Benzoic acid fermentation from starch and cellulose via a plant-like β-oxidation pathway in Streptomyces maritimus.

Abstract

Benzoic acid is one of the most useful aromatic compounds. Despite its versatility and simple structure, benzoic acid production using microbes has not been reported previously. Streptomyces are aerobic, Gram-positive, mycelia-forming soil bacteria, and are known to produce various kinds of antibiotics composed of many aromatic residues. S. maritimus possess a complex amino acid modification pathway and can serve as a new platform microbe to produce aromatic building-block compounds. In this study, we carried out benzoate fermentation using S. maritimus. In order to enhance benzoate productivity using cellulose as the carbon source, we constructed endo-glucanase secreting S. maritimus.
After 4 days of cultivation using glucose, cellobiose, or starch as a carbon source, the maximal level of benzoate reached 257, 337, and 460 mg/l, respectively. S. maritimus expressed β-glucosidase and high amylase-retaining activity compared to those of S. lividans and S. coelicolor. In addition, for effective benzoate production from cellulosic materials, we constructed endo-glucanase-secreting S. maritimus. This transformant efficiently degraded the phosphoric acid swollen cellulose (PASC) and then produced 125 mg/l benzoate.
Wild-type S. maritimus produce benzoate via a plant-like β-oxidation pathway and can assimilate various carbon sources for benzoate production. In order to encourage cellulose degradation and improve benzoate productivity from cellulose, we constructed endo-glucanase-secreting S. maritimus. Using this transformant, we also demonstrated the direct fermentation of benzoate from cellulose. To achieve further benzoate productivity, the L-phenylalanine availability needs to be improved in future.

Full text

Page 1

RESEARCHOpen Access
Benzoic acid fermentation from starch and
cellulose via a plant-like β-oxidation pathway in
Streptomyces maritimus
Shuhei Noda1, Eiichi Kitazono2, Tsutomu Tanaka1, Chiaki Ogino1*and Akihiko Kondo1
Abstract
Background: Benzoic acid is one of the most useful aromatic compounds. Despite its versatility and simple
structure, benzoic acid production using microbes has not been reported previously. Streptomyces are aerobic,
Gram-positive, mycelia-forming soil bacteria, and are known to produce various kinds of antibiotics composed of
many aromatic residues. S. maritimus possess a complex amino acid modification pathway and can serve as a new
platform microbe to produce aromatic building-block compounds. In this study, we carried out benzoate
fermentation using S. maritimus. In order to enhance benzoate productivity using cellulose as the carbon source, we
constructed endo-glucanase secreting S. maritimus.
Results: After 4 days of cultivation using glucose, cellobiose, or starch as a carbon source, the maximal level of benzoate
reached 257, 337, and 460 mg/l, respectively. S. maritimus expressed β-glucosidase and high amylase-retaining activity
compared to those of S. lividans and S. coelicolor. In addition, for effective benzoate production from cellulosic materials,
we constructed endo-glucanase-secreting S. maritimus. This transformant efficiently degraded the phosphoric acid
swollen cellulose (PASC) and then produced 125 mg/l benzoate.
Conclusions: Wild-type S. maritimus produce benzoate via a plant-like β-oxidation pathway and can assimilate various
carbon sources for benzoate production. In order to encourage cellulose degradation and improve benzoate
productivity from cellulose, we constructed endo-glucanase-secreting S. maritimus. Using this transformant, we also
demonstrated the direct fermentation of benzoate from cellulose. To achieve further benzoate productivity, the
L-phenylalanine availability needs to be improved in future.
Keywords: Streptomyces, Benzoic acid, Endo-glucanase, Cellulose
Background
In the past few decades, chemicals and fuel production
from renewable resources have attracted attention due to
global warming and limited supplies of fossil fuels [1-3].
The aromatic series include a large number of industri-
ally important materials, and production of aromatic
compounds using microorganisms is an active research
area, as well as production of bio-fuel and other build-
ing-block compounds [4]. Phenol production using
Pseudomonas putida and p-hydroxy cinnamic acid pro-
duction using P. putida and Escherichia coli have been
successfully demonstrated [5-7].
Benzoic acid is one of the most useful aromatic com-
pounds, and can be converted to terephthalic acid by the
Henkel reaction [8], epsilon-caprolactam by the Snia Vis-
cosa process [9], and phenol by a decarbonation reaction
[10]. Terephthalic acid is used to make polyethylene ter-
ephthalate and aramid, epsilon-caprolactam is a main
component of nylon 6, and phenol is used to make poly-
carbonate. Benzoic acid is chemically synthesized via an
oxidation reaction of toluene in the presence of potas-
sium permanganate; however, this process is energy in-
tensive. Despite its versatility and simple structure,
benzoic acid production using microbes has not been
reported previously.
* Correspondence: ochiaki@port.kobe-u.ac.jp
1Department of Chemical Science and Engineering, Graduate School of
Engineering, Kobe University, 1-1 Rokkodai, Nada, Kobe 657-8501, Japan
Full list of author information is available at the end of the article
© 2012 Noda et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
Noda et al. Microbial Cell Factories 2012, 11:49
http://www.microbialcellfactories.com/11/1/49

Page 2

Streptomyces are aerobic, Gram-positive, mycelia-form-
ing soil bacteria, and are known to produce various
kinds of antibiotics composed of many aromatic resi-
dues [11,12]. Piel et al. characterized the biosynthesis
pathway of the polyketide bacteriostatic agent enterocin
in the sediment-derived bacterium, Streptomyces mariti-
mus [13]. S. maritimus possess a complex amino acid
modification pathway and can serve as a new platform
microbe to produce aromatic building-block compounds.
Through a process involving β-oxidation of cinnamoyl-
CoA into benzoyl-CoA, S. maritimus produce benzoyl-
CoA in a plant-like manner from L-phenylalanine during
the biosynthesis of the polyketide (Figure 1) [14]. S. mar-
itimus is known to convert benzoic acid to polyketide via
benzoyl-CoA using the gene encoding benzoyl-CoA lig-
ase [15]. However, the conversion of benzoyl-CoA to
benzoic acid has not been reveled in S. maritimus.
In the present study, we identified S. maritimus as a
benzoate producer and succeeded in benzoate fermenta-
tion using a plant-like β-oxidation pathway in S. mariti-
mus. Bio-production using a biomass resource with
Streptomyces as a host has been an area of recent focus
due to its great advantages in biomass assimilation [16-18].
In order to examine the carbon-assimilating ability of S.
maritimus and apply it to benzoate fermentation, we used
various carbon sources. We succeeded in benzoate fermen-
tation using glucose, cellobiose, and starch. We compared
the carbon-assimilating ability of S. maritimus with that of
other model Streptomyces, and the versatility of S. mariti-
mus as a host strain was also demonstrated. In benzoate
fermentation from cellulose using wild-type S. maritimus,
the amount of produced benzoate was considerably low,
compared to using other carbon sources. The cellulose
degradation to glucose and cello-oligosaccharide is one of
rate-limiting steps. Here, to enhance benzoate product-
ivity from cellulose, we constructed endo-glucanase
(EG)-secreting S. maritimus. The engineered S. mari-
timus expressed EG from Thermobifida fusca YX and
efficiently degraded cellulose. Using this strain, we
successfully demonstrated the direct fermentation of
benzoate cellulose.
Results and discussion
Identification and fermentation of benzoate using S.
maritimus
After cultivation of S. maritimus/WT using TSB medium,
benzoate produced in S. maritimus was identified using a
co-chromatography method and UV spectrophotometer
[19,20]. Figure 2(A) shows a chromatogram of a standard
sample of benzoic acid solution (Lane 1) and the culture
supernatant of S. maritimus/WT containing produced
benzoic acid (Lane 2). The peak of benzoic acid in the
standard sample was found at about 8 min (Lane 1),
which was also observed in the culture supernatant of S.
maritimus/WT (Lane 2). The addition of L-phenylalan-
ine into the initial culture medium increased the peak
areas of benzoic acid (data not shown). The UV spec-
tra of the benzoic acid fraction separated from the
culture supernatant of S. maritimus/WT exhibit two major
absorption peaks in the region of 190–300 nm, similar to
the standard sample of benzoic acid (Figure 2(B)). In
addition, we carried out MS spectra analysis. MS and
tandem MS analysis show that the peak of benzoic
acid dehydrogenated and decarboxylated was observed
Figure 1 Proposed biosynthesis pathway from L-phenylalanine to benzoic acid.
Noda et al. Microbial Cell Factories 2012, 11:49
http://www.microbialcellfactories.com/11/1/49
Page 2 of 10

Page 3

around m/z 121.10 and 77.00, respectively. These results
demonstrated that S. maritimus/WT produces benzoate in
the culture supernatant.
In order to enhance benzoate productivity, we tested
benzoate fermentation using 3% glucose or xylose as the
carbon source using S. maritimus/WT. Figure 3(A)
shows time-courses of dry cell weight. S. maritimus/WT
consumed glucose or xylose within 3 days (data not
shown), and the maximum dry cell weight of S. mariti-
mus/WT using xylose was slightly higher than that using
glucose (Figure 3(A)). Figure 3(B) shows the amount of
produced benzoate. Although it had similar cell-growth
ability, the maximal amount of produced benzoate from
glucose was 260 mg/l, which was 3-fold higher than that
using xylose (Figure 3(B)). Figure 3 shows benzoate pro-
duction started after cell growth reached the maximal
level. This indicates that the cell growth before benzoate
production is one of key factor in benzoate fermentation
using S .maritimus. In this study, S. maritimus com-
pletely consumed 3% glucose after 3 days cultivation,
and the cell of S. maritimus reached the maximal level
after 3 days. The cell growth is a key factor of benzoate
fermentation using S. maritimus. These imply that sugar
uptake is one of rate-limiting step in cell growth and
benzoate fermentation using S. maritimus. In Strepto-
myces, a lot of the genes encoding sugar transporter were
identified. The sugar consumption may be improved by
introducing those genes, and the more rapid cell growth
and production of benzoate may be achieved. The
addition of L-phenylalanine improved benzoate product-
ivity from xylose (data not shown), which indicates that
L-phenylalanine availability in S. maritimus using xylose
as the carbon source is not enough for benzoate
production.
In enterocin biosynthesis pathway of S. maritimus,
benzoyl-CoA, which is a precursor of benzoate, is con-
verted to enterocin by multiple enzymes. Here, to im-
prove benzoate productivity, we carried out inactivation
of encABCL closely concerning enterocin production.
We successfully disrupted encABCL in S. maritimus,
and the engineered strain was named S. maritimus/
ΔencABCL. Using S. maritimus/ΔencABCL, benzoic acid
fermentation was tested. Inactivation of encABCL had no
negative effect on the cell growth of S. maritimus, how-
ever, benzoic acid productivity drastically decreased,
compared to wild-type S. maritimus (data not shown).
Figure 2 (A) HPLC traces of benzoic acid analysis. Lane 1: Standard sample of benzoic acid in acetonitrile: phosphate buffer (50 mM, pH 2.5)
(30:70) solution. Lane 2: The culture supernatant of S. maritimus. (B) UV spectra of benzoic acid. 2.5 mg/l Standard sample of benzoic acid in
acetonitrile: phosphate buffer (50 mM, pH 2.5) (30:70) solution (dotted line), 4 mg/l benzoic acid fraction separated from the culture supernatant
of S. maritimus by HPLC (solid line).
Noda et al. Microbial Cell Factories 2012, 11:49
http://www.microbialcellfactories.com/11/1/49
Page 3 of 10

Page 4

Benzoate production from cellobiose and starch using
wild-type S. maritimus
Although some Streptomyces are known to assimilate
various carbon sources, such as oligosaccharides and
starch [18,21], the carbon-assimilating ability and growth
profile of S. maritimus have not been investigated.
Figure 4(A) shows the time-course of dry cell weight
using 3% cellobiose, and the maximum dry cell weight of
S. maritimus/WT using cellobiose is higher than that
using glucose. As shown in Figure 4(B), the maximal
level of produced benzoate using 3% cellobiose was
323 mg/l after 6 days of cultivation, and the estimated
yield for benzoate was 1.76 Cmol% (Table 1). In the
present study, S. maritimus/WT almost completely con-
sumed 3% cellobiose within 2 days (data not shown). We
compared the β-glucosidase (BGL) activity of S. mariti-
mus to that of S. lividans and S. coelicolor, which are
model strains in Streptomyces. Figure 4(C) shows the
time-courses of BGL activity in the intracellular fractions
of S. maritimus, S. lividans, and S. coelicolor. The intra-
cellular BGL activity detected in S. maritimus reached
46.1 U/g-drycell, which was 5-fold higher than that in S.
lividans and S. coelicolor (Figure 4(C)) after 4 days of
cultivation. Although the BGL activity expressed in S.
maritimus was vastly higher than those in S. lividans or
S. coelicolor, the cellobiose consumption rates among
these strains were almost the same. This result indicates
that cellobiose uptake is the rate-limiting step, and that
overexpression of the sugar transporter may improve the
cellobiose consumption and benzoate productivity in S.
maritimus. Although the benzoate productivity using
xylo-oligosaccharide was lower, similar to that using xy-
lose, the cell growth was similar to that using other car-
bon sources (data not shown).
We carried out benzoate fermentation using starch.
Figure 4(A) shows the time-course of dry cell weight
using 3% starch. The cell growth of S. maritimus using
starch as the carbon source was similar to that using glu-
cose (Figure 3(A), 4(A)). As shown in Figure 4(B), the
maximal level of produced benzoate using 3% starch was
460 mg/l after 6 days of cultivation. The estimated yield
of benzoate using starch as the carbon source was 2.64
Cmol%, which was 1.77 times higher than that using glu-
cose (Table 1). Surprisingly, benzoate production using
starch as the carbon source was more efficient than
when using glucose. This indicates that S. maritimus can
directly assimilate starch for benzoate production more
effectively than glucose. Although the cell growth of S.
maritimus using cellobiose was higher than that of using
starch, the maximal level of produced benzoate using
starch was about 1.5 times greater, compared to using
cellobiose. This may indicate that starch as the carbon
source encourages the carbon flux to flows smoothly to
benzoate synthesis pathway in S. maritimus. Here, to
demonstrate that additional carbon sources were used
for benzoate formation and the cell growth of S. mariti-
mus, we carried out benzoate fermentation in S. mariti-
mus using TSB medium with additional 5% tryptone
(without additional each carbon source). The maximal
level of produced benzoate reached about only 100 mg/L,
and the cell growth of S. maritimus was drastically low,
compared to fermentation with additional carbon source.
Hence, we conclude that additional glucose, cellobiose,
and starch were used for the cell growth, and encouraged
benzoate formation.
After the fermentation, the amount of produced cin-
namic acid, one of the easily detected intermediates, was
170 mg/l (from 3% starch) and 73 mg/l (from 3% glu-
cose), respectively. These results show that the ratio of
cinnamic acid to benzoic acid (mol/mol) was increased
along with the increased the amount of produced benzo-
ate, suggesting that optimization of the carbon flux may
improve benzoate productivity. Enhancing L-phenylalan-
ine availability in the cell may also lead to further benzo-
ate productivity, because the amount of produced
benzoate in fermentation using TSB medium with 1.5%
glucose, 1.5% tryptone, and additional 100 mM of
0
5
10
15
0
100
200
300
400
500
0123456
Dry cell weight (g/l)
Benzoate concentration (mg/l)
Cultivation time (days)
Figure 3 Time-courses of dry cell weight using glucose as the
sole carbon source: S. maritimus/WT in modified TSB medium
with 5% tryptone and 3% glucose (open circles); S. maritimus/
WT in modified TSB medium with 5% tryptone and 3% xylose
(open triangles). Time-courses of produced benzoate in culture: S.
maritimus/WT in modified TSB medium with 5% tryptone and 3%
glucose (closed circles); S. maritimus/WT in modified TSB medium
with 5% tryptone and 3% xylose (closed triangles). The dry cell
weight and benzoate concentration were determined in the same
culture. Each data point shows the average of three independent
experiments, and error bars represent standard deviation.
Noda et al. Microbial Cell Factories 2012, 11:49
http://www.microbialcellfactories.com/11/1/49
Page 4 of 10

Page 5

Figure 4 (A) Time-courses of dry cell weight using glucose as the sole carbon source: S. maritimus/WT in modified TSB medium with
5% tryptone and 3% starch (open circles); S. maritimus/WT in modified TSB medium with 5% tryptone and 3% cellobiose (open
diamonds). (B) Time-courses of produced benzoate in culture: S. maritimus/WT in modified TSB medium with 5% tryptone and 3% starch (closed
circles); S. maritimus/WT in modified TSB medium with 5% tryptone and 3% cellobiose (closed diamonds). (C) Time-courses of β-glucosidase
activity in the intracellular fractions of S. maritimus/WT, S. lividans, and S. coelicolor using modified TSB medium with 5% tryptone and 3%
cellobiose: S. maritimus/WT (closed circles), S. lividans (closed squares), S. coelicolor (closed triangles). (D) Time-courses of α-amylase activity in the
culture supernatant of S. maritimus/WT, S. lividans, and S. coelicolor modified TSB medium with 5% tryptone and 3% starch: S. maritimus/WT
(closed circles), S. lividans (closed squares), S. coelicolor (closed triangles). The dry cell weight and benzoate concentration were determined in the
same culture. Each data point shows the average of three independent experiments, and error bars represent standard deviation.
Table 1 Various parameters in benzoate fermentationa
StrainCarbon sourceMaximum dry
cell weight (g/l)
Benzoate
produced (mg/l)
Yieldb
(Cmol%)
S. maritimus/WT Glucose (3%)10.8±0.29257±62.4 1.47
S. maritimus/WT Xylose (3%)12.9±1.36 90.0±28.90.52
S. maritimus/WTCellobiose (3%)13.6±0.52 337±65.51.83
S. maritimus/WTCorn starch (3%)11.4±0.88460±36.82.64
S. maritimus/WTPASC (1%)-23.3±8.170.40
S. maritimus/ps-tfu0901
PASC (1%)-125±8.962.15
S. maritimus/ps-tfu1074
PASC (1%)-103±0.94 1.77
aValues represent means±standard deviation for three independent experiments.
bmol of carbon involved in produced benzoate per mol of carbon involved in consumed carbon source.
Noda et al. Microbial Cell Factories 2012, 11:49
http://www.microbialcellfactories.com/11/1/49
Page 5 of 10