Archives of Microbiology (2021) 203:5993–6005

https://doi.org/10.1007/s00203-021-02554-6

ORIGINAL PAPER

Statistical optimisation of polyhydroxyalkanoate production

in Bacillus endophyticus using sucrose as sole source of carbon
M. Geethu1 · H. Raghu Chandrashekar2 · M. S. Divyashree1

Received: 14 July 2021 / Revised: 4 August 2021 / Accepted: 22 August 2021 / Published online: 22 September 2021
© The Author(s) 2021

Abstract
Microorganisms have been contemplated as a promising source for the inexhaustible synthesis of many novel materials utiliz-
ing renewable sources. Among many of such products, polyhydroxyalkanoate (PHA) remains as an essential biodegradable
polymer with functions similar to conventional plastics. Bacillus endophyticus is capable of accumulating biopolymer PHA
in nutrient limiting conditions with excess of carbon source. Screening and optimizing the parameters for increased PHA
production was done statistically. The optimized medium gave a maximum yield of 46.57% which was in well agreement
with the given predicted value provided by response surface methodology model yield of 47.02%. Optimal media conditions
when extrapolated in bioreactor gave an even higher production percentage of 49.9. This is the first report highlighting 49%
of polyhydroxybutyrate statistically using sucrose as a source. The main highlight of the study was the use of wild type strain
for producing high quality PHA using simple carbon source which can be a starting platform for using this strain for large
scale PHA production industrially. FTIR and 1HNMR analysis confirmed the polymer produced.

Keywords Bacillus endophyticus · Biopolymer · Bioreactor · Polyhydroxyalkanoate (PHA) · Shake flask cultivation

Abbreviations and industrial sector. Total demand for the biodegradable

PHA Polyhydroxyalkanoate polymers in North America, Western Europe and Asia
CCD Central composite design reached around 85,000 metric tons and the estimated con-
PB Plackett–Burman design sumption of biodegradable polymers has increased in these
RSM Response surface methodology years (Akaraonye et al. 2010). Microorganisms are consid-
PM Production medium ered as a promising source for the inexhaustible synthesis
of many novel materials utilizing renewable sources. Many
eco-friendly materials have been produced as reserve mol-
Introduction ecules under stress which can be used to resolve the putative
problems with several industrial applications. Among many
Worldwide interest to use bio-based polymers has acceler- of such products, polyhydroxyalkanoates (PHAs) remain as
ated as an eco-friendly alternative which resembles plastic in an essential biodegradable polymer with functions similar
their physio-chemical properties to overcome the increased to conventional plastics (Amache et al. 2013). A sequence of
demands in the rapidly developing fields like biomedical complex stages starting from biomass accumulation, poly-
mer synthesis and extraction of polymer to recovery is asso-
Communicated by Erko Stackebrandt. ciated with PHA production which significantly influences
PHA quality (Mohammed et al. 2019). In recent years, an
* M. S. Divyashree inclination towards efficient wild strain isolation and screen-
[email protected] ing with quality PHA accumulation capable of utilizing
1
Department of Biotechnology, Manipal Institute unconventional substrates and altered metabolic pathways
of Technology, Manipal Academy of Higher Education, was explored widely.
Manipal, Udupi, Karnataka 576104, India PHA accumulation and bioprocess optimization is very
2
Department of Pharmaceutical Biotechnology, Manipal significant to ensure the quality and quantity of biopolymer
College of Pharmaceutical Sciences, Manipal Academy synthesized via wild type strains. Industries mainly relay
of Higher Education, Manipal, Udupi, Karnataka 576104, on strains like Cupriavidus necator and Alcaligenes latus
India

13
Vol.:(0123456789)
5994 Archives of Microbiology (2021) 203:5993–6005

due to their high production rate and PHA yield. In strains Materials and methods
like Corynebacterium, Nocardia and Rhodococcus are the
only wild type strains reported to synthesize commercially Microorganism and culture maintenance
important copolymers from simple carbon sources like glu-
cose, which allowed a decrease in the production cost of Bacillus endophyticus, organism of interest was previously
the polymer (Kaur and Roy 2015). Decreased yield of the isolated and confirmed in the lab and was frequently sub-
polymer significantly contribute to the overall economy of cultured on nutrient agar medium (Himedia, Mumbai, India)
the polymer as less bioconversion require increased substrate and stored at 4 °C. Commonly used staining methods were
availability to uplift the PHA production (Prados and Mai- performed to confirm the microorganism. Primary staining
cas 2016). This leads to increase in the overall production methods like gram staining and endospore staining were per-
rate and to avoid this rapid improvement in the econom- formed to confirm Bacillus spp.
ics of PHA production process need to be implemented by
incorporating significant process designing and utilization of Screening and optimization of PHA production
cheap substrate thereby increasing PHA yield. Optimization media components using statistical experimental
of the bacterial requirement for maximal PHA production design
enabled by incorporating cheap readily available substrates
in Gram positive stains has witnessed a trend recently (Chen Plackett–Burman design (PBD)
and Jiang 2018).
PHA production by microorganism mainly depends on PBD are predominantly used to screen the factors in an
the carbon source and other nutrients provided. A strategy experiment. It is necessary to optimize the production
can be introduced to induce pressure on PHA producers to medium for PHA biosynthesis by B. endophyticus using
convert available carbon source to biopolymer. This condi- sucrose as major carbon source. PB design is largely used
tion can be achieved by optimising the ecosystem and by to detect the main effects economically and helps in estimat-
maintaining a feast and famine condition to organism (Ara- ing the variances of the factors using minimum experiments.
mvash et al. 2015). As conventional method of optimization 95% of relative significance level was chosen to carry out the
consumes lot of time and the relative interactions of differ- experimental trials with five factors by experiments designed
ent parameters cannot be readily studied, statistical experi- by Minitab 17. High (+ 1) and Low (− 1) level were coded
mental design can be employed to avoid the limitations of by PBD and real values of the independent variables were
one factor at a time method (Tripathi et al. 2013; Din et al. represented in Table 1.
2014). Screening of significantly influencing parameters Biopolymer production was done in cotton plugged
and analysing the interaction between them enable to depict 250 mL Erlenmeyer flask with 50 mL working volume of
their combined effect through factorial design followed by production medium (PM) which was varied based on the
response surface methodology (RSM). Statistical methods composition obtained by Minitab software which included
of optimisation remains an important strategy in achieving (NH4) 2 SO 4 , ­K H 2 PO 4 , ­Na 2 HPO 4 ­2 H 2 O, ­M gSO 4 7H 2 O
optimal PHA concentration before stepping into large scale Sucrose. All the shake flask experiment was carried out in
production in the presence of cheap carbon sources (Kaur duplicates. The pH of the PM was maintained at 7 before
and Roy 2015). inoculation. 10% of the inoculum was added to production
This work mainly focuses on providing experimental medium which was later incubated at 32 °C for 72 h.
insight towards hypothesis linking PHA accumulation and
the pattern of utilising sucrose as a major carbon source. Sta-
tistical optimization of various production parameters during Table 1  Different levels of experimental factors selected for PHA
PHA biosynthesis in Bacillus endophyticus, utilizing sucrose production by Bacillus endophyticus using Plackett–Burman design
as a sole source of carbon was carried out. Plackett–Burman No Components (factors) Low level (− 1) High level
design (PBD) and central composite design (CCD) were g/L (+ 1) g/L
predominantly used in optimising and screening the most
1 Na2HPO4 0.5 5
influencing parameter in increasing PHA production. The
2 KH2PO4 1.5 15
correlation of the increase in cell density and PHA produc-
3 (NH4)2SO4 1.5 15
tion was also evaluated during this study. The novelty of the
4 MgSO4·7H2O 0.2 2
work lies in the higher accumulation of PHA by wild type
5 Sucrose 20 40
B. endophyticus strain by utilising simple carbon source for
polymer biosynthesis.

13
Archives of Microbiology (2021) 203:5993–6005 5995

Central composite design (CCD) wall disruption was found to be negligible even after expos-
ing to temperature as high as 80 °C (Russell 2003).
Central composite design was carried out for PHA optimisa-
tion studies using sucrose as carbon source. Different com- Extraction of PHA
binations of three significant variables (­ Na2HPO4, ­KH2PO4
and sucrose) found to influence PHA production by PBD The PHA content of the cells was estimated by subjecting
was used in CCD. CCD with three variables at 5 levels (− 2, dry cells into hydrolysis for 1 h using 4% Sodium hypochlo-
− 1, 0, + 1, + 2) with 20 experiments was used to fit the sec- rite solution at 40 °C (Slepecky and Law 1960) followed
ond order polynomial model. The study was performed to with water and acetone wash for the centrifuged hydrolysate
understand the influence of these parameters on responses the remained residue obtained after wash was dissolved in
like biomass and PHA production. The variables and levels solvent chloroform. Gravimetric quantification of obtained
for CCD were represented in Table 2. The parameters for the PHA film was done routinely after each experimental trial
study were selected on literature which was used as compo- (Lakshman and Shamala 2006).
nents for PHA production medium (Mohammed et al. 2019).
FTIR and 1HNMR analysis of PHA
Bioreactor studies
The PHA film obtained after dissolving in chloroform was
The initiation of PHA synthesis and the maximal production pelletized using potassium bromide (KBr). FTIR spectrum
can be represented as time course study in bioreactor. Batch was recorded for each pelletized sample in the sample cham-
fermentation to evaluate initiation of PHA biosynthesis was ber of FTIR spectrophotometer and was exposed to infra-red
studied in a 3 L bioreactor (Bioflo 110-New Brunswick sci- radiation. The obtained spectrum using Shimadzu 8400 S
entific). The optimised basal medium obtained after statis- was in the spectral range of 4000–400 ­cm−1. 5 mg of PHA
tical optimisation was ran in 3 L bioreactor. The inoculum film was used for 1H NMR and characterized in Bruker
(10% for I L of PM) was cultured until 24 h at 32 °C with Ascend 400 NMR spectrometer at 22 °C.
150 rpm were aseptically transferred to 3 L of production
medium aseptically. Dissolved oxygen and pH values were
pre-set and regulated by corresponding probes attached to Results
bioreactor system. 200 rpm was the agitation rate maintained
and addition of 0.1 N polyethylene glycol was carried out Sequential statistical optimisation of production
occasionally as antifoaming agent during foaming. Samples medium
(15 mL) were drawn at regular intervals to study biomass
and PHA production rate of B. endophyticus. Plackett–Burman design

Estimation of cell dry mass Software driven experimental trial was developed (Table 3)
using Minitab 17 and experiments were carried out sequen-
The culture broth was collected and centrifuged at tially. The main aim of this screening study was to identify
10,000 rpm (Plastocraft SSR-V/FM) for 10 min and the the parameters that were highly significant during PHA
supernatant was analyzed for residual sugar by dinitrosali- production. The responses, Biomass and PHA yield was
cylic acid method (Miller 1959) through spectrophotometric represented in g/L and the relative significance of five fac-
analysis at 540 nm. The harvested cell pellet was washed tors ­(Na2HPO4, ­KH2PO4, ­(NH4)2SO4, ­MgSO4·7H2O and
with distilled water and then dried at 80 °C to a constant concentration of sucrose) at two levels (low and high). The
weight. As it was gram positive strain, the chances of cell significance level was 95% and 12 experiments with 3 centre
points were carried out in shake flask. The experimental
and predicted values of biomass and PHA are represented in
Table 2  Low and high levels for 3 factors screened for PHA yield by Table 3. Statistical analysis of PB design on biomass (g/L)
B. endophyticus and PHA production (g/L) was represented in Tables 4 and
5 respectively as ANOVA table and the effects were repre-
No Components (factors) Units Low level High
(− 1) level sented in Pareto chart for the same in Fig. 1 (a and b).
(+ 1) The graphical representation of this phenomenon is given
in Pareto chart Fig. 2a, b, which details the relevance of
1 Na2HPO4 (g/L) g/L 1 2
main effect to determine how far these factors are different
2 KH2PO4 (g/L) g/L 1 2
from zero. The horizontal column in the graph represents the
3 Sucrose (g/L) g/L 20 60
value for 95% confidence level and gave a T value of 2.18 as

13
5996 Archives of Microbiology (2021) 203:5993–6005

Table 3  Experimental model of Plackett–Burman design on biomass and PHA yield by B. endophyticus using sucrose as carbon source
Run Na2HPO4 (g/L) KH2PO4 (g/L) (NH4)SO4 MgSO4·7H2O Sucrose g/L Biomass (g/L) PHA (g/L)
Experimental Predicted Experimental Predicted

1 5.00 1.50 15.00 0.2 20 2.130 2.21033 0.220 0.21533

2 5.00 15.00 1.50 2.0 20 3.028 2.98367 0.662 0.84133
3 0.50 15.00 15.00 0.2 40 2.558 2.31233 0.962 0.90200
4 5.00 1.50 15.00 2.0 20 2.206 2.06100 0.266 0.22400
5 5.00 15.00 1.50 2.0 40 3.894 3.83100 1.766 1.42000
6 5.00 15.00 15.00 0.2 40 3.900 4.01500 1.186 1.23800
7 0.50 15.00 15.00 2.0 20 1.200 1.31567 0.316 0.33200
8 0.50 1.50 15.00 2.0 40 1.126 1.20567 0.428 0.46667
9 0.50 1.50 1.50 2.0 40 1.114 1.17100 0.486 0.64000
10 5.00 1.50 1.50 0.2 40 2.966 3.02300 0.806 0.96733
11 0.50 15.00 1.50 0.2 20 1.308 1.43033 0.338 0.49667
12 0.50 1.50 1.50 0.2 20 0.602 0.47300 0.360 0.05267
13 2.75 8.25 8.25 1.1 30 2.858 2.90933 1.694 1.68267
14 2.75 8.25 8.25 1.1 30 2.940 2.90933 1.698 1.68267
15 2.75 8.25 8.25 1.1 30 2.930 2.90933 1.656 1.68267

Table 4  Statistical analysis Source df Adj SS Adj MS F value p value Main effect Std. error
of Plackett–Burman design
on biomass yield with five Model 6 14.9854 2.49756 118.35 0
fermentative parameters
Linear 5 13.6711 2.73422 129.57 0
Na2HPO4 1 8.6972 8.69722 412.14 0 1.7027 0.0419
KH2PO4 1 2.7495 2.74946 130.29 0 0.9573 0.0419
(NH4)2SO4 1 0.0036 0.00361 0.17 0.69 0.0347 0.0419
MgSO4·7H2O 1 0.0669 0.0669 3.17 0.113 − 0.1493 0.0419
Sucrose 1 2.1539 2.15392 102.07 0 0.8473 0.0419
Curvature 1 1.3142 1.31424 62.28 0
Error 8 0.1688 0.0211
Lack-of-fit 6 0.1648 0.02747 13.73 0.069
Pure error 2 0.004 0.002
Total 14 15.1542

df degrees of freedom, SS sum of squares

S = 0.14526; R2 = 98.89%; R2(adj.) = 98.05%; F = ­MS(Factor) /MS(Error)

Table 5  Statistical analysis Source df Adj SS Adj MS F value p value Main effect Std. error
of Plackett–Burman design
on PHA yield with five Model 6 4.58603 0.76434 18.41 0
fermentative parameters
Linear 5 2.02502 0.405 9.75 0.003
Na2HPO4 1 0.33869 0.33869 8.16 0.021 0.3360 0.0588
KH2PO4 1 0.59141 0.59141 14.24 0.005 0.4440 0.0588
(NH4)2SO4 1 0.09013 0.09013 2.17 0.179 − 0.1733 0.0588
MgSO4·7H2O 1 0.00023 0.00023 0.01 0.943 0.0087 0.0588
Sucrose 1 1.00457 1.00457 24.19 0.001 0.5787 0.0588
Curvature 1 2.56101 2.56101 61.68 0
Error 8 0.33217 0.04152
Lack-of-fit 6 0.33109 0.05518 102.7 0.01
Pure error 2 0.00107 0.00054
Total 14 4.9182

13
Archives of Microbiology (2021) 203:5993–6005 5997

Fig. 1  Standardized effects on the fermentative parameters shown in Pareto chart for a biomass (g/L) and b PHA (g/L) production by B. endo-
phyticus

Fig. 2  2D contour plot showing

the interactive effects of three
variables showing the interac-
tion effect of Biomass g/L of
B.endophyticus using sucrose as
carbon source

a vertical line in the plot. The T value indicates the minimum Designing the screening experiment was to evaluate the
statistically significant effect for 95% significant level. From factors that highly influence PHA production. Before spe-
Pareto chart Fig. 1a, b, it was observed that the concentra- cific screening of the highly influencial parameters on the
tion of three factors ­(Na2HPO4, ­KH2PO4 and concentration production, Plackett–Burman studies were carried out which
of sucrose) was positively affecting biomass and PHA yield. gave an idea on which the production may relay. The fac-
Independent variables above 95% were considered as signifi- tors used for the experiments were selected that had close
cantly influencing biomass and PHA yield. link with the production during shake flask cultivation done

13
5998 Archives of Microbiology (2021) 203:5993–6005

prioirly. This was clearly visualised in pareto chart which interaction of the parameters through factorial experimen-
showed the standardised lineance of the production towards tal design. The maximum PHB yield was 49.47% of bio-
each factors. The cumulative inference is that the concentra- mass after 72 h of cultivation which was not reported by any
tion of N
­ a2HPO4, ­KH2PO4 and sucrose influences the overall Bacillus strain so far utilising sucrose as sole carbon source
yield of biomass and PHA production during the course of (Mohammed et al. 2019).
growth. Sucrose concentration readily had an influence on
biomass and the production of biopolymer. During biomass Central composite design and response surface
increment, the role of ­KH2PO4and ­Na2HPO4 was found to methodology
be significant which indicates the trivial role of potassium
and sodium ions for different transport system and cytoplas- Central composite design was carried out for PHA optimisa-
mic signalling (Wang and Bakken 1998; Bora 2013). This tion studies using sucrose as carbon source. Different com-
enable cells to divide and increase in cell mass. Whereas binations of three significant variables (­ Na2HPO4, ­KH2PO4
during PHA production, the role of ­(NH4)2SO4 was found and sucrose) found to influence PHA production by PBD
to be insignificant which justifies the correlation of limited were used in CCD. CCD with three variables at five levels
nitrogen source for enhancing PHA production. (− 2, − 1, 0, + 1, + 2) with 20 experiments was used to fit the
Regression equation for the Model (Biomass g/L): second order polynomial model. The study was performed
to understand the influence of these parameters on responses
Biomass g∕L = −0.657 + 0.3784 Na2 HPO4 + 0.07091 KH2 PO4
( ) like biomass and PHA production. The variables and levels
+ 0.00257 NH4 2SO4 − 0.0830 MgSO4
for CCD were represented in Table 6 and the design table
+ 0.04237 Sucrose + 0.7400 Ct Pt. obtained by Minitab 17 is represented in Table 7.
Regression equation for the Model (PHA g/L): Biomass g∕L = 4.573 − 3.541Z1 − 1.729 Z2 + 0.02468 Z3
PHA g∕L = −0.594 + 0.0747 Na2 HPO4 + 0.03289 KH2 PO4 + 1.0907 Z1 × Z1 + 0.5217 Z2 × Z2
( )
− 0.01284 NH4 2 SO4 + 0.0048 MgSO4 − 0.000296 Z3 × Z3 + 0.140 Z1 × Z2
+ 0.02893 Sucrose + 1.033 Ct Pt. + 0.01140 Z1 × Z3 − 0.00510 Z2 × Z3 .
Effect of carbon source, sucrose concentration (0–4%) on
PHA biosynthesis were studied in B. endophyticus in PHA PHA g∕L = 1.754 − 1.003 Z1 − 1.137 Z2 + 0.01335 Z3
production medium. The correlation of biomass and PHA + 0.1945 Z1 × Z1 + 0.1835 Z2 × Z2
in the excess presence of sucrose was studied as batch cul- − 0.000237 Z3 × Z3 + 0.288 Z1 × Z2
tivation using statical optimisation tool. The biomass accu-
+ 0.00400 Z1 × Z3 + 0.00205 Z2 × Z3 .
mulation is very essential, as there is a direct relationship
between the biomass and PHA content as PHA is directly
accumulated within the cytoplasm. Residual biomass is the where Z1 is the concentration of ­Na2HPO4, Z2 is the con-
catalytic component responsible for the metabolic activity, centration ­KH2PO4 g/L, Z3 is the concentration of Sucrose.
so increase in the biomass determine the quantity of PHA The linear models of the regression analysis (Z1, Z2 and
that was produced potentially. So the cell growth balance Z3) demonstrated the significance of the parameters rela-
is necessary during the growth phase to avoid incomplete tively. Similarly the quadratic models Z12, Z22 and Z32 and
PHA accumulation. High biomass % is directly proportional Z1Z3 two way interaction model were found to be significant.
to PHA concentration so it is necessary to decide the har-
vesting time to attain maximum PHA production (Pijuan Fit of model
et al. 2009). Commercial production of PHA relay mainly on
fed batch and batch cultivation. Taking this into considera- The design matrix in terms of coded and real values along
tion, batch cultivation in shake flask was carried out and the with the response (Biomass and PHA g/L) is represented in
optimal concentration of components obtained via statistical Table 6. The maximum biomass g/L obtained was 2.59 g/L
optimisation was extrapolated in bioreactor. It was necessary and maximum PHA g/L was 0.88 g/L (Table 6). The com-
to carry out RSM which helped in specifically determine the petence of the CCD model and the fitness evaluation was
second order behaviour of the parameters (Chen and Jiang performed by ANOVA and the statistical significance of the
2018). It also enables the correlation between the factors model was predicted based on the F test and P test for analy-
and responses with the best combination of the parameters sis of variance (ANOVA) is given in Tables 7 and 8. The
can be used by the model to obtain maximum yield. Using model F value of 51.75 (Biomass g/L in Table 7) and 7.95
factorial experimental design, the errors generated gets nul- (PHA g/L in Table 8) indicates the significance of the model
lified and the overall effects of variable provides mutual and the lack of fit value of 0.0133 and 0.79 for biomass and

13
 Table 6  Central composite experimental design for optimising three parameters
Run Na2HPO4 (g/L) KH2PO4 (g/L) Sucrose Biomass (g/L) PHA (g/L) PHA (%)
Coded value Real value Coded value Real value Coded value Real value Experimental Predicted Experimental Predicted
Archives of Microbiology (2021) 203:5993–6005

1 −1 1.0 −1 1.0 −1 20 1.73 1.70 0.54 0.58 31.21

2 1 2.0 1 2.0 −1 20 1.89 1.81 0.61 0.56 32.27
3 1 2.0 −1 1.0 1 60 2.59 2.54 0.74 0.71 28.57
4 −1 1.0 1 2.0 1 60 1.58 1.66 0.41 0.43 25.94
5 0 1.5 0 1.5 0 40 1.54 1.64 0.55 0.57 35.71
6 0 1.5 0 1.5 0 40 1.66 1.64 0.57 0.57 34.34
7 1 2.0 −1 1.0 −1 20 1.91 1.84 0.56 0.57 29.30
8 −1 1.0 1 2.0 −1 20 1.56 1.61 0.31 0.37 19.87
9 −1 1.0 −1 1.0 1 60 1.95 2.03 0.56 0.64 28.71
10 1 2.0 1 2.0 1 60 2.36 2.40 0.88 0.86 37.28
11 0 1.5 0 1.5 0 40 1.72 1.68 0.80 0.61 46.51
12 0 1.5 0 1.5 0 40 1.74 1.68 0.55 0.61 31.6
13 −2 0.5 0 1.5 0 40 1.96 1.87 0.60 0.52 30.61
14 2 2.5 0 1.5 0 40 2.58 2.67 0.81 0.87 31.39
15 0 1.5 −2 0.5 0 40 1.82 1.86 0.79 0.75 43.4
16 0 1.5 2 2.5 0 40 1.58 1.54 0.60 0.61 37.97
17 0 1.5 0 1.5 −2 0 0.22 0.29 0.00 − 0.02 Nil
18 0 1.5 0 1.5 2 80 1.20 1.12 0.27 0.26 22.5
19 0 1.5 0 1.5 0 40 1.18 1.18 0.46 0.50 39.31
20 0 1.5 0 1.5 0 40 1.17 1.18 0.46 0.50 39.31

13
5999
6000 Archives of Microbiology (2021) 203:5993–6005

Table 7  Analysis of variance for biomass (g/L) with three factors confirming the significance of the model in predicting the
Source df Adj SS Adj MS F value p value
overall PHA yield with respect to biomass.

Model 11 5.23547 0.47595 51.75 0 Interaction of variables on responses

Blocks 2 1.02819 0.5141 55.9 0
Linear 3 1.42927 0.47642 51.8 0 The two dimensional (2D) contour plot are usually used
Z1 1 0.62885 0.62885 68.37 0 to visually represent the regression equations to study the
Z2 1 0.09986 0.09986 10.86 0.011 interaction effect of independent and dependent variable. In
Z3 1 0.70057 0.70057 76.17 0 these plots, two response factors are presented keeping other
Square 3 2.93903 0.97968 106.52 0 factor at constant and the extent of interaction is usually
Z12 1 1.78433 1.78433 194.01 0 determined by the shape of the plot. To study the interaction
Z22 1 0.4082 0.4082 44.38 0 effect between three factors that is significantly affecting
Z32 1 0.33607 0.33607 36.54 0 biomass and PHA accumulation was well studied using 2D
2-Way interaction 3 0.13458 0.04486 4.88 0.033 plots as represented in Figs. 2 and 3 respectively. The cir-
Z1 ­Z2 1 0.0098 0.0098 1.07 0.332 cular contour plot Fig. 2a between ­KH2PO4 and ­Na2HPO4
Z1 ­Z3 1 0.10397 0.10397 11.3 0.01 shows that there is only less interaction between these fac-
Z2 ­Z3 1 0.02081 0.02081 2.26 0.171 tors in increasing biomass accumulation.
Error 8 0.07358 0.0092 The factorial design enabled to restrict the number of
Lack-of-fit 5 0.06691 0.01338 6.02 0.085 experiment to 15 and gave an idea about the higher order
Pure error 3 0.00667 0.00222 interaction among the factors. All the experiments were
Total 19 5.30905 conducted in a randomized manner. The synergistic effect
df degrees of freedom, SS sum of squares, MS mean square of 2 or more factors could be thoroughly analyzed. These
S = 0.0959016; R2 = 98.61%; R2(adj) = 96.71%; F = ­MS(Factor) /MS(Error)
significant parameters were taken up for optimization using
central composite design. For further confirmation and to
obtain the right combination, those factors which showed
PHA respectively was found to be less than the F values influence was taken for central composite design and helped
in concluding the right combination of parametrs. The coef-
Table 8  Analysis of variance for PHA (g/L) with three factors
ficient of determination R2 value determines fit of the model
Source df Adj SS Adj MS F value p value which was 0.9162 (p < 0.05), 91.62% for PHA and 0.9861
(p < 0.05), 98.61% for biomass production. This enables in
Model 11 0.750316 0.068211 7.95 0.003
identifying the variability of the response which can be pre-
Blocks 2 0.041597 0.020799 2.42 0.15
dicted. The observed model was relevant in predicting the
Linear 3 0.216226 0.072075 8.4 0.007
optimal growth and PHA biosynthesis in B. endophyticus
Z1 1 0.119025 0.119025 13.87 0.006
using sucrose as carbon source based on the predicted values
Z2 1 0.021025 0.021025 2.45 0.156
given by the statistical CCD model.
Z3 1 0.076176 0.076176 8.88 0.018
The elliptical shape of contour plot as in Fig. 2a indi-
Square 3 0.431045 0.143682 16.74 0.001
cates a negative effect on PHA production with respect
Z12 1 0.056745 0.056745 6.61 0.033
to higher concentration of K ­ H2PO4 and ­Na2HPO4 which
Z22 1 0.050508 0.050508 5.89 0.041
decreases PHA production. This was in well agreement
Z32 1 0.21603 0.21603 25.17 0.001
with the necessity of nutrient limiting condition for organ-
2-Way interaction 3 0.057634 0.019211 2.24 0.161
ism to initiate PHA production. In our another study, it was
Z1 ­Z2 1 0.041472 0.041472 4.83 0.059
reported that PHA production in the presence of inexpensive
Z1 ­Z3 1 0.0128 0.0128 1.49 0.257
carbon source like sugarcane molasses (4%) with increased
Z2 ­Z3 1 0.003362 0.003362 0.39 0.549
agitation rate enhanced PHA production by 10.7 g/L with
Error 8 0.068657 0.008582
15.37 g/L of cell mass statistically. This experimental model
Lack-of-fit 5 0.039085 0.007817 0.79 0.618
created using CCD reported that increase in sugar concentra-
Pure error 3 0.029572 0.009857
tion (> 4%) had a negative influence on biopolymer produc-
Total 19 0.818973
tion (Geethu et al. 2019).
df degrees of freedom, SS sum of squares, MS mean square Mohanrasu (2020) reported the importance of nitrogen,
S = 0.0926396; R2 = 91.62%; R2(adj) = 80.09%; F = ­MS(Factor) /MS(Error) potassium and sodium concentration in the basal medium for
cell growth for the PHA producers which later on limiting

13
Archives of Microbiology (2021) 203:5993–6005 6001

Fig. 3  2D contour plot showing

the interactive effects of three
variables and on PHA produc-
tion by B.endophyticus using
sucrose as carbon source

the nutrients enhance the production. 2.74 g/L of PHA was Bioreactor studies for the optimised medium
reported by the same authors in B. megaterium strain uti-
lizing glucose as carbon source (Mohanrasu et al. 2020). PHA production in the excess presence of sucrose by B.
Similarly in this study as well, the significance and depend- endopyticus was further scaled up in 3 L bioreactor (New
ency of the strain on N, P and Na source were clearly shown Brunswick Scientific, Bioflo 110) using optimal media com-
as a prompt requirement to increase the cell mass. Later for ponents suggested by statistical model. A working volume
PHA production, limitation of the nitrogen source was found of 1 L of production media with optimal concentration of
to be significant which shows the limitation of nutrients ­KH2PO4, ­Na2HPO4 and sucrose with 1.5, 1.5 g/L and 4%
except carbon enhances PHA production (Narayanan et al. was used respectively for batch cultivation. As per statis-
2021). The conclusive finding of the study is that the basal tical analysis (CCD) the obtained PHA % obtained using
medium can be enriched with essential nutrients for fast this optimal condition was 46.5 (11th run Table 6). 250 rpm
cell growth and later during the late log phase reducing the agitation rate with the desired dissolved oxygen of 100% was
nitrogen source can amplify PHA production. The insignifi- maintained during 72 h of batch cultivation. In this scale
cant correlation obtained during statistical optimization can up study, it was found that at 72 h the PHA production was
be correlated in this case. Similar results were reported by 49.9% which was even higher than the production obtained
another group where in growth of B. aryabhattai were found during the statistical experimental run. So this optimised
to excess in nutrient rich medium given initially with extra media was in well accordance with the suggested model
nitrogen source which on depletion of the same enhanced and can be used further for increased production. A similar
PHA production (Pillai et al. 2017). Another study reported statistical study was conducted previously under controlled
that PHB was cell-growth-associated and PHA production medium condition with initial 4% sugar content provided by
started with log phase around 36 h and increase to maximum SCM at 250 rpm by test organism which gave upto 68.5%
around 48 h with minimal nitrogen sources (Yao et al. 2012). of PHA. This result was published and was also found that

13
6002 Archives of Microbiology (2021) 203:5993–6005

PHA accumulation started during the late log phase utilising due to intracellular utilisation of PHA. PHA seems to get
the sugar content as expected and gave a maximum produc- reduced after 76 h due to reutilisation of the energy storage
tion of 5.85 g/L from 8.53 g/L of biomass at 72 h (Geethu and hydrolysis reaching stationary phase wich suggests the
et al. 2019). critical time to harvest the biomass for the optimal gain of
The 1 h interval time course growth study of B. endophyt- PHA. In the low sugar environment provided by the medium
icus during growth in the bioreactor revealed that the organ- during late stationary phase, or in the sugar depleting con-
ism directly utilise sucrose and produce PHA during the dition, hydrolysis happens which is mediated by enzymes
growth phase (Fig. 4). Biomass was increasing constantly especially produced by PhaZ gene within the strain (Mans-
till the nutrients got exhausted till 72 h and about 1.72 g/L field et al. 1995; de Andrade Rodrigues et al. 2000). The
of biomass and 0.8 g/L of PHA (49.47% of PHA) was gravi- culture pH tend to decrease from initial pH of 7.2–4.8, which
metrically analysed during cultivation and started to reduce inturn effects the cell growth during the stationary stage. But
after 72 h. The optimal production rate was observed during in this condition, the accumulation of PHA can be assumed
68–72 h, indicating that late log phase is the ambient cultur- to be maximum in the existing cells due to the harsh condi-
ing time to extract the polymer before getting utilised by the tions provided in the stationary phase. The main fact is that
same strain owing to favourable conditions that may prevail PHA acts as the carbon source leading to the reduction in
(Fig. 4). The production was found to be reduced as part of the cytoplasmic PHA during the late phases of production
the reutilisation by the organism which gave only 38.81% itself. The specific optimal production phase of the strain
of PHA. Based on the result depicted in Fig. 4, 70–72 h was found to be during late log phase/stationary phase. So
were considered to be optimum for the maximum production any stratergies that can prolong growth phase of the organ-
of PHA in B. endophyticus, beyond which the production ism enable to increase the PHA yield per biomass.
terminates due to sporulation and cessation of cell growth
due to intracellular utilisation of PHA (Valappil et al. 2007; FTIR and 1HNMR analysis of the obtained polymer
Nair et al. 2014). Application of different statistical design
enables proper planning to study the interactions between Characterisation of extracted PHA film was confirmed by
different parameters. Difference in the PHA production rate Fourier transform Infrared (FTIR) and 1H nuclear magnetic
is mainly due to the substrate variability and variability of resonance (NMR) spectroscopy (Kulkarni et al. 2011).
parameter during the production phase (Mohanrasu et al. Characterization of PHA obtained from B. endophyticus
2020). The efficient interaction of various parameters with was represented in Fig. 5b with a characteristic peak C=O
minimal experimental design enables to generate maximal stretch at 1726 ­cm−1 corresponding to the ester group. The
process response. strong adsorption band around 1379, 1458, 2929, 1649 and
Based on the result, 70–72 h, late stationary phase was 3749 ­cm−1 corresponds to –CH3, –CH2, CH, C–O, and O–H
considered to be optimum for the maximum production of groups respectively, which was reported to be similar with
PHA in B. endophyticus, beyond which the production ter- pure PHB (Fig. 5a).
1
minates due to sporulation and cessation of cell growth, or H NMR is routinely used as an analytical technique to
determine PHA produced. In this study, PHA film obtained
after each trial was characterised by 1H NMR. Similar spec-
tral pattern with the standard PHB spectra was obtained
(Fig. 6) (Chaijamrus and Udpuay 2008). Characteristic dou-
blet at 1.299 ppm indicated the presence of methyl group
was found coupled with single proton. The peak at 2.57 ppm,
Doublet of quadruplet at corresponding methylene group
adjacent to an asymmetric carbon bearing single atom was
observed. The methylene group characteristic Multiplet was
found precisely on 5.27 was determined from the 1H NMR
spectra (Narayanan et al. 2021).

Conclusing remarks

The process development stratergy for B. endophyticus, an

aerobic PHA producer mainly relay on two basic princi-
Fig. 4  Growth of B. endophyticus with 1 L optimal production media ples like oxygen utilisation behaviour and its capability to
in a 3 L bioreactor PHA cope up with the fermentation parameters economically

13
Archives of Microbiology (2021) 203:5993–6005 6003

Fig. 5  FTIR spectra of a stand-

ard PHB and b PHA extracted
from B. endophyticus using 4

to synthesise PHA. Experimental studies were conducted of this study. This study has led to the conclusion in corre-
to identify and optimise factors that significantly influ- lating the biomass and PHA production by optimizing two
ence PHA production in bioreactor. The different param- different media conditions so as to increase cell mass during
eters influencing the biomass accumulation which in turn the initial stages and PHA production in the latter stages
increases the PHA synthesis has to be identified and opti- by altering the media ingredients. The optimization stud-
mized for maximum PHA production. For this, conventional ies showcased the requirement of potassium, and sodium
optimization process was carried through bioreactor. PHA requirement of the strain for growth and maintenance of
producing B. endophyticus was found to be an excellent cell, whereas deficiency of nitrogen source increased PHA
strain to utilize simple sugars like sucrose was an outcome production. PHB of 49% was obtained from the lab scale

13
6004 Archives of Microbiology (2021) 203:5993–6005

otherwise in a credit line to the material. If material is not included in

the article's Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

References
Akaraonye E, Keshavarz T, Roy I (2010) Production of polyhydroxy-
alkanoates: the future green materials of choice. J Chem Technol
Biotechnol 85(6):732–743. https://​doi.​org/​10.​1002/​jctb.​2392
Amache R, Sukan A, Safari M, Roy I, Keshavarz T (2013) Advances
in PHAs production. Chem Eng. https://​doi.​org/​10.​3303/​CET13​
32156
Fig. 6  1H NMR spectra of PHA extracted from B. endophyticus using Aramvash A, Shahabi ZA, Aghjeh SD, Ghafari M (2015) Statistical
4% sucrose as carbon source physical and nutrient optimization of bioplastic polyhydroxybu-
tyrate production by Cupriavidus necator. Int J Environ Sci Tech-
nol 12:2307–2316. https://​doi.​org/​10.​1007/​s13762-​015-​0768-3
bioreactor which was higher than any other PHB producing Bora L (2013) Polyhydroxybutyrate Accumulation in Bacillusmega-
Bacillus strain utilizing sucrose as sole source of carbon. terium and optimization of process parameters using response
surface methodology. J Polym Environ 21:415–420. https://​doi.​
Further cheap feed stocks rich in sucrose content can be org/​10.​1371/​journ​al.​pone.​02304​43
replaced to reduce the overall production rate of PHB in Chaijamrus S, Udpuay N (2008) Production and characterization of
large scale for industrial production. polyhydroxybutyrate from molasses and corn steep liquor pro-
duced by Bacillus megaterium ATCC 6748. Agric Eng Int CIGR
Acknowledgements The authors are grateful to the Department of J. https://​cigrj​ournal.​org/​index.​php/​Ejoun​ral/​artic​le/​view/​1216
Biotechnology (DBT), Government of India, for sponsoring the work Chen G-Q, Jiang X-R (2018) Engineering microorganisms for improv-
(PR 18430/BIC/101/703/2016) and the Department of Biotechnology, ing polyhydroxyalkanoate biosynthesis. Curr Opin Biotechnol
Manipal Institute of Technology, Manipal University, India; for provid- 53:20–25. https://​doi.​org/​10.​1016/j.​copbio.​2017.​10.​008
ing the facilities to carry out the research work. Rodrigues MF, Valentin HE, Berger PA, Tran M, Asrar J, Gruys KJ,
Steinbüchel A (2000) Polyhydroxyalkanoate accumulation in
Burkholderia sp.: a molecular approach to elucidate the genes
Author contributions MG performed the experimental part of this involved in the formation of two homopolymers consisting of
research work as part of her Ph.D. thesis project and dissertation. MSD short-chain-length 3-hydroxyalkanoic acids. Appl Microbiol
provided supervision of the research and writing this article (review Biotech 53:453–460. https://​sfamj​ourna​ls.​onlin​elibr ​ary.​wiley.​
and editing) and the manuscript. HRC provided the necessary facilities com/​doi/​pdfdi​rect/​10.​1111/​jam.​13469
required to initiate the research work. Md Din MF, Ponraj M, Van Loosdrecht M, Ujang Z, Chelliapan S,
Zambare V (2014) Utilization of palm oil mill effluent for polyhy-
Funding Open access funding provided by Manipal Academy of droxyalkanoate production and nutrient removal using statistical
Higher Education, Manipal. This study was supported by The Depart- design. Int J Environ Sci Technol 11:671–684. https://​doi.​org/​10.​
ment of Biotechnology (DBT), Government of India, for sponsoring 1007/​s13762-​013-​0253-9
the work (PR 18430/BIC/101/703/2016). Geethu M, Vrundha R, Raja S, Raghu Chandrashekar H, Divyashree
MS (2019) Improvement of the production and characterisation of
Declarations polyhydroxyalkanoate by Bacillus endophyticus using inexpensive
carbon feedstock. J Polym Environ 27:917–928. https://​link.​sprin​
ger.​com/​artic​le/​10.​1007/​s10924-​019-​01397-z
Conflict of interest There is no conflict of interest involved in this re- Kaur G, Roy I (2015) Strategies for large-scale production of polyhy-
search work. droxyalkanoates. Chem Biochem Eng Q 29:157–172. https://​doi.​
org/​10.​15255/​CABEQ.​2014.​2255
Ethics approval and consent to participate Not applicable. Kulkarni SO et al (2011) Characterisation of copolymer, poly
(hydroxybutyrate-co-hydroxyvalerate) (PHB-co-PHV) produced
Consent for publication Not applicable. by Halomonas campisalis (MCM B-1027), its biodegradability
and potential application. Bioresour Technol 102:6625–6628.
Availability of data and materials It will be provided on request and https://​doi.​org/​10.​1016/j.​biort​ech.​2011.​03.​054
there is no supplementary data provided. Lakshman K, Shamala TR (2006) Extraction of polyhydroxyalkanoate
from Sinorhizobium meliloti cells using Microbispora sp. culture
Open Access This article is licensed under a Creative Commons Attri- and its enzymes. Enzyme Microb Technol 39:1471–1475. https://​
bution 4.0 International License, which permits use, sharing, adapta- doi.​org/​10.​1016/j.​enzmi​ctec.​2006.​03.​037
tion, distribution and reproduction in any medium or format, as long Mansfield DA, Anderson AJ, Naylor LA (1995) Regulation of PHB
as you give appropriate credit to the original author(s) and the source, metabolism in Alcaligenes eutrophus. Can J Microbiol 41:44–49.
provide a link to the Creative Commons licence, and indicate if changes https://​doi.​org/​10.​1139/​m95-​166
were made. The images or other third party material in this article are Miller GL (1959) Use of dinitrosalicylic acid reagent for determination
included in the article's Creative Commons licence, unless indicated of reducing sugar. Anal Chem 31:426–428

13
Archives of Microbiology (2021) 203:5993–6005 6005

Mohammed S, Behera HT, Dekebo A, Ray L (2019) Optimization of Russell A (2003) Lethal effects of heat on bacterial physiology and
the culture conditions for production of Polyhydroxyalkanoate and structure. Sci Prog 86 (1–2):115–137. https://​doi.​org/​10.​3184/​
its characterization from a new Bacillus cereus sp. BNPI-92 strain, 00368​50037​83238​699
isolated from plastic waste dumping yard. Int J Biol Macromol. Slepecky RA, Law JH (1960) A rapid spectrophotometric assay of
https://​doi.​org/​10.​1016/j.​ijbio​mac.​2019.​11.​138 alpha, beta-unsaturated acids and beta-hydroxy acids. Anal Chem
Mohanrasu K et al (2020) Optimization of media components and cul- 32 (12):1697–1699. https://doi.org/10.1021/ac60168a046
ture conditions for polyhydroxyalkanoates production by Bacillus Tripathi AD, Srivastava SK, Singh RP (2013) Statistical optimization
megaterium. Fuel 271:117522 of physical process variables for bio-plastic (PHB) production by
Nair AM, Annamalai K, Kannan SK, Kuppusamy S (2014) Charac- Alcaligenes sp. Biomass Bioenerg 55:243–250. https://​agris.​fao.​
terization of polyhydroxyalkanoates produced by Bacillus subtilis org/​agris-​search/​search.​do?​recor​dID=​US201​50015​5357
isolated from soil samples. Malaya J Biosci 1:8–12 Valappil SP, Boccaccini AR, Bucke C, Roy I (2007) Polyhydroxy-
Narayanan M et al (2021) Optimization and production of polyhy- alkanoates in Gram-positive bacteria: insights from the genera
droxybutyrate from sludge by Bacillus cereus categorized through Bacillus and Streptomyces. Anton Leeuw 91:1–17. https://​doi.​
FT-IR and NMR analyses. J Environ Chem Eng 9:104908. https://​ org/​10.​1007/​s10482-​006-​9095-5
doi.​org/​10.​1016/j.​jece.​2020.​104908 Wang J, Bakken L (1998) Screening of soil bacteria for poly-β-
Pijuan M, Casas C, Baeza JA (2009) Polyhydroxyalkanoate synthe- hydroxybutyric acid production and its role in the survival of
sis using different carbon sources by two enhanced biological starvation. Microb Ecol 35:94–101. https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 0024​
phosphorus removal microbial communities. Process Biochem 89900​063
44(1):97–105. https://​doi.​org/​10.​1016/j.​procb​io.​2008.​09.​017 Yao Z, Davis RM, Kishony R, Kahne D, Ruiz N (2012) Regulation of
Pillai AB, Kumar AJ, Thulasi K, Kumarapillai H (2017) Evaluation of cell size in response to nutrient availability by fatty acid biosyn-
short-chain-length polyhydroxyalkanoate accumulation in Bacil- thesis in Escherichia coli. PNAS 109 (38):E2561–E2568. https://​
lus aryabhattai. Braz J Microbiol 48:451–460. https://​doi.​org/​10.​ doi.​org/​10.​1073/​pnas.​12097​42109
1016/j.​bjm.​2017.​01.​005
Prados E, Maicas S (2016) Bacterial production of hydroxyalkanoates Publisher's Note Springer Nature remains neutral with regard to
(PHA). Univ J Microbiol Res 4:23–30. https://​doi.​org/​10.​13189/​ jurisdictional claims in published maps and institutional affiliations.
ujmr.​2016.​040104

13