Accepted Manuscript

Title: Kinetic modeling of hyaluronic acid production in

<!–<query id="Q2">Your article is registered as a regular item
and is being processed for inclusion in a regular issue of the
journal. If this is NOT correct and your article belongs to a
Special Issue/Collection please contact
a.arulanandaraj@elsevier.com immediately prior to returning
your corrections.</query>–>palmyra palm (Borassus
flabellifer) based medium by Streptococcus zooepidemicus
MTCC 3523

Authors: Rohit Sheshrao Ghodke, Jyoti Prasad Kakati, Subbi

Rami Reddy Tadi, Naresh Mohan, Senthilkumar Sivaprakasam

PII: S1369-703X(18)30194-3
DOI: https://doi.org/10.1016/j.bej.2018.06.011
Reference: BEJ 6975

To appear in: Biochemical Engineering Journal

Received date: 9-2-2018

Revised date: 12-5-2018
Accepted date: 9-6-2018

Please cite this article as: Rohit SG, Jyoti PK, Subbi RRT, Naresh M, Senthilkumar S,
Kinetic modeling of hyaluronic acid production in palmyra palm (Borassus flabellifer)
based medium by Streptococcus zooepidemicus MTCC 3523, Biochemical Engineering
Journal (2018), https://doi.org/10.1016/j.bej.2018.06.011

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Kinetic modeling of hyaluronic acid production in palmyra palm
(Borassus flabellifer) based medium by Streptococcus
zooepidemicus MTCC 3523

Rohit Sheshrao Ghodke, Jyoti Prasad Kakati, Subbi Rami Reddy Tadi, Naresh Mohan,
Senthilkumar Sivaprakasam*

PT
BioPAT Laboratory, Department of Biosciences and Bioengineering, Indian Institute of
Technology Guwahati, Guwahati, Assam – 781039, India

RI
Authors:

SC
Rohit Sheshrao Ghodke (rohit.2014@iitg.ernet.in)
Jyoti Prasad Kakati (jyotiprasadkakati@gmail.com)

U
Subbi Rami Reddy Tadi (t.subbi@iit.ernet.in)
Naresh Mohan (l.naresh@iitg.ernet.in) N
Senthilkumar Sivaprakasam* (senthilkumar@iitg.ernet.in)
A
M

*Contact Details of Corresponding Author*:

Dr.Senthilkumar Sivaprakasam,
D

Associate Professor,
TE

Department of Biosciences and Bioengineering,

IIT Guwahati,
Guwahati, Assam, India-781039
EP

Email: senthilkumar@iitg.ernet.in
Tel: +913612582226
CC

Fax: +913612582249
A

1
Highlights

 First report on production of high value added HA using nutritionally rich PJ

 HA production from PJ based medium comparable to synthetic medium
 Kinetic studies on the influence of PJ on cell growth and HA production
 Techno-economic feasibility analysis of HA production

PT
RI
Abstract

SC
In this study, we evaluated palmyra palm sugar (PJ) formulated medium for HA production

U
using Streptococcus zooepidemicus MTCC 3523. Concurrently, effect of various carbon and
nitrogen sources, an analysis of the effect of initial sugar concentration, on HA production were
N
examined. The PJ - soya peptone based medium was found to be the most effective, and resulted
A
in 0.54 ± 0.08 gL-1 of HA titer and specific growth rate of 0.5 ± 0.07 h-1 in shake-flask
M
experiments. Kinetic experiments were performed in a bioreactor, and the initial PJ
concentrations were varied from 10 - 50 gL-1. Maximum HA and LA titers (1.22 gL-1 and 20. 2
gL-1) were obtained at 30 gL-1 PJ concentration in bioreactor. Han–Levenspiel substrate
D

inhibition model exhibited best fit for the experimental data. The growth inhibitory PJ
TE

concentration was found to be at 90.5 gL-1. Medium based process economic analysis proved PJ
and soya peptone based medium as most economical for the production of HA. The present
EP

investigation validates the potential of PJ and soya peptone based medium for enhanced
economical production of HA. This study will also provide new product dimensions and market
CC

avenues, to the marginalized palmyra palm farmers and toddy tapping community, who are
unable to secure the right price for their product.
A

Key words: Hyaluronic acid, Streptococcus zooepidemicus, Kinetic modeling, Substrate

inhibition, Designed biomass approach, Biopolymer

2
Nomenclature:

R2 Correlation coefficient (dimensionless)

Ks Monod saturation constant (gL-1)
Ki Inhibition constant (gL-1)
ms Maintenance coefficient (gg-1h-1)

PT
t Time (h)
rx (vx) Volumetric biomass growth rate (gL-1h-1)

RI
X0 Initial biomass concentrations (gL-1)
X Biomass concentration (gL-1)

SC
Xm Maximum biomass concentration at time (gL-1)
S Substrate concentration (gL-1)

U
S0 Initial substrate concentration (gL-1)
YX/S Biomass yield on substrate consumed (gg-1)
rs
N
Volumetric substrate consumption rate (gg-1h-1)
A
HA Hyaluronic acid concentration (gL-1)
M

LA Lactic acid concentration (gL-1)

rHA Volumetric HA production rate (gL-1h-1)
Volumetric LA production rate (gL-1h-1)
D

rLA
YP/X Product yield on biomass produced (gg-1)
TE

YP/S Product yield on substrate consumed (gg-1)

n, m Dimensionless numbers
EP

S* Inhibitory substrate concentration (gL-1)

SSE Sum of square error (dimensionless)
CC

EY Economic yield ($ of HA $ of nutrient-1)

EP Economic productivity ($ of HA $ of nutrient-1t-1)
Greek Letters
A

µ Specific growth rate (h-1)

µmax Maximum specific growth rate (h-1)
αHA Luedeking- piret model growth associated constant for HA (gg-1)
βHA Luedeking- piret model Non-growth associated constant for HA (gg-1h-1)
αLA Luedeking- piret model growth associated constant for LA (gg-1)

3
βLA Luedeking- piret model Non-growth associated constant for LA (gg-1h-1)
λx Log period of culture (h)

1. Introduction

PT
Hyaluronic acid (HA) is a linear hetero polysaccharide with recurrent repeating β D - glucuronic

RI
acid and β N-acetylglucosamine attached by β 1-3 and β 1-4 glycosidic linkages. This natural
non-sulfated glycosaminoglycan, found in connective and epithelial tissues, is part of signaling

SC
pathways and joint lubricants [1]. HA engenders extensive applications in biomedical and health
care industries through its high water-holding capacity, visco-elasticity and biocompatibility.

U
The biological activity and utility of HA is mainly dictated by its molecular weight [2]. Thus the
high molecular weight HA (> 2x 106 Da) is utilized in pharmaceuticals as a filling agent in
N
arthritis treatment, ophthalmic surgery and adhesion prevention after abdominal surgery.
A
Whereas the low molecular weight HA (0.8 -8 x 105 Da) is used in the cosmetic industry for
M

moisturizing and wound healing [3]. HA in conjunction with chondroitin and glucosamine, is
employed as a food supplement. HA bolstered by its plasticity, stability, bioavailability and
lower cytotoxicity, has extensive and preferential application in drug formulations and targeted
D

drug delivery [4].

TE

The utility of HA is compromised by its elevated risk of viral contamination and spiraling cost
of procurement. Thus HA extracted from animal sources such as rooster comb and bovine
EP

vitreous humor, poses a higher risk of viral contamination [5]. The steady increase in market
demand for HA, and the large supply to demand gap has contributed to its increasing cost.
CC

Pharmaceutical grade HA accounts for less than a ton, out of total demand for bulk HA; and is
marketed at USD 60,000/kg [6]. According to HA market analysis report 2016 (Transparency
A

market research, US) the estimated global HA market was 5.36 billion USD in 2012, and is
expected to reach 9.52 billion USD by 2019 [7].

In the last two decades, microbial production of HA has become the preferred alternative,
reinforced by lack of viral contamination, higher product yield, cheaper production cost and
highly purified product. Many wild microbial strains naturally produce HA as capsular
4
exopolysaccharide, as a part of their self defence against invading hosts. Examples of such
Strepococcus species are Streptococcus zooepidemicus, Streptococcus pyogenes, Streptococcus,
equisimilis, Streptococcus thermophilus and other organisms such as Pasteurella multocida and
Cryptococcus neoformans [8-12]. HA production has also been assessed after incorporating key
genes for HA enzymes into Bacillus subtilis, Lactococcus lactis, Escherichia coli,
Corynebacterium glutamicum, and Pichia pastoris [13-18]. However, the current industrial

PT
production of HA is mainly facilitated by Streptococcus species owing to its high HA titer (6 - 7
gL-1) under optimal conditions [19]. HA production from microbial sources is primarily

RI
influenced by the cost of the raw material. Consequently, exploration of economically viable
alternatives for renewable feedstocks and bioprocess technology platforms have gained

SC
momentum in industrial and academic sectors [20].
An aspect that is infrequently addressed is the Kinetic modeling of economically viable HA

U
production from potential agricultural feedstocks [6, 21-27]. Kinetic modeling is essential to

N
develop a fermentation process with optimal operational conditions for the production of the
metabolite of interest. In a bioprocess, the kinetic model is a set of relationships between
A
biomass growth, substrate utilization and product formation; employed to understand microbial
M

kinetics and to predict kinetic parameters. In large scale HA production, these kinetic variables
will become the key process parameters to be monitored and controlled though some reports
D

address kinetic modeling of the HA production using either pure sugars or renewable and
TE

economically viable feedstocks [21, 28-31]; very few reports substantiate effect of substrate
inhibition on the biomass growth.
To address these issues, we employed Palmyra palm jaggery (PJ) as potential feedstock for the
EP

microbial production of HA. PJ is traditionally made by the concentrating the sap extracted from
the palmyra palm trees (Borassus flabellifer). Naturally obtained PJ contains 65-85 % (w/w)
CC

sucrose and 5-15 % (w/w) reducing sugars, and is consumed directly or used for preparation of
sweet confectionary items and in traditional medicine. It is rich in vitamins (Nicotinic acid -
A

5.24, Thaiamin - 24.0, Riboflavin - 432.0 and Vitamin C - 11.0, (all expressed as mg/100g), and
minerals (Calcium -0.06 % (w/w), Phosphorus - 0.06 % (w/w), Iron – 0.0025 (gg-1)), vital to the
growth of an organism [32]. PJ has been employed in the production of value added products
such as polyhydroxybutyrate, citric acid, ethanol, D (-) lactic acid through microbial
fermentation [33-36]. PJ is a pre-treatment free feedstock, and has the potential to serve as an

5
alternative economic substrate for microbial production of high value products such as HA.
Further, utilization of PJ for the production of high value products would generate new revenue
options for the marginalized palm tapping farmers. It would thus create a valuable socio
economic dividend.
The aim of this study is to formulate a cost-effective medium employing palm sugar as a carbon
source for the production of HA. As an immediate and significant corollary, it would yield

PT
definitive insights on biomass growth, HA production, substrate consumption and substrate
inhibition in PJ based production medium. The kinetic parameters gleaned from this study

RI
would be immensely useful for effective process design leading to enhanced HA production.

SC
2. Materials and Methods

U
2.1 Raw material

N
PJ was obtained from cottage industries, Erode, Tamil Nadu, India. All the chemicals and
medium components used in this study were purchased from M/s Himedia Laboratories,
A
Mumbai.
M

2.2 Microorganism and Inoculum preparation

D

Streptococcus zooepidemicus MTCC 3523 was procured from Microbial Type Culture
TE

Collection (MTCC) and Gene Bank, Chandigarh, India. Streptococcus thermophilus NCIM 2904
was purchased from National Collection of Industrial Microorganisms (NCIM), Pune, India and
Streptococcus thermophilus STI- 12 was a kind gift from M/s. Chr. Hansen’s Laboratory, India.
EP

Inoculum preparation was carried out based on the method described for Streptococcus spp.
[37]. All the cultures were preserved in 25 % (v/v) glycerol at −80°C as stock. The preculture
CC

was prepared by inoculating 1 ml of glycerol stock into a 100 ml reaction volume of the M17
glucose medium [9]. The composition of M17 glucose medium used for preculture preparation
A

is provided in supplementary Table A. The flask was incubated in a rotary shaker at 37°C; 180
rpm for overnight. Inoculum (5% v/v) with an optical density ranging 0.6 – 0.9 was used to
inoculate in the production medium.
2.3 Growth medium and culture conditions

6
2.3.1 Shake flask experiments: Modified M17 medium used for the elite strain selection
studies had the following composition (gL-1): peptic digest of animal tissue, 2.5; casein enzyme
hydrolysate, 2.5; peptic digest of soyabean meal, 5.0; yeast extract, 2.5; beef extract, 5.0;
ascorbic acid, 0.5; magnesium sulphate, 0.25; disodium β glycerophosphate, 19.0; whereas
sucrose and PJ are used as alternative carbon sources at, 5.0 gL-1. In nutrient screening studies,
the combinatorial influence of nitrogen substrates viz., soya peptone, beef extract, casein

PT
enzyme hydrolysate, yeast extract and brain heart infusion were investigated with the alternate
carbon substrates (Glucose/Sucrose/PJ) as shown in supplementary Figure S1. The other

RI
mineral components of M17 medium were kept unchanged. Optimal carbon and nitrogen
substrates were identified for bioreactor and kinetic experiments. All the experiments were

SC
performed in 250 ml baffled Erlenmeyer flask containing 100 ml of the medium. Before
autoclaving initial pH of the medium was adjusted to 7.0. All the experiments were performed at

U
37 oC, 180 rpm and 1% (v/v) inoculum was used.

N
2.3.2 Bioreactor experiments: Kinetic experiments were carried out in a 3.5 L bioreactor
(Biojenik Engineering, Chennai, India) with a working volume of 2 L. Composition of
A
production medium for bioreactor experiments was as follows (gL-1): Optimal carbon source,
M

10.0 - 50.0; Optimal nitrogen source, 17.5; Ascorbic acid, 0.5; Magnesium sulphate, 0.25;
Disodium β glycerophosphate, 19. The pH was kept at 7.0 using 3M NaOH and 3M HCl. The
D

agitation rate and reactor temperature were set at 200 rpm and 37oC respectively. Dissolved
TE

oxygen value was maintained above the critical value (> 10% air saturation) and aeration rate
was maintained at 1 vvm for all the experiments. The production medium used in both shake
flask and bioreactor studies was autoclaved at 121oC for 20 min. The carbon source was
EP

autoclaved separately and mixed aseptically with other medium components before inoculation.
2.4 Kinetic model development
CC

2.4.1 Cell growth kinetics

The rate of biomass formation can be explained by the equation (1)

A

𝒅𝒙
𝑟𝑥 = = 𝜇𝑋 (1)
𝒅𝒕

Where, µ is the specific growth rate (h-1). In general, simple unstructured non-segregated kinetic
models like Monod’s and Logistic equations are used to describe biomass growth. Substrate

7
independent logistic model or Verhulset model is commonly used to study the sigmoidal
behavior of microbial growth, which describes the biomass growth in terms of carrying capacity.
Verhulset model has been extensively used to describe the growth in various fermentation
processes [38] as given in equation (2).

PT
𝒅𝒙 𝑿𝒎 −𝑿
𝑟𝑥 = = µ𝑚 𝑋 ( ) (2)
𝒅𝒕 𝑿𝒎

RI
Where, Xm is the maximum biomass (gL-1) and µm is the maximum specific growth rate (h-1).
Integrating equation (2),

SC
𝑿𝟎 𝑿𝒎 𝒆µ𝒎 𝒕
𝑋=𝑿 µ 𝒕 (3)
𝒎 −𝑿𝟎 +𝑿𝟎 𝒆 𝒎

U
Where, X0 is initial biomass concentration (gL-1),

N
Vázquez JA et al.[39] have further simplified the equation (3) by considering log period of
A
culture, λx, (h) and the maximum growth rate vx (gL-1h-1) resulted in equation (4)
M

𝑿𝒎
𝑋= 𝟒𝒗 (4)
𝟏+𝒆𝒙𝒑[𝟐+ 𝒙 (𝝀𝒙 −𝒕)]
𝑿𝒎
D

2.4.2 Substrate consumption kinetics

TE

In microbial fermentation processes, the substrate consumed is utilized for biomass growth,
product formation and cell maintenance activities. The overall mass balance equation for the
substrate utilization in a batch fermentation process is given in equation (5):
EP

−𝒅𝒔 𝟏 𝒅𝒑 𝟏 𝒅𝒙
-𝒓𝒔 = =𝒀 +𝒀 + 𝑚𝑠 𝑋 (5)
𝒅𝒕 𝑷/𝑺 𝒅𝒕 𝑿/𝑺 𝒅𝒕
CC

In the present study substrate flux towards HA production is low and hence, the product term is
neglected [21, 29] in the equation (5), which will reduce to
A

𝒅𝒔 𝟏 𝒅𝒙
= −𝒀 − 𝑚𝑠 𝑋 (6)
𝒅𝒕 𝑿/𝑺 𝒅𝒕

8
Where, YP/S product yield coefficient (gg-1), YX/S biomass yield coefficient (gg-1) and ms is the
maintenance coefficient (gg-1h-1). Integrating the equation (6) and substituting X (equation 4)
resulted the final expression for substrate consumption as a function of time is as follows:

4𝑣𝑥
𝑡
2 𝑋0 (𝑒 𝑋𝑚 −1)+𝑋𝑚
1 𝑋𝑚 𝑚𝑠 𝑋𝑚
𝑆 = 𝑆0 − 𝑌 [ 4𝑣 − 𝑋0 ] − ( ) 𝑙𝑛 [ ] (7)
𝑋/𝑆 1+𝑒𝑥𝑝[2+ 𝑥 (𝜆𝑥 −𝑡)] 4𝑣𝑥 𝑋𝑚
𝑋𝑚

PT
Where, S0 is initial substrate concentration (gL-1).

RI
2.4.3 Product formation kinetics

SC
The rate of HA and LA production can be explained by Leudeking – Piret equation

𝑑𝑝 𝑑𝑥
𝑟𝑝 = = 𝛼 + 𝛽𝑥 (8)

U
𝑑𝑡 𝑑𝑡

N
Where α and β are the Leudeking – Piret constants of growth associated and non-growth
associated product formation. Integrating the equation (8) and substituting the X (equation 4)
A
resulted in the final expression for product formation as a function of time is as follows:
M

4𝑣𝑥
𝑡
2 𝑋0 (𝑒 𝑋𝑚 −1)+𝑋𝑚
𝛼𝐻𝐴 𝑋𝑚 𝛽𝐻𝐴 𝑋𝑚
𝐻𝐴 = [ 4𝑣 − 𝛼𝐻𝐴 𝑋0 ] + ( ) 𝑙𝑛 [ ] (9)
D

1+𝑒𝑥𝑝[2+ 𝑥 (𝜆𝑥 −𝑡)] 4𝑣𝑥 𝑋𝑚

𝑋𝑚
TE

4𝑣𝑥
𝑡
𝑋0 (𝑒 𝑋𝑚 −1)+𝑋𝑚
𝛼𝐿𝐴 𝑋𝑚 𝛽𝐿𝐴 𝑋𝑚 2
𝐿𝐴 = [ − 𝛼𝐿𝐴 𝑋0] + ( ) 𝑙𝑛 [ ] (10)
EP

4𝑣
1+𝑒𝑥𝑝[2+ 𝑥 (𝜆𝑥 −𝑡)] 4𝑣𝑥 𝑋𝑚
𝑋𝑚

Where αHA and αLA are growth associated Leudeking – Piret constants (gg-1) of HA and LA
CC

respectively, βHA and βαLA are non -growth associated Leudeking – Piret constants (gg-1h-1) of
HA and LA respectively.
A

2.5 Kinetic parameters estimation

Data obtained from a set of batch fermentation experiments performed in bioreactor with
varying initial substrate concentration (10 - 50 gL-1) was used to estimate the kinetic parameters.
9
Here in the sum of squared errors (SSE) between models simulated data and experimental data
was minimized using non-linear least square method. All the kinetic parameters estimation and
fitting procedures were carried out using Microsoft Excel add on solver tool [39].

𝑆𝑆𝐸 = ∑𝑛𝑖=1(𝑋𝑠𝑖𝑚 − 𝑋𝑒𝑥𝑝 )2 (11)

Where xexp is the experimental x value, xsim is the model simulated x value and n is the number

PT
of observations.

RI
2.6 Analytical Methods

SC
The samples were drawn from shake flask and bioreactor at regular time intervals for the estimation of

U
biomass, substrate and HA concentrations. Biomass was estimated by UV spectrophotometer (Gene
Quant, GE Healthcare, UK) by using OD600 method. Measured OD values were correlated to Dry Cell
N
Weight (DCW) values using the observed relation (1 OD = 2.77DCW gL-1). Total sugar concentration
A
was estimated by Anthrone method. Glucose concentration was estimated by glucose oxidase-peroxidase
method using GOD-POD kit (Tulip Pharmaceuticals, Mumbai). Capsular HA was stripped from the
M

fermentation broth by the addition of an equal volume of 0.1% (v/v) SDS (Sodium dodecyl sulfate) and
incubated at 37ºC for 10 min, followed by a centrifugation at 5000×g for 10 min. and the supernatant was
D

retained. HA was precipitated by mixing with 4 - 6 volumes of absolute ethanol to the supernatant and
incubated at 4o C for 1 h. It was then centrifuged at 5000×g for 10 min to obtain HA pellet, which was
TE

washed with ethanol for 2 – 3 times and stored in saline at 4ºC overnight for solubilization. HA
concentration was estimated by CTAB (Cetyl tri methyl ammonium bromide) method [40]. LA
EP

concentration was estimated by HPLC (Shimadzu, Japan) methodology using refractive index detector.
Rezex RHM-monosaccharide H column (8%) (300×7.8 mm ID; 8 µm; Phenomenex Inc., Torrance, CA)
CC

was eluted by 0.5 mM H2SO4 at a flow rate of 0.5 ml/min. HA molecular weight was estimated by HPLC
gel permeation methodology using polysep-gfc-p-6000 (Phenomenex Inc., Torrance, CA) and refractive
index detector. Mobile phase of 0.1 M NaNO3 concentration was passed through the column at a flow
A

rate of 0.5 ml/min. Hyaluronic acid of various molecular weight standards viz., 0.6, 1.5 & 1.8 MDa
(Sigma–Aldrich, USA) were employed to obtain calibration plot representing the logarithmic function of
molecular weight against retention time.

3. Results and discussion

3.1 Strain selection

10
Designed biomass study was adopted to select an elite strain for HA production, using PJ as a
substrate. Natural HA producing Streptococcus strains S. zooepidemicus MTCC 3523, S.
thermophilus NCIM 2904 and S. thermophilus STI- 12 were initially used in the strain selection
study. Effect of PJ on HA production was also evaluated on strain performance by constituting
PJ-based medium (Supplementary Table A). Specific growth rate (h-1), HA yield (gg-1), HA titer
(gL-1) and biomass yield (gg-1) were considered for elite strain selection. Streptococcus spp are

PT
fastidious microbes, requiring complex nutrients for their growth; therefore PJ enriched in
nutrients and vitamins should support better microbial growth than pure sucrose based medium.

RI
This was evident from the experimental results (Table 1), where Streptococcus species have
shown relatively higher specific growth rate and HA titers on PJ based medium, in comparison

SC
to pure sucrose based medium. S.zooepidemicus grew at a specific growth rate of 0.32 ± 0. 12 h-
1
and 0.54 ± 0.1 h-1 on pure sucrose and PJ based medium respectively. S.thermophilus strains

U
exhibited relatively lower specific growth rate ranging from 0.15 – 0.19 h-1 when grown on the

N
same media. The divergence in growth rates is in accordance with reports on S.thermophilus
[41] and S.zooepidemicus [2] .S.zooepidemicus MTCC 3523 showed higher HA productivity,
A
(0.41 ± 0.15gL-1 ) compared to S.thermophilus NCIM 2904 (0.21 ± 0.13 gL-1 ) and
M

S.thermophilus STI -12,( 0.180 ± 0.09 gL-1) under optimal growth conditions. S. thermophilus is
a well known dairy industry starter culture, where it is an acknowledged as higher producer of
D

lactic acid than HA [12]. Similar values were reported for S.thermophilus (0.208 gL-1) grown on
TE

skimmed milk [12, 41]. The lower HA titer incase of S.thermophilus could be attributed to the
diversion of carbon flux towards LA pathway, which is the conventional byproduct of the
organism. Furthermore, S. thermophilus is known to produce different types of hetero
EP

polysaccharides composed of N-acetylglucosamine [42], which is one of the precursors of HA.

The competition of this precursor for the growth, production of HA and exopolysaccharide
CC

could also led to low HA titer.

But, the has operon of S. zooepidemicus consists of five genes hasABCDE, which are involved
A

in precursor synthesis of UDP-Glucuronic acid (hasB and hasC) and UDP-N-acetylglucosamine

(hasD and hasE). Further, hasA codes for hyaluronan synthase, which binds both these
precursors leading to HA synthesis. But, Streptococcs thermophillus consists of has operon
encoding only hasA, hasB and hasC. This suggests that the high HA titer observed for S.
zooepidemicus correlates with the availability of UDP-sugar precursors [5]. The HA yield

11
values obtained in this study using S.zooepidemicus are well corroborated with the previous
reported literature on different raw feed stocks such as whey (0.13 gL-1), soya protein
hydrolysate (0.17 gL-1), cashew apple juice (0.89 gL-1) and pure glucose (0.85 gL-1) based media
[6]. Based on HA production and product yield (Table 1) S.zooepidemicus MTCC 3523 found
to be an elite strain for the production of HA using PJ based medium.
3.2 Effect of carbon and nitrogen substrate on HA productivity

PT
The selected elite strain S. zooepidemicus MTCC 3523 was further used to identify a suitable

RI
nitrogen source for improved HA yield. Frequently, cited nitrogen substrates for HA production
like soya peptone, yeast extract, beef extract, casein enzyme hydrolyte and brain heart infusion

SC
were employed for screening experiment [6, 43]. Influence of carbon source from PJ and sucrose
on HA production with individual nitrogen source was also evaluated (supplementary Figure

U
S1). Specific growth rate (h-1), residual substrate concentration (gL-1) and HA titer (gL-1) values
were used to identify the optimal carbon and nitrogen substrate.
N
A
Among all the nitrogen and carbon substrates investigated, glucose - brain heart infusion
combination resulted in relatively high specific growth rate, 0.59 ± 0.13 (supplementary Table
M

B) in comparison to HA titer of 0.54 ± 0.08 gL-1 obtained for PJ –soya peptone combination.
The medium formulated using casein enzyme hydrolysate resulted in very poor growth rate and
D

HA titer for all the carbon sources tested. These major differences in the kinetics of biomass
TE

growth and HA titer could be due to the different type of nitrogen sources used in the HA
fermentation. Nitrogen sources act as supplements to the different vitamins, amino acids and
EP

peptides, required for efficient trans membrane transport and biomass growth. Moreover peptide
transport (e.g. di, tri or tetrapeptides) could be more efficient than transport of the individual
amino acids [44].Fastidious lactic acid bacteria like S.zooepidemicus, has a complex nutritional
CC

requirement for growth and HA production. Armstrong et al.[9] have identified that 11 amino
acids are very vital for S. zooepidemicus growth. Liu et al. [45] have reported enhanced cell
A

growth and HA production with arginine and lysine supplemented medium.

Biomass yield and HA were found to be relatively higher for PJ based medium, than for sucrose
and glucose based medium. This could be due to the vitamins and minerals originally present in
the PJ. Improved HA titers were reported with the medium supplemented with mineral salts of

12
Sodium and Calcium [46]. S. zooepidemicus effectively metabolize PEP-dependent
phosphotransferase system involving ScrA, ScrB and ScrK genes for conversion of 1 mole of
sucrose into 1 mole of each glucose-6-phosphate and fructose-6-phosphate, which are then
converted into HA precursors. The low HA titer is attributed to the requirement of two moles of
glucose, required for achieving similar conversion [2, 27, 47, 48]. Patil KP et al [3] have
reported enhanced HA production in shake flask for S. zooepidemicus MTCC 3523, using

PT
glucose and soya peptone in the medium. The statistically optimized medium resulted in 0.8 gL-1
of HA production, whereas an extended supplement of soya peptone did not increase HA yield.

RI
Kotra SR et al [49] have also optimized glucose and yeast extract for the HA production using
wild and mutated strains of S. zooepidemicus MTCC 3523; observed an improved HA

SC
productivity of 0.98 and 1.32 gL-1 respectively. In this study, S. zooepidemicus MTCC 3523
yielded HA titer of 0.50 ± 0.05 and 0.38 ± 0.09 gL-1 for glucose –soya peptone & glucose yeast

U
extract respectively. Thus, the high HA yielding combination of PJ – soya peptone was selected

N
for kinetic study under controlled condition in a bioreactor.
A
3.3 Kinetic modeling
M

3.3.1 Microbial growth

In order to understand the growth of S. zooepidemicus on aerobic fermentation of PJ, a series of

D

batch reactor experiments were carried, out by varying initial substrate (PJ) concentrations
TE

ranging from 10 – 50 gL-1. Experimental data obtained was fitted to simplified biomass
expression(equation 4), and the estimated kinetic parameter values were presented in Table 2.
EP

The highest biomass concentration obtained was 8.99 gL-1, which corresponded to a PJ
concentration of 30 gL-1. Simulated profiles of biomass growth at different intial PJ
concentrations were depicted in Figure 1A-E. An increase in the lag time from 3.75 h to 5.8 h,
CC

in relation to increased initial PJ concentrations was also observed. Vázquez JAet al [21, 22]
have also reported a lag time of similar magnitude for S. zooepidemicus cultivated on different
A

renewable marine derived complex substrates. Don MM et al [29] have reported only a 2 h lag
phase for S. zooepidemicus grown on pure glucose. It is slightly higher in case of PJ based
medium, which could be attributed to the complex nature of the substrate and the time taken by
the organism to synthsize new enzymes to metabolize the complex substrate. There is an
increase in the lag time from low to high PJ concentrations used, which could be due to the

13
inhibitory effect of substrate on growth. Aftera lag phase of 3 – 5.8 h, an exponential growth
phase follwed till 10 hr; subsequently, a deceleration phase was observed from 10 – 12 h.
Growth rate decreased in the stationary phase with small flucations from 12 - 16 h. Correlation
coefficient (R2) and SSE were used to verify the goodness of the fit of the model to the
experimental data. The higher the discrepancy between experimental data and simulated data
resulted in high SSE and smaller the R2 [50]. From Table 2 it is evident that R2 = 0.99 and it

PT
observed to have a good agreement between experimental and simulated biomass data.

RI
3.3.2 Product formation

Effect of initial substrate concentration on HA and LA titer (gL-1) was studied using logistic

SC
equation incorporated Leudeking – Piret model (equation 9 & 10). Experimental data obtained
was used to estimate the model kinetic parameters. Satisfactory fitting results were obtained

U
with correlation coefficient (R2), which is above 0.96. The estimated non-growth associate terms

N
of HA (βHA) and LA(βLA) were found to be almost zero (Table 2) This indicates that both are
good growth associated products and corroborate well with other reports [28, 29]. The growth
A
associated constants for both the products αHA and αLA can be considered as their respective
M

yield factors based on biomass (YHA/X and YLA/X). The growth associated constant for HA (αHA)
varied from 0.13 -0.08 gg-1, whereas αLA varied from1.98 to2.44 gg-1. When initial PJ
D

concentration was increased above 20 gL-1, no significant increase in LA production was

TE

observed (Table 2).This could be due to the saturation of glycolytic enzymes and rerouting of
carbon flux towards HA synthesis [29]. The maximum HA titer obtained was 1.22 gL-1 for 30
gL-1 initial palm sugar concentration, above which there was a fall in both HA titer and growth
EP

associated constant (αHA = YHA/X). It indicates that there was an inhibitory effect of initial
substrate concentration on HA production. A 3.5 to 5.8 h delay in HA production corresponding
CC

to lag period (λx, h-1) of growth can be seen in Figure 1 A-E. Don MM et al [29] have reported a
time delay of 2.16 h-1 for HA production using S. zooepidemicus on glucose. Increase in the lag
A

period could be due to the complexity of the carbon source present in the medium, in a microbial
system it is an established fact that simple sugars get assimilated faster than the complex.
Variation in the HA molecular weight with respect to initial PJ concentration was shown in
Figure. 2. Molecular weight of HA slightly increased from 0.93 MDa to 0.96 MDa for an
increase in PJ concentration from 10 - 20 gL-1. Further increase in PJ concentration resulted in a

14
sharp lump in molecular weight of HA. A similar trend was observed for HA fermentation by
S.zooepidemicus when grown on glucose and other raw feedstocks [9, 29]. Low HA molecular
weight also could be due to the presence of ascorbate in the culture medium. Hydroxyl radicals
generated from ascorbate could involve in depolymerization of HA [51]. S. zooepidemicus
metabolic activity is highly susceptible to the external factors such as pH, temperature, aeration,
agitation and type of carbon and nitrogen substrate used, which resulted in production of HA of

PT
varying molecular weights [2, 9]. Fall in HA molecular weight could be also attributed to higher
osmotic pressure above 20 gL-1 PJ. This could be due to the channeling of the key precursors

RI
into other metabolic pathways. Balancing the precursors (UDP–GA, UDP-GluNac)
concentration is vital to control HA molecular weight. Appropriate balance between the

SC
precursors resulted in higher molecular weight HA, where as their excess did not favor
production of high molecular weight HA [40, 52]. UDP-GluNac has a pronounced positive

U
effect on HA molecular weight at low concentration, but at high concentrations it influences

N
negatively. Phosphoglucoisomerase and carbon flux that drives the process towards LA
production are the major bottle necks in high molecular weight HA production [53].
A
M

3.3.3 Substrate consumption

In this present study, PJ is the primary energy source for S. zooepidemicus growth, maintenance
D

and also provides the precursors for HA production. Batch fermentation data obtained was used
TE

to fit in the simplified substrate consumption expression (equation 7). The estimated kinetic
parameters were presented in Table 2. Simulated substrate consumption profiles were depicted
in Figure 1 A-E. In HA fermentation, high PJ (> 30 gL-1) and LA (> 20 gL-1) concentrations lead
EP

to inhibitory effect on biomass growth, which could be attributed to low substrate substrate
uptake. Biomass yield on substrate (YX/S) is not constant during the fermentation process, a fall
CC

in biomass yield from 0.48 gg-1 to 0.19 gg-1was observed for change in initial PJ conc. from 10
gL-1 to 50 gL-1. Moreover a seven fold increase in the cellular maintenance energy from 0.01 to
A

0.07 gg-1h-1 was also observed (Table 2). This indicates that a significant inhibitory effect of
initial PJ concentration on S. zooepidemicus specific growth rate. Experimental biomass data
was used to calculate specific growth rates at different initial PJ concetrations. The estimated
specific growth rate values for 10, 20, 30, 40 and 50 gL-1 of initial PJ concentration are 0.27,
0.31, 0.34, 0.31 and 0.26 h-1 respectively. The fall in specifc growth rate values with increase in

15
the initial substare concentration and decrease in HA yield also substantiates inhibitory effect of
substrate on biomass growth (supplementary Figure S2). Estimated specific growth rate values
were further used to study the substrate inhibition on growth employing various substrate
inhibiton growth models (Tabel 3). Kinetic parametres for these substrate inhibition models
were estimated and presented in supplementary Table C. Goodness of the fit was estimated
using R2 and SSE values. Han - Lenvenspiel model shown higher R2 value of 0.91 and least SSE

PT
of 4.87 x 10-4, whereas competitive inhibition and Moser inhibition models had shown very poor
fit (supplementary Figure S2 A-J). Further, the values of Han - Lenvenspiel model constants (n,

RI
m) are almost equal (0.8) and greater than zero. It shows that the inhibition is uncompetitive

SC
type ((m = n) > 0) [53]. In addition to this, the estimated kinetic parameters of Han- Lenvenspiel
model, maximum specific growth rate 0.54 h-1 is corroborated with the experimentally
determined specific growth rate. Using Han- Lenvenspiel model, the inhibitory concentration of

U
PJ is estimated to be 90.45 gL-1. Don MM et al [29] studied the inhibitory effect of glucose on
N
HA production and S. zooepidemicus growth; identified Aiba and Han- Lenvenspiel as the best
A
models to study the inhibitory effect of glucose on S. zooepidemicus. The inhibitory
concentration of glucose was also obsreved as 86.86 and 71.92 gL-1 respectively. Jagannath S et
M

al [2] also repoted the fall in the S. zooepidemicus specific growth rate when initial substrate
(glucose and sucrose) concentration was doubled from 30 gL-1 to 60 gL-1. Moreover a decrease
D

in the specific LA production rate and increase in specific acetic acid production rate were also
TE

obserevd with increase in the initial sucrose concentration, which resulted in low specific growth
rate.
EP

3.4 Economics of HA production

Techno –economic feasibility of HA fermentation [54] was analyzed by predicting economic

CC

yield (EY) and economic productivity (EP). EP and EY of HA production on different carbon
and nitrogen substrate based media were calculated using equations 12 and 13 from the data
A

presented in Table 2.
𝒀
𝐸𝑐𝑜𝑛𝑜𝑚𝑖𝑐 𝑦𝑖𝑒𝑙𝑑, 𝑬𝒀 = ∑(𝑪𝑵) 𝑰 (12)
𝒀
𝐸𝑐𝑜𝑛𝑜𝑚𝑖𝑐 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑣𝑖𝑡𝑦, 𝑬𝑷 = 𝑰 (13)
𝒕 ∑(𝑪𝑵)

16
Where, Y is HA concentration gL-1, I is selling price of HA ($.g-1), C is the cost associated with
nutrients $.g-1, N is the nutrient concentration gL-1, t is fermentation time, h. Since carbon and
nitrogen substrates are the major cost factors in the formulated medium, they were considered in
economic analysis. The detailed cost information of individual medium components and HA
were taken from the online bulk chemical trading site (Alibaba.com) and presented in
supplementary Table D. Calculated economic yields and productivities were depicted in Figure

PT
3A and 3B. It is evident that medium formulated with PJ has shown high EP and EY over pure
glucose and sucrose based medium. Casein enzyme hydrolysate has resulted in very poor EP and

RI
EY values, whereas significantly high EP and EY values were found for soya peptone in
combination with glucose, sucrose and PJ. Soya peptone - PJ combination resulted in highest

SC
EY (3.02 $ of HA ($ of nutrient)-1) in shake flask experiments. Whereas an economic yield (EY)
of 1.418 $ of HA ($ of nutrient)-1) was observed from the literature report for HA production in

U
statistically optimized glucose – soya peptone based medium using S. zooepidemicus MTCC

N
3534 [55]. Yeast extract – PJ and soya peptone – PJ based medium resulted in an EP of 0.4 and
0.38 $ of HA $ of nutrient-1h-1 respectively. EP values of casein enzyme hydrolysate were less
A
than 1 in all the carbon sources employed in this study, which substantiates that HA production
M

using casein enzyme hydrolysate is not economical.

Hence, the calculated EP and EY values would provide insight on the selection of raw material
D

for economical HA production. In order to study the change in EP during the course of the
TE

fermentation, dynamic HA data obtained during the kinetic study (Figure 1C) was used and
results were shown in Figure 4. Extended fermentation resulted in reduced EP. Though there is
a slight improvement in HA between 10 h to 12 h, a drastic fall in EP can be seen after 10 h.
EP

Optimal EP (0.515 $ of HA $of nutrient-1h-1 ) was obtained at 10 h, therefore beyond 10 h

period, the HA production process is not economical. Therefore based on these considerations, a
CC

fermentation time of 10 h and the fermentation medium containing PJ-soya peptone combination
for the production of HA using S. zooepidemicus MTCC 3534 would be economical.
A

4. Conclusion

One of the vital economic factors in the large scale production of HA is the cost of the raw
material. In this context, an economical medium was formulated successfully using PJ and soya
peptone for the production of HA. The logistic model derived substrate consumption and HA

17
production models were adequately fitted to the experimental data with varying initial PJ
concentration (10 -50 gL-1). Han- Lenvenspiel models showed the best fit for the substrate
inhibition study. The estimated growth inhibitory PJ concentration was 90.45 gL-1. This kinetic
study would be useful for optimal medium formulation and improved process design for
enhanced HA production at the technical scale. Further optimization of physical process
parameters (temperature, agitation rate, dissolved oxygen) and developing effective process

PT
strategies for high cell density cultivation would facilitate cost effective and enhanced
production of HA in large scale operation.

RI
Acknowledgements

SC
Authors are would like to thank Union Ministry of Human Resources and Development
(MHRD), Govt. of India, New Delhi, India for stipend. Authors are very much grateful to the

U
Council of Scientific and Industrial Research (CSIR),Govt. of India, New Delhi, India for their

N
kind grant (Grant No: 22(0657)/14/EMR-11) to install bioreactor facility in our laboratory.
Authors extended their gratitude for the support rendered by the Department of Biosciences and
A
Bioengineering, IIT Guwahati. Authors acknowledge Prof. Gaurangi Maitra for proof reading
M

the manuscript.
D
TE
EP
CC
A

18
References

[1] C. Schiraldi, A. La Gatta, M. De Rosa, Biotechnological production and application of

hyaluronan, in: Biopolymers, InTech, 2010, pp. 387–412.
[2] S. Jagannath, K. Ramachandran, Influence of competing metabolic processes on the
molecular weight of hyaluronic acid synthesized by Streptococcus zooepidemicus, Biochem.

PT
Eng. J., 48 (2010) 148-158.
[3] K.P. Patil, D.K. Patil, B.L. Chaudhari, S.B. Chincholkar, Production of hyaluronic acid from
Streptococcus zooepidemicus MTCC 3523 and its wound healing activity, J. Biosci. Bioeng.,

RI
111 (2011) 286-288.
[4] G. Kogan, L. Šoltés, R. Stern, P. Gemeiner, Hyaluronic acid: a natural biopolymer with a
broad range of biomedical and industrial applications, Biotechnol. Lett., 29 (2007) 17-25.

SC
[5] C.G. Boeriu, J. Springer, F.K. Kooy, L.A. van den Broek, G. Eggink, Production methods
for hyaluronan, Int. J. Carbohydr. Chem, 2013 (2013).
[6] A.M. Pires, A.C. Macedo, S.Y. Eguchi, M.H. Santana, Microbial production of hyaluronic

U
acid from agricultural resource derivatives, Bioresour. Technol., 101 (2010) 6506-6509.
[7] J.D. Oliveira, L.S. Carvalho, A.M.V. Gomes, L.R. Queiroz, B.S. Magalhães, N.S. Parachin,
N
Genetic basis for hyper production of hyaluronic acid in natural and engineered microorganisms,
Microb. Cell Fact., 15 (2016) 119.
A
[8] P.L. DeAngelis, Enzymological characterization of the Pasteurella multocida hyaluronic acid
synthase, Biochemistry, 35 (1996) 9768-9771.
M

[9] D. Armstrong, M. Cooney, M. Johns, Growth and amino acid requirements of hyaluronic-
acid-producing Streptococcus zooepidemicus, Appl. Microbiol. Biotechnol., 47 (1997) 309-312.
[10] V.L. Tlapak-Simmons, B.A. Baggenstoss, T. Clyne, P.H. Weigel, Purification and Lipid
D

Dependence of the Recombinant Hyaluronan Synthases from Streptococcus pyogenes and

Streptococcus equisimilis, J. Biol. Chem., 274 (1999) 4239-4245.
TE

[11] A. Jong, C.-H. Wu, H.-M. Chen, F. Luo, K.J. Kwon-Chung, Y.C. Chang, C.W. LaMunyon,
A. Plaas, S.-H. Huang, Identification and characterization of CPS1 as a hyaluronic acid synthase
contributing to the pathogenesis of Cryptococcus neoformans infection, Eukaryotic Cell, 6
(2007) 1486-1496.
EP

[12] N. Izawa, T. Hanamizu, R. Iizuka, T. Sone, H. Mizukoshi, K. Kimura, K. Chiba,

Streptococcus thermophilus produces exopolysaccharides including hyaluronic acid, J. Biosci.
Bioeng., 107 (2009) 119-123.
CC

[13] E. Jeong, W.Y. Shim, J.H. Kim, Metabolic engineering of Pichia pastoris for production of
hyaluronic acid with high molecular weight, J. biotechnol., 185 (2014) 28-36.
[14] H. Yu, G. Stephanopoulos, Metabolic engineering of Escherichia coli for biosynthesis of
hyaluronic acid, Metab. Eng., 10 (2008) 24-32.
A

[15] F. Cheng, Q. Gong, H. Yu, G. Stephanopoulos, High titer biosynthesis of hyaluronic acid
by recombinant Corynebacterium glutamicum, Biotechnol. J, 11 (2016) 574-584.
[16] L.-J. Chien, C.-K. Lee, Hyaluronic acid production by recombinant Lactococcus lactis,
Appl. Microbiol. Biotechnol., 77 (2007) 339-346.
[17] B. Widner, R. Behr, S. Von Dollen, M. Tang, T. Heu, A. Sloma, D. Sternberg, P.L.
DeAngelis, P.H. Weigel, S. Brown, Hyaluronic acid production in Bacillus subtilis, Appl.
Environ. Microbiol., 71 (2005) 3747-3752.

19
[18] Z. Mao, R.R. Chen, Recombinant synthesis of hyaluronan by Agrobacterium sp,
Biotechnol. Prog., 23 (2007) 1038-1042.
[19] L. Liu, Y. Liu, J. Li, G. Du, J. Chen, Microbial production of hyaluronic acid: current state,
challenges, and perspectives, Microb. Cell Fact., 10 (2011) 99.
[20] J.H. Sze, J.C. Brownlie, C.A. Love, Biotechnological production of hyaluronic acid: a mini
review, 3 Biotech, 6 (2016) 67.
[21] J.A. Vázquez, M.I. Montemayor, J. Fraguas, M.A. Murado, Hyaluronic acid production by
Streptococcus zooepidemicus in marine by-products media from mussel processing wastewaters

PT
and tuna peptone viscera, Microb. Cell Fact., 9 (2010) 46.
[22] J.A. Vázquez, L. Pastrana, C. Piñeiro, J.A. Teixeira, R.I. Pérez-Martín, I.R. Amado,
Production of hyaluronic acid by Streptococcus zooepidemicus on protein substrates obtained

RI
from Scyliorhinus canicula discards, Mar Drugs, 13 (2015) 6537-6549.
[23] A.C. de Macedo, M.H.A. Santana, Hyaluronic acid depolymerization by ascorbate-redox
effects on solid state cultivation of Streptococcus zooepidemicus in cashew apple fruit bagasse,

SC
World J. Microbiol. Biotechnol., 28 (2012) 2213-2219.
[24] L.J. Benedini, M.H.A. Santana, Effects of soy peptone on the inoculum preparation of
Streptococcus zooepidemicus for production of hyaluronic acid, Bioresour. Technol., 130 (2013)
798-800.

U
[25] N. Mohan, R. Balakrishnan, S. Sivaprakasam, Optimization and effect of dairy industrial
waste as media components in the production of hyaluronic acid by Streptococcus thermophilus,
Prep. Biochem. Biotechnol., 46 (2016) 628-638. N
[26] I.R. Amado, J.A. Vázquez, L. Pastrana, J.A. Teixeira, Cheese whey: A cost-effective
A
alternative for hyaluronic acid production by Streptococcus zooepidemicus, Food Chem., 198
(2016) 54-61.
M

[27] I.R. Amado, J.A. Vázquez, L. Pastrana, J.A. Teixeira, Microbial production of hyaluronic
acid from agro-industrial by-products: Molasses and corn steep liquor, Biochem. Eng. J., 117
(2017) 181-187.
D

[28] W.-C. Huang, S.-J. Chen, T.-L. Chen, Modeling the microbial production of hyaluronic
acid, J. Chin. Inst. Chem. Eng., 38 (2007) 355-359.
TE

[29] M.M. Don, N.F. Shoparwe, Kinetics of hyaluronic acid production by Streptococcus
zooepidemicus considering the effect of glucose, Biochem. Eng. J., 49 (2010) 95-103.
[30] L. Liu, M. Wang, J. Sun, G. Du, J. Chen, Application of a novel cavern volume controlled
EP

culture model to microbial hyaluronic acid production by batch culture of Streptococcus

zooepidemicus, Biochem. Eng. J., 48 (2010) 141-147.
[31] Z.-W. Lai, R.A. Rahim, A.B. Ariff, R. Mohamad, Biosynthesis of high molecular weight
CC

hyaluronic acid by Streptococcus zooepidemicus using oxygen vector and optimum impeller tip
speed, J. Biosci. Bioeng., 114 (2012) 286-291.
[32] P. Vengaiah, D. Ravindrababu, G. Murthy, K. Prasad, Jaggery from Palmyrah palm
(Borassus flabellifer L.)-Present status and scope, Indian J. Tradit. Know., 12 (2013) 714–717.
A

[33] G. Sathiyanarayanan, G. Saibaba, G.S. Kiran, J. Selvin, Process optimization and

production of polyhydroxybutyrate using palm jaggery as economical carbon source by marine
sponge-associated Bacillus licheniformis MSBN12, Bioprocess Biosyst Eng, 36 (2013) 1817-
1827.
[34] S.R. Reddy Tadi, A. EVR, A.M. Limaye, S. Sivaprakasam, Enhanced production of
optically pure d (–) lactic acid from nutritionally rich Borassus flabellifer sugar and whey

20
protein hydrolysate based–fermentation medium, Biotechnol. Appl. Biochem., 64 (2017) 279-
289.
[35] P. Ambati, C. Ayyanna, Optimizing medium constituents and fermentation conditions for
citric acid production from palmyra jaggery using response surface method, World J. Microbiol.
Biotechnol., 17 (2001) 331-335.
[36] B. Ratnam, S.S. Rao, D.R. Mendu, M.N. Rao, C. Ayyanna, Optimization of medium
constituents and fermentation conditions for the production of ethanol from palmyra jaggery
using response surface methodology, World J. Microbiol. Biotechnol., 21 (2005) 399-404.

PT
[37] J.-H. Kim, S.-J. Yoo, D.-K. Oh, Y.-G. Kweon, D.-W. Park, C.-H. Lee, G.-H. Gil, Selection
of a Streptococcus equi mutant and optimization of culture conditions for the production of high
molecular weight hyaluronic acid, Enzyme Microb. Technol., 19 (1996) 440-445.

RI
[38] R. Glaser, J. Venus, Model-based characterisation of growth performance and L-lactic acid
production with high optical purity by thermophilic Bacillus coagulans in a lignin-supplemented
mixed substrate medium, N Biotechnology, 37 (2017) 180-193.

SC
[39] J.A. Vázquez, M.A. Murado, Unstructured mathematical model for biomass, lactic acid and
bacteriocin production by lactic acid bacteria in batch fermentation, J Chem Technol Biotechnol,
83 (2008) 91-96.
[40] Y.-H. Chen, Q. Wang, Establishment of CTAB Turbidimetric method to determine

U
hyaluronic acid content in fermentation broth, Carbohydr Polym, 78 (2009) 178-181.
[41] N. Izawa, T. Hanamizu, T. Sone, K. Chiba, Effects of fermentation conditions and soybean
N
peptide supplementation on hyaluronic acid production by Streptococcus thermophilus strain
YIT 2084 in milk, J Biosci Bioeng., 109 (2010) 356-360.
A
[42] P. Ruas-Madiedo, J. Hugenholtz, P. Zoon, An overview of the functionality of
exopolysaccharides produced by lactic acid bacteria, Int Dairy J, 12 (2002) 163-171.
M

[43] N.C. Pan, H.C.B. Pereira, M.d.L.C. da Silva, A.F.D. Vasconcelos, M.A.P.C. Celligoi,
Improvement Production of Hyaluronic Acid by Streptococcus zooepidemicus in Sugarcane
Molasses, Appl. Biochem. Biotechnol., 182 (2017) 276-293.
D

[44] G. Rice, F. Stewart, A. Hillier, G. Jago, The uptake of amino acids and peptides by
Streptococcus lactis, J. Dairy Res, 45 (1978) 93-107.
TE

[45] L. Liu, J. Sun, W. Xu, G. Du, J. Chen, Modeling and optimization of microbial hyaluronic
acid production by Streptococcus zooepidemicus using radial basis function neural network
coupling quantum‐behaved particle swarm optimization algorithm, Biotechnol Prog, 25 (2009)
EP

1819-1825.
[46] A.M.B. Pires, S.Y. Eguchi, M.H.A. Santana, The influence of mineral ions on the microbial
production and molecular weight of hyaluronic acid, Appl. Biochem. Biotechnol., 162 (2010)
CC

2125-2135.
[47] X. Zhang, M. Wang, T. Li, L. Fu, W. Cao, H. Liu, Construction of efficient Streptococcus
zooepidemicus strains for hyaluoronic acid production based on identification of key genes
involved in sucrose metabolism, AMB Express, 6 (2016) 121.
A

[48] N. Caldas Pan, J.A. Vignoli, C. Baldo, H.C. Braga Pereira, R.S. dos Santos Ferreira da
Silva, M.A. Pedrine Colabone Celligoi, Effect of fermentation conditions on the production of
hyaluronic acid by Streptococcus zooepidemicus ATCC 39920, Acta Sci Biol Sci, 37 (2015).
[49] S.R. Kotra, T. Venkateswarulu, B. John, Cost effective media optimization for the
enhanced production of hyaluronic acid using a mutant strain Streptococcus zooepidemicus
3523–7: a statistical approach, Int J Adv SciTechnol, 60 (2013) 83-96.

21
[50] A.D. Nandasana, S. Kumar, Kinetic modeling of lactic acid production from molasses using
Enterococcus faecalis RKY1, Biochem. Eng. J., 38 (2008) 277-284.
[51] L. Liu, G. Du, J. Chen, Y. Zhu, M. Wang, J. Sun, Microbial production of low molecular
weight hyaluronic acid by adding hydrogen peroxide and ascorbate in batch culture of
Streptococcus zooepidemicus, Bioresour. Technol., 100 (2009) 362-367.
[52] W.Y. Chen, E. Marcellin, J.A. Steen, L.K. Nielsen, The role of hyaluronic acid precursor
concentrations in molecular weight control in Streptococcus zooepidemicus, Mol. Biotechnol.,
56 (2014) 147-156.

PT
[53] K. Han, O. Levenspiel, Extended Monod kinetics for substrate, product, and cell inhibition,
Biotechnol. Bioeng., 32 (1988) 430-447.
[54] G. Bustos, A. Moldes, J. Alonso, M. Vázquez, Optimization of D-lactic acid production by

RI
Lactobacillus coryniformis using response surface methodology, Food Microbiol., 21 (2004)
143-148.
[55] K.P. Patil, K.K. Kamalja, B.L. Chaudhari, Optimization of medium components for

SC
hyaluronic acid production by Streptococcus zooepidemicus MTCC 3523 using a statistical
approach, Carbohydr. Polym., 86 (2011) 1573-1577.
[56] J.F. Andrews, A mathematical model for the continuous culture of microorganisms utilizing
inhibitory substrates, Biotechnol. Bioeng., 10 (1968) 707-723.

U
[57] S. Aiba, M. Shoda, M. Nagatani, Kinetics of product inhibition in alcohol fermentation,
Biotechnol. Bioeng., 10 (1968) 845-864.
N
[58] M.L. Shuler, F. Kargi, Bioprocess engineering: basic concepts, Prentice Hall Upper Saddle
River, NJ, 2002.
A
[59] V.H. Edwards, The influence of high substrate concentrations on microbial kinetics,
Biotechnol. Bioeng., 12 (1970) 679-712.
M

[60] T. Yano, T. Nakahara, S. Kamiyama, K. Yamada, Kinetic studies on microbial activities in

concentrated solutions. Part. I, Agric. Biol. Chem., 30 (1966) 42-48.
[61] J.L. Webb, Enzyme and metabolic inhibitors, 1963.
D

[62] J. Luong, Generalization of Monod kinetics for analysis of growth data with substrate
inhibition, Biotechnol. Bioeng., 29 (1987) 242-248.
TE

[63] S.M.T. Gharibzahedi, S.H. Razavi, M. Mousavi, Kinetic analysis and mathematical
modeling of cell growth and canthaxanthin biosynthesis by Dietzia natronolimnaea HS-1 on
waste molasses hydrolysate, RSC Advances, 3 (2013) 23495-23502.
EP
CC
A

22
FIGURE CAPTIONS

Figure 1: Experimental data (points) and simulation (continuous lines) of batch fermentation of
S.zooepidemicus MTCC 3523 on modified M17 with different initial sugar concentrations A) 10

PT
gL-1, B) 20 gL-1, C) 30 gL-1, D) 40 gL-1, E) 50 gL-1 ; DCW (Triangle), Substrate (Circle), HA
(Cross) and LA (Square)

RI
Figure 2: Effect of initial palm sugar concentration on HA molecular weight

SC
Figure 3: Economic yield (A) and Economic productivity (B) of HA on different carbon and

U
nitrogen sources

N
Figure 4: Dynamic profiles of economic productivity, EP (Square) and HA (Circle) studied for
A
the kinetic experiment at 30 gL-1 of initial PJ concentration
M
D
TE
EP
CC
A

23
Tables:

Table 1: Selection of an elite strain for high HA productivity

YP/S YX/S μmax HA

PT
(gg-1) (gg-1) (h-1) (gL-1)
Sucro Sucro

RI
Strain PJ PJ Sucrose PJ Sucrose PJ
se se
S.thermophilus 0.05 ± 0.04 ± 0.50 ± 0.49 ± 0.15 ± 0.24 ± 0.21 ±

SC
0.19 ± 0. 01
NCIM 2904 0.01 0.01 0.13 0.09 0.05 0.08 0.13
S.thermophilus STI 0.03 ± 0.04 ± 0.54 ± 0.58 ± 0.13 ± 0.15 ± 0.18 ±
0.18 ± 0.06
12 0.01 0.01 0.1 0.15 0.08 0.13 0.09

U
S.zooepidemicus 0.06 ± 0.09 ± 0.59 ± 0.52 ± 0.54 ± 0.31 ± 0.41 ±
0.32 ± 0. 12
MTCC 3523
aProduct
0.01 0.02 0.1 0.12 N 0.1 0.09 0.15
Yield (YP/S, gg-1) was calculated as a ratio of HA produced (g) to substrate consumed (g). b Biomass Yield (YX/S, gg-1)
A
was calculated as a ratio of biomass produced (g) to substrate consumed (g). cµmax (h-1), maximum specific growth rate, was
calculated during exponential growth phase. All the cultures were grown in 250 ml baffled Erlenmeyer flask contains 100 ml
medium, at 37 oC, 180 rpm.
M
D

Table 2: Estimated kinetic parameters for biomass growth, substrate consumption and
TE

product formation at different initial substrate concentrations

Kinetic parameter Initial substrate concentration (gL-1)

10 20 30 40 50
EP

Biomass formation
X0 (gL-1) 0.18 0.16 0.14 0.13 0.13
Xm (gL-1) 4.59 5.96 9.00 6.40 5.67
CC

vx (gL-1h-1) 0.71 0.95 1.43 1.17 0.94

λx (h) 3.75 3.46 4.26 4.94 5.80
R2 0.99 0.99 0.99 0.99 0.99
SSE 0.20 0.07 0.52 0.56 0.13
A

Sugar consumption
S0 (gL-1) 10.93 20.54 29.65 40.20 5.67
YX/S (gg-1) 0.48 0.29 0.34 0.22 0.19
mS (gg-1h-1) 0.01 0.01 0.01 0.07 0.10
R2 0.98 0.99 0.98 0.98 0.99
SSE 3.16 4.77 22.41 34.60 31.68

24
 Hyaluronic acid production
αHA(gg-1) 0.13 0.13 0.15 0.08 0.10
βHA (gg-1h-1) 0.00 0.00 0.00 0.00 0.000
R2 0.96 0.97 0.99 0.96 0.98
SSE 0.02 0.05 0.01 0.05 0.01

Lactic acid production

PT
αLA(gg-1) 2.42 2.91 1.98 2.44 2.44
βLA (gg-1h-1) 0.00 0.00 0.00 0.00 0.05
R2 0.98 0.94 0.96 0.99 0.92

RI
SSE 4.04 41.84 23.98 4.24 67.79
Batch fermentation experiments were performed in bioreactor at 200 rpm, 37 °C and 7.0 pH

SC
Table 3: Substrate inhibition kinetic models used in this study
Name of the Model Model equation Reference
𝜇𝑚𝑎𝑥 ∗ 𝑆

U
Andrews 𝜇= 𝑆2 [56]
(𝐾𝑆 + 𝑆 + 𝐾 )

(𝜇m𝑎𝑥 ∗ 𝑆) −𝑆
𝐼
N
A
[ ]
Aiba 𝜇= ∗ 𝑒𝑥𝑝 𝐾𝐼 [57]
(𝐾𝑆 + 𝑆)
M

𝜇𝑚𝑎𝑥 ∗ 𝑆
Competitive 𝜇= 𝑆 [58]
(𝐾𝑆 ∗ (1 + 𝐾 ) + 𝑆)
D

𝐼
𝜇𝑚𝑎𝑥
Non-competitive 𝜇= 𝑆 𝐾𝑆 [58]
(1 + 𝐾 ) ∗ (1 + )
TE

𝐼 𝑆
−𝑆 −𝑆
Edward [ ] [ ]
𝜇 = 𝜇𝑚𝑎𝑥 ∗ [𝑒𝑥𝑝 𝐾𝐼 − 𝑒𝑥𝑝 𝐾𝑆
] [59]
(Tipo-Teisser type)
EP

(μmax ∗ S)
Yano μ= S S2 [60]
S + K S + [(1 + K) ∗ (K )]
I
CC

S
μmax ∗ S ∗ (1 + K)
μ= S2
[61]
Webb S + K S + (K )
I
(μmax ∗ S) S n
A

μ= ∗ [1 − ] [62]
Luong (K S + S) Sm
S n S
μ = μmax (1 − ∗ ) ∗
S S m
Han-Levenspiel (S + K S (1 − S∗) ) [53]

(μmax ∗Sn )
Moser μ= (KS +Sn )
, n>0 [63]

25
 A Figures:

CC
EP
TE
D
M

26
A
N
U
SC
RI
PT
 A
CC
EP
TE
D
M

27
A
N
U
SC
RI
PT
 PT
RI
SC
U
Figure 1: Experimental data (points) and simulation (continuous lines) of batch
N
fermentation of S.zooepidemicus MTCC 3523 on modified M17 with different initial sugar
A
concentrations A) 10gL-1, B) 20 gL-1, C) 30 gL-1, D) 40 gL-1, E) 50 gL-1; DCW (Triangle),
M

Substrate (Circle), HA (Cross) and LA (Square).

D
TE
EP
CC
A

28
Figure 2: Effect of initial palm sugar concentration on HA molecular weight

PT
RI
SC
U
N
A
M
D
TE
EP
CC
A

Figure 3: (A) Economic yield and (B) Economic productivity of HA on different carbon
and nitrogen sources

29
 PT
RI
SC
U
N
A
Figure 4: Dynamic profiles of economic productivity, EP (Square) and HA (Circle) studied
M

for the kinetic experiment at 30 gL-1 of initial PJ concentration.

D
TE
EP
CC
A

30