Production of glyceric acid by gluconobacter sp. NBRC3259 using raw glycerol

Article abstract of DOI:10.1271/bbb.90163

Gluconobacter sp. NBRC3259 converted glycerol to glyceric acid (GA). The enantiomeric composition of the GA produced was a mixture of DL-forms with a 77% enantiomeric excess of d-GA. After culture conditions, such as initial glycerol concentration, types

Full text of DOI:10.1271/bbb.90163

Biosci. Biotechnol. Biochem., 73 (8), 1799–1805, 2009
Production of Glyceric Acid by Gluconobacter sp. NBRC3259
Using Raw Glycerol
y
Hiroshi HABE,1; Yuko SHIMADA,2 Tokuma FUKUOKA,1 Dai KITAMOTO,1 Masayuki ITAGAKI,2
1
Kunihiro WATANABE,2 Hiroshi YANAGISHITA,1 and Keiji SAKAKI
1Research Institute for Innovation in Sustainable Chemistry, National Institute of Advanced Industrial Science
and Technology (AIST), Central 5-2, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan
2Department of Chemical Industry, Science University of Tokyo, 2641 Yamazaki, Noda, Chiba 278-8510, Japan
Received March 2, 2009; Accepted May 14, 2009; Online Publication, August 7, 2009
[doi:10.1271/bbb.90163]
Gluconobacter sp. NBRC3259 converted glycerol to
glyceric acid (GA). The enantiomeric composition of the
GA produced was a mixture of DL-forms with a 77%
enantiomeric excess of D-GA. After culture conditions,
such as initial glycerol concentration, types and amounts
of nitrogen sources, and initial pH, were optimized,
Gluconobacter sp. NBRC3259 produced 54.7 g/l of GA
as well as 33.7 g/l of dihydroxyacetone (DHA) from
167 g/l of glycerol during 4 d of incubation in a jar
fermentor with pH control. GA production from raw
glycerol samples, the main by-product of the trans-
esterification process in the biodiesel production and
oleochemical industries, was also evaluated after proper
pretreatment of the samples. Using a raw glycerol
sample with activated charcoal pretreatment, 45.9 g/l of
GA and 28.2 g/l of DHA were produced from 174 g/l of
glycerol.
for several chemical compounds. In the chemical
synthesis of GA, a racemic mixture of DL-GA can be
obtained by metal-catalytic oxidation of the primary
hydroxyl groups of glycerol.8–10) On the other hand, little
is known about GA production by microorganisms
(Fig. 1), although GA is known as a by-product during
the production of dihydroxyacetone (DHA), the main
active ingredient in all sunless tanning skincare prepa-
rations, by Gluconobacter oxydans.11) According to a
Japanese patent application in 1987 (JP0751069, Daicel
Chemical Industries), four Gluconobacter strains pro-
duced 38.3 g/l of D-GA (on average) from 100 g/l of
glycerol in a fermentor, but detailed investigation has
not been done. Also, in our recent report, Acetobacter
tropicalis NBRC16470 produced 22.7 g/l of D-GA from
200 g/l of glycerol in a fermentor.12)
Gluconobacter strains oxidize a broad range of
substrates, including alcohols, sugars, sugar acids, and
sugar alcohols, and the corresponding oxidative products
are accumulated in the culture. These bacteria have
numerous membrane-bound dehydrogenases and oxi-
doreductases, and thus, energy-consuming transport of
substrates into the cell and of products out of the cell is
not required.13) Hence, we searched for an efficient
producer of GA among various Gluconobacter strains
(including type strains), and investigated the production
of GA and the by-production of DHA by the strains
selected. Also, we tried to use raw glycerol as a
feedstock for GA production.
Key words: glycerol use; glyceric acid; dihydroxyace-
tone; acetic acid bacteria; Gluconobacter sp.
Due to recent concern about renewable resources
other than petroleum, much attention has been paid to
biorefineries, which enable the production of biofuels as
well as building-block chemicals from biomass.1,2) One
of the renewable resources from biomass, glycerol
(1,2,3-propanetriol), is thought to be a promising and
abundant carbon source for industrial microbiology.3)
Glycerol can be obtained up to 14% weight as a by-
product of biodiesel fuel (BDF) production through
transesterification of vegetable oils and animal fats. In
several European countries, the production of glycerol
has increased significantly due to BDF uptake. As a
result, the price has fallen, and chemical synthesis of
glycerol from petrochemical feedstock has shut down
owing to a surplus.4) This indicates that glycerol is an
attractive feedstock for the production of various
chemicals.
Materials and Methods
Bacterial strains. Seven Gluconobacter strains, G. albidus
NBRC3250T (type strain), G. frateurii NBRC3262, G. condonii
NBRC3266T, G. cerinus NBRC3267T, G. thalilandicus NBRC100600T,
G. oxydans NBRC14819T, and Gluconobacter sp. NBRC3259 were
obtained from the National Institute of Technology and Evaluation
(NITE), Japan. Stock cultures were cultivated at 30 ꢀC on an agar
medium containing 5 g/l of polypepton (Nihon Pharmaceutical,
Tokyo), 5 g/l of yeast extract (Difco Laboratories, Detroit, MI), 5 g/l
Research has been carried out to find new applications
of glycerol in both chemical processes5–7) and biopro-
cesses.3) In this study, we focused on the production of a
promising glycerol derivative, glyceric acid (GA), via a
bioprocess that has the potential to be a building block
.
of glucose, and 1 g/l of MgSO4 7H2O (pH 6.5). Bacterial strains were
precultivated in 5 ml of the above glucose medium at 30 ꢀC for 48 h
in test tubes (200 mm ꢁ ꢀ18 mm), and the seed cultures (1.5 ml)
were transferred to 300-ml Erlenmeyer flasks containing 30 ml of
the medium (pH 6.5) consisting of 10% (v/v) glycerol, 5 g/l of
y
To whom correspondence should be addressed. Fax: +81-29-861-6247; E-mail: hiroshi.habe@aist.go.jp
Abbreviations: BDF, biodiesel fuel; DHA, dihydroxyacetone; GA, glyceric acid; HPLC, high-performance liquid chromatography; NMR, nuclear
magnetic resonance; RT, retention time

1800
H. HABE et al.
(iii) Effect of initial polypepton concentration. Seed cultures were
transferred to 30 ml of the medium (pH 6.5) consisting of 10% (v/v)
CH2OH
CH2OH
CHO
COOH
Glycerol
dehydrogenase
?
?
HC=O
CHOH
CH2OH
Glycerol
CHOH
CH2OH
CHOH
CH2OH
glycerol, 5, 10, 15, 20, 25, or 30 g/l of polypepton, 1 g/l of yeast
.
CH2OH
extract, and 1 g/l of MgSO4 7H2O.
(iv) Effect of pH. Seed cultures were transferred to 30 ml of medium
(pH 3, 4, 5, 6, 7, 8, or 9) consisting of 10% (v/v) glycerol, 25 g/l of
Dihydroxyacetone
Glyceraldehyde
Glyceric acid
Fig. 1. Proposed Pathway for the Conversion of Glycerol to Glyceric
Acid by Acetic Acid Bacteria.
Bioconversion of glycerol to dihydroxyacetone is also repre-
sented. The ‘‘?’’ indicates that no enzymes have been reported yet.
Parentheses represent a compound that has not been found yet as an
intermediate during GA production.
.
polypepton, 1 g/l of yeast extract, and 1 g/l of MgSO4 7H2O.
(v) Effect of methanol. Seed cultures were transferred to 30 ml of
medium (pH 6.5) consisting of 10% (v/v) glycerol, 0.2, 0.4, 0.6, 0.8, or
1% (v/v) methanol, 25 g/l of polypepton, 1 g/l of yeast extract, and
.
1 g/l of MgSO4 7H2O.
(vi) Effect of NaCl or Na2SO4. Seed cultures were transferred to
30 ml of medium (pH 6.5) consisting of 10% (v/v) glycerol, 1, 3, 5, 10,
or 15 g/l of NaCl or Na2SO4, 25 g/l of polypepton, 1 g/l of yeast
polypepton (Nihon Pharmaceutical), 5 g/l of yeast extract, and 1 g/l
ꢀ
.
of MgSO4 7H2O, and were incubated at 30 C on a rotary shaker
.
extract, and 1 g/l of MgSO4 7H2O.
BR-23FP (200 rpm; TAITEC, Saitama) for 4 d. After removal of the
cells by centrifugation, the supernatant was filtrated with a 0.45-mm
cellulose filter. To quantify both glycerol and GA, a 20-ml sample of the
supernatant was analyzed by high-performance liquid chromatography
(HPLC).
Jar fermentor experiments with pure glycerol. GA production by
Gluconobacter sp. NBRC3259 was carried out in a 1-liter jar fermentor
(Model MDL; B.E. Marubishi). Jar fermentor experiments were
performed as follows: Gluconobacter sp. NBRC3259 was cultivated in
five test tubes each containing 5 ml of glucose medium (total 25 ml of
culture) for 2 d (30 ꢀC, 200 rpm). All seed cultures were transferred to a
1-liter jar fermentor containing 500 ml of medium (pH 6.5) consisting
of 150–170 g/l of glycerol, 25 g/l of polypepton, 1 g/l of yeast extract,
Structural analysis of GA produced by Gluconobacter sp.
NBRC3259. Seed culture (1.5 ml) of Gluconobacter sp. NBRC3259
precultivated with the above glucose medium was transferred to a flask
containing 30 ml of a medium (pH 6.5) consisting of 10% (v/v)
glycerol, 5 g/l of polypepton, 5 g/l ofꢀ yeast extract, and 1 g/l of
.
0.9 g/l of KH2PO4, 0.1 g/l of K2HPO4, and 1 g/l of MgSO4 7H2O,
and incubated for 4 d. During the jar fermentor experiments, the
aeration rate and agitation speed were set to 1.0 vvm and 500 or
700 rpm. Temperature was maintained at 30 ꢂ 1 ꢀC. When necessary,
pH was controlled with 5 M NaOH so as not to be under pH 5.
.
MgSO4 7H2O, and was incubated at 30 C on a rotary shaker for 4 d.
After removal of the cells by centrifugation, 25 ml of the supernatant
was filtrated and adjusted to pH 5 with 10 M NaOH. GA was separated
from glycerol by ion exchange chromatography with DOWEX 1-X8
(Dow Chemical, Midland, MI) in a 20 ꢁ 300 mm column. The
supernatant (25 ml) was applied to the column, and chromatography
was carried out with 0.5 M HCl. The eluate was fractionated and
checked for GA amounts by HPLC. An appropriate amount of
Pretreatment of raw glycerol samples. Raw glycerol sample A,
obtained from the transesterification of tryglyceride (kindly provided
by Sun Care Fuels Corporation),14) contained the following compo-
nents (analyzed by our group): glycerol, 66.4% (w/v); methanol,
30.9% (w/v); and sodium salt, 0.54% (w/v); pH 12. A second type of
raw glycerol sample B (kindly provided by LION Cooperation) of a
higher grade contained the following components (analyzed by our
group): glycerol, 88.4% (w/v); and sodium salt, 0.19% (w/v); pH 8.
The methanol was not detected in raw glycerol sample B. Methanol in
raw glycerol sample A was removed by evaporation in vacuo. The pH
of glycerol sample A was adjusted to pH 7 to 8 with HCl or H2SO4.
When necessary, raw glycerol samples were treated with granular,
activated charcoal (Wako Pure Chemicals) after the samples were
diluted with the same volume of water. Approximately 20% (w/v)
activated charcoal was added directly to the raw glycerol samples, and
the samples were incubated statically at room temperature for more
than 8 h. The charcoal was removed from the treated samples by
filtration, and the resulting raw glycerol samples were used in further
experiments.
.
CaCl2 2H2O (half moles of GA) was added to the fractions containing
GA (concentration, about 30 g/l; pH adjusted to pH 5), and GA
calcium salt was precipitated by adding ethanol to the crude GA
solution. The purified GA calcium salt was dissolved in deuterium
oxide (D2O), and 1H and 13C NMR analysis was performed using a
Varian INOVA 400 (400 MHz). The enantiomeric composition of the
GA calcium salt was analyzed by HPLC consisting of an LC-20AD
HPLC pump (flow rate, 1.0 ml/min) and an SPD-20AV UV/VIS
detector (detection, 254 nm; Shimadzu, Kyoto) equipped with two
tandemly-linked CHIRALPAKꢀ MA(þ) Columns (Daicel Chemical
Industries, Osaka). A mobile phase of 0.45 mM CuSO4 solution was
chosen as the eluent. During analysis, the column temperature was kept
at 21 ꢀC. DL-GA calcium salt dihydrate (Wako Pure Chemicals,
Osaka), D-GA calcium salt dihydrate (Sigma-Aldrich, St. Louis, MO),
and L-GA calcium salt dihydrate (Sigma-Aldrich) were used as
standard samples.
Jar fermentor experiments with raw glycerol samples. Jar fermentor
experiments with raw glycerol were carried out similarly to the jar
fermentor experiments with pure glycerol described above. Seed
cultures were transferred to a 1-liter jar fermentor containing 500 ml of
medium (pH 6.5) consisting of raw glycerol samples (150–170 g/l of
glycerol at final concentration), 25 g/l of polypepton, 1 g/l of yeast
extract, 0.9 g/l of KH2PO4, 0.1 g/l of K2HPO4, and 1 g/l of
Growth characteristics of Gluconobacter sp. NBRC3259 on
glycerol. The seed culture of Gluconobacter sp. NBRC3259 was
transferred to 30 ml of a medium (pH 6.5) consisting of 1% (v/v)
glycerol, 5 g/l of polypepton, 5 g/l of yeast extract, and 1 g/l of
.
MgSO4 7H2O (nutrient-rich condition) or 1% (v/v) glycerol, 1 g/l of
.
yeast extract, 2 g/l of (NH4)2SO4, and 1 g/l of MgSO4 7H2O
(nutrient-poor condition), and was incubated at 30 ꢀC on a rotary
shaker (200 rpm) for 4 d. Samples (0.5 ml) were taken from the cultures
at regular intervals. After removal of the cells by centrifugation, the
respective supernatants were analyzed by HPLC.
.
MgSO4 7H2O, and incubated for 4 d. During the jar fermentor
experiments, the pH was controlled with 5 M NaOH so as not to be
under pH 5.
Quantification of glycerol, DHA, and GA. The concentrations of
glycerol, DHA, and GA in culture broth were analyzed by HPLC, with
an LC-20AD HPLC pump (flow rate, 1.0 ml/min) and an RID-10A
detector (Shimadzu) equipped with a Shodexꢀ SC1011 Column
(Showa Denko, Tokyo) for glycerol and DHA and a Shodexꢀ
SH1011 Column (Showa Denko) for GA. A mobile phase of pure
water and 5 mM H2SO4 solution was chosen for the respective columns
as the eluent. During analysis, the column temperature was kept at
80 ꢀC and 60 ꢀC for the respective columns. DL-GA calcium salt
dihydrate (Wako Pure Chemicals) or DL-GA (40% in water; Tokyo
Chemical Industry, Tokyo) and DHA (MP Biomedicals, Santa Ana,
CA) was used to determine the standard curve for GA and DHA
quantification, respectively.
Optimization of culture conditions for GA production. All cultures
of Gluconobacter sp. NBRC3259 were incubated at 30 ꢀC on a rotary
shaker (200 rpm) for 4 d. After removal of the cells by centrifugation,
the respective supernatants were analyzed by HPLC.
(i) Effect of initial glycerol concentration. Seed cultures were
transferred to 30 ml of a medium (pH 6.5) consisting of 5, 10, or 15%
(v/v) glycerol, 5 g/l of polypepton, 5 g/l of yeast extract, and 1 g/l of
.
MgSO4 7H2O.
(ii) Effect of nitrogen sources. Seed cultures were transferred to
30 ml of medium (pH 6.5) consisting of 10% (v/v) glycerol, 1 g/l of
.
yeast extract, 1 g/l of MgSO4 7H2O, and 9 g/l of nitrogen compounds
(polypepton, peptone, yeast extract (NH4)2SO4, and NaNO3).

Raw Glycerol Conversion to Glyceric Acid by Gluconobacter sp.
Table 1. Production of Glyceric Acid and Dihydroxyacetone from
1801
A
Glycerol by Gluconobacter Strains
mV
30
25
20
15
10
5
Glyceric
acid
(g/l)
Dihydroxyacetone
(g/l)
Strain
Gluconobacter albidus
NBRC3250T
9.1
10.4
4.1
48.2
18.9
74.4
44.6
37.9
108.2
41.0
0
Gluconobacter frateurii
NBRC3262
-5
0.0
2.5
5.0
7.5
10.0
12.5
15.0
17.5
min
Gluconobacter condonii
NBRC3266T
B
Gluconobacter cerinus
NBRC3267T
18.0
2.3
mV
25
Gluconobacter thalilandicus
NBRC100600T
20
15
10
5
Gluconobacter oxydans
NBRC14819T
1.4
Gluconobacter sp.
NBRC3259
18.2
0
0.0
2.5
5.0
7.5
10.0
12.5
15.0
17.5
min
TType strain
Retention time (min)
Fig. 2. Enantiomeric Separation of DL-GA.
Chemicals. All chemicals used were the purest commercially
available (i.e., 98–100%; Sigma-Aldrich, Kanto Chemical, Wako
Pure Chemicals, Nacalai Tesque, Tokyo Chemical Industry, MP
Biomedicals).
The first peak (13.2 min) and the second (15.8 min) represent L-
and D-GA, respectively. Two tandemly-linked CHIRALPAKꢀ
MA(þ) Columns (Daicel Chemical Industries, Osaka) were used
with a mobile phase 0.45 mM CuSO4 solution as the eluent. A, GA
sample produced by Gluconobacter sp. NBRC3259 (approximately
0.5 g/l); B, Authentic sample, DL-GA calcium salt (0.5 g/l).
Results
Structural characterization of GA produced by Glu-
conobacter sp. NBRC3259
Gluconobacter sp. NBRC3259 grew on glycerol and
reached stationary phase within 2 d. Approximately 91%
of glycerol was degraded in 2 d of cultivation; however,
not even a trace amount of GA was detected during this
period (data not shown). In contrast, the strain reached
stationary phase within 24 h in nutrient-rich conditions,
and approximately 0.6 g/l GA was detected within this
period. In both cases, no DHA production was observed
(data not shown). Based on these results, Gluconobacter
sp. NBRC3259 can use glycerol as a carbon source
under both culture conditions, but produces GA when
cultivated under nutrient-rich conditions. We then
investigated the effect of the initial glycerol concen-
tration on GA production under nutrient-rich conditions,
and found that the GA concentration in the culture
reached a maximum (20.3 g/l) with the addition of 10%
(v/v) glycerol. Therefore, both nutrient-rich conditions
and 10% (v/v) glycerol were used as glycerol media in
the following experiments.
The effects of types and amounts of nitrogen sources
on GA production were investigated. Figure 3A illus-
trates GA production according to nitrogen source. The
addition of polypepton exhibited the highest productiv-
ity among several nitrogen sources. In addition, the
effect of polypepton concentration on GA production
was examined. Maximum yield of 27.9 g/l was attained
with the addition of 25 g/l polypepton (Fig. 3B). Then
GA production was investigated over a range of initial
pH values. Approximately 25 g/l of GA was produced at
the same level from pH 4 to pH 7 (maximum at pH 5),
but decreased at pH 3 and pH 8 (Fig. 3C). In addition,
the effect of NaCl or Na2SO4 on GA production was
investigated. It was found that an increase in the amount
of Clꢃ affected GA productivity as well as the strain’s
growth (Fig. 3D).
Seven Gluconobacter strains including type strains
were examined for their ability to produce GA and DHA
from 10% (v/v) glycerol (Table 1). Preliminary exami-
nation to determine culture periods revealed that GA
production reached a maximum within 4 d (data not
shown). Among the strains tested, Gluconobacter sp.
NBRC3259 was the most efficient GA producer at 10%
(v/v) glycerol concentration, and hence was used in
further studies as the representative strain.
The culture supernatant of Gluconobacter sp.
NBRC3259 was collected and roughly purified by ion
exchange chromatography with DOWEX 1-X8. Using
the resulting sample, GA calcium salt was precipitated
1
and further purified. By H and 13C NMR analyses, we
confirmed that the 1H and 13C spectra of the purified GA
calcium salt were identical to those of three authentic
samples, DL-GA calcium salt dihydrate (Wako Pure
Chemical), D-GA calcium salt dihydrate (Sigma-
Aldrich), and L-GA calcium salt dihydrate (Sigma-
Aldrich).12)
The enantiomeric composition of the GA produced by
Gluconobacter sp. NBRC3259 was investigated. As
shown in Fig. 2, two peaks in retention time (RT)
occurred at 13.2 and 15.8 min in the authentic DL-GA
(L- and D-GA respectively; Fig. 2B), and the purified
GA calcium salt also had two peaks corresponding
to L- and D-GA (Fig. 2A). This result indicates that
Gluconobacter sp. NBRC3259 produced both D- and
L-GA, and yielded D-GA with 77% enantiomeric excess.
Optimization of culture conditions for GA production
We first tested the growth of Gluconobacter sp.
NBRC3259 on 1% (v/v) glycerol under both nutrient-
poor and -rich conditions. In nutrient-poor conditions,

1802
H. HABE et al.
A
B
30
25
20
15
10
5
30
25
20
15
10
5
0
0
5
10
15
20
25
30
Polypepton concentration (g/l)
Nitrogen sources
D
C
30
25
20
15
10
5
4
3
2
30
25
20
15
10
5
1
0
0
0
4
3
5
6
7
8
9
0
2.5
5
10
15
pH
NaCl / Na2SO4 concentration (g/l)
Fig. 3. Effects of Nitrogen Sources (A), Polypepton Concentration (B), Initial pH (C), and NaCl or Na2SO4 (D) on GA Production by
Gluconobacter sp. NBRC3259.
A, B, and C, Error bars represent the standard deviation calculated from three independent experiments. D, The data represent the average of
two independent experiments. White and gray bars represent GA production with the addition of NaCl and Na2SO4 respectively and and
represent the OD600 with NaCl and Na2SO4 respectively.
B
180
A
180
60
60
4
3
4
3
2
1
0
150
120
90
50
40
30
20
10
150
50
40
120
90
2
1
0
30
20
10
0
60
60
30
30
0
0
0
1
2
4
3
3
1
4
2
Cultivation time (d)
Cultivation time (d)
C
60
50
40
4
3
180
150
120
90
30
20
10
2
1
0
60
30
0
0
1
2
3
4
Cultivation time (d)
Fig. 4. Typical Time Course of GA Production from Pure Glycerol by Gluconobacter sp. NBRC3259.
A, pH was not controlled, 500 rpm. B, pH was controlled with 5 M NaOH so as not to be under pH 5, 500 rpm. C, pH was controlled so as not
to be under pH 5, 700 rpm. Symbols: solid circles, glycerol concentration; open circles, GA concentration; solid squares, optical density; open
squares, DHA concentration.
Time course of GA production
trolled, and 500 rpm of agitation speed showed that
approximately 24.6 g/l of GA was produced during 4 d
of incubation (Fig. 4A). The pH of the medium (initially
6.3) decreased to 3.2. By contrast, Fig. 4B illustrates the
typical time course of GA production under pH control
(pH 5) and 500 rpm of agitation speed. GA at 54.9 g/l
Under the optimized conditions (approximately
150 g/l of initial glycerol and the addition of 25 g/l of
polypepton), Gluconobacter sp. NBRC3259 was culti-
vated using a jar fermentor. The time course of GA
production with 150–170 g/l of glycerol, pH uncon-

Raw Glycerol Conversion to Glyceric Acid by Gluconobacter sp.
1803
A
B
25
20
15
10
5
25
20
15
10
5
0
0
Fig. 5. Effects of Activated Charcoal Pretreatment of Raw Glycerol Samples on GA Production.
A, Raw glycerol sample A neutralized with H2SO4. B, Raw glycerol sample B. Error bars represent the standard deviation calculated from
three independent experiments.
was obtained after 4 d of incubation, and productivity
was enhanced approximately 2-fold with pH control. At
the same time, the amount of DHA in the culture was
analyzed by HPLC, resulting in 33.7 g/l of DHA
production after 4 d of incubation (Fig. 4B). Considering
that glycerol consumption was 92.3 g/l, GA and DHA
yields of 0.52 and 0.37 mol/mol-glycerol respectively
were obtained. Further cultivation of Gluconobacter sp.
NBRC3259 did not increase GA production (data not
shown). In the case of an agitation speed of 700 rpm
with pH control (pH 5), the productivity of GA by the
strain decreased to 40.2 g/l after 4 d in culture (Fig. 4C).
Instead, the amount of DHA increased to 49.6 g/l. These
results indicate that agitation speed is also an important
factor for efficient GA production.
Time course of GA production with raw glycerol
Using charcoal-pretreated raw glycerol samples, jar
fermentor experiments were performed. Figure 6 illus-
trates the typical time course of GA production using
raw glycerol sample A neutralized with HCl (Fig. 6A)
or H2SO4 (Fig. 6B). The agitation speed was set to
500 rpm and the pH was kept at pH 5. When neutralized
with HCl (final Clꢃ concentration, 5.2 g/l), GA was
produced, but the amount was only 15.1 g/l after 4 d of
incubation. In contrast, when neutralized with H2SO4
2ꢃ
(final SO4
concentration, 4.5 g/l), the maximal
amount of GA (45.9 g/l) was reached after 4 d of
incubation, and productivity was 84% of the results with
pure glycerol. The amount of DHA was 28.2 g/l after
4 d of incubation (Fig. 6B). Since glycerol consump-
tion was 87.1 g/l, GA and DHA yields of 0.46 and
0.33 mol/mol-glycerol respectively were obtained.
Figure 7A represents the typical time course of GA
production with raw glycerol sample B, pH control
(pH 5), and 500 rpm. The maximum amount of GA
(39.6 g/l) was reached after 4 d of incubation, and
productivity was 72% of the results with pure glycerol.
The amount of DHA was 22.7 g/l after 4 d of incubation
(Fig. 7A). As glycerol consumption was 63.3 g/l, GA
and DHA yields of 0.54 and 0.36 mol/mol-glycerol
respectively were obtained. At 700 rpm (Fig. 7B), GA
production decreased to 28.0 g/l, but the amount of
DHA increased to 31.1 g/l after 4 d of incubation.
Effect of raw glycerol pretreatment on GA production
With the use of raw glycerol sample A as a feedstock,
the effects of methanol on GA production were inves-
tigated. As compared to GA productivity in no methanol
culture (24.4 g/l), the addition of 0.4% (v/v) methanol
drastically decreased to 1.9 g/l of production, although
the optical density (OD600) decreased slightly to 2.1
from 2.9. When 1% (v/v) methanol was added, GA
production and OD600 were 0.2 g/l and 0.5 respectively.
According to the above results, methanol must be
removed from raw glycerol samples. Therefore, meth-
anol was removed from raw glycerol sample A to less
than 0.1 g/l by evaporation.
Figure 5 illustrates the effect of activated charcoal
pretreatment of raw glycerol samples on GA production.
In the case of raw glycerol sample A (Fig. 5A), 35% of
GA productivity was obtained without activated char-
coal treatment as compared to the control samples using
pure glycerol, while 90% of GA productivity was
obtained with activated charcoal treatment. The OD600
of the cultures using intact and pretreated raw glycerol
samples were 1.5, and 2.1 respectively. In contrast, in
the case of raw glycerol sample B (Fig. 5B), no GA was
produced without activated charcoal treatment. With
activated charcoal treatment, 86% of GA productivity
was obtained as compared to the control. At this time,
the OD600 of the cultures using intact and pretreated raw
glycerol samples were 0 and 2.4, respectively. This
result indicates that pretreatment by activated charcoal is
very effective for GA production with raw glycerol
samples as the feedstock.
Discussion
We found that Gluconobacter sp. NBRC3259 pro-
duced 45.9 g/l of GA and 28.2 g/l of DHA from raw
glycerol sample A (containing 174 g/l of glycerol in
culture) during 4 d of incubation. This is the first report
describing the production of GA with raw glycerol as a
feedstock. According to this and previous studies,14–16)
the impurities of raw glycerol were mainly composed
of methanol, sodium (or/and potassium) salts, heavy
metals-lignin, other organic materials, and so on.
However, their compositions depended on respective
transesterification processes for the BDF and oleochem-
ical industries as well as the differences in the raw
materials themselves (various vegetable oils, animal
fats, and so on). In the use of both raw glycerol samples,
pretreatment with activated charcoal was effective for
GA production, but in raw glycerol sample A, 35% GA
productivity was observed without activated charcoal

1804
H. HABE et al.
B
180
A
180
60
3
60
50
40
3
2
150
120
90
50
40
150
2
120
90
30
20
30
20
10
0
60
1
0
60
1
0
30
10
0
30
0
0
4
1
3
1
2
2
3
4
Cultivation time (d)
Cultivation time (d)
Fig. 6. Time Course of GA Production by Gluconobacter sp. NBRC3259 Using Raw Glycerol Sample A Neutralized with HCl (A) and H2SO4 (B).
Symbols: solid circles, glycerol concentration; open circles, D-GA concentration; solid squares, optical density; open squares, DHA
concentration.
A
180
B
180
60
60
3
2
3
2
150
120
90
60
30
0
50
40
30
150
120
90
50
40
30
20
10
0
1
0
20
10
0
60
30
1
0
0
1
2
3
4
1
2
3
4
Cultivation time (d)
Cultivation time (d)
Fig. 7. Time Course of GA Production from Raw Glycerol Sample B by Gluconobacter sp. NBRC3259.
A, 500 rpm. B, 700 rpm. Symbols: solid circles, glycerol concentration; open circles, GA concentration; solid squares, optical density; open
squares, DHA concentration.
pretreatment (Fig. 5). This result suggests that raw
glycerol sample B contained higher amounts of un-
known impurities, with a negative effect on the growth
of the strain. Also, we found that the type of acid for
neutralization is important in GA production by the
strain. As shown in Figs. 3D and Fig. 6, Clꢃ showed a
more negative effect on the enzymatic conversion of
glycerol to GA than SO42ꢃ. It is necessary to do prior
purification properly, depending on the raw glycerol
sample.
As shown in Figs. 4, 6, and 7, most of the glycerol
consumed by Gluconobacter sp. NBRC3259 was con-
verted to GA and DHA, and both GA and DHA
production increased concomitantly with growth of the
strain. When the strain’s growth reached a stationary
state after 2 d, GA and DHA production appeared to
have ceased. In this stationary state, the pH of the
culture became pH 3.2 (initial pH 6.3). When pH was
controlled so as not to fall under pH5 during cultivation
(initial pH 6.3), GA productivity was enhanced approx-
imately 2-fold (Fig. 4). Also, an agitation speed of
700 rpm (higher aeration conditions) resulted in lower
GA production than 500 rpm (lower aeration condi-
tions), whereas it led to increasing DHA production. In
contrast, an agitation speed of 250 rpm resulted in very
poor growth of the strain over 4 d (data not shown).
These results indicate that both pH control and agitation
speed are important factors in efficient GA production.
During all the experiments, a possible intermediate,
glyceraldehyde (Fig. 1), was not found.
was able to produce not only D-GA (88.4%) but also
L-enantiomer (11.6%, Fig. 1). In 1987, a Japanese patent
application (JP0751069, Daicel Chemical Industries)
described that the GA produced by four Gluconobacter
strains was D-GA. This result indicates the possibility
that the enantiomeric composition of GA produced by
Gluconobacter strains is different in different strains.
Until now, concerning the production mechanisms of
GA from glycerol by acetic acid bacteria at the
molecular level, no reports have been published. Adachi
et al.17,18) reported that purified membrane-bound alco-
hol dehydrogenase and membrane-bound aldehyde de-
hydrogenase from G. suboxydans did not exhibit spe-
cific activities toward glycerol and DL-glyceraldehyde
respectively. Therefore, to elucidate the production
mechanisms of D-GA and of L-GA, it is important to
purify and characterize the enzymes involved in GA
production.
Acknowledgments
This work was supported by the Industrial Technol-
ogy Research Grant Program in 08A26202c from the
New Energy and Industrial Technology Development
Organization (NEDO) of Japan.
References
1) Ohara O, Appl. Microbiol. Biotechnol., 62, 474–477 (2003).
2) Kamm B and Kamm M, Appl. Microbiol. Biotechnol., 64, 137–
145 (2004).
The enantiomeric composition of the GA produced by
Gluconobacter sp. NBRC3259 was analyzed. The strain
3) da Silva GP, Mack M, and Contiero J, Biotechnol. Adv., 27, 30–
39 (2009).

Raw Glycerol Conversion to Glyceric Acid by Gluconobacter sp.
1805
4) Dharmadi Y, Murarka A, and Gonzalez R, Biotechnol. Bioeng.,
94, 821–829 (2006).
Microbiol. Biotechnol., 81, 1033–1039 (2009).
13) Matsushita K, Toyama H, and Adachi O, Adv. Microb. Physiol.,
36, 247–301 (1994).
5) Claude S, Fett/Lipid, 101, 101–104 (1999).
6) Pagliaro M, Ciriminna R, Kimura H, Rossi M, and Della Pina C,
Angew. Chem. Int. Ed., 46, 4434–4440 (2007).
7) Zhou C-H, Beltramini JN, Fan Y-X, and Lu GQ, Chem. Soc.
Rev., 37, 527–549 (2008).
14) Asad-ur-Rehman, Wijesekara RGS, Nomura N, Sato S, and
Matsumura M, J. Chem. Technol. Biotechnol., 83, 1072–1080
(2008).
15) Papanikolaou S, Fick M, and Aggelis G, J. Chem. Technol.
Biotechnol., 79, 1189–1196 (2004).
8) Garcia R, Besson M, and Gallezot P, Appl. Catal. A, 127, 165–
176 (1995).
´
16) Gonzales-Pajuelo M, Andrade JC, and Vasconcelos I, J. Ind.
Microbiol. Biotechnol., 31, 442–446 (2004).
9) Porta F and Prati L, J. Catal., 224, 397–403 (2004).
10) Bianchi CL, Canton P, Dimitratos N, Porta F, and Prati L, Catal.
Today, 102, 203–212 (2005).
17) Adachi O, Tayama K, Shinagawa E, Matsushita K, and
Ameyama M, Agric. Biol. Chem., 42, 2045–2056 (1978).
18) Adachi O, Tayama K, Shinagawa E, Matsushita K, and
Ameyama M, Agric. Biol. Chem., 44, 503–515 (1980).
ˇ
ˇ
´
11) Svitel J and Sturdık E, J. Ferment. Bioeng., 78, 351–355 (1994).
12) Habe H, Fukuoka T, Kitamoto D, and Sakaki K, Appl.