic materials, while in an antibiotic fermenta-
are any recalcitrant materials. The use of laboratory tion the effluent may represent in excess of 95%.
continuous-culture vessels fitted with sludge-return Data for a variety of fermentation effluents are
pumps and settling tanks can provide detailed infor- summarized in Table 11.2. The BODs of many of
mation (Ramathan and Gaudy, 1969). Proposed these samples are much higher than that of domestic
large-scale operating conditions for feed and aera- sewage and some may be comparable with strong
tion rates can be tested and their effectiveness asses- effluents such as sulphite paper mill liquor. It is
sed. The results from all these experiments may help evident from these data that fermentation effluents
in the design of a full-scale plant. may present serious potential pollution problems
If the survey is comprehensive it should be possi- and may be expensive to dispose ofunless well-plan-
ble to plan an overall treatment programme for a ned processes are used. A number of steps may be
factory site and establish: taken to reduce BODs in a process. Some of these
will be discussed in this chapter. Careful selection of
1. Water sources which can be combined for
raw materials may have a significant effect on the
reuse.
type, and quantities, of effluent being produced.
2. Concentrated waste streams which contain
The cheapest raw material which meets the nutrient
valuable wastes to be recovered as food, ani-
requirements of the micro-organisms may not be
mal feed, fertilizer or fuel.
ideal if product yield, recovery cost, effluent-dis-
3. Toxic effluents needing special treatment, or
posal cost and possible by-product value are consi-
acids or alkalis needing neutralizing.
dered together. The high BOD value for fungal
4. The effluent loading expected under
mycelium (40,000 to 70,000 mg dm~*) would indi-
maximum production conditions.
cate that any biomass should normally be kept
5. The effluent(s) which might be discharged
separate from the remainder ofan effluent and some
directly, without treatment, to rivers or the sea
and not cause any pollution.
of it may be sold as a by-product. It may also be
worth while to concentrate liquid fractions, for
6. The effluent(s) which might be discharged
example industrial alcohol and distillery stillages
into municipal sewers.
(10,000 to 25,000 mg dm~*) will both produce dried
When all the relevant information has been obtained solubles fractions which can be sold.
one can predict the size and type offactory effluent-
treatment plant, its operational costs, the cost of
water used and sewage works charges.
TREATMENT AND DISPOSAL
OF EFFLUENTS
222
Effluent Treatment
which it might be done (Fisher, 1977). The range of In such a case there may be little preliminary treat-
effluent-disposal methods which can be considered ment and one relies solely on the degree of dilution
are: in the sea water. The pipe may be installed by the
factory or the authorities, if a public sewer. If
1. The effluent is discharged to a river or the sea
effluents are to be discharged into a river they must
in an untreated state.
meet the requirements of the local river or drainage
2. The effluent is discharged to the ground, a
authorities. In Britain there is a Royal Commission
lagoon, a disused mine or well.
standard requiring a maximum BOD (5 days) of 20
3. Part of the effluent is untreated and discharged
mg dm? and 30mg dm_° ofsuspended solid matter.
as in | or 2, the remainder is treated at a
There are, as well, often stringent upper limits for
sewage works or at the factor before discharge.
toxic metals and chemicals which might kill the
4. All of the effluent is sent to the sewage works
fauna (particularly fish) and flora, e.g. sulphites,
for treatment although there might be reluc-
cyanides, phenols, copper, zinc, cadmium, arsenic,
tance by the sewage works to accept it, and
éte:
might result in some preliminary factory treat- Until early this century, Scottish whisky distillery
ment being required, and discharge rates and wastes were normally discharged in an untreated
effluent composition defined. state to rivers. This practice has now ceased. A
5. All the effluent is treated at the factory before survey of Scottish distilleries by Mackel (1976)
discharge. showed that 27% were still permitted to dispose of
their spent waste at sea, 10% used public sewers and
5% sprayed it on to derelict land.
DISPOSAL
The simplest way of disposal will be ona sea coast Lagoons, holding ponds, oxidation ponds, etc.,
or in a large estuary where the effluent is discharged may be used by a number of industries if land is
through a pipeline extending below low-tide mark. available at a reasonable cost. It is a method which
223
Principles of Fermentation Technology
considerably reduce the BOD of the resulting removed by coagulation and/or flocculation
effluent. In nearly all fermentation processes the (Cooper, 1975; see also Chapter 10). Coagulation is
cells are separated from the liquid fraction in recov- essentially instantaneous whereas flocculation
ery processes (Chapter 10). Obviously, biomass requires some more time and gentle agitation to
processes need not be considered. Yeast cells from achieve the ‘aggregation’ ofthe particles. Ferrous or
other processes may be a marketable product, but ferric sulphate, aluminium sulphate (alum) or cal-
microbial cells may not always be marketable, par- cium hydroxide (lime) are often used as chemical
ticularly when contaminated with filter aid. In these coagulants. A solution of coagulant of the approp-
instances, when the cells and filter aid are a waste, riate strength for effective treatment is added to the
then the recovered material may be dealt with in two effluent in a mixing tank, a precipitate or floc forms
basic ways: which carries down the suspended solids to form a
sludge. This sludge may be drawn off, mechanically
1. The waste is disposed without any further
dewatered and subjected to further treatment.
treatment.
2. The waste bulk is reduced by mechanical
dewatering with a filter press, centrifuge, rot- Biological treatment
ary vacuum filter or filter belt press. The
compressed waste is then incinerated (Grieve,
Most organic-waste materials may be degraded
1978) or dumped.
biologically. This process may be achieved aerobi-
The apparently simple procedure of incineration at cally or anaerobically in a number of ways. The
a factory or sewage works may be severely controlled most widely used processes are the trickling filter
by the Alkali Inspectorate in Britain and compara- (and its modifications of towers), rotating discs or
ble authorities elsewhere (Carter, 1959). rotating drums and the activated sludge process and
Solid wastes are produced in some processes its modifications. The anaerobic processes are diges-
before inoculation. In breweries, where malted grain tion or filtration.
is still used, coarse screens or ‘whirlpool’ centrifuges
may be used to remove spent grain from the wort
after it is mashed. About 5 kg (wet weight) of grain Aerobic processes
are produced per barrel (180 dm’) of beer. If hops
are used, rather than hop extracts, they will also be TRICKLING FILTERS
recovered on screens in a ‘hop back’. This residue A simple trickling filter usually consists ofa cylin-
may amount to 250 g per barrel. Both the spent drical concrete tank 2 to 3 m indepth and 8 to 16m
grain and hop waste may then be mechanically in diameter, or it may be ina rectangular shape. The
dewatered before being sold or dumped. tank is packed with a porous bed ofslag, clinker or
The stillage (after distillation) in whisky distil- stone, the bed being underlaid with drains. The
leries may be passed through screens (1 mm open- diameters of the packing material particles should
ings). These screenings are then removed, mechani- be 50 to 100 mm to give a specific surface (m* m7 *)
cally dewatered, and dried in rotary driers to yield a of 100 to 200, and the material should be packed to
potentially marketable residue known as Distillers’ give a voidage (% air space of total bed volume) of
grains. Ina survey of Scotland about half the whisky 45 to 55%, which should minimize the risk of the
distilleries were evaporating the spent waste to a spaces between the packing material becoming
syrup containing 45% solids, mixing with spent blocked by the microbial film. The effluent, from
grain, drying and selling the final product, ‘Distil- which the suspended solids have been removed, is
lers’ Dark Grains’, as a low-grade cattle food (Mac- fed intermittently on to the upper surface of the bed
kel, 1976). by spray nozzles or mechanical distributor arms
(McKinney, 1962; Higgins, 1968). The effluent
trickles gradually through the bed and a slime layer
Chemical treatment
of biologically active material (bacteria, fungi,
Fine suspended particles in an effluent may be algae, worms, flies, etc.) forms on the surface of the
225
Principles of Fermentation Technology
support material. The large surface area created in It is also possible to modify the trickling filter to
the bed permits close contact between air flowing increase the capacity for organic loading. One way
upwards through the bed, the descending effluent is to use two sets offilters and settling tanks in series.
and the biologically active growth. The bacteria in Effluent is applied to the first filter at a high rate,
the biological film remove the majority of the while the second filter only receives a light loading.
organic loading. Complex organic materials are After a while the sequence is reversed. Loading rates
broken down and utilized; nitrogenous matter and of 0.32 to 0.47 kg BOD m~* day| have been claimed
ammonia are slowly oxidized to nitrates and sul- (Forster, 1977). Alternatively, enclosed deep beds of
phides and other compounds are similarly oxidized. 3.5 to 5.5 m depth may be used in which air is blown
The higher organisms, particularly the worms and down the beds by fans. Loading rates up to 12 times
larvae, coritrol the accumulation of the biological that of the ordinary filter have been claimed (Abson
film and improve the settling characteristics of the and Todhunter, 1967).
solids (humus) discharged with the filter effluent. Cook (1978) has stressed the need to consider the
The scouring action of the hydraulic load normally possible intermittency ofa factory waste-treatment
has a minor role in removal of any loose microbial process. It was shown that starving ofa laboratory-
film. The active slime takes time to develop and can scale trickling filter beyond 48 hours resulted in near
be poisoned by the addition of toxic chemicals. The failure of the filter. This indicated the need to supple-
simple filter is inefficient when operated at abnor- ment or artificially load the filter to maintain a
mally high loading rates. Initially, there is a very viable biomass in the system.
rapid build up of bacteria, fungi and algae on the
surface of the support material, which cannot be TOWERS
controlled by the resident population of worms and Because trickling filters do not have both a high
larvae. The voids, therefore, block up resulting in specific area and a high voidage, they are less suit-
untreated effluent accumulating on the surface of able for the treatment of large volumes of strong
the filter bed. industrial effluents (Table 11.2). Large areas ofland
A trickling filter bed should remove 75 to 95% of would be required for the expensive and extensive
the BOD and 90 to 95% of the suspended solids. A traditional filter beds. Towers, 6 to 9 m in height,
basis for normal sewage-type effluents is to provide packed with light-weight plastic multi-faced mod-
1 m? of filter medium for each 90 g day”! of 5-day ules or small random packing units (Chapter 10)
BOD for conventional trickling filtration. When the have provided to be a space saving and relatively
effluent is being circulated with an equal volume of inexpensive solution to the problem. These packings
treated effluent the loading can be increased to 180 have a relatively open structure for oxygen transfer
g day 'of5day BOD m™. (specific surface of 100 to 300) and large voids (90 to
Lamb and Owen (1970) proposed the following 98%), but are expensive compared with the conven-
formula for calculating the design parameters for a tional filter packings. They are capable of coping
trickling filter: with high BOD loadings. At a loading of 3.2kg BOD
B, 7.0 F 7 907-15) m * day! a 50% BOD removal may be achieved,
and at 1.5 kg m * day! 70% removal is possible
B, -By VS (Ripley, 1979). The biological film is similar to that
where B, = inflowing effluent BOD (mg dm *) b
formed on the conventional packing. Scouring, how-
treated effluent BOD (mg dm ‘°), ever, is due to the hydraulic load rather than the
= flow rate (m* day"'), worms and larvae.
= volume offilter medium (m’),
Sy
Cy)= specific surface of filter medium (m 2 ROTATING DISCS
mee) In this treatment method a unit composed of
oa II temperature of the sewage in the filter. closely spaced discs (2 to 3 m diameter), on acentral
In designs based on this formula, it was suggested drive shaft are rotated slowly (0.5 to 15 rpm)
that a minimum temperature of 8° was to be through the effluent so that 40 to 50% of the disc
assumed. surfaces are submerged (Borchardt, 1970; Pretori-
226
Effluent Treatment
ACTIVATED SLUDGE
The basic activated-sludge process consists of
aerating and agitating the effluent in the presence of
a flocculated suspension of micro-organisms on par-
ticulate organic matter. This process was first
reported by Arden and Lockett (1914). The raw
effluent enters a primary sedimentation tank where Fic. 11.1. Deep-shaft effluent treatment plant (Hemming ef al.,
coarse solids are removed. The partially clarified 1977).
effluent passes to a second vessel, which can ke ofa
variety of designs, into which air or oxygen is
injected by bubble diffusers, paddles, stirrers, etc. Shaft’ (Fig. 11.1) consists ofa shaft 50 to 150 m deep,
Vigorous agitation is used to ensure that the effluent separated into a down-flow section (down-comer)
is in contact with the activated sludge. After a and an up-flow section (riser). The shaft may be 0.5
predetermined residence time of several hours, the to 10 m diameter depending on capacity. Fresh
effluent passes to a sedimentation tank to remove effluent is fed in at the top of the ‘Deep Shaft’ and air
the flocculated solids. Part of the sludge from the is injected into the down-flow section at a sufficient
settlement tank is recycled to the aeration tank to depth to make the liquid circulate at 1 to2ms_!.
maintain the biological activity. A clear-water over- The driving force for circulation is created by the
flow is obtained from the settlement tank which then difference in density (due to air bubble volume)
can be used in the factory or discharged to a river between the riser and down-flow sections. For start-
without further treatment. The excess sludge is ing up, circulation of liquid is stimulated by inject-
dewatered and dried, to be sold as a fertilizer, burnt ing air at the same depth in the riser. Air injection is
or dumped. A number of modifications of the pro- then gradually all transferred to the air injection
cess are used in industrial treatment. These include point in the down-comer. Because of the pressure
multiple units with various flow patterns, different created in the down-comer, oxygen-transfer rates of
air-distribution patterns in long tanks, stepwise 10 kg O, m™? h™! can be achieved and bubble
addition of the raw effluent to create a more even contact times of3 to 5 minutes are possible instead
oxygen demand or intermediate clarifiers (Winkler of 15 seconds in conventional diffused air systems.
and Thomas, 1978). Using this technique, it has been possible to remove
One vessel type which is being installed is the 92% of the BOD from a raw effluent with a strength
‘Deep Shaft’ (Hemming et al., 1977), which is quite of 110 to 350 mg d~* in an average residence time of
distinctive from the other aeration tanks and has 80 minutes. Sludge production was found to be
been developed from the ICI Ltd SCP process much less than for conventional sewage-treatment
(Taylor and Senior, 1978; Chapter 7). The ‘Deep processes.
227
PFT-P
Principles of Fermentation Technology
The activated sludge process should be a straight- (Pfeffer, 1966). The digester tanks used for this
forward operation with good removal of organic process may be up to 12,000 m? volume and equip-
matter from the effluent. Loadings of 0.16 to 4.00 kg ped with heating coils for accurate temperature
BOD m™*° day| are possible depending on the type control (30° to 37°) to increase the rate of digestion,
of process being operated. Unfortunately, it is a which is also improved by mechanical agitation
process which is ideally suited to a steady flow of an (Loll, 1977). The residual solids, after 10 to 30 days’
effluent of a fairly uniform composition. Extra nitro- digestion, can be disposed of by dewatering, drying
gen and phosphorus may have to be added to give and incinerating, dumping, etc.
the theoretical optima of |mg dm~®of nitrogen to 20
mg dm~? of BOD and 1 mg dm? of phosphorus to ANAEROBIC FILTERS
60 mg dm’ of BOD. In anaerobic filters, as with the aerobic filters,
there is a microbial film growing on an inert support
(Chian and DeWalle, 1977). During the anaerobic
Anaerobic treatment. digestion, the acid-fermenting bacteria degrade the
waste to free volatile fatty acids, mainly acetic and
proprionic acid, which are then converted by the
Anaerobic treatment of waste organic materials methane-fermenting bacteria to methane and car-
originated with the use of septic tanks and Imhoff
bon dioxide. This technique produces low biomass,
tanks, which have now been replaced by a variety of
hence surplus sludge and power requirements are
high-rate digestors (Pohland and Ghosh, 1971).
lower than for aerobic systems.
Loehr (1968) has listed the following reasons for
using anaerobic processes for waste treatment:
1. Higher loading rates ofdigestors or filters can BY-PRODUCTS
be achieved than are possible for aerobic treat-
ment techniques.
2. Lower power requirements may be needed per The marketing of wastes from fermentation pro-
unit of BOD treated. cesses has been established at least 200 years. By the
3. Useful end-products such as digested sludge 1700s, brewers’ grains, spent hops and surplus yeast
and/or combustible gases may be produced. from larger breweries were accumulating in
4. Organic matter is metabolized to a stable sufficient quantities for specialized trades to develop
form. (Mathias, 1959). Around London, cattle and pigs
5. There is an alteration of water-binding charac- were fattened on ‘wash’ and brewers’ grains. Excess
teristics to permit rapid sludge dewatering. yeast was supplied to bakers and gin distillers. —
6. In anaerobic processes the reduced amount of In any fermentation recovery process there is a
microbial biomass leads to easier handling of need to recognize if there is a potential marketable
sludge. waste and if necessary develop a market. Obviously
7. Low levels of microbial growth will decrease the marketability and cost of reclamation of by-pro-
the possible need for supplementary nutrients ducts will be very important in deciding a policy of
with nutritionally unbalanced wastes. waste recovery. Under favourable market condi-
tions it has been claimed that the profit on animal
ANAEROBIC DIGESTION feed by-products from a completely integrated distil-
Large volumes of wet sludge are produced in lery may almost pay the cost of the grain (Blaine,
physical, chemical and biological treatment proces- 1965). This claim would be now more difficult to
ses which may have to be reduced in volume before substantiate, due to fuel costs and capital outlay on
disposal. This volume of sludge can be reduced by plant (Quinn and Marchant, 1980; Sheenan and
sludge digestion in an anaerobic digester, which Greenfield, 1980).
leads to the conversion of a large proportion of the Some microbial processes yield residues which
organic matter into methane and carbon dioxide by are difficult or impossible to sell (Blaine, 1965). In
anaerobic and facultatively anaerobic bacteria these processes it may be possible, and worth while,
228
Effluent Treatment
to change to low residue raw materials such as recovered from the mash tun and is then sold as
refined sugars or other pure compounds as sources animal feed either after pressing in a wet state or
of soluble nutrients. after drying in rotary driers. Alternatively, the wet
grain may be used in the preparation of silage for
Distilleries cattle. The possible markets for hops are restricted.
Some is used as a fertilizer or as a low-grade fuel.
In grain-based distilleries it has been common Yeast can be separated from beer by filtration or
practice to recover spent grain and stillage, the centrifuging. The yeast slurry is then dried in drum
waste liquor after the alcohol has been distilled off driers. Some of the yeast may be mixed with brewers’
(Boruff, 1953; Blaine, 1965). The stillage is first spent grains to produce a feed material with a
passed through screens with l1-mm openings. The slightly higher protein content than normal brewers’
screenings are then dewatered with mechanical filter grains. The yeast may also be used directly as a
presses and dried in rotary driers (Chapter 10). This source of vitamins. Ifit is to be used as a human food
product is termed Distillers’ Dried Grains (light it must be debittered to remove the hop bitter
grains). The screened stillage is concentrated in substances absorbed on to the yeast cells. The cells
evaporators to give 25 te 35% solids in a thick syrup. are then washed in an alkaline solution, washed
Stillages which have been prefiltered may be con- with water and drum dried. Although bakers’ yeast
centrated to 35 to 50% solids. This syrup can then was originally obtained as a brewery by-product,
be mixed with the pressed screenings and dried to this market has diminished considerably. Most bak-
give Distillers’ Dried Grains with Solubles (dark ers’ yeast is now produced directly by a distinct
grains). Alternatively the evaporated stillage may production process.
be dried completely in drum driers to produce Distil-
lers’ Dried Solubles. Dark and light grains and dried
solubles have all been used as animal-feed supple- Amino acid wastes
ments. Dried solubles have also been used as a
medium adjunct in the preparation of antibiotics The main wastes from glutamic acid or lysine
(Chapter 4). fermentations are cells, a liquor with a high amino
In distilleries using cane molasses as a feedstock, acid content which can be used as an animal-feed
evaporated spent wash has been used as a fuel for supplement, and the salts removed from the liquor
boilers (Sheenan and Greenfield, 1980). It has by crystallization, which is a good fertilizer (Re-
proved worth while to recover potassium salts from naud, 1980).
sugar-beet stillage. The market for the evaporated
product must be considered within a range of 50 km
of the evaporation plant (Lewicki, 1978). REFERENCES
Boruff (1953) cited an unpublished 1949 survey of
American distilleries which showed that 85% of the Asson,J.W. and Topuunter, K. H. (1967) Effluent treatment.
stillage solids were being recovered as dried feeds, In Biochemical and Biological Engineering Science, vol. 1,
pp. 309-343 (Editor Blakebrough, N.). Academic Press,
14% solids as wet grains and only 1% was waste. In London.
recent years the traditional distillers’ by-products AMERICAN SOCIETY FOR TESTING AND MATERIALS (1966) Manual
from stillage have become increasingly uneconomic. on Industrial Waste Water, 2nd edition. Philadelphia.
ArpDEN, E. and Locxerr, W. T. (1914) Experiments on the
A number of processes to produce SCP from stillage oxidation of sewage without the aid of filters. Surveyor, 45,
using Geotrichum candidum, Candida utilis or C. tropicalis 610-620.
have been evaluated (Quinn and Marchant, 1980; Bauuincer, D. G. and Lisuxa, R. J. (1962) Reliability and
precision of BOD and COD determinations. Journal of the
Sheenan and Greenfield, 1980). Water Pollution Control Federation, 34, 470-474.
BrainE, R. K. (1965) Fermentation products. In Jndustrial Waste
Water Control, pp. 147-166 (Editor Gurnham, C. F.).
Breweries
Academic Press, New York.
Borcuarot,J. A. (1970) Biological waste treatment using rotat-
The three marketable wastes in breweries are ing discs. In Biological Waste Treatment. Biotechnol. Bioeng.
spent grain, spent hops and yeast. The spent grain 1s Symp. 2, 131-140.
229
Principles of Fermentation Technology
Borurr, C. S. 53) The fermentation industries. In Industrial MackeEL, C. J. (1976) A Study of the Availability of Distillery
Wastes. Their Disposal and Treatment, pp. 99-131 (Editor By-products to the Agricultural Industry, 46 pp. Report of the
Rudolfs. W.). Reinhold, New York. School of Agriculture, University of Aberdeen, Scotland.
Carter, J. S. (1959) The Alkali, Etc., Works Regulation Act Maruias, P. (1959) The Brewing Industry in England, 1700-1830.
1906 and Alkali, Etc. Works Orders 1928-1958. In Chemical University Press, Cambridge.
Engineering Practice, Vol. 2, pp. 223-264 (Editors Cremer, McKinney, R. E. (1962) Complete mixing activated sludge
H. W. and Watkins, S. B.). Butterworths, London. treatment ofantibiotic wastes. Biotech. Bioeng. 4, 181-195.
Cuan, E. S. K. and DEWALLE, F. B. (1977) Treatment of
high Me ccuer, R. R. (1962) Pharmaceutical waste disposal by soil
strength acidic waste water with a completely mixed injection. Biotech. Bioeng. 4, 147-151.
anaerobic filter. Water Research, 11, 295-304. Mo tor, A. H. (1962) Pharmaceutical waste treatment by the
Covovos, G. C. and TinkLenser, N. (1962) Land disposal of trickling filter, activated sludge and compost processes.
pharmaceutical manufacturing wastes. Biotech. Bioeng. 4, Biotech. Bioeng. 4, 197-209.
153-160. Orro, R., BarKER, W., Scowarz, D. and Tyarksen, B. (1962)
Cook, E. E. (1978) Effects of long term endogenous respiration Laboratory testing of pharmaceutical wastes for biological
(fasting) on the organic removal capacity ofa trickling filter. control. Biotech. Bioeng. 4, 139-145.
Biotech. Bioeng. 20, 293-296. PrerFER, J. (1966) Application of the anaerobic process to waste
Cooper, P. (1975) Physical and chemical methods of sewage water treatment. Dev. Ind. Microbiol. 7, 134-143.
treatment. Review of present state of technology. Water PoHLAND, F. G. and Guosu, S. (1971) Developments in
Pollution Control, 74, 303-311. anaerobic treatment processes. In Biological Waste Treatment.
Davis, E. M. (1971) BOD vs COD vs TOG vs TOD. Water and Biotech. Bioeng. Symp. 2, 85-106 (1971).
Wastes Engineering, 8(2), 32-34, 38. Pretorious, W. A. (1973) The Rotating Disc unit. A waste
ERICHSEN Jones, J. R. (1964) Fish and River Pollution. Butter- treatment system for small communities. Water Pollution
worths, London. Control, 72, 721-724.
Fisuer, N. S. (1977) Legal aspects of pollution. In Treatment of Quinn, J. P. and Marcuant, R. (1980) The treatment of malt-
Industrial Effluents, pp. 18-29 (Editors Callely, A. G., Forster, whisky distillery waste using the fungus Geotrichum candidum.
C.F. and Stafford, D. A.). Hodder and Stoughton, London. Water Research, 14, 545-551.
Forster, C. F. (1977) Bio-oxidation. In Treatment of Industrial RaMATHAN, M. and Gaupy, A. F. (1969) Effect of high substrate
Effluents, pp. 65-87 (Editors Callely, A. G., Forster, C. F.and concentration and cell feed-back on kinetic behaviour of
Stafford, D. A.). Hodder and Stoughton, London. heterogeneous populations in completely mixed systems.
Grieve, A. (1978) Sludge incineration with particular reference Biotech. Bioeng. 11, 207—237.
to the Coleshill plant. Water Pollution Control, 77, 314-321. RENAUD, C. (1980) Treatment ofeffluent from glutamic acid and
Hemmine, M. L., Oussy,J.C., Plowricut, D. R. and WALKER, lysine fermentation. In Effluent Treatment in the Biochemical
J. (1977) ‘Deep Shaft’—Latest position. Water Pollution Con- Industries, Process Biochemistry’s 3rd International Confer-
trol, 76, 441-451. ence.
Hiceins, P. M. (1968) Waste treatment by aerobic techniques. Riptey, P. (1979) Process engineering aspects of the treatment
Dev. Ind. Microbiol. 9, 146-159. and disposal of distillery effluent. Proc. Biochem. 14(1), 8-10.
Hirt, F. (1980) Effluent treatment in the Bakers’ yeast industry. SHEENAN, G. J. and GREENFIELD, P. F. (1980) Utilisation, treat-
In Effiuent Treatment in the Biochemical Industries, Process ment and disposal of distillery waste water. Water Research,
Biochemistry’s 3rd International Conference. 14, 257-277.
Jackson, J. (1960) The treatment of distillery and antibiotics Simpson, J. R. and Anperson, G. K. (1967) Large-volume
wastes. In Waste Treatment, pp. 226-239 (Editor Isaac, respirometers with particular reference to waste-treatment.
P. C. G.). Pergamon Press, Oxford. Prog. Ind. Microbiol. 6, 141-167.
Koziorowsk1, B. and Kucuarsxt, J. (1972) Industrial Waste Taytor, I. J. and Senror, P. J. (1978) Single cell proteins: a new
Disposal. Pergamon Press, Oxford. source of animal feeds. Endeavour (N.S.), 2, 31-34.
Lamp, R. and Owsn, S. G. H. (1970) A suggested formula for the WaTER PoLiutTion RESEARCH CENTRE (1972) Water Pollution
process of biological filtration. Water Pollution Control, 69, Research: report of the Director of Water Pollution Research, 1971.
209-220. HMSO, London.
Lewicxi, W. (1978) Production, application and marketing of Winker, W. A. and Tuomas, A. (1978) Biological treatment of
concentrated molasses-fermentation-effluent (vinasses). aqueous wastes. In Topics in Enzyme and Fermentation
Proc. Biochem. 14(6), 12-13. Biotechnology, Vol. 2, pp. 200-279 (Editor Wiseman, A.).
Loenr, R. C. (1968) Anaerobic treatment of wastes. Dev. Ind. Ellis Horwood, Chichester.
Microbiol. 9, 160-174. Zajc,J.E. (1971) Deep well disposal. In Water Pollution, Disposal
Loti, U. (1977) Engineering, operation and economics of and Re-use, Vol. 2, pp. 574-589. Marcel Dekker, New York.
biodigestion. In Microbial Energy Conversion, pp. 361-378
(Editors Schlegel, H. G. and Barnes, J.). Pergamon Press,
Oxford.
230
CHAPTER 12
Fermentation Economics
INTRODUCTION The consideration of so many criteria means that
there may have to be acompromise for the particular
set of circumstances relating to an individual pro-
Ifa fermentation process is to yield a product at a
cess. In any process it is important to know the cost
competitive price, the chosen micro-organism
breakdown (Table 12.1), so that it may be seen
should give the desired end-product in predictable,
where the biggest potential savings may be
and economically adequate, quantities. A number
achieved. In a review ofa number ofprocesses, Nyiri
of basic objectives are commonly used in developing
and Charles (1977) concluded that there were four
a successful process:
basic components contributing to the process cost in
1. The capital investment in the fermenter and the following decreasing order: raw materials, fixed
ancillary equipment should be confined to a costs, utilities, labour. When raw materials are a
minimum, provided that the equipment is reli- major part of the total cost it is obvious that media
able and may be used in a range of fermenta- and microbial strain-improvement research should
tion processes. form a major part of a development programme.
2. Raw materials should be as cheap as possible Development work on components contributing
and utilized efficiently. A search for possible little to the cost of the product could not bejustified.
alternative materials might be made even Costs quoted in this chapter have been taken from
when a process is operational. the source of information and are only applicable at
3. The highest-yielding strain of micro-organism the time of that publication. During the last few
should be used. years there have been steep rises in all costs, and to
4. There should be a saving in labour whenever bring costs to a common basis for the present time
possible and automation should be used where by applying increases ofdifferent proportions to the
it is feasible. various components would be a formidable task and
5. When a batch process is operated, the growth one which would not be accurate because of the
cycle should be as short as possible to obtain approximate nature ofthe information.
the highest yield of product and allow for Caution is necessary when examining some exam-
maximum utilization of equipment. To ples of potential fermentation processes, since the
achieve this objective it may be possible to use capital and operating costs normally have been
fed-batch culture (Chapters 2 and 4). estimated using chemical engineering costing prin-
6. Recovery and purification procedures should ciples. Accurate detailed costing of industrial fer-
be as simple and rapid as possible. mentations are rarely published.
7. The effluent discharge should be kept to a Government aid or taxation can determine the
minimum. viability of many fermentation processes. United
8. Heat and power should be used efficiently. States of America agricultural aid programmes
9. Space requirements should be kept to a made available low-cost supplies of grain and
minimum, but there should be some allowance potatoes and enabled fermentations to be operated
for potential expansion. when they would not have been economically viable
231
Principles of Fermentation Technology
in a free market (Perlman, 1970). In 1980 carbohyd- isolation programmes are unfortunately time con-
rates as molasses or cane juice could be obtained in suming, expensive and may be something of agam-
sugar-producing countries at half the price of molas- -ble, particularly when searching for new antibiotics.
ses in the European Economic Community (Meers, Pfizer Ltd. spent £1,430,000 on a screening prog-
1980). Government support in construction of plant, ramme for a broad spectrum antibiotic producer
development and production programmes of during which terramycin was detected (Reeves,
acetone-butanol and penicillin during wartime led 1952). Merck, Sharp and Dohme found one useful
to production of compounds at an earlier stage than micro-organism in 10,000 (Woodruff and Mac-
might normally be regarded as economically feasible Daniel, 1958), while Eli Lilly and Company Ltd.
(Hastings, 1971, 1978). discovered three new antibiotics in 10 years while
Some very useful reviews of process economics screening 400,000 micro-organisms (Nelson, 1961).
have been produced for penicillin G (Swartz, 1979), Because of extensive soil screening a significant
gibberellic acid (Vass and Jefferys, 1979), biomass proportion of the microbial population capable of
from natural gas (Hamer, 1979), biomass from whey synthesizing antibiotics has been exploited. It is
(Meyrath and Beyer, 1979), biomass from waste evident that as early as 1961 the cost effectivessness
carbohydrates (Mateles, 1975) and biomass from of antibiotic screening combined with the decreasing
cane and coffee processing by-products (Rolz, possibility ofa significant discovery had been criti-
1975). A number ofgeneral reviews of fermentation cally considered in the United Sates of America
economics are also available (Whitaker, 1973; Nyiri (Foster, 1961). At least one American pharmaceuti-
and Charles, 1977; Bartholomew and Reisman, cal company had already discontinued its. own
1979; Stowell and Bateson, 1984). screening programmes and was relying entirely on
contract work done in Japanese laboratories. This
change in policy had been influenced by the compe-
tence of the Japanese workers to discover new anti-
ISOLATION OF MICRO-ORGANISMS OF
POTENTIAL INDUSTRIAL INTEREST biotics by systematic screening at a cost of5 to 10
times less than in the United States.
Arima (1977) quotes a list of antibiotics which
The most appropriate micro-organism for a were discovered, between 1929 and 1966, and used
potential process is usually found by isolation from for medical purposes in Japan. About 40% of these
a variety of sources, most commonly soil. These had been discovered in Japan. Yamada (1977) has
232
Fermentation Economics
4. Its content of high nutritional-value protein expertise such as Cetus Ltd. and Panlabs Inc. who
and the absence oftoxins and pathogenic prop- will perform strain-development programmes or
make cultures available for commercial clients. An
erties.
example of work by Panlabs Inc.’s improvements in
yields of penicillin from cultures of P. chrysogenum
In strain selection for enzyme production, Aunstrup
during 1973 to 1976 is shownin Table 12.2 (Queener
(1977) specified:
and Swartz, 1979). This data makes it possible to
compare yields from the same culture cultured at
1. High enzyme yields.
various scales.
2. Genetic stability.
3. The organism could produce the enzyme con- MARKET POTENTIAL
stitutively.
4. The organism could utilize media components The fermentation technologist should be aware of
efficiently. the problems ofassessing market potential although
233
Principles of Fermentation Technology
1973 P-] V 6,100 7,500 7,100 5 14,500 17,600 15,900 140 11,700 14,400 13,560 130
- V 11,000 13,100 12,000 8 13,900 20,700 18,000 180 14,200 19,100 16,500 180
V 17,400 21,300 . 19,800 9 15,400 21,800 18,200 182
G 17,600 17,900 17,700 132
p-4 Vv 23,300 25,400 24,400 10 92,100 182 19,700 20,700 20,120 182
G 15,800 20,600 17,900 182
1974 P-5 V 27,500 31,200 29,800 10 22600 23,700 23,150 182 23,600 25,900 24,700 182
G 21,300 24,600 23,150 182
P-7 V 29,200 33,900 32,100 10 25,100 33,400 30,200 210 23,400 29,600 27,300 190
G 25,200 29,000 27,700 185 23,700 28,200 25,600 190
1975 P-8 Vv 34,700 36,900 36,000 10 31,500 33,000 32,250 185 29,200 34,800 32,400 203
G 27,900 30,800 29,600 203
P-10 Vv 36,800 42,600 38,700
1976 P-10 Vv 32,500 36,500 33,700 185
P-12 Vv 40,911 42,800 41,900 9 32,300 38,000 35,000 185
G
P-13 Vv 33,700 39,000 37,200 7 41,500 185
G 42,250 185
he/she may not be primarily involved in collating or supplies of carbohydrate were available. In an
assessing the necessary data. Some aspects have ‘example based on 1977 costs, if crude oil cost $100
been considered by MacLennan (1976), Hepner ton', raw sugar feedstock would have to cost $109
(1977) and Lawson and Sutherland (1978). ton | or less to produce a competitive product and
Fermentation products may be classified as low molasses would have to cost a maximum of $75
cost-high volume or high cost—low volume. Only a ton ' (Hepner, 1978). In Brazil in 1975, an ethanol-
limited number of low-cost—high-volume products production programme using sugar cane as a sub-
have been produced by fermentation. This group strate was started so that the petroleum imports
includes solvents, organic acids and single-cell pro- could be reduced. During 1977, 7 x 10° dm? of
tein. Among the 100 most important organic chemi- ethanol were produced and it was hoped to have
cals (by production volume), microbial syntheses increased production to 1.5 X 10° dm? during 1979.
are known for ethanol, n-butanol, iso-propanol, The ethanol, although subsidized by the Brazilian
acetone, glycerol and acetic acid. In each case the government, was selling at $1 per 4.5 dm? which was
microbial process is less economical than the chemi- more than the cost of petrol refined from imported
cal synthesis (Pape, 1977). Microbial processes may oil (Hammon, 1978). Other aspects of potential
be economically viable for compounds that have a ethanol production have been discussed by Bu’ Lock
complicated structure, are chemically or thermally (1979).
unstable, and for which a multi-stage chemical syn- It is necessary to estimate the size of the present
thesis would be expensive. These processes include and potential market and the increase in demand for
the production of antibiotics, corticosteroids, L- a compound. This type ofexercise has been underta-
amino acids, vitamins and citric acid, and are class- ken for single-cell protein (MacLennan, 1976;
ified as high-cost—low-volume products. Taylor and Senior, 1978). Taylor and Senior (1978)
Hepner (1977, 1978) has examined the factors gave summaries of major single-cell protein plants
that determine the feasibility of large-scale ethanol which were operational or planned, estimates of
production by fermentation. He considered that single-cell protein production in 1980 and 1985 and
ethanol produced by fermentation would only be predicted world supply and demand for high-quality
competitive with synthetic ethanol from crude oil if protein meal. It was estimated that by 1985 there
the fermentation plant was in an area where cheap would be a market of 5 X 10° tonnes annum™! for
234
Fermentation Economics
single-cell protein world wide, whereas production which it will not be possible to remove enough heat
would only be 2.9 x 10° tonnes annum7!. Such to maintain constant temperature unless the cooling
predictions may support the establishment of such a capacity is increased by incorporating internal cool-
process. ing coils or by using external heat exchangers. These
The life expectancy of acompound will then have modifications may prove costly or else interfere with
to be predicted even when covered by a patent. This mixing in the vessels. There is possibly the alterna-
is sometimes difficult, as has been demonstrated tive of using micro-organisms with higher optimum
with the industrial enzyme market (Aunstrup, temperatures. The oxygen requirements ofa process
1977). In about 1965 detergent enzymes became may limit the size of vessel. In acetic acid produc-
widely used, which led to a general increase in sales tion, where very efficient aeration is vital, in 1977
of microbial enzymes. In 1971 allergic symptoms the maximum acetator capacity was about 50,000
were discovered in workers handling enzymes in a dm? (Pape, 1977).
detergent factory which resulted in the removal of There is an upper size limit for a custom-built
enzymes from most detergent powders and a sudden fermenter that can be transported to a site, while
drop in enzyme sales. Enzyme sales have again much larger vessels may be built on site. Most
recovered after the introduction of improved process countries limit the maximum size ofunit that can be
techniques. transported on the road. In 1979 the first ICI Ltd.
production vessel for single-cell protein was instal-
led which had a volume of 1.5 X 10° dm?, was 80 m
tall and cost £6 X 10° (Tonge, 1979. Personal com-
PLANT AND EQUIPMENT
munication). It was constructed in France, floated
on a barge to the River Tees, and transported a very
It is most logical to build equipment as large as short distance on land. At the time oferection it was
possible because of the economy ofscale. There is an the largest fabricated fermenter ever to be trans-
empirical relationship between cost and size of an ported.
item of equipment. According to this relationship, In brewing, when cylindrico-conical vessels are
as facility size increases, its cost increases thus: used there would appear to be no economic advan-
tage in scaling up above 108,000 dm’ capacity (Hog-
cost, _ /size, \"
gan, 1977), although vessels of 360,000 dm* have
cost, a
been installed. In some breweries a number of smal-
where n is an exponent or scale factor. In brewing, ler vessels have been installed to allow for fluctuat-
Pratten (1971) estimated the scale factor to be 0.6. ing demand ofdifferent beers.
The scale factor for a single-cell protein plant has Because ofthe capital investment and operational
been estimated to be between 0.7 and 0.8 (Hum- costs there is now a trend in fermenter design to
phrey, 1975; MacLennan, 1976). consider unconventional designs of simple construc-
However, there are a number of restraints which tion with very efficient oxygen transfer to be used for
have to be considered before deciding on the scale of specific purposes, particularly single-cell protein. In
operation. Such restrictions include cooling and this context, Schugerl et al. (1978) have considered
aeration requirements and the method of fermenta- the case of a single-cell protein plant for 100,000
tion vessel construction. The need for cooling provi- tonnes annum! using methanol as the main sub-
sions in many fermentations has already been strate. The plant investment was only 20% of the
described (Chapter 7). At this stage it is important production cost. In this investment cost only 20 to
to remember that the volume of a fermenter is 25% was due to the fermenter. Of this, the vessel
proportional to ? (where r is the fermenter radius), accounted for 5 to 10%, the stirrer and aerator 5 to
whereas the increase in surface area is proportional 10% and cooling 10 to 15%. Energy, water, aeration
to r. Therefore the scaling up of a vessel will lead to and auxiliaries were estimated to be a further 10% of
a decrease in the surface area to volume ratio and the production costs. Since the fermenter investment
therefore a decrease in the effectiveness of acooling costs are approximately 4 to 5% and the operating
jacket. There may be a fermenter volume above costs are 10%, it was concluded that the type of
235
Principles of Fermentation Technology
vessel could only influence 15% of the total produc- fermentation plants
forowns
Tas.e 12.3. Capital cost breakd
Se
tion costs. If atower fermenter were used, the main for five 225,000-dm? fermenters
(a) Penicillin plant, estimated
advantage would appear to be lower operating costs, with ancillary equipment (Swartz, 1979)
particularly reduced energy and cooling-water Item % of total
costs, as the removal of mechanical stirring would Process equipment 23.6
diminish the cooling-water requirements. Installation Dad
Insulation leo
A useful guide has been prepared by the Institu- Dah
Instruments
tion of Chemical Engineers (1977) which outlined in Piping 11.8
reasonably quantitative terms many factors making Electrical 15.8
Building WES
up a check list which must be incorporated into the
Utilities 23
final capital costing of achemical plant project. It is Site ae
possible from a knowledge of the proposed plant Laboratory equipment 3.8
Spare parts 0.5
location, a sketch of the chemical plant process flow
sheet, the size of the major items of equipment and (b) Norprotein plant (Morgren, 1979)
the service requirements, to estimate capital and Item % of total
operating costs to 15% (Backhurst and Harker, Raw materials storage 10
1973). Unfortunately, comparable literature for fer- Media preparation and utilities 17
Fermentation 4]
mentation plants is not available.
Cell recovery and drying 22
Some details of the cost of equipment used in Product storage 10
fermentation processes have been discussed by
Whitaker (1973), Humphrey (1975), Mateles (c) ICI Ltd. Single-cell protein plant (Smith, 1980)
(1975), Rolz (1975) and Nyiri and Charles (1977). Item % of total
and mixing of media should not be neglected. Pow- the material, impurities which make downstream
ders must be kept in dry conditions because of the processing more difficult, high water content mak-
possibility of substances becoming rocklike or ing transport costly, geographical location, quan-
glutinous. Some bulk liquids with a high solids tities produced and limited seasonal availability.
content need to be kept warm to prevent them The economics of biomass production from whey
solidifying, e.g. glucose and corn-steep liquor. If have been reviewed by Meyrath and Bayer (1979).
storage temperatures are too high there could be The possibile use of waste substrates may depend on
degradation. It is also vital for workers to follow the cost of alternative methods ofdisposal or on the
batching instructions for media preparation very availability of government grants to diminish pollu-
carefully to prevent ‘lumpy’ media, etc. (Corbett, tion (Perlman, 1970).
1980). Mineral components normally constitute a smal-
In the search for suitable nutrient components for ler part of the cost of media, e.g. they account for 4
media, six criteria were considered in Chapter 4 to 14% of the manufacturing cost of single-cell
which ought to be satisfied whenever possible. Rat- protein (Cooney et al., 1977). Although feed grades
ledge (1977) has stressed the need to note the of phosphates are more expensive than fertilizer
amount of carbon in a substrate when costing. grades, they do not contain impurities such as iron,
Higher yields might be achieved by changing from a arsenic and fluoride. This is an important considera-
carbohydrate to an alternative carbon source, but tion in the production offoods and drugs (Litchfield,
increased aeration and/or agitation rates may be 1977). The hydroxides and sulphates of potassium,
necessary with the change in substrate. The cost of magnesium, manganese, zinc and iron are preferred
this extra provision must therefore be less than the to the chlorides to minimize corrosion of stainless
savings from the change ofsubstrate if the process is steel. The source of basic materials can cause consid-
‘to be feasible. When cheapness is added as a further erable variation in product yield. Corbett (1980)
restraint the number of potential major nutrients compared six samples of calcium carbonate, and
which could be used on an industrial scale is limited. found that five of them reduced the titre of penicillin
Carbohydrates from beet, sugar cane or grain are G ina production medium.
the major carbon and energy sources in most media
and comply with the requirements of economy and
AIR STERILIZATION
those stated in Chapter 4.
In recent years there has been a ‘flirtation’ with The problems associated with producing large
petrochemicals as carbon sources, primarily as sub- volumes of sterile air for aerobic fermentations are
strates for single-cell protein production. Dimmling unique (see Chapter 5). Although sterilization by
(1977) claimed that as far as it was possible to heating is technically possible, it has generally been
predict at the time of writing, methane and regarded as too costly for full-scale operation
methanol were the most economical future (Cherry et al., 1963). Richards (1968a) has given the
feedstocks for large-scale fermentation processes. outline of a plant for the sterilization of air by
He admitted that a reduction in the price of peiro- heating. He estimated that the heat input to treat
chemicals could not be expected. Since that time the 283 m? min™! of air would be 33,000 CHUs min!
price of petrochemicals has risen considerably. ICI with a corresponding cooling load. The overall
Ltd. are one of the few companies who can at present operating costs were expected to be £5 to £10 h!.
operate an economic single-cell protein process Stark and Pohler (1950) investigated heat-of-com-
because of the long-term price they negotiated for pression for air sterilization. They claimed that
natural gas as a feed stock. reciprocating compressors at 7.03 kg cm? without
A variety of waste materials would seem to be filters would be most economical in the range 56.6 to
potential cheap carbon sources. Unfortunately, it 141.5 m?® min7!, Single-stage compressors with
has been shown that their use is very restricted filters were most economical for the range 141.5 to
because they cannot compete economically with 566 m? min ' and turbocompressors were reported
conventional substrates. This may be due to a to give maximum savings for the range 707 to 849 m?®
number of possible reasons including variability of min
237
Principles of Fermentation Technology
Filtration of air through deep beds of fibrous or The economics of sterilization have already been
granular material is the commonest method used for considered earlier in this chapter. Cooling require-
air sterilization. It is simpler and economically pref- ments will be influenced by the size and type of an
erable to alternative methods (Humphrey and individual process (Chapter 7). British Petroleum
Gaden, 1955). The capital expenditure on filter Ltd. have estimated that the cooling requirements
plant is dependent on its size, chiefly on its cross-sec- for a 100,000 tonnes annum! single-cell protein
tional area, while the operating costs for a given size plant to be 110 X 10° kcal h7! using n-alkanes or
are dependent upon the pressure drop across the gas-oil as the primary substrate (Litchfield, 1977).
filter. As the cross-sectional area of the filter is To reduce cooling requirements, the specific energy
increased, so are the capital costs. However, the input may be minimized through the use of airlift
superficial velocity and the pressure drop are fermenters (Schugerl et al., 1978). Cooling equip-
reduced which result in lower operating costs (Hum- ment has been estimated at 10 to 15% ofthe invest-
phrey, 1960). Some designs offilter require less floor ment cost for single-cell protein (Moo-Young, 1977;
space than others. White and Smith (1960) reported Schugerl et al., 1978). Another way to minimize
that to filter 8,470 m? min! with wool resin canisters cooling costs is to use micro-organisms with higher
required a floor space of 325 m?; Porton-type asbes- optimum growth temperatures, ifit is feasible. The
tos cloth required 97.7 m*; while asbestos paper-type selection and use of thermophiles and thermo-
‘absolute’ filters needed only 38.3 m?. tolerant organisms would have obvious advantages
Operating costs will be based on the estimated life to reduce cooling demands (Chapter 7).
of the filters. Factors to consider include the cost of
replacement filters or filter materials, servicing and
labour. Even if the filters could be cleaned there AERATION AND AGITATION
must be an allowance for depreciation due to normal
wear and tear. Savings may also be made by intro- Nearly all fermentations require some form of
ducing series filtration whereby the major part ofthe mixing to maintain a constant environment, and
foreign matter from the air stream is taken out by many also need aerating (see also Chapter 9). Fer-
varying degrees of coarse filtration, thus reducing mentations may be broadly classified into:
renewals of the more expensive high-efficiency filter
media. 1. Fermentations which are anaerobic where
oxygen is undesirable, e.g. acetone-butanol.
2. Fermentations which have a minimal oxygen
HEATING AND COOLING demand, e.g. ethanol.
3. Fermentations which have a high oxygen
Ideally there should be no heating or cooling at demand, e.g. antibiotics, acetic acid, single-
any stage in a fermentation process, but because this cell protein.
is virtually impossible, heat should be conserved
In categories | and 2, aeration is not generally
and cooling minimized by careful process design. A
regarded as a major economic consideration. During
fermentation may include the following heating or
an acetone—butanol fermentation carbon dioxide
cooling stages:
and hydrogen are evolved. Once this gas production
1. Sterilization or boiling of the medium to 100° starts it will help to maintain anaerobic conditions
or above followed by cooling to 35° or below. and stir the mass of broth without the need for
2. Heating the fermenter and ancillary equip- mechanical agitation. Anaerobic conditions are ini-
ment to sterilize it followed by cooling. tially achieved in a production fermenter by main-
3. Heat may be generated during the fermenta- taining a positive pressure offiltered carbon dioxide
tion. This heat output has to be removed by and hydrogen obtained from another established
cooling to maintain the growth temperature of fermentation (Beesch, 1952).
the micro-organism within prescribed limits. For ethanol production, the yeast inoculum in the
4. After harvesting, heat may be required to vessel is initially dispersed in the medium by com-
remove water from the product. pressed air or by mechanical stirring. Aeration or
238
Fermentation Economics
Tasce 12.4. Effect ofsubstrate and yield coefficients on SCP operating costs (Abbott and Clamen, 1973)
Substrate costs
O, transfer Heat removal Combined
clb7! clb™! costsclb ! costs clb! costs clb7!
Substrate substrate cells cells cells cells
agitation is stopped once the biomass concentration absorbed oxygen. Swartz (1979) has reported that
reaches a predetermined level. A vigorous anaerobic the mixing costs in a penicillin fermentation are
fermentation commences, and the evolution of car- >15% of the total production costs.
bon dioxide bubbles stirs the contents of the vessel In single-cell protein processes, the carbon sub-
and disperses the cells in the medium so that strate yield coefficient is the most critical physiolog-
mechanical agitation is not necessary. In this pro- ical factor (Hamer, 1979). Itisalso well documented
cess aeration and agitation are considered to be a that much higher carbon substrate yield coefficients
minor component of the total production costs. are obtained with methane or n-alkanes instead of
Fermentations having a high oxygen demand carbohydrates (see Table 12.4). Unfortunately, cells
must be agitated with sufficient power to maintain a grown on hydrocarbons. have greater oxygen
uniform environment and to disperse the stream of requirements. The oxygen requirements of ahydro-
air introduced by aeration. In an early reference it carbon yeast fermentation is almost triple that of a
was stated that the cost of energy necessary to yeast fermentation producing an equal quantity of
compress air for yeast production proved that a cells (Darlington, 1964; Chapter 4). Therefore, if
considerable amount (10 to 20%) of the total pro- there is to be effective utilization of a hydrocarbon
duction expenses was due to aeration (de Becze and substrate, which can account for over 50% of total
Liebmann, 1944). Hastings and Jackson (1965) production costs (Litchfield, 1977), the production
reported that the cost of agitation for a penicillin fermenter must have a high oxygen-transfer capac-
fermentation could vary from 0.9 to 1.4p h7! 455 ity. The demands on fermenter design are further
dm“? with air flow rates of 0.06 to 1.0 volumes ofair complicated by the hydrocarbon fermentation being
per volume of medium per minute. To obtain highly exothermic which necessitates the provision
optimum conditions, a power input to the impeller of good cooling facilities ifa constant temperature is
of 1.85 to 2.20 joules 455 dm~? and an air-flow rate to be maintained in the fermenter.
of 0.25 to 0.5 volumes of air per volume of medium A few companies have developed processes using
per minute was required. In this type of fermenta- mechanically stirred fermenters with sparged air.
tion the cost of mixing and oxygen transfer was still BP Ltd. constructed vessels of up to 1000 m* capac-
only 1.9% of the total production cost (Ryu and ity for their n-alkane process in their Sardinian Ital
Oldshue, 1977). Even though this is only a small protein project (Levi et al., 1979). In the Swedish
percentage of the total production cost, these work- Norprotein process it has been estimated that the
ers recognized that the supply of oxygen was a total utilities costs which included aeration and
limiting factor. They cited an example for an anti- agitation (1978 prices) for 100,000 tons year”! of
biotic fermentation where the agitator motor size SCP would only be 16% of total production costs
was changed from 745 to 1860 joules to increase the (Mogren, 1979).
oxygen-transfer rate. Cost savings for aeration and A number of companies including ICI Ltd.
agitation of 25% were claimed using the larger (Taylor and Senior, 1978) and Hoechst (Knecht et
1860-joules motor, for an antibiotic which was being al., 1977) decided to develop fermenters based on
produced at a cost of 30 cents (USA) per mol of the air-lift principle (Chapter 7; Hamer, 1979; Levi
239
Principles of Fermentation Technology
et al., 1979). The main advantages of these fermen- — Higher level of raw materials and energy
_..-. Lower level of raw materials and energy
ters are simpler design and reduced energy and
cooling water costs. Since the energy supplied to an
air-lift fermenter is only supplied with the air, it is
crucial to obtain a fermenter design which
minimizes the energy requirement for biomass pro-
duction yet creates high oxygen-transfer facilities to
ensure efficient substrate utilization. In the ICT Ltd.
process, the estimated manufacturing costs for all
utilities were 14%, with aeration accounting for
70% of fermentation utilities costs (Moo-Young,
1977).
P= I x;
2
— In + tp tin tt, CONTINUOUS CULTURE
oes X
It will be possible from this equation to determine At the present time, very few large-scale con-
the effects of process changes on the overall produc- tinuous-culture plants are being operated and these
tivity. A larger initial inoculum would increase XQ are primarily for the production of microbial
and shorten the process tme. Actively growing biomass. It is appropriate at this stage to compare
inocula would reduce the lag me é,. Aspects of this batch-culture productivity and continuous culture
240
Fermentation Economics
TABLE 12.5. Comparison of productivity in batch and continuous culture efficiencies can be achieved. Unfortunately in some
(Wang et al., 1979)
eee processes the final product concentration in the
Continuous productivity effluent broth from a continuous culture will be less
Maximum specific growth rate than that obtained in a batch process, which will
in batch culture (hr!) Batch productivity
create a need for greater concentration at the recov-
0.05 0.09 ery stage.
0.10 0.21 In the production of antibiotics and other secon-
0.20 0.53
dary metabolites, major costs may be associated
0.40 1.5
0.80 4.6 with the recovery stage. The final product concen-
1.0 6.8 tration can have a big effect on the cost of the
12 DES)
product since the size and operational cost of a
recovery system is inversely proportional to the
productivity. Wang et al. (1979) have derived an volume ofliquid that must be handled.
equation to quantify this relationship: The continued increase in efficiency of batch-cul-
ture processes for antibiotics and other non-growth-
Continuous-culture productivity
associated products makes manufacturers reluctant
Batch-culture productivity to make radical alterations to established processes
such as the introduction of continuous culture (see
ln Xn + Mme also Chapter 2).
= wb esse nasDY
xXm —? Xo i SS
xX m
RECOVERY COSTS
where #4, = maximum specific-growth rate,
initial cell density,
maximum cell density, The costs of product recovery and purification are
turn-around time, rarely quoted, though in some processes they are
critical dilution rate, obviously considerable. Stowell and Bateson (1984)
beMS = Cellular yield coefficient for the limit- identified a number of factors contributing to these
ing nutrient. costs:
In an example they used an inocula size of 5%
1. Yield losses, even if only modest, are certain to
(X,/X,, = 0.05), a process turn-around time of 10
occur at each stage of the recovery process.
hours, a cellular yield of 0.5 g cells per g of substrate
2. High energy and maintenance costs associated
and a final cell concentration of 30 g dm~°. Produc-
with running filtration and centrifugation
tivity was then calculated for a series of maximum
equipment.
specific growth rates (Table 12.5). It is clear from
3. High costs of solvents and other raw materials
this data that the faster the organism can grow the
used in recovery and refining of products.
more favourable is a continuous process over a
batch process. Atkinson and Mavituna (1983) gave losses of 8% for
When assessing feasibility of acontinuous process citric acid and 4% for penicillin G before conversion
for product formation it is necessary to know: vol- to the potassium salt in process flow charts. They
umetric productivity, conversion yield of product also stressed the importance oftrying to reduce the
from the most expensive substrate in the medium number of downstream stages as much as possible to
and the product concentration (Wang et al., 1979). reduce capital and operating costs.
In processes for cell biomass, some alcohols and It is thought that depreciation, return on capital
organic acids where the major production cost is at and maintenance can account for over 80% of the
the fermentation stage, volumetric productivity and overall cost for a large-scale rotary filtration or
conversion yield will be most important. A continu- centrifugation plant (Atkinson and Mavituna,
ous process may be more economic than a batch 1983). However, it is considered that removing cells
process if higher productivities at high conversion by filtration will be less energy consuming than by
241
Principles of Fermentation Technology
centrifuging. Iffilter aids are to be used, in the most overall consumption. Ina bacitracin plant described
economical way, this will still add £9 ton! for a by Inskeep et al. (1951) the water from the mash
product at a 10% concentration in the broth. cooler was collected and reused to charge the mash-
When a product may be made by a microbial ing vessels and wash the fermenters. Water from the
process or obtained from an alternative source there cooler coils was used to wash down the discharge
are cost limits on product recovery to be considered. cake from the filter presses. Bernstein et al. (1977)
Atkinson and Mavituna (1983) have estimated: have designed a ‘closed-loop’ system for fermenting
cheese whey in which effluents were completely
1. A limit of about £40 ton! for ethanol selling at
£220 ton |produced at 7% w/v in broth from recycled. In large-scale SCP processes, recycle of
a local source of carbohydrate.
the process water is planned to minimize water
2. A limit of about £100 ton”! for SCP selling at consumption (10° m* day') for a 100,000-ton-
£300 ton”! produced from a petroleum-based year 1 plant, to reduce effluent-treatment costs,
substrate to give a 3% w/v yield. improve utilization of nutrients and improve overall
3. A limit of about £300 ton! for organic acids or yield (Ash and Topiwala, 1976; Hamer, 1979).
glycol selling at £700 ton~', produced at 10%
concentration in a carbohydrate medium.
EFFLUENT TREATMENT
When high-value end-products have been pro-
duced it has been acceptable to use relatively large
weights of filter aid to achieve initial clarification to In the majority of fermentation processes it is
remove small amounts of solids. The solvents used impossible to dispose of effluents at zero cost.
in subsequent extraction have then been recovered Whether the waste is incinerated, dumped on waste
in a high energy-consuming distillation plant (At- land, or discharged to sewers, rivers or tidal waters,
kinson and Sainter, 1982). In this case, the manufac- some expenditure will be necessary for treatment
turer has only to make his product as economically that ensures that minimal harm is done to the
as other fermentation companies. environment. In the United Kingdom, effluents
may be discharged directly into a river provided
that the biological oxygen demand is less than 20 mg
WATER USAGE AND RECYCLING dm** in 5 days, the suspended solids are less than 30
mg dm7’ and there is at least an 8-fold dilution of
the
discharged effluent (Royal Commission on Sewage
Many fermentations have a high daily water Disposal, 8th Report, 1912). It is obvious that very
usage (Table 12.6). As charges for water increase, few untreated fermentation effluents meet these
many of these processes will become vulnerable to stringent requirements.
cost escalation because of the relatively large vol- The various alternative disposal procedures may
umes of water required per unit volume of product. now be compared using economic considerations.
There is now a widespread interest in reducing Pape (1977) claimed that the cheapest treatment
242
Fermentation Economics
method was controlled dumping, followed by waste ArkInson, B, and Maviruna, F. (1983) Biochemical Engineering
incineration or dumping in salt mines. The most and Biotechnology Handbook, Chapter 12, pp. 890-931. Mac-
millan, London.
expensive method was biological degradation in a ATKINSON, B. and SainTer, P. (1982) The development of down-
waste-water-treatment plant. The last method has stream processing.J.Chem. Tech. Biotechnol. 32, 100-108.
often to be used because the effluents usually contain Aunstrup, K. (1977) Production of industrial enzymes. In
Biotechnology and Fungal Differentiation, FEMS Symp. 4, pp.
only a few percent of organic matter which would be 157-171 (Editors Meyrath, J. and Bu’Lock, J. D.).
costly to separate, concentrate and incinerate. Academic Press, London.
The possibility of direct disposal into the sea is Backuurst,J. R. and Harker,J. H. (1973) Process Plant Design.
Heinemann, London.
probably very limited although many of the large BARTHOLOMEW, W. H. and Retsman, H. B. (1979) Economics of
fermentation plants in the United Kingdom are in fermentation processes. In Microbial Technology (2nd edition),
coastal locations. In 1972 Jackson and Lines stated Vol. 2, pp. 463-496 (Editors Peppler, H. J. and Perlman,
D.). Academic Press, London.
that a pipeline of over 2.8 km overland and 2.8 km bE Brecze, G. and Lirsmann, A. J. (1944) Aeration in the
on the sea bed at a cost of £350,000 would be needed production of compressed yeast. Jnd. Eng. Chem. 36, 882-890.
to dispose of 3000 dm? day! of untreated antibiotic Brescu, S. C. (1952) Acetone—butanol fermentation of sugars.
Ind. Eng. Chem. 44, 1677-1682.
waste. The other options are to discharge the BERNSTEIN, S., 1zENG, T. H. and Sisson, D. (1977) The commer-
effluent direct to the sewers and pay a charge, to cial fermentation of cheese whey for the production of
treat the waste in the plant, or to operate a combina- protein and/or alcohol. In Single Cell Protein from Renewable
and Non-Renewable Resources, Biotech. Bioeng. Symp.’7, 1-10.
tion of the two. Sewage works’ charges for treating Bu’Lock,J. D. (1979) Industrial alcohol. In Microbial Technology:
effluents have increased 1000% in less than 5 years Current State, Future Prospects, Soc. Gen. Micro. Symp. 29,
in some instances (Forage, 1978). Those plants pp. 309-325 (Editors Bull, A. T., Ellwood, D. C. and
Ratledge, C.). University Press, Cambridge.
which treat all or part of their effluents have disco- Ca.am, C. T. (1967) Media for industrial fermentations. Proc.
vered that energy costs have risen and sludge dis- Biochem. 2(6), 19-22, 46.
Cavam, C. T. (1969) Automation in screening. In Fermentation
posal is more costly and difficult. Ripley (1979) has
Advances, pp. 34-41 (Editor Perlman, D.). Academic Press,
estimated the costs for a treatment plant for a New York.
whisky distillery producing 4,500,000 dm? year| of Cuerry, G. B., Kemp, S. D. and Parker, A. (1963) The steriliza-
proof whisky to have a capital cost of £75,000 and tion of air. Prog. Ind. Microbiol. 4, 37-60.
Cooney, C. L. and Maxicucui, N. (1977) An assessment of
operating costs of £9000 year'. Power was calcu- single cell protein from methanol-grown yeast. In Single Cell
lated at 0.9 kW kg| of BOD removed while dosage Protein from Renewable and Non-Renewable Resources, Biotech.
Bioeng. Symp.’ 7, 65-76.
of nutrients was £0.03 kg! ofeffluent BOD. CorsetT, K. (1980) Preparation, sterilization and design of
Before deciding on the most economic form of media. In Fungal Biotechnology, Brit. Mycol. Soc. Symp. 3, pp.
treatment, the water volume, the organic and solids 25-41 (Editors Smith,J. E., Berry, D. R. and Kristiansen,
B.). Academic Press, London.
loading, range of pH variation, nutrient level, tem- Daruincton, W. A. (1964) Aerobic hydrocarbon fermentation—
perature fluctuation (Ripley, 1979), company a practical evaluation. Biotech. Bioeng. 6, 241-242.
finance policies, the site location and government Dawson, P. S. S. (1977) Continuous fermentations. Ann. Rep.
243
PFT-Q
Principles of Fermentation Technology
Hamwonp, A. L. (1978) Energy: elements of a Latin American Meyratu, J. and Bayer, K. (1979) Biomass from whey. In
policy. Science, 200, 753-754. Microbial Biomass, Economic Microbiology, Vol. 4, pp. 207-269
Hastincs, J. J. H. (1971) Development of the fermentation (Editor Rose, A. H.). Academic Press, London.
industries in Great Britain. Adv. Appl. Microbiol. 16, 1-45. Mocren, H. (1979) SCP from methanol—the Norprotein pro-
Hastines, J. J. H. (1978) Acetone-butanol fermentation. In cess. Proc. Biochem. 14(3), 2-4, 7.
Primary Products of Metabolism, Economic Microbiology, Vol. 2, Moo-Younc, M. (1977) Economics of SCP production. Proc.
pp. 31-45 (Editor Rose, A. H.). Academic Press, London. Biochem. 12(4), 6-7, 9-10.
Hastincs, J. J. H. and Jackson, T. (1965) The technical use of Netson, T. C. (1961) Mutation and Plant Breeding, p. 331, Publ.
biochemical processes. In Chemical Engineering, Vol. 8, No. 891. Nat. Acad. Sci. Nat. Res. Council, Washington.
pp. 406-409 (Editors Cremer, H. W. and and Watkins, Nyirt, L. K. and Cuarzes, M. (1977) Economic status of
S. B.). Butterworths, London. fermentation processes. Ann. Rep. Ferm. Processes, 1, 365-381.
Hepner, L. (1977) Feasibility of producing basic chemicals by Pape, M. (1977) The competition between microbial and chemi-
fermentation. In Microbial Energy Conversions, pp. 531-554 cal processes for the manufacture of basic chemicals and
(Editors Schlegel, H. G. and Barnea, J.). Pergamon Press, intermediates. In Microbial Energy Conversion, pp. 510-530
Oxford. (Editors Schlegel, H. G. and Barnea, J.). Pergamon Press,
Hepner, L. (1978) The feasibility of basic chemicals by fermenta- Oxford.
tion processes. Engineering and Process Economics, 3, 17-23. PERLMAN, D. (1970) Some prospects for the fermentation indus-
Hocean, J. (1977) Aspects of fermentation in conical vessels.J. tries. Wallerstein Comm. 33, 165-173.
Inst. Brewing, 83, 133-138. PratTEN, C. F. (1971) Economies of Scale in Manufacturing Industry,
Humpureey, A. E. (1960) Air sterilization. Adv. Appl. Microbiol.. 2, Department of Applied Economics Occasional Papers 28.
301-311. University Press, Cambridge.
Humpureey, A. E. (1975) Product outlook and technical feasibil- QUEENER, S. and Swartz, R. W. (1979) Penicillins: biosynthetic
ity of SCP. In Single Cell Protein II, pp. 1-23 (Editors and semisynthetic. In Secondary Products of Metabolism,
Tannenbaum, S. R. and Wang, D. I. C.). MIT Press, Economic Microbiology, Vol. 3, pp. 35-122 (Editor Rose,
Massachusetts. A. H.). Academic Press, London.
Humpureey, A. E. and Gapen, E. L. (1955) Air sterilization by RaTLeDGE, C. (1975) The economics ofsingle cell protein produc-
fibrous media. Ind. Eng. Chem. 47, 924-930. tion; substrates and processes. Chem. Ind. (21), 918-920.
Humpnureey, A. E., Moreira, A., ARMIGER, W. and ZasrRIskKIE, D. RATLEDGE, C. (1977) Fermentation substrates. Ann. Rep. Ferm.
(1977) In Single Cell Protein from Renewable and Non-Renewable Processes, 1, 49-71.
Resources; Biotech. Bioeng. Symp. 7, 45-64. Reeves, R. V. (1952) Terramycin—from dirt to drug. Chem. Eng.
Inskeep, G. C., Bennett, R. E. Dubey,J.F. and SHEPARD, M. (New York), 59, 145-147.
W. (1951) Bacitracin, product of biochemical engineering. Ricwarps, J. W. (1968a) Introduction to Industrial Sterilisation.
Ind. Eng. Chem. 43, 1488-1498. Academic Press, London.
INSTITUTION OF CHEMICAL ENGINEERS (1977) A New Guide to Ricuarps,J.W. (1968b) Economics offermenter operation. Part
Capital Cost Estimation. Institute of Chemical Engineers, 1. Proc. Biochem. 3(5), 28-31.
London. Ricuarps,J.W. (1968c) Economics of fermenter operation. Part
Jackson, C. J. and Lings, G. T. (1972) Measures against water 2. Proc. Biochem, 3(6), 56-58.
pollution in the fermentation industries. In Industrial Waste RiGHELATO, R. C. (1980) Anaerobic fermentation: alcohol pro-
Water, pp. 381-393 (Editor Goransson, B.). Butterworths, duction. Phil. Trans. Roy. Soc. (London), B, 290, 303-312.
London. Riptey, P. (1979) Process engineering aspects of the treatment
Knecut, R., Prave, P., SEIPENBUSCH, R. and Sukatscu, D. A. anddisposal of distillery effluent. Proc. Biochem. 14(1), 8-10.
(1977) Microbiology and biotechnology of SCP produced Roxz, C. (1975) Utilization of cane and coffee processing by-pro-
from n-paraffin. Proc. Biochem. 12(4), 11-14. ducts as microbial protein substrates. In Single Cell Protein I,
Lawson, C. J. and SUTHERLAND, I. W. (1978) Polysaccharides. pp. 273-313 (Editors Tannenbaum, S. R. and Wang, D. I.
In Primary Products of Metabolism, Economic Microbiology, Vol. C.). MIT Press, Massachusetts.
2, pp. 337-392 (Editor Rose, A. H.). Academic Press, Royat Commission on SewacE Disposa, 8th Report (1912)
London. HMSO, London.
Levi,J. D., SHENNAN,J. L. and Epson, G. P. (1979) Biomass Ryu, D. Y. and Oxpsnue,J. Y. (1977) A reassessment of mixing
from liquid n-alkanes. In Microbial Biomass, Economic Micro- cost in fermentation processes. Biotech. Bioeng. 19, 621-629.
biology, Vol. 4, pp. 361-419 (Editor Rose, A. H.). Academic SCHIERHOLT, J. (1977) Fermentation processes for production of
Press, London. citric acid. Proc. Biochem. 12(9), 20-21.
LitcHFiELp, J. H. (1977) Comparative technical and economic ScHuGERL, K., Lucke,
J., LEHMANN,J.and Wacner, F. (1978)
aspects of single cell protein processes. Adv. Appl. Microbiol. Application oftower biorectors in cell mass production. Adv.
22, 267-305. Biochem. Bioeng. 8, 63-131.
Lockwoop, L. B. and Streets, B. W. (1966) Potentials in strain Smitu, S. R. L. (1980) Single cell protein. Phil. Trans. Roy. Soc.
development. Dev. Ind. Microbiol. 7, 74-78. (London, B, 290, 341-354.
MacLennan, D. G. (1976) Single cell protein from starch. In Spivey, M. J. (1978) The acetone/butanol/ethanol fermentation.
Continuous Culture 6: Applications and New Fields, pp. 69-84 Proc. Biochem. 13(11), 2-4, 25.
(Editors Dean, A.C. R., Ellwood, D. C., Evans, C.G. T. and Stark, W. H. and Pouter, G. M. (1950) Sterile air for industrial
Melling, J.). Ellis Horwood, Chichester. fermentations. Ind. Eng. Chem. 42, 1789-1792.
Mate tes, R. I. (1975) Production of SCP in Israel. In Single Cell STOWELL, J. D. and Bateson, J. B. (1984) Economic aspects of
Protein II, pp. 208-222 (Editors Tannenbaum, S. R. and industrial fermentations. In Bioactive Microbial Products, Vol.
Wang, D. 1. C.). MIT Press, Massachusetts. 2, Development and Production, pp. 117-139 (Editors Nisser,
Meerrs, J. L. (1980) Comment included in discussion of Smith L. J. and Winstantey, D.J.). Academic Press, London.
(1980). Swartz, R. W. (1979) The use of economic analysis ofpenicillin
244
Fermentation Economics
G manufacturing costs in establishing priorities for fermen- Wuirtaker, A. (1973) Fermentation economics. Proc. Biochem.
tation process improvement. Ann. Rep. Ferm. Processes, 3, 8(9), 23-26. .
75-110. Wuire, P. A. F. and Smrru, S. E. (1960) Removal of radioactive
Taytor, I. J. and Senror, P. j. (1978) Single cell proteins: a new particulates from air. Research, 8, 228-233.
source. Endeavour (N.S.), 2, 31-34. Wooprurr, H. B. and McDanir1, L. E. (1958) The antibiotic
Vass, R. C. and Jerrerys, E. G. (1979) Gibberellic acid. In approach. In The Strategy of Chemotherapy, Symp. Soc. Gen.
Secondary Products of Metabolism, Economic Microbiology, Vol. 3, Microbiol. 8, pp. 29-50 (Editors Cowan, S. T. and Rowatt,
pp. 421-433 (Editor Rose, A. H.). Academic Press, London. E.). University Press, Cambridge.
Wane, D. I. C., Cooney, C. L., Demain, A. L., DUNNILL, P. and Yamapa, K. (1977) Japan’s Most Advanced Industrial Fermentation
Litey, M. D. (1979) Fermentation and Enzyme Technology. Technology and Industry. International Technical Information
Wiley, New York. Institute, Tokyo.
245
-_
aval i q
j
' qi
ray DN, ‘
’ i= w by
« 12
a / :
'
' 2
i
e-* = ‘ ‘~-
i. r
. i
.
1 x
: ,
:
—
'
Index
249
Index
for recovery and partial purification of penicillin G 194 for streptomycetes 112
for recovery and purification of citric acid 195 for yeast 108
Fluorimeters' 163 Inosine monophosphate production 42
Foam Instrumentation 145-67
effect on oxygen transfer 190 Insulin 8, 65
non-foaming microbial strains 67 Interferon 8, 65
sensing probe 155 Isoenzyme control 36
Foam separation, for recovery of microbial cells, proteins and Isolation of micro-organisms 26
colloids 195-6 economics of 232-3
Foaming, patterns in fermentations 86-7 enrichment culture 27
Frings generator 138 solidified media, use of 29
Froude number 183 Itaconic acid, medium for production of 75
TEP Ltd 1S. 21, 65,140; 227, 233, 235, 237 Magnetic drives, for fermenters 127
Idiophase 4, 13 Maintenance phase 13
Idiotrophs 70 Market potential 233
Impellers 124 single-cell protein 234
tip speed 183 Mass spectrometers 163
types of Maximum specific growth rate
disc turbine 125 see Specific growth rate
marine 125 Measurement of process variables
multi-rod 125 agitator shaft power 155
open turbine 125 dynamometer 155
vaned disc 125 strain gauges 155
Improvement of micro-organisms 33-73 by fluorimeters 163
economics of 233 by mass spectrometers 163
producing primary metabolites 35 dissolved oxygen 158-9
producing enzymes 47 flow of gases 151-2
producing secondary metabolites 49 rotameter 151], 152
66 thermal mass flowmeter 152
of properties other than yield
47 flow of liquids 152-3
Induction
116, 131 electrical flow transducer 152
Inoculation of fermenters
Inoculum development ; load cells 153
109 - foaming probe 155
for bacteria
for fungi 111 pH 161
251
Index
252
Index
253
Index
254
Index
255
—_ 5
4 he we
bee") (qecmt
nt
eatin tap, te! of1 ine Rich) 2 —«Ss
é _
i rt eae - oa ~ » —
oe rola
Th ne ce Dern eal
é
“ti eeweeeres|
) paOanels
ayy
re Denaee
4raflda
aah
Wf
vi
y le TE, OF mien
\ 2a! Gare As ) «.? ill
off tueew yr Shaver eee oe det mT A ~h\en*? eee 6 ’ a
“ 19> cow4 1 on sul me] WG 2G Spee aeuee ag vata paherel
bison rename bing|
Pie
Nighi: 5 be, (que ay (ew VotW
€ :
em rieg: ree
» ar
=
2050
eas Lh,
cmarvarte
gua
tala
hfe ve vd
n iy
oy at ae
Tho! AP Serdalii)
ie the) »
‘i J
' i
mi 3 ¥
- ~—
‘=
as
= biwe =e t.
oan.7 7
ysis
S lo ensho deed ded
= ee) 68 eau ks
it haw o 4 kD) le Gre
a Parner sag So ee
) y - oon@? ite
+9 epmactd
2oer~e 1664 1 Le To iy mp
“a
yan
Principles. of
eT LLCLUCLL
Technology
The aim of the book is to provide an in-depth study of the
principles of fermentation technology. This is achieved by
considering the common features of fermentation systems
Mcoliai= mialeiamia’=e\-1(ellMelme ms= (= “melmiale|\i(e Ue] e)(e. eons
It will be of interest to final year and post-graduate students
o)f eee Biology, Biotechnology, Microbiology, Biochemical
and Chemical Engineering.
SO NEN IRS
PNamiayicere Uleiileamrom (= (an(-allelleaM eleeccns.)
Wilerro)e}fe] me |(e\uial (alse
ilarsmse)(elitelaym e)(= > 1n.e] fea me lao mmiag)|©)ce\(-108\-1a1me)!
axe lUiitre]| Manlio ree)rerelai saat
Wilsxelfo iro)aller (Uh itre|incocaat-aireliielays
Sterilization
alat=We(=\/=\(e) olaar= lal melm aeceu(om(omiareleiiie] fermentations
aBY-J(e|amelmem (= (ael-iai(=i
fasivlancairelloakelaekeo nice
PX=io)f(oame|are ele lire]s(o\n)
iets) recovery (o|ato o)0|)(exe) (ela)mi 1(aa\~ 1010] foal e\(ere|Uleiss
Effluent treatment
Fermentation economics