Professional Documents
Culture Documents
21butt 1
21butt 1
D.K. Button
Institute of Marine Science and Department of Chemistry, University of Alaska, Fairbanks, Alaska 99775, USA
Abstract
Traditional concepts of nutrient uptake and growth kinetics as linked by cell yield are presented. Phenomena
affecting the kinetics are examined along with a discussion of those which lead to ambiguity. Concepts of flux
control are presented to help understand the distribution of material along metabolic pathways. Specific
affinity is described to relate nutrient accumulation rates to transporter density. It is shown to be a primary
kinetic constant and the best available index of nutrient collection ability. As an aid to understanding, specific
affinity is reexpressed in terms of membrane permeability. Formulations of nutrient transport rate as a func-
tion of cellular composition, particularly transporter and enzyme content and known as janusian kinetics, are
described as an improvement to specific affinity theory. Procedures for quantified unidirectional fluxes are
reviewed to identify the difference between gross and net transport rates of substrate. Collision frequency
theory is used to show that in addition to total biomass, cell size and transporter density should also be in-
cluded in rate equations describing microbial growth. Theory diversity suggests that one reason for microbial
metabolic is that the likelihood of additional collisions of substrate molecules with a cell surface, after an
initial collision, requires only a sparse distribution of transporter sites for maximal rate, leaving room for
additional transporters able to collect other substrate types.
Growth kinetics and nutrient transport port and growth may be filtered through macro-
molecular processes having metabolic control that
Formulations of the kinetics of microbial growth varies with the specific rate of growth.
began with Monod (1950). He observed that the As techniques for measuring substrate uptake
specific rate of growth rate increased in a hyperbol- developed, it was found that these kinetics were hy-
ic way with the an external concentration of limiting perbolic as well, and transport was also described
nutrient. By analogy with enzyme kinetics the rela- by the Michaelis Menten analogy:
tionship is:
VlnaxA
v- - - (2)
~maxA (1) Kv+A
g- KM+A
where v and Vmax describe rates and maximal rates
where gmaxis the maximal growth rate, A is sub- of transport rather than growth and K a-is the associ-
strate concentration, and K Mis the Monod or half- ated Michaetis constant. Growth is then coupled to
maximal rate associated concentration (see p. 234 transport by the relationship
for definitions). The relationship is empirical and
vY
no theoretical base was proposed. The main defi- ~=--2- (3)
ciencies are that nutrient limitation implies kinetic
control by transport, and the linkage between trans-
226
where Y is the yield of cells of biomass X from sub- of the cell must be satisfied before substrate can
strate which gives be diverted to growth and, if true, render the ba-
sic form of most kinetic formulations insuffi-
(K-~--~A) (4) cient,
2. metabolic costs of nutrient limitation can shift
The theoretical base was presumed to be that trans- with substrate concentration,
porters limited the rate of growth and behaved like 3. changes in transporter content can change the
enzymes. Measurements of cell yield have shown rate of substrate transport at low concentration
that Y is not a constant, but a term that varies with while K T is unaffected because of a correspond-
the rate of growth. The change in yield with growth ing change in Vmax which may mask important
rate has been explained by endogenous metabo- changes in culture kinetics,
lism: maintenance energy devoted to active trans- 4. the whole-cell Michaelis constant K Mmay sim-
port, macromolecule replacement, replacement of ply reflect the ratio of transporters to enzymes,
leaked constituents, synthesis of materials to alle- and
viate cell damage during starvation, and metabolic 5. the relevant flux for describing growth in com-
control (Chesbro 1988). For carbon, this change in plex dilute systems is the sum of those flows con-
yield, extrapolated to a growth rate of zero, is some- tributing limiting material rather than from a
times used to evaluate endogenous metabolism. single substrate.
Yield is a strong function of growth rate for sub-
strates such as phosphate, silica, and vitamins where
the cells can continue to grow with reduced Saturation phenomena
amounts of the limiting nutrient in their biochem-
ical makeup. This observation led to the cell quota As the permeases of microorganisms and other en-
concept which has growth rate a function of the zymes of metabolism become occupied with sub-
amount of limiting nutrient within the cells (Droop strate, saturation phenomena ensue. For aquatic
1968). These formulations are inappropriate for systems ambient substrate concentrations seldom
carbon limitation because much of the transported attain concentrations sufficient to saturate these
carbon is excreted and usual changes in cell-carbon transporters, however experimental additions that
per unit biomass with growth rate are not large. For achieve saturation are useful to probe mechanisms.
mineral nutrients such as phosphate, the change in In the case of simple enzymes, the shape of the ki-
yield leads to a changing requirement for limiting netic curve is perfectly hyperbolic in theory and of-
nutrient per unit mass of cells produced. In addi- ten in practice as well. For whole cells, departure
tion, the accumulation of phosphate from dilute so- from this ideal shape has been used to evaluate a
lutions often leads to an increase in tranporters for number of phenomena such as natural substrate
the substrate which results in high transport capac- levels (Wright & Hobbie 1965) and multiple trans-
ity (Vmax).The large Vmaxvalues lead to large values porters (Azam & Hodson 1981). The kinetics for
for the transport constant K T (which depends on each of two individual transporters with different
Vm,x) relative to the Monod constant for growth K M Michaelis constants for transport K 1and K 2 can be
(which depends on ~l,max), Consequences include lin- evaluated as a simultaneous operation of a pair of
ear uptake with concentration up to the maximal transporters:
growth rate, and growth kinetic curves with inflec-
tions that lead to an apparent Michaelis constant
{ ~ [Vmax2A~
v+ + AI +
K Mthat is independent of saturation (Button et al.
1973). Other difficulties with these approaches to If the curve is biphasic, the individual kinetic con-
growth kinetics are that stants can be extracted (Neal 1972). However the
1. thresholds for growth are specified by the com- resulting kinetics which show more gentle curva-
mon assumption that endogenous requirements ture in the region of saturation, are similar to a corn-
227
bination of diffusion limitation and saturation ki- complete pathways can be deduced from those of
netics (Martinez et al. 1992). component enzymes by the simplified procedures
of Cleland (1975). For example, if k 2 << kl, k 3 the
Michaelis constant is simply K A = k ] k l as usually
Flux through pathways presumed. However, if k 3 << k 2 < k 1then K A = k3/kl,
and would (W.W. Cleland, pers. comm.) be analo-
Linkage between nutrient transport rate and gous to a situation where an oligotroph had many
growth rate may depend not only on nutrient accu- transporters but only few molecules of catabolic en-
mulation by transporters, but on the rate of utiliza- zymes. This leads to the conclusion that
tion of the accumulated nutrient through enzymatic 1. kinetic constants for long pathways may reflect
sequences as well. One might presume that overall the properties of a number of enzymes,
kinetics of microbial growth are controlled by the 2. changing the kinetic behaviour of a whole orga-
kinetics of formation of a particular intermediate in nism is difficult to effect by mutational or other
scant supply. Further that the formation rate of that adjustment of a single enzyme,
component is controlled by the slowest step in an 3. individual enzymes in short pathways are more
enzymatic sequence. Such control led to the idea of influential in controlling overall flux than in long
flux-generating enzymes (Newsholme & Crabtree ones, and
1981), enzymes that control flow at the beginnings 4. large ratios of transporter to limiting metabolic
or branch points of pathways and are particularly enzymes lead to small values of the transport
receptive to metabolic control. Another view is that constant Kr.
flux control is broadly distributed among pathway
components (Kell & Westerhoff 1986). This sharing
is quantified by flux control coefficients C which re- Specific affinity theory
late the influence of each enzyme on total pathway
flux. Thus the conversion of substrate to product This theory was derived to express the concept that
with various rate constants the affinity (ability of organisms to sequester) nutri-
ents reflected the density of transporters on the cell
tl c, c2 c3 surface, and that the resulting rate of growth de-
pended on the associated mass of cells, i.e. the sur-
A --~ E t --~ E z --+ E 3 ---) P
face to volume ratio, as well. Further considerations
I[J kl k2 k3
showed that specific affinity, as formulated below,
Scheme 1
rather than the Michaelis constant for transport KT,
along with maximal velocity for substrate accumu-
leads to some steady state flux j across the mem- lation, V .... are the primary kinetic constants which
brane. Without complexities such as those associ- describe that ability. The general case is that sub-
ated with the phosphotransferase mechanism, the strate uptake is given by the second-order rate
sum of the flux control coefficients Z C is often uni- equation
ty, according to the flux control summation theo-
rem, and the value of each coefficient here is 1/3 v = aAX A (6)
(Van Dam et al. 1993). It follows that individual en-
zymes in long pathways are less effective in control- which defines the specific affinity a a of the cells for
ling overall fluxj in a sequence than those in a short substrate A as the rate of substrate collection v/(X
pathway. These considerations reveal strategies for A). As A approaches zero, a n approaches the limit
the distribution of protein resources among path- a0
A and is the initial slope of equation 2 as well as the
ways and ultimately affect properties such as DNA/ initial slope of any v/X vs A curve whether hyper-
protein ratio and cell size. bolic or not.
Rate constants and overall kinetic constants for That specific affinity is a primary kinetic constant
228
If specific affinity from equation 6 is taken as the Transporter/enzyme ratio and janusian kinetics
forward rate c o n s t a n t k I between substrate and
cells, the kl is given by the collision rate between Increased transporter content of cells increases up-
molecules and cells (Abbot & Nelsetuen 1988), take as given by specific affinity theory. However,
down-stream enzymes can restrict flow if, for exam-
/4AD]
kl = \ 1-T0-0-O! n r x (14) ple, k2E2 of Scheme 1 is small. In laboratory experi-
ments a decrease in uptake rate over time for both
where r x is the radius of the cell. Both, substrate natural populations and for the typical marine oli-
concentration requirements and the maximal spe- gobacterium isolate RM 1 was observed indicating
cific affinitycan thus be computed directly from col- the importance of this phenomenon (unpubl. data,
lision frequency. To do this one assumes that the cell this laboratory). To quantify this effect one may as-
provides a perfectly absorbing surface. Then for a sume that the immediate result of transport is in-
typical marine bacterium that is approximately ternal or pool-nutrient concentration, that trans-
spherical, 0.3 gm in diameter and growing on glu- port is Michaelian, and that utilization is Michae-
230
first term of the denominator giving a cell-size de- utilize the products as fast as they are formed. Such
pendent rate constant. When small, it drops from losses result in a difference between gross transport
the equation giving the rate constant as a property rate as may be measured in short-term uptake ex-
of the transporter site. System properties described periments and net transport which contributes
by the relationships (Table 1) show that the basic more directly to growth. Rate constants connecting
kinetic formulations for whole cells are cell-size de- these unidirectional fluxes are shown in equation
pendent. set (18)
An inherent assumption in both specific affinity
theory and janusian kinetics is that the collision lim- A kl" Apool (18)
it does not apply because cell size is not accounted
for, only the number of transporters T. Thus if T/X
is small and T, which is constrained to the number of
hpool k3" Vpool
free sites, is decreased by increasing substrate, then
kf increases with rate because the number of free Vpool k5~ Vcell
transporters is reduced. Reduction of the number k"-:-~-
of available transporters can cause the kinetics to k7...
revert to transporter-dependent rates. Because of Ppooi .k 8 P
recollisions, maximal rates, called the collisional
limit, may be attained at relatively low site density The flux associated with the active transport of sub-
and rate can be formulated as a function of cell size. strate A is identified by Jl; that associated with sub-
Thus while transporters operate at maximal effi- strate leakage is J2. Pool products may be either used
ciency at low density, the overall nutrient capture for growth, Js, or leaked from the cell, JT-These uni-
rate is low. By increasing transporter density, the directional fluxes can be determined in continuous
rate is increased but diffusion time reduces the spe- culture while maintaining nutrient-limited growth
cific affinityfrom what it might otherwise have been at steady state (Button & Kinney 1987). For exam-
and increases KT. Experimental verification was ple, to determine the rate of substrate leakage and
obtained by using an Escherichia coli system in therefore net transport Jl-J2, isotope relaxation
which both the porin and internal enzyme concen- methods can be employed. The continuous culture
tration could be controlled (Martinez et al. 1992). reactor is supplied by two reservoirs of medium,
For this reason kinetic constants depended on identical in all respects, except that the substrate in
transporter and other enzyme content, which is in- one is radiolabeled. If the steady state is achieved
consistent with conventional theory, and curve from the radiolabeled substrate, and the supply is
shapes became nonhyperbolic at high enzyme con- suddenly switched, radioactivity in the medium is
centrations, presumably due to limitation by diffu- quickly lost. The loss rate depends on the biomass
sion. This helps to verify janusian kinetics and docu- in the reactor as controlled by the amount of sub-
ments limitations of conventional theory. strate A 0in the supply, and on the specific affinity of
the cells for the substrate. This rate of radioactivity
loss is known to the extent that the growth rate is
Leakage and relaxation methods equal to the dilution rate, a central assumption of
continuous culture, and the accuracy with extracel-
Important aspects of cell-leakage include i) loss of lular substrate concentration A is measured. Such
substrate actively deposited into the metabolic concentrations, it turns out, can be quite difficult to
pools, and ii) loss of substrate-derived metabolic measure (Law et al. 1976). A small and known
products from the pools to the environment either amount of substrate is lost due to washout from the
due to failure of the cell membranes to completely reactor as well. If the specific affinity is large, A is
retain these metabolites over the steep concentra- small, its residence time is short and of the order of
tion gradient produced, or inability of the cells to seconds, even if A 0 is small. But the residence time
232
of the cells is much longer, and equal to the rate of Use of kinetic data
dilution of the reactor. Thus the extracellular sub-
strate radioactivity soon depends almost entirely on Interpretation of kinetic data from microbial sys-
leakage Jr The set of equations (19) for determining tems is more difficult than for enzymatic systems
the rate of leakage of pool substrate, Apool , and pool because of additional constraints:
p r o d u c t , Ppool, formulated in terms of the specific 1. Rate limitation may involve diffusion either to
rate of growth, is given by an organism buried in a viscous matrix or to a
limiting enzyme buried within the organism.
(19) 2. Limiting steps may shift with concentration. For
g = X(j~-j2)/(Ao-A)
example at low concentrations, collision with
g = - X(js-jT)/P transporters may be limiting; at intermediate
B = [01 + J4)-(J2 + J3)] R A concentrations, collisions of transported nutri-
ent with internal enzymes may be limiting; and at
[.t - [03 + J6 + Js)-(J4 + J5 + J7)] R Ppool
high concentrations the turnover number of an
g = (is-J6)/R Pcell enzyme or the activity of a subcellular particle
such as a ribosome may control rate.
where the ratio R converts the mass of cells to the 3. As shown above, the nature of the rate equation
fraction occupied by metabolic pools. For the trans- may change with cell size and transporter densi-
port of phosphate by Rhodotorula rubra, pool or- ty.
thophosphate substrate and organic phosphate 4. Evaluation of the responsible microbial popula-
product leakage during phosphate-limited growth tion is necessary to give affinities for natural
was up to 5% and 10% of the phosphate transport populations a basis in biomass.
rate respectively, depending on the rate of growth Difficulties in evaluating microbial biomass have
(Robertson and Button 1980). In the case of toluene constrained environmental kinetics because micro-
metabolism, leakage of metabolic products such as bial populations and cell size have been hard to
2-hydroxy-6-oxo-heptadienoic acid can amount to measure. Flow cytometry is helpful for minimizing
most of the substrate accumulated (Button & Ro- this difficulty in the case of bacteria (Button et al.
bertson 1987), presumably because of inability of 1992), and good progress is underway for phyto-
the cells to either retain this amphipathic metabo- plankton (Olson et al. 1985; Chisholm 1992). Mole-
lite or actively transport it back in. In such cases the cularbiological methods offer potential for helping
difference between the measured rate of transport identify population components responsible for
and the net substrate flux is quite large. particular activities (Schmidt et al. 1991).
Janusian kinetics allow the formulation of kinetic
constants for microbial processes in molecular
terms (Button 1991) which are useful for under-
Table 2. Response of kinetic parameters to changes in composi- standing morphological constraints. For example
tion. they can be used to anticipate the effect of whole-
cell composition on kinetic properties (Table 2).
Compositional parameter Kinetic parameter It can be seen, for example, that increasing the
responding
transporter-to-biomass ratio might be useful for
Transporter mass/cell mass collecting nutrients (line 1), but that accumulation
Limiting enzyme x Transporter content Wmax capacity would suffer (line 2). Simplifications ig-
Limiting enzyme/Transporter content nore the kinetic interactions discussed above as
Biomass/Transporter content Allofunctional well as the influences of metabolic control. How-
overhead ~
Oligotrophic capacity log a0A/KT ever, it is suggested that kinetic constraints imposed
by quantities of catalytic components do reflect the
1Biomass devoted to functions other than nutrient collection. ability of organisms to collect and process nutrients,
233
and as such help describe the balance between dis- sites can reach the collisional limit when rather
solved nutrients and particulate material in the sparesely distributed within the cell envelope, the
form of microorganisms that process these nutri- space may advantageously be used by other types of
ents. sites. A diversity of sites (TA, TB, Tc.... ) may there-
fore collect a diversity of dilute substrates (A, B,
C...) faster than an equal number of total sites con-
Microbial ecology strained to a single molecular species. Multiple
transporter types can lead to a diversity of species.
A central question is the extent of microbial diversi- Since each species may focus on a particular combi-
ty in the presence of a single major limiting nutrient. nation of dilute substrates, the number of possible
We know from oligotrophic strategy that good oli- combinations exceeds the total number of nutrients
gotrophs lie at the extreme end of a log a0A -log K r collected. Multiple transporters must be connected
plot (Button 1991). This can be seen from the defini- to multiple metabolic systems which are costly to
tion of specific affinity where usual hyperbolic ki- build. However multiple systems can lead to higher
netics apply: nutrient-collection efficiency if the additional costs
of the extra transport and metabolic systems are
a ° = Vmax/KT (20) outweighed by increased nutrient flux. In addition
there can be cost advantage to transporting re-
and the assumption that the variation in Vmax is quired biochemical monomers over de novo syn-
small with respect to variation in K T. Logic of the thesis. Other costs of transport site diversity include
assumption is that required rates of major carbon/ small maximal growth rates from a single substrate,
energy source accumulation vary only approxi- pool imbalance, and substrate accelerated death
mately as much as the difference in maximal growth upon nutrient addition. However, these costs are
rates among species. Then equation 19 specifies a unlikely for many organisms in oligotrophic sys-
reciprocal relation between specific affinity and Mi- tems. When transporters are dilute, rates of trans-
chaelis constant and leads to a definition of oligo- port v of various substrates are given by their re-
trophic capacity (Table 2). The interpretation is that spective forward rate constants, transporter con-
good oligotrophs are small and have large transpor- centations, and substrate concentrations
ter/limiting-enzyme ratios for operation at small
limiting nutrient concentrations. VA + VB + Vc = kATAA + k BTBB + k c TcC >
For example Vibrio $14 with a specific affinity of 3kAA, 3kBB, 3kcC (21)
563 liters/g-cells- h for glucose and a transport con-
stant of 9.9 x 10-5 g/L (Akagi & Taga 1989) has an which leads to an overall average rate constant k s
0
oligotrophic capacity, log aA/KT, of 6.7. On the other for the total substrate concentration S. The inequal-
hand Pseudomonas sp. RP-386, a 0A= 2.5 and K T= 6.4 ity of equation 21 states the observed preference by
x 10-4 (Albertson et al. 1990) has an oligotrophic ca- oligotrophic organisms for simultaneous multisub-
pacity of only 3.5. Perhaps this pseudomonad de- strate collection and the collision frequency analy-
votes more material to the metabolism of glucose as sis gives a plausible reason. Namely, if additional ca-
well as the metabolism of other substrates relative tabolic system construction costs are outweighed by
to the number of glucose transporters formed. the additional substrate collection rate of a diverse
An advantage of diversity in nutrient transport cell surface, then a mixed diet is of substantial ad-
ability may lie in the colision frequency equations. vantage.
Consider that once a nutrient molecule collides
with a cell surface, it remains in the vicinity and of
the order to 1000 additional collisions are likely as
discussed above. Thus a colliding molecule is
marked and distinct from others. If the density of
234
Raw RAT, Robertson BR & Button DK (1976) On describing tinuous culture of Rhodotorula rubra: kinetics of transport,
microbial growth kinetics from continuous culture data. Some leakage and growth. J. Bacteriol. 138:884-895
general considerations, observations and concepts. Microb. Robertson BR & Button DK(1987) Toluene induction and up-
Ecol. 2:261-283 take kinetics and their inclusion in the specific-affinity rela-
Martinez MB, Sehendel FJ, Flickiger MC & Nelsetuen GL tionship for describing rates of hydrocarbon metabolism.
(1992) Kinetic properties of enzyme regulation in vivo: Alka- Appl. Environ. Microbiol. 53:2193-2205
line phosphatase of the E. coli periplasm. Biochemistry 31: Schmidt TM, DeLong EF & Pace NR (1991). Analysis of a ma-
11500-11509 rine picoplankton community by 16S rNA gene cloning and
Michaelis L & Menten MM (1913) Die Kinetik der Invertinwir- sequencing. J. Bacteriol. 173:4371-4378
kung. Biochem. Z. 49:333-369 Sen K, Hellman J & Nikaido H (1980) Porin channels in intact
Monod J (1950) Technique de culture continue. Theorie et appli- ceils of Escherichia coli are not affected by Donnan potentials
cations. Ann. Inst. Pasteur 79:390-410 across the outer membrane. J. Biol. Chem. 263:1182-1187
Neal J (1972) Analysis of Michaelis kinetics for two independent, Van Dam K, Postma P, Ruijter G, Ruthers M, van der Vlag J &
saturable membrane transport functions. J. Theor. Biol. 35: Westerhoff HV (1993) Control and regulation of metabolic
113-118 fluxes in microbes by substrates and enzymes. Antonie van
Newsholme EA & Crabtree B (1981) Flux-generating and regu- Leeuwenhoek (this issue)
latory steps in metabolic control. Trends Biochem. Sci. 6: 53- Wright RT & Hobbie JE. (1965) The uptake of organic solutes in
56 lake water. Limnol. Oceanogr. 10:22-28
Olson RJ, Vaulot D & Chisholm SW (1977) Marine phytoplank- Zimmermann RC, Kremer JN & Dugdale RC (1987). Acceler-
ton distributions measured using shipboard flow cytometry. ation of nutrient uptake by phytoplankton in a coastal upwell-
Deep-Sea Res. 32:1273-1280 ing ecosystem: A modeling analysis. Limnol. Oceanogr. 32:
Robertson BR & Button DK (1979) The Phosphate-limited con- 359-367