Journal of Cleaner Production 182 (2018) 272e279

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Cleaner production of vanillin through biotransformation of ferulic

acid esters from agroresidue by Streptomyces sannanensis
Pritam Chattopadhyay a, b, Goutam Banerjee c, d, Sukanta K. Sen a, *
a
Department of Botany, Visva Bharati, West Bengal PIN-731235, India
b
Department of Biotechnology, Gauhati University, Assam PIN-781014, India
c
Department of Zoology, Visva Bharati, West Bengal, PIN-731235, India
d
Department of Biochemistry, University of Calcutta, West Bengal PIN-700019, India

a r t i c l e i n f o a b s t r a c t

Article history: Vanillin (4-hydroxy-3-methoxybenzaldehyde; C8H8O3) is the major organoleptic compound of vanilla
Received 11 April 2017 flavor originally obtained from Vanilla planifolia. Attempt has been made for production of vanillin
Received in revised form from ferulic acid esters present in wheat bran, the agro-residue. Streptomyces sannanensis MTCC 6637
31 December 2017
was used for biotransformation of ferulic acid esters. The key enzymes involving vanillin production
Accepted 3 February 2018
Available online 8 February 2018
were assayed and the products were estimated using thin layer chromatography and high pressure
liquid chromatography. Vanillin production was optimized through response surface methodology.
Optimum vanillin production (708 mg L
1) was achieved with de-starched wheat bran (10% w/v),
Keywords:
Agroresidue
sucrose (0.2% w/v), peptone (1% w/v) at pH 7.5, agitation 220 rpm, and temperature 28  C and
Ferulic acid esters fermentation continued for a period of 5 days. The experimental strain converts ferulic acid ester into
Biotransformation ferulic acid with the help of ferulic acid esterase. Ferulic acid was catabolized through Coenzyme-A
Vanillin dependent noneb-oxidation (retro-aldol reaction) of ferulic acid. Feruloyal Coenzyme-A synthetase
Streptomyces sannanensis and Enoyl-Coenzyme-A hydratase/aldolase were involved in conversion of ferulic acid into vanillin.
Response surface methodology Transient formation of vanillic acid from vanillin was found due to steady state expression of vanillin
dehydrogenase. To the best of authors’ knowledge this is the first report of vanillin production directly
from agroresidue by S. sannanensis.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction identical vanillin can be achieved either by microbial or enzymatic

Vanillin is the major organoleptic compound of natural vanilla
obtained from beans of Vanilla planifolia, an orchid. Vanillin is one
of the most demanding food additives also used in beverages,
perfumes and pharmaceuticals (Priefert et al., 2001). The natural
production from Vanilla bean, beside its high cost, cannot satisfy the
total vanillin demand of the world. Annually, more than 12,000 tons
of vanillin is produced out of which only 1% originates
from V. planifolia (Sindhwani et al., 2017). Considering the
increased interest in natural products, the production of flavors via
biotransformation process offers a viable alternative to the natural
and chemical sources (Sindhwani et al., 2017). According to
European commission (EC) legislation, production of nature
* Corresponding author. Department of Botany, Visva-Bharati University, Santi-
niketan, 731235 West Bengal, India.
E-mail address: sksenvb@rediffmail.com (S.K. Sen).
transformation of a precursor or by microbial de novo synthesis,
that may well be referred as biovanillin (Zamzuri and Abd-Aziz,
2013).
Biotechnologically vanillin production may be based on
microbial fermentation processes or on bioconversion of natural
precursor using tailored microbial cells or enzymes. Ferulic acid [4-
hydroxy-3-methoxycinnamate (FA)] is a phenolic compound,
linked to the plant cell wall through ester bonds and can be
released by enzymatic hydrolysis of the ester bonds (di Gioia et al.,
2007). The cell wall of agroresidue like wheat bran contains
approximately 95% (w/w) phenolic compounds (Parker et al.,
2005). The major drawback of microbial production of vanillin is
it's the high price of its raw material i.e. FA. Green productions of
vanillin from wood chips (Modahl et al., 2015) and bamboo
(Harshvardhan et al., 2017) through bacterial consortium are also
attempted that could produce vanillin through lignin biodegrada-
tion. The search for possible eco-friendly single-step vanillin pro-
duction system through different mechanism (biotransformation/

https://doi.org/10.1016/j.jclepro.2018.02.043
0959-6526/© 2018 Elsevier Ltd. All rights reserved.
 P. Chattopadhyay et al. / Journal of Cleaner Production 182 (2018) 272e279
bioconversions/de novo synthesis) using various biological agents
(microorganisms/plant cells/isolated enzymes) is still going on.
Agroresidues like wheat bran are cheap and contains ferulic acid
esters (FAE) that may directly be used as substrate for cleaner
production of vanillin. Employing agroresidue as the raw material
for vanillin production represents an interesting way of disposing
the wastes and at the same time value addition to them. To achieve
this, the qualified strain has to be survived well in presence of
phenolic acids by hydrolyzing ester linkage present in agroresidue
through ferulic acid esterase (FAEase), and to convert FA into
vanillin. Present study describes biotransformation of FAE of de-
starched wheat bran (DSWB) into vanillin by Streptomyces
sannanensis (MTCC 6637). The present paper also deals with the
process optimization to check the potentiality of this system for
commercial vanillin production in future.
2. Materials and methods
2.1. Strain used
Streptomyces sannanensis MTCC 6637 was procured from the
Microbial Type Culture Collection (MTCC), Institute of Microbial
Technology (IMTECH), Chandigarh, India. The strain was main-
tained in Arginine Glycerol Salt (AGS) slants at 4  C.
2.2. Agroresidue used
Wheat bran is a reach source of FAE (Kim et al., 2006). Wheat
(Triticum aestivum) bran samples were collected from local market
at Santiniketan, India. Collected wheat bran was de-starched
following the method of Sarangi and Sahoo (2010). Physical
properties of wheat bran (particle size, intrinsic toughness, intrinsic
strength) vary with the cell wall composition and milling
quality (Liu, 2013). However, physical properties become identical
in de-starched wheat bran (DSWB). DSWB was used in all the ex-
periments as primary source of carbohydrate unless otherwise
stated.
2.3. Fermentation of DSWB
S. sannanensis MTCC 6637 was grown in liquid minimal medium
(Muheim and Lerch, 1999) containing wheat bran (1%, w/v) as a sole
carbon source in 250 mL flasks containing 50 mL medium. The
initial pH of the medium was adjusted to 7.0 and sterilized by
autoclaving for 15 min at 15 lbs. Fermentation was carried at 35  C
and analyzed daily up to 14 days.
2.4. Growth curve determination
To obtain the dry weight the biomass was separated from fer-
mented broth by centrifugation (1000xg for 20 min at 4  C) and
dried in a hot air oven (45  C) unless the weight reaches a constant
value. To determine the growth, biomass concentration (dry
weight) was monitored throughout the fermentation period at
every 3 h interval.
2.5. Extraction and analysis of biotransformed products
The fermented broth was separated from biomass by
centrifugation (1000xg for 20 min at 4  C). The supernatant was
acidified (pH 1.0e2.0) with 1 M hydrochloric acid (HCl), and
extracted with equal volume of ethyl acetate. The ethyl acetate
was evaporated in a vacuum evaporator and the residue was
dissolved in 50% methanol. In this process the fermented broth
was subjected to thin layer chromatography (TLC) and high
273
pressure liquid chromatography (HPLC) (Sachan et al., 2004).
HPLC data were analyzed and quantified based on custom
software.
2.6. Preparation of crude cell extracts
For preparation of cell extract S. sannanensis MTCC 6637 was
grown as described with on DSWB as sole carbon source. Cell pellet
was collected (by centrifugation at 12,000 rpm for 20 min), washed
(in cold TriseHCl buffer, 50 mM, pH 7.8), and sonicated (in short
bursts of 30 s with a total exposure time of 5 min). Cell extract was
again centrifuged and the supernatant was collected, concentrated
(using rotary vacuum evaporator), sterilized (through 0.45-mm
membrane filter) and was used as crude extract for enzyme assay
and in vitro conversion study.
2.7. Assay of key enzymes of DSWB biotransformation
To monitor the biotransform process, four key enzymes of the
FAE biotransformation catabolic route (Gallage and Moller, 2015)
were estimated throughout the fermentation process. For assay of
extracellular FAEase, the centrifuged fermented broth was
extracted with equal volume of ethyl acetate and FA was measured
spectrophotometrically at 310 nm against the standard curve
(Kaur et al., 2013). To estimate enoyl-Coenzyme-A hydratase/
aldolase (ECH), crude cell extract (0.5 mL) was mixed with equal
volume reaction mixture [90 mM sodium phosphate buffer (pH
7.0), 3 mM Magnesium chloride (MgCl2), and 0.2 mM 4-hydroxy-
3-methoxyphenyl-bhydroxypropionyl-Coenzyme-A], and incu-
bated at 30  C (Overhage et al., 1999). The resultant vanillin was
estimated through HPLC. For feruloyl-Coenzyme-A synthetase
(FCS) assay crude cell extract (0.5 mL) was mixed with equal
volume reaction mixture [100 mM potassium phosphate buffer
(pH 7.0), 2.5 mM MgCl2, 0.7 mM ferulic acid, 2 mM Adenosine
three phosphate (ATP), 0.4 mM Coenzyme-A (CoA)], and
incubated at 30  C. The resultant feruloyl-CoA was measures
spectrophotometrically at 345 nm against the standard curve
(Overhage et al., 1999). For measuring vanillin dehydrogenase
(VDH) activity, cell free fermented broth was mixed with equal
volume reaction mixture [vanillin 1 gL
1, triammonium citrate
2 gL
1, sodium acetate 2 gL
1, magnesium sulphate 50.1 gL
1,
manganous sulphate 0.05 gL
1, di-potassium hydrogen phosphate
2 gL
1; pH 5.6] and incubated at 37  C for 24 h (Kaur et al., 2013).
Thereafter 1 mL of reaction mixture carefully mixed with 5 mL of
HCl (24% v/v) and 2 mL of thiobarbituric acid (1% w/v) and heated
in a 55  C for 10 min and allowed to cool at room temperature for
20 min. The absorbance was recorded at 434 nm. The specific ac-
tivity in each case was expressed in term of product produced
min
1 mg
1 protein (U mg
1 protein).
2.8. Determining the metabolic route of DSWB biotransformation
2.8.1. In vivo conversion study
In order to explore the metabolic route of DSWB biotransfor-
mation by S. sannanensis, DSWB two separate set of experiments
were carried out. In first set, ferulic acid, vanillin and vanillic acid
were independently supplied to the medium as a sole carbon
source. The fermented broth was analyzed up to eight days for the
detection of transformed product(s). To explore the CoA thioesters-
dependent route of biotransformation, in second set, S. sannanensis
was grown on minimal medium supplemented with a metabolic
inhibitor [3,4 (Methylenedioxy)-cinnamic acid (MDCA)], and
changes in the accumulation of products were compared to control
as previously described by Ghosh et al. (2007).
274 P. Chattopadhyay et al. / Journal of Cleaner Production 182 (2018) 272e279
2.8.2. In vitro conversation study
In order to study in vitro conversion of FAE, FA, vanillin, and
vanillic acid (VA), four set of experiments were carried out where
each set contains three separate experiments. In each of the set the
only variable was 1 mM solo carbon source (FAE/FA/vanillin/VA). In
brief, 1.5 mL of reaction mixture [260 mM potassium phosphate
buffer (pH 7.0), 3 mM ATP, 3 mM MgCl, 1e3 mM Nicotinamide
adenine dinucleotide (NAD), 0.25 mM CoA] and crude cell extract
(200 pg protein) was incubated at 25  C for 4 h (Narbad and Gasson,
1998). Then the samples were analyzed by HPLC.
2.9. Process optimization for vanillin production
Process optimization was carried out in 250 mL flasks as 50 mL
batch culture following the model ‘one-factor-by-another’ for
maximum accumulation of vanillin in the fermented broth within a
time period of 14 days. After selecting the minimal medium, me-
dium component (carbon, nitrogen and surfactant) along with
physical parameters (inoculum growth phase, fermentation period,
temperature, pH, and agitation) were optimized. Finally optimiza-
tion was done with four independent variables (DSWB, sucrose,
peptone and tween 80) by central composite design (CCD) of
response surface methodology (RSM) using the statistical software
package DESIGN-EXPERT® 10.0.3 (StatEase, Inc., Minneapolis, USA).
Fig. 1. Detection of phenolic derivatives by HPLC-UV analysis of fermented broth on 4th, 8th, and 12th days of incubation of S. sannanensis MTCC 6637 where DSWB was the only
carbon source.
3. Results
3.1. Fermentation of DSWB by S. sannanensis MTCC 6637
The conversion of FAE present in DSWB into vanillin by
S. sannanensis MTCC 6637 is checked through TLC and HPLC up to
two weeks, at two days interval (14 days) (Fig. 1). TLC chromato-
grams represented three major phenolics, corresponding to FA,
vanillin and VA as compared with standards. HPLC analysis also
confirmed the presence of FA, vanillin and low amount of VA as the
major metabolites resulted from fermentation of DSWB after eight
days of incubation period (Fig. 1).
3.2. Relation among bacterial growth, bio-transformed products
and key enzymes
Maximum accumulation of FA is achieved on 4th day, highest
concentration of vanillin is observed on 8th day, and maximum
accumulation of VA is recorded on 12th day (Figs. 1 and 2a). The
highest rate of FA biotransformation was observed on 7th day of
fermentation (Fig. 2a). The bacterial growth curve correlates the
sequential utilization of FAE and FA, indicated by the appearance of
two lag phases (Fig. 2a).
In order to verify the biotransformation events, the key en-
zymes involve in FAE transformation are evaluated. FAEase
 P. Chattopadhyay et al. / Journal of Cleaner Production 182 (2018) 272e279 275

Fig. 2. Relation of bacterial growth curve with (a) phenolic derivatives and (b) key enzymes.

converts FAE into FA. Maximum FAEase specific activity of
0.294 ± 0.035 U mg
1 protein was exhibited by S. sannanensis
MTCC 6637 after 2 days of fermentation (Fig. 2b). Thereafter
FAEase activity was gradually reduced. FAEase activity can be
correlated with the first step of growth curve (Fig. 2b). On the
other hand, FCS converts FA into feruloyal Co-A. Maximum FCS
activity (0.719 ± 0.025 U mg
1 protein) is recorded on 4th day
(Fig. 2b). ECH converts feruloyal Co-A into vanillin. Maximum ECH
activity (1.192 ± 0.025 U mg
1 protein) is recorded on 6th day and
decreased thereafter (Fig. 2b). Thus, FCS and ECH activity can be
correlated with second step of growth curve. However, the
concentration of VDH was remained constant throughout the
experimental period irrespective of the vanillin concentration
(Fig. 2b). It might be due to transient VA formation up to 8 day
(1.3 gL
1) of fermentation. VA accumulation reached its maximum
level at 10th day of fermentation, probably due to the exhaustion
of carbon sources except vanillin.
3.3. Metabolic route of DSWB biotransformation and key enzymes
To explore the metabolic route of DSWB biotransformation, in-
vivo and in-vitro biotransformation experiments were carried out
and the results are presented in Table 1. In both the system,
biotransformation of FAE, FA, vanillin and transient VA formation
were observed [Table 1, Exp. No. A(i), and C(i)]. When FA was used
as a sole carbon source, vanillin and VA were observed [Table 1,
Exp. No. A(ii), and D(i)]. It was also found to convert vanillin into
VA [Table 1, Exp. No. A(iii), and E(i)]. Protocatechuic acid or
guaiacol was not formed upon VA degradation [Table 1, Exp. No.
A(iv), and F(i), (ii), (iii)]. Therefore, S. sannanensis MTCC 6637 in-
dicates the capacity to utilize FAE, FA and vanillin as sole carbon
sources, but cannot utilize VA. These results may suggest that
vanillin has formed from FAE (present in DSWB) via FA.
While S. sannanensis was grown on minimal medium containing
DSWB as a sole source of carbon along with MDCA inhibitor
276 P. Chattopadhyay et al. / Journal of Cleaner Production 182 (2018) 272e279

Table 1
Results of in-vitro and in-vivo biotransformation of FAE, FA, vanillin and VA by S. sannanensis MTCC 6637.

Bio-transformation Substrate Exp. Ex. Fermentation cond. Biotransformation product (g L
1)

type used set no.
FA Vanillin VA Prottocatechuic
acid and/or
Guaiacol

In-vivo FAE (DSWB) A i Bacteria þ MBS for 8 day 0.224 ± 0.004 0.376 ± 0.004 0.004 ± 0.001 e
FA ii Bacteria þ MBS for 8 day e 0.403 ± 0.004 0.007 ± 0.001 e
Vanillin iii Bacteria þ MBS for 8 day e e 0.015 ± 0.003 e
VA iv Bacteria þ MBS for 8 day e e e e
FAE (DSWB) B i Bacteria þ MBS þ MDCA for 8 day 0.225 ± 0.004 e e e
FA ii Bacteria þ MBS þ MDCA for 8 day e e e e
Vanillin iii Bacteria þ MBS þ MDCA for 8 day e e e e
VA iv Bacteria þ MBS þ MDCA for 8 day e e e e
In-vitro FAE (DSWB) C i Cell extract þ PBS þ MgCl2 þ ATP þ NAD þ CoASH for 0.024 ± 0.005 0.058 ± 0.005 0.025 ± 0.004 e
6h
ii Cell extract þ PBS þ MgCl2 þ ATP þ CoASH for 6 h 0.022 ± 0.003 0.060 ± 0.005 0.004 ± 0.001 e
iii Cell extract þ PBS þ MgCl2 þ ATP þ for 6 h 0.024 ± 0.003 e e e
FA D i Cell extract þ PBS þ MgCl2 þ ATP þ NAD þ CoASH for e 0.109 ± 0.008 0.024 ± 0.004 e
6h
ii Cell extract þ PBS þ MgCl2 þ ATP þ CoASH for 6 h e 0.110 ± 0.008 0.005 ± 0.001 e
iii Cell extract þ PBS þ MgCl2 þ ATP þ for 6 h e e e e
Vanillin E i Cell extract þ PBS þ MgCl2 þ ATP þ NAD þ CoASH for e e 0.086 ± 0.005 e
6h
ii Cell extract þ PBS þ MgCl2 þ ATP þ CoASH for 6 h e e 0.005 ± 0.001 e
iii Cell extract þ PBS þ MgCl2 þ ATP þ for 6 h e e 0.004 ± 0.002 e
VA F i Cell extract þ PBS þ MgCl2 þ ATP þ NAD þ CoASH for e e e e
6h
ii Cell extract þ PBS þ MgCl2 þ ATP þ CoASH for 6 h e e e e
iii Cell extract þ PBS þ MgCl2 þ ATP þ for 6 h e e e e

‘-’ indicates not detected.

(25 mM), only FA was produced as biotransformed product [Table 1,
Exp. No. B(i)]. Since, FAEase plays the crucial role in conversion of
FAE into FA, FAEase is not inhibited by MDCA. Rather, MDCA was
also found to inhibit formation of vanillin and VA from FA [Table 1,
Exp. No. B(ii)]. MDCA is a potential inhibitor of hydroxycinnamate
CoA-ligase (4CL), like FCS. FCS converts FA into feruloyal Co-A
(Fig. 2b). On the other hand, in in-vitro experiment trans-
formation of FAE and FA into vanillin was not achieved in the
absence of CoASH [Table 1, Exp. No. C(iii) and D(iii)]. Therefore,
this biotransformation process is a CoA-dependent one. Feruloyal
Co-A is then converted into 4-hydroxy-3-methoxyphenyl-b-
hydroxypropionyl-CoA by ECH as a transient intermediate and
subsequently converted into vanillin through a retro-aldol reaction
(Fig. 2b).
In in-vitro biotransformation experiments, in the presence of
CoASH, MgCl2, ATP and NADþ, the cell free extract was found to
convert DSWB into FA, vanillin, and VA [Table 1, Exp. No. C(i)];
whereas FA into vanillin, and VA [Table 1, Exp. No. D(i)]. In repeat
experiment without NADþ [Table 1, Exp. No. C(ii) and D(ii)], very
low amount of VA was produced, indicating necessity of NADþ for

Fig. 3. Detail of the catabolic route of DSWB biotransformation by S. sannanensis MTCC 6637 (Based on Gallage and Moller, 2015 and present work).
 P. Chattopadhyay et al. / Journal of Cleaner Production 182 (2018) 272e279 277

Table 2 incubation period (Table S1). To determine the optimal inoculum

Optimized condition for biotransformation of ferulic acid esters from DSWB into condition, bioconversion assays were performed with cells
vanillin.
collected from cultures at different phase of growth. Serial dilutions
Optimization parameter Resultant condition were carried out to get readable c.f.u. and results were obtained by
Minimal medium MBS multiplication with dilution factor. The results showed that the
Carbon source DSWBa highest bio-conversion yield was obtained with cells grown at 2nd
Supplementary carbon source Sucrosea log phase of growth curve (8th day). A 2-fold decrease in vanillin
Nitrogen source Peptonea
formation was observed when cells were harvested at stationary
Surfactant Tween 80a
Inoculum condition during log phase of 2nd stage phase (12th day) or early log phase (2nd day) of growth curve
of growth curve (Fig. S.1). When sucrose (0.1% w/v) was used as supplementary
Temperature 28  C carbon source, an increase in production of vanillin (56 mg L
1) was
pH 7.5
observed compared to the isolate grown in MBS medium contain-
Agitation 120 rpm
Fermentation time 5 day
ing 1% DSWB with or without supplemented with other carbohy-
a
drates (0.1% w/v) (Fig. S.2). Peptone (5% w/v) was found as best for
Optimum concentration was determined through RSM.
vanillin production by S. sannanensis from DSWB (Fig. S.3). Tem-
perature was found to affect the vanillin production markedly. At
Table 3 28  C temperature highest vanillin production (110.45 mg L
1) was
Central Composite Design showing production of S. sannanensis MTCC 6637 at achieved (Fig. S.4). pH level was determined as 7.5 for optimal
different variable combinations. production of vanillin. pH below 6.5 compromise vanillin produc-
Run Original level Coded level Vanillin tion and pH above 7.5 resulted into VA accumulation rather than
production vanillin production (Fig. S.5).The production rate in shake flasks
(gL
1) was always greater than that of static culture, though agitation at
DSWB Sucrose Peptone Tween A B C D Observed Pre- more than 220 rpm was proven detrimental for vanillin production
(gL
1) (gL
1) (gL
1) 80 dicted (Fig. S.6). Over agitation may be responsible for hyphal shearing,
(gL
1) which resulted into poor biotransformation.
1 100 2 10 0 þ1 þ1 þ1 -a 0.708 0.723 When, DSWB (10%, w/v) is used as main carbon source along
2 0 0 0 0 -a -a -a -a 0.017 0 with sucrose (0.1%, w/v) and peptone (1%, w/v) in MBS medium (pH
3 0 1 5 0.05 -a
1
1
1 0 0.033
7.5), at 220 rpm agitation and inoculum was used from 2nd log
4 50 3 5 0.05
1 þa
1
1 0.242 0.336
5 150 1 5 0.05 þa
1
1
1 1.126 1.269 phase, optimum vanillin production time was found to be curtailed
6 0 0.1 10 0.1 -a
1 þ1 þ1 0.11 0.084 from 8 day to 5 day by S. sannanensis (Table 2). Further optimization
7 50 2 5 0.15
1 þ1
1 þa 0.277 0.411 was done using CCD statistical model taking minimal medium,
8 100 1 0 0 þ1
1
a
a 0.652 0.617 physical parameters and production time.
9 100 0 0 0 þ1
a
a
a 0.674 0.573
10 0 2 10 0
a þ1 þ1
a 0.084 0.05
11 0 1 0 0.1
a
1
a þ1 0.054 0.002 3.5. Product optimization through RSM
12 50 1 5 0.05
1
1
1
1 0.376 0.376
13 100 1 0 0.1 þ1
1
a þ1 0.613 0.529 Among several factors, DSWB, sucrose, peptone and tween 80
14 50 0 0 0.05
1 -a
a
1 0.302 0.429
were considered as important for vanillin production. Each factor
15 100 1 0 0.1 þ1
1
a þ1 0.581 0.531
16 50 2 5 0.05
1 þ1
1
1 0.376 0.318 was taken at five different levels (ea,
1, 0, þ1 and þa). A set of 30
17 100 1 0 0.1 þ1
1
a þ1 0.605 0.504 experiments were carried out (Table 3). Vanillin production was
18 100 2 10 0.1 þ1 þ1 þ1 þ1 0.611 0.569 taken as the dependent variable or response (Y). This resulted in a
19 0 1 10 0.1
a
1 þ1 þ1 0.139 0.121 quadratic model that related the response measured to the in-
20 50 0 15 0.05
1
a þa
1 0.367 0.419
21 0 1 0 0.1
a
1
a þ1 0.109 0.034
dependent variables of the experiment. For a four factor system,
22 50 0 5 0.05
1
a
1
1 0.376 0.392 the quadratic model was fitted to a polynomial equation as
23 50 1 5 0.05
1
1
1
1 0.376 0.392 follows:
24 50 1 5 0.05 þ1
1
1
1 0.376 0.392
25 0 1 0 0
a
1
a
a 0.018 0
Y ¼ b0 þ b1A þ b2B þ b3C þ b4D þ b12AB þ b13AC þ b14AD þ b23
26 100 2 10 0 þ1 þ1 þ1
a 0.66 0.594
27 50 0 5 0
1
a
1
a 0.289 0.333 BC þ b24BD þ b34CD þ b11A2 þ b22B2 þ b33C2 þ b44D2 (1)
28 50 1 5 0.05
1
1
1
1 0.248 0.333
29 0 0 10 0
a
a þ1
a 0.058 0.048 where Y denotes response; b0 denotes model constant A, B, C and D
30 50 1 5 0.05
1
1
1
1 0.376 0.392 were the variable; b1, b2, b3 and b4 are linear interaction; b12, b13,
b14, b23, and b24 are cross product interaction; whereas b11, b22, b33,
and b44 are squared interaction. Using the above model, the
optimum concentration of the medium components was used to
in vitro conversion of vanillin into VA. It is because of the fact that
generate response surface graphs (Fig. 4).
NADþ is an important co-factor for the enzyme VDH that converts
Result of the analysis of variance (ANOVA) is fitted into a
vanillin into VA. Detail catabolic route of bio-transformation was
quadratic model (Table 4). The statistical significance of the
depicted in Fig. 3.
model, as indicated by F-value is observed as 66.15. Among the
experimental parameters, DSWB is found to be significant model
3.4. Media optimization term (Prob > F value is less than 0.05). The lack of fit of F value of the
model is 1.16, which is not significant and may be considered as
Process optimization for vanillin production by S. sannanensis pure error. The coefficient of determinant (R2) that indicates the
from DSWB was carried out in 250 mL flasks as batch culture variability of the model and real relationship between variables is
following the model ‘one-by-one’. All the test media supported found to be 0.9841. The difference between predicted (0.9841) and
ferulic acid production, but maximum vanillin production adjusted (0.9902) R2 value is less than 0.2, which adds to the
(8.63 mgL
1) was obtained with MBS medium after 8 days of acceptance of the model. The degree of precision of the model
278 P. Chattopadhyay et al. / Journal of Cleaner Production 182 (2018) 272e279
Fig. 4. Response surface graph of vanillin production by S. sannanensis MTCC 6637. The effect of different factors on vanillin production was represented as: (a) DSWB and sucrose,
(b) DSWB and peptone, (c) DSWB and tween 80, (d) sucrose and peptone, (e) sucrose and tween 80, and (f) peptone and tween 80.
Table 4
Analysis of variance (ANOVA) for response surface quadratic model.
Values Vanillin production
R2
0.9841
Adjusted R2 0.9692
Predicted R2 0.9345
Adequate precision 33.717
Model F-value 66.15
Lack of fit F-value 1.16
indicated by the coefficient of variance (C.V.) is found to be 13%. The
adequate precision value of the experimental model is 33.71, which
simply indicates the signal to noise ratio (Table 4). The desired
adequate precision value must be greater than 4 this model can be
used to enhance the production. Multiple regression analysis of the
experimental data, followed by polynomial equation (2) is found to
describe the interaction between DSWB (A), sucrose (B), Peptone
(C) and tween 80 (D) for vanillin production. Vanillin ¼ e
(6.27870E-003) þ (5.11367E-003) DSWB þ (0.041410) Sucrose þ
(8.21074E-003) Peptone þ (0.99321) Tween80 e (4.36014E-004)
DSWB*Scrose þ (2.41495-005) DSWB*Peptone e (0.010159)
DSWB*Tween80 þ (3.10790E-004) Sucrose*Peptone e (0.16891)
Sucrose*Tween80 þ (2.42865E-003) Peptone*Tween80 þ
(1.67372E-005) DSWB2 e (0.014011) Sucrose2 e (3.55557E-004)
Peptone2e (2.98325) Tween802.
A validation of the model and regression equation was done by
taking DSWB (100 g L
1), sucrose (2.0 g L
1) and tween 80
(0.0 g L
1) in the experiment. The predicted response for vanillin
production was 0.723 g L
1 and the actual response is 0.708 g L
1,
which, prove the model validity.
4. Discussion
Actinomycetes are well known for biotransformation of
different hydroxycinnamic acids into value added products. Strep-
tomyces are reported to produce FA from FAE of agroresidue. Such
conversion is reported from S. avermitilis CECT 3339 (Garcia et al.,
1998a) and S. avermitilis UAH 30 (Garcia et al., 1998b),
S. thermophile ATCC 34628 (Levasseur et al., 2004), Streptomyces sp.
S10 (Mukherjee et al., 2007), S. ambofaciens (Kheder et al., 2009).
On the other hand, several Streptomyces spp. were reported to
convert FA into vanillin. For example Streptomyces isolate S10
(Sarangi and Sahoo, 2009), Streptomyces sp. strain V-1 (Yang et al.,
2013), and S. setonii (Ma and Daugulis, 2014; Salgado et al., 2012).
Conversion of trans-FA to VA by S. setonii via a CoA-independent
noneb-oxidative reaction was also reported (Sutherland et al.,
1983). Production of vanillin from FA through VA using S. setonii
was also reported (Salgado et al., 2012). But, to the best of author's
knowledge, no organism has been reported to use FAE from agro-
residue as a sole carbon source to produce vanillin as bio-
transformed product in a single step fermentation. In this
investigation, the strain S. sannanensis MTCC 6637 is able to convert
trans-FA into vanillin via CoA-dependent noneb-oxidative reaction
sequence, known to occur via a CoA-dependent retro-aldol
mechanism.
Microbial production of vanillin is often as an intermediate
product and is highly sensitive for oxidation into VA (Ghosh et al.,
2007). VA may further be converted into protocatechuic acid
(Sutherland et al., 1983), or guaiacol (Chow et al., 1999). Present
study observes a stable accumulation of vanillin with transient
production of VA. But, further conversion of VA was not observed.
Detail catabolic routes for biotransformation are deduced (Fig. 3).
 P. Chattopadhyay et al. / Journal of Cleaner Production 182 (2018) 272e279
Searching of industrial residual substance as substrate for
biotransformation is an ongoing process. Byproducts of agro in-
dustries (e.g. wheat and maize) may be the ideal source material for
biotechnological exploitation. Optimum production of vanillin
(708 mg L
1) is observed, in medium containing DSWB (10% w/v),
sucrose (0.2% w/v) and peptone (1% w/v) at pH 7.5, agitation
220 rpm, and temperature 28  C for 5 days of fermentation.
5. Conclusion
Bioconversion of FAE of wheat bran (agroresidue) into vanillin
has been explored. The strain S. sannanensis MTCC 6637 is found to
convert FAE into vanillin through CoA-dependent noneb-oxidative
reaction sequence to occur via CoA-dependent retro-aldol mecha-
nism. Commonly available wheat bran may bring a positive impact
to the industries to keep peace with uptrend global demand of
natural vanillin through bioconversion. Since, the strain does not
convert VA into other products, thus, the product vanillin becomes
stable for a considerable period. It is believed that use of such re-
sidual material for the production will definitely minimize the
market price of biovanillin. Thus, the working strain is having
enough potentiality for biotechnological exploitation.
Author contributions
The manuscript is written through contributions of all authors.
All authors have given approval to the final version of the
manuscript.
Conflict of interest
It is hereby declared that none of the authors has any conflict of
interest.
Acknowledgements
Authors convey their indebtedness to Department of Botany
(DST-FIST and UGC-DRS sponsored) for administrative support.
Author (PC) also conveys his sincere thanks to Prof. Nirmalya
Banerjee, Department of Botany, Visva-Bharati and Prof. Pratap J.
Handique, Department of Biotechnology, Gauhati University for
providing necessary support to carry out the work.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.jclepro.2018.02.043.
References
Chow, K.T., Pope, M.K., Davies, J., 1999. Characterization of a vanillic acid non-
oxidative decarboxylation gene cluster from Streptomyces sp. D7. Microbi-
ology 145, 2393e2403.
di Gioia, D., Sciubba, L., Setti, L., Luziatelli, F., Ruzzi, M., Zanichelli, D., Fava, F., 2007.
Production of biovanillin from wheat bran. Enzym. Microb. Technol. 41 (4),
498e505.
Gallage, N.J., Moller, B.L., 2015. Vanillinebioconversion and bioengineering of the
279
most popular plant flavor and its de novo biosynthesis in the vanilla orchid.
Mol. Plant 8, 40e57.
Garcia, B.L., Ball, A.S., Rodriguez, J., Perez-Leblic, M.I., Arias, M.E., Copa-Patino, J.L.,
1998a. Production and characterization of ferulic acid esterase activity in crude
extracts by Streptomyces avermitilis CECT 3339. Appl. Microbiol. Biotechnol. 50
(2), 213e218.
Garcia, B.L., Ball, A.S., Rodriguez, J., Perez-Leblic, M.I., Arias, M.E., Copa-Patina, J.L.,
1998b. Induction of ferulic acid esterase and xylanase activities in Streptomyces
avermitilis UAH30. FEMS Microbiol. Lett. 158, 95e99.
Ghosh, S., Sachan, A., Sen, S.K., Mitra, A., 2007. Microbial transformation of ferulic
acid to vanillic acid by Streptomyces sannanensis MTCC 6637. J. Ind. Microbiol.
Biotechnol. 34 (2), 131e138.
Harshvardhan, K., Suri, M., Goswami, A., Goswami, T., 2017. Biological approach for
the production of vanillin from lignocellulosic biomass (Bambusa tulda).
J. Clean. Prod. 149, 485e490.
Kaur, B., Chakraborty, D., Kumar, B., 2013. Phenolic biotransformations during
conversion of ferulic acid to vanillin by lactic acid bacteria. BioMed Res. Int.
2013 https://doi.org/10.1155/2013/590359.
Kheder, F., Delaunay, S., Abo-Chameh, G., Paris, C., Muniglia, L., Girardin, M., 2009.
Production and biochemical characterization of a type B ferulic acid esterase
from Streptomyces ambofaciens. Can. J. Microbiol. 55 (6), 729e738.
Kim, K.H., Tsao, R., Yang, R., Cui, S.W., 2006. Phenolic acid profiles and antioxidant
activities of wheat bran extracts and the effect of hydrolysis conditions. Food
Chem. 95 (3), 466e473.
Levasseur, A., Page s, S., Fierobe, H.P., Navarro, D., Punt, P., Belaïch, J.P., Asther, M.,
Record, E., 2004. Design and production in Aspergillus niger of a chimeric pro-
tein associating a fungal feruloyl esterase and a clostridial dockerin domain.
Appl. Environ. Microbiol. 70 (12), 6984e6991.
Liu, Y., 2013. Physicochemical Properties of Starch in Bran and Endosperm and
Relationship between Bran Starch and Bran Characteristics of Selected Soft
Wheats Grown in Michigan. Michigan State University.
Ma, X.K., Daugulis, A.J., 2014. Transformation of ferulic acid to vanillin using a fed-
batch solideliquid two-phase partitioning bioreactor. Biotechnol. Prog. 30 (1),
207e214.
Modahl, I.S., Brekke, A., Valente, C., 2015. Environmental assessment of chemical
products from a Norwegian biorefinery. J. Clean. Prod. 94, 247e259.
Muheim, A., Lerch, K., 1999. Towards a high-yield bioconversion of ferulic acid to
vanillin. Appl. Microbiol. Biotechnol. 51 (4), 456e461.
Mukherjee, G., Singh, R.K., Mitra, A., Sen, S.K., 2007. Ferulic acid esterase production
by Streptomyces sp. Bioresour. Technol. 98 (1), 211e213.
Narbad, A., Gasson, M.J., 1998. Metabolism of ferulic acid via vanillin using a novel
CoA-dependent pathway in a newly-isolated strain of Pseudomonas fluo-
rescens. Microbiology 144 (5), 1397e1405.
Overhage, J., Priefert, H., Steinbüchel, A., 1999. Biochemical and genetic analyses of
ferulic acid catabolism in Pseudomonas sp. strain HR199. Appl. Environ.
Microbiol. 65 (11), 4837e4847.
Parker, M.L., Ng, A., Waldron, K.W., 2005. The phenolic acid and polysaccharide
composition of cell walls of bran layers of mature wheat (Triticum aestivum L.
cv. Avalon) grains. J. Sci. Food Agric. 85 (15), 2539e2547.
Priefert, H., Rabenhorst, J., Steinbüchel, A., 2001. Biotechnological production of
vanillin. Appl. Microbiol. Biotechnol. 56 (3e4), 296e314.
Sachan, A., Ghosh, S., Mitra, A., 2004. An eYcient isocratic separation of hydrox-
ycinnamic acids and their corresponding benzoates from microbial and plant
sources by HPLC. Biotechnol. Appl. Biochem. 40, 197e200.
Salgado, J.M., Max, B., Rodríguez, R., Pe rez, N., Domínguez, J.M., 2012. Extraction of
ferulic acid from agro-industrial wastes and evaluation of bioconversion of
ferulic acid to vanillin by Streptomyces setonii. In: 1er Congreso Iberoamericano
sobre Biorrefinerais (1-CIAB). Universidad de Guanajuato, pp. 451e457.
Sarangi, P.K., Sahoo, H.A., 2009. Standardization of cultural conditions for maximum
vanillin production through ferulic acid degradation. Sci. Publ. 1, 49e51.
Sarangi, P.K., Sahoo, H.P., 2010. Ferulic acid production from wheat bran using
Staphylococcus aureus. N. Y. Sci. J. 3, 79e81.
Sindhwani, G., Ilyas, U.K., Aeri, V., 2017. Microbial transformation of eugenol to
vanillin. J. Microbiol. Biotechnol. Res. 2 (2), 313e318.
Sutherland, J.B., Crawford, D.L., Pometto, A.L., 1983. Metabolism of cinnamic, p-
coumaric and ferulic acids by Streptomyces setonii. Can. J. Microbiol. 29,
1253e1257.
Yang, W., Tang, H., Ni, J., Wu, Q., Hua, D., Tao, F., Xu, P., 2013. Characterization of two
Streptomyces enzymes that convert ferulic acid to vanillin. PLoS One 8 (6),
e67339.
Zamzuri, N.A., Abd-Aziz, S., 2013. Biovanillin from agro wastes as an alternative food
flavour. J. Sci. Food Agric. 93 (3), 429e438.