Bioresource Technology 156 (2014) 170–175

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Enhanced production of 3-hydroxypropionic acid from glycerol by

modulation of glycerol metabolism in recombinant Escherichia coli
Kwangwook Kim a, Sun-Ki Kim b, Yong-Cheol Park c,⇑, Jin-Ho Seo a,b,⇑
a
Interdisciplinary Program of Bioengineering, Seoul National University, Seoul 151-921, South Korea
b
Department of Agricultural Biotechnology and Center for Food and Bioconvergence, Seoul National University, Seoul 151-921, South Korea
c
Department of Bio and Fermentation Convergence Technology, Kookmin University, Seoul 136-702, South Korea

h i g h l i g h t s

 3-Hydroxypropionic acid (3-HP) is a valuable biochemical with high potential.

 The E. coli glycerol metabolism was modulated to enhance 3-HP production.
 Deletion of glpK and yqhD highly increased the glycerol flux to 3-HP production.
 A new semialdehyde dehydrogenase was expressed to reduce 3-HPA accumulation.
 The optimal E. coli strain produced 57 g/L 3-HP in a fed-batch culture.

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

Article history: 3-Hydroxypropionic acid (3-HP) is a valuable biochemical with high potential for bioplastic manufacturing.
Received 21 October 2013 The endogenous glycerol metabolism and by-product formation pathway in Escherichia coli were modu-
Received in revised form 1 January 2014 lated to enhance 3-HP production from glycerol. Double deletion of glpK and yqhD directed the glycerol flux
Accepted 3 January 2014
to 3-HP biosynthesis and reduced the formation of 1,3-propanediol. Since 3-hydroxypropionaldehyde (3-
Available online 11 January 2014
HPA), a precursor of 3-HP, is toxic to cell growth, the gene encoding Pseudomonas aeruginosa semialdehyde
dehydrogenase (PSALDH) highly active on 3-HPA was expressed in E. coli. Finally, fed-batch culture of
Keywords:
recombinant E. coli BL21star(DE3) without glpK and yqhD, and expressing Lactobacillus brevis DhaB-DhaR,
3-Hydroxypropionic acid
Glycerol
and P. aeruginosa PSALDH resulted in 57.3 g/L 3-HP concentration, 1.59 g/L-h productivity and 0.88 g/g
Glycerol metabolism yield. In conclusion, modulation of the glycerol metabolism in combination with enhanced activity of 3-
Escherichia coli HPA dehydrogenation improved the production of 3-HP from glycerol.
Fed-batch fermentation Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Glycerol, a by-product of biodiesel production using plant oils, is

3-Hydroxypropionic acid (3-HP) is a three-carbon carboxylic
acid containing a b-hydroxyl group. 3-HP has been selected as
one of the top value-added chemicals produced from biomass by
US Department of Energy (Bozell and Petersen, 2010) and has a
high potential to produce several chemicals such as acrylic acid,
acrylonitrile, malonic acid, and 1,3-propanediol (1,3-PDO) (Kumar
et al., 2013).
For microbial production of 3-HP, bacterial strains of Escherichia
coli, Klebsiella pneumoniae and Pseudomonas denitrificans were met-
abolically engineered to overexpress essential enzymes engaged in
the 3-HP biosynthesis because natural 3-HP producers did not
possess a high capacity for 3-HP production (Kumar et al., 2013).
⇑ Corresponding authors. Tel.: +82 2 910 5462; fax: +82 2 910 5739 (Y.-C. Park),
tel.: +82 2 880 4855; fax: +82 2 873 5095 (J.-H. Seo).
E-mail addresses: ycpark@kookmin.ac.kr (Y.-C. Park), jhseo94@snu.ac.kr (J.-H.
Seo).
an attractive carbon source for biochemical production (Durnin
et al., 2009). Currently, two metabolic pathways for 3-HP produc-
tion from glycerol were developed: the CoA-dependent and
CoA-independent pathways (Kumar et al., 2013). In the CoA-inde-
pendent pathway (Fig. 1), glycerol dehydratase (DhaB) and
aldehyde dehydrogenase (Aldh) are two essential enzymes for
glycerol conversion to 3-HP. Glycerol dehydratase is a coenzyme
B12-dependent diol dehydratase (EC 4.2.1.30) catalyzing dehydra-
tion of glycerol to 3-hydroxypropionaldehyde (3-HPA). Since glyc-
erol dehydratase undergoes irreversible inactivation when its
substrate is glycerol (Toraya et al., 1976), the inactivated glycerol
dehydratase should be reactivated by expression of reactivating
proteins (Kwak et al., 2013; Mori et al., 1997). The second enzyme
of aldehyde dehydrogenase converts 3-HPA to 3-HP and is specific
for NAD+ cofactor. To date, aldehyde dehydrogenases with 3-HPA
oxidation activity have been identified in some microorganisms
(Kumar et al., 2013; Valdehuesa et al., 2013). Since coenzyme B12

0960-8524/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2014.01.009
 K. Kim et al. / Bioresource Technology 156 (2014) 170–175
Fig. 1. Metabolic pathway for 3-HP production in recombinant E. coli BL21star(DE3)
strain expressing glycerol dehydratase and aldehyde dehydrogenase. The name of
metabolites are abbreviated as follows: 3-HP, 3-hydroxypropionic acid; 3-HPA, 3-
hydroxypropionaldehyde; 1,3-PDO, 1,3-propanediol; Gly3P, glycerol-3-phosphate;
DHA, dihydroxyacetone; DHAP, dihydroxyacetone phosphate. Italicized letters
indicate metabolic enzymes: DhaB, glycerol dehydratase; Ald, aldehyde dehydro-
genase; YqhD, 1,3-propandiol oxidoreductase; GlpF, glycerol symporter; GlpK,
glycerol kinase; GldA, glycerol dehydrogenase; GlpD/GlpAC, glycerol-3-phosphate
dehydrogenases; DhaLKM, dihydroxyacetone kinase.
is required for dehydration of glycerol to 3-HPA, a natural coen-
zyme B12 producer of K. pneumoniae was engineered to produce
3-HP. Since K. pneumoniae produces 1,3-propanediol (1,3-PDO) as
a main metabolite of glycerol, the endogenous dhaT and yqhD were
disrupted and its aldehyde dehydrogenase genes of puuC and dhaB
were overexpressed. A fed-batch culture with 5% dissolved oxygen
content resulted in 28 g/L 3-HP concentration and 40% g/g yield
(Ashok et al., 2013b). By-products of 3.3 g/L 1,3-PDO, 10.6 g/L ace-
tate and 7.4 g/L 2,3-butanediol also accumulated in the culture
broth. Recombinant K. pneumoniae expressing E. coli K12 aldH pro-
duced 48.9 g/L 3-HP and 25.3 g/L 1,3-PDO simultaneously in a
microaerobic fed-batch fermentation while by-products of lactate,
acetate and ethanol were also produced in the range of 3–25 g/L
(Huang et al., 2013). Since E. coli does not have the 3-HP biosyn-
thetic enzymes, genes for both glycerol dehydratase and aldehyde
dehydrogenase should be introduced. Recombinant E. coli SH-BGK1
expressing K. pneumoniae DhaB123 and GdrAB, and Azospirillum
brasilense KGSADH produced 38 g/L 3-HP in a 71 h fed-batch
culture with 35% yield and 0.54 g/L-h productivity, in which
by-products of 9.7 g/L 1,3-PDO and 7.2 g/L acetate were produced
(Rathnasingh et al., 2009).
In our previous report (Kwak et al., 2013), new glycerol dehy-
dratase and its reactivase from Lactobacillus brevis were identified.
Their coding genes and the E. coli K12 aldH gene were overexpres-
sed in E. coli. The two-step fed-batch process using dual substrates
of glycerol and glucose resulted in 14.3 g/L 3-HP concentration,
0.26 g/L-h productivity and 0.14 g/g yield. Since the 3-HP yield
based on glycerol consumed is only 15% relative to the theoretical
yield, more metabolic engineering should be done to increase the
glycerol flux to 3-HP production. As shown in Fig. 1, a wild type
of E. coli has two glycerol metabolic pathways toward the central
carbon metabolism, which are activated or deactivated by the
171
presence of electron acceptors (Durnin et al., 2009). Recombinant
E. coli stains expressing the 3-HP biosynthetic enzymes produced
a by-product of 1,3-PDO by the action of endogenous oxidoreduc-
tase YqhD (Kwak et al., 2013; Zhang et al., 2006).
In this study, two genes (glpK and gldA) engaged in the central
glycerol metabolism and the yqhD gene encoding an endogenous
propanediol oxidoreductase were disrupted combinatorially in
recombinant E. coli overexpressing the 3-HP biosynthetic enzymes
in order to enhance 3-HP production. Since 3-HPA, an intermediate
in the 3-HP biosynthesis, accumulated because of low activity of
aldehyde dehydrogenase (Rathnasingh et al., 2009), new bacterial
aldehyde dehydrogenases highly active on 3-HPA were identified
and overexpressed in recombinant E. coli. Finally, a two-step fed-
batch fermentation of the optimal E. coli strain was carried out to
increase the performance of 3-HP production.
2. Methods
2.1. Strains and plasmids
E. coli DH5a and BL21star(DE3) (Invitrogen Co., Carlsbad, CA,
USA) strains were used for plasmid construction and 3-HP produc-
tion, respectively. Plasmids pELDRR expressing the L. brevis glyc-
erol dehydratase complex (DhaB1B2B3) and DhaB reactivase
(DhaR1R2) genes and pCEa harboring the E. coli aldH gene (Kwak
et al., 2013) were used for 3-HP production. Plasmids pKD46,
pKD13 and pCP20 (Datsenko and Wanner, 2000) were used for
deletion of the chromosomal glpK, gldA and yqhD genes, respec-
tively. Each gene coding for putative semialdehyde dehydrogen-
ases was cloned in plasmid pCDFDuet-1 (Novagen Co., Madison,
WI, USA) with the T7 promoter.
2.2. Genetic manipulation
Deletion of the chromosomal genes of glpK, gldA and yqhD genes
followed the previous report with minor modification (Datsenko and
Wanner, 2000). Briefly, each gene deletion cassette with the kana-
mycin resistance gene was amplified by a polymerase chain reaction
(PCR) with the corresponding DNA oligomers (Table 1) and a tem-
plate of plasmid pKD13: D1glpK and D4glpK for glpK, D1gldA and
D4gldA for gldA, D1yqhD and D4yqhD for yqhD. Each PCR product
was transformed into E. coli competent cells harboring plasmid
pKD46 expressing k recombinase. After selection of the transfor-
mants on LB medium with kanamycin and ampicilin, plasmid
pCP20 expressing the yeast FLP recombinase was introduced into
the transformants to remove the kanamycin resistance gene. The
plasmids were removed by incubation of the transformants at 43 °C.
The genes encoding putative aldehyde dehydrogenases were
PCR-amplified from the genomic DNA of P. aeruginosa ATCC10145
or K. pneumoniae 342 with the corresponding PCR primers (Table 1
and Supplement 2). For example, two PCR primers of F-PSPA3072
and R-PSPA3072 were used for amplification of the P. aeruginosa
semialdehyde dehydrogenase (PSALDH) gene (PSPA7_3072, NCBI
Gene ID: 5354745). After digestion of the PCR product with BamHI
and HindIII restriction enzymes, the nucleotides were ligated with
plasmid pCDFDuet-1, resulting in the construction of plasmid
pCPa72. Construction of other plasmids (pCKa19, pCKa47, pCKa57,
pCKa99 and pCPa53) was described in Supplement 1.
Transformation of E. coli followed the CaCl2 treatment or elec-
troporation method and the sequences of DNA oligomers used in
this study are listed in Table 1 and Supplement 2.
2.3. Culture conditions
E. coli was grown in LB medium (1% yeast extract, 2% bacto-
tryptone and 1% NaCl) with appropriate antibiotics for genetic
172 K. Kim et al. / Bioresource Technology 156 (2014) 170–175

Table 1
List of DNA oligomers used in this study.

Name Nucleotide sequence (50 –30 ) Target gene

D1glpK CCGGATGCGGCATAAACGCTTCATTCGGCATTTACAGTGTAGGCTGGAGCTGCTTC glpK
D4glpK TGGTGAATTGCTGTTTGGTACGGTTGATACGTGGCTTATTCCGGGGATCCGTCGACC
D1gldA TTCAAACTCCCGGACAAGCCGGGAGTTTGGAGTAGGTTAGTGTAGGCTGGAGCTGCTTC gldA
D4gldA CCGATTTGGCACTACTCATCTCTAAAGGAGCAATTATGATTCCGGGGATCCGTCGACC
D1yqhD TTCTCTGCCCTCATATTGGCCAGCAAAGGGAGCAAGTAGTGTAGGCTGGAGTGCTTC yqhD
D4yqhD CGAAATGCCCGAAAACGAAAGTTTGAGGGTAAAAAGCATTCCGGGGATCCGTCGACC
F-PSPA3072 CGGGATCCCATGACAGCCATCCTCGGACACAACT P. aeruginosa PSALDH gene
R-PSPA3072 CCCAAGCTTTCAAGCAACCGCCGCCCGCGTC

Underlined sequences present homologous recombination regions of glpk, gldA or yqhD in the E. coli chromosome.
Italicized sequences indicate the recognition sites of the corresponding restriction enzymes.

manipulation and seed culture. To produce 3-HP, a fed-batch cul-
ture was performed in a 2.5-L jar fermentor (Kobiotech, Incheon,
Korea) with a 1-L working volume of defined R medium (Kim
et al., 2011) at 37 °C and pH 6.8. After the depletion of 20 g/L glu-
cose initially added, a feeding solution of 800 g/L glucose (Samyang
Genex, Incheon, Korea) was fed by the pH-stat mode of operation
throughout the fed-batch cultivation (Kwak et al., 2013). At the
mid-exponential growth phase (18 h culture time), IPTG, coen-
zyme B12 and glycerol were added once at final concentrations of
0.2 mM, 20 lM and 57–60 g/L, respectively. The culture tempera-
ture was changed to 25 °C and the jar was covered with aluminum
foil. When glycerol was consumed completely, glycerol was added
to a final concentration of around 35 g/L. To maintain a dissolved
oxygen level above 20%, agitation speed and aeration rate were
set at 1200–1300 rpm and 1 vvm of air supply, respectively. To
investigate the effect of 3-HPA on cell growth, E. coli BL21star(DE3)
was cultured in a baffled flask containing 100 mL of R medium
with 20 g/L glucose at 230 rpm and 37 °C. When the cell mass
reached about 0.5 g/L, various concentrations of 3-HPA was added
into the culture broth.
2.4. Determination of cell and metabolite concentrations
Optical density (OD) of the cells was measured at 600 nm using
a spectrophotometer (UV-1601, Shimadzu, Japan). Dry cell mass
was obtained by multiplication of OD with a pre-determined con-
version factor, 0.365 g/L/OD.
The concentrations of glucose, glycerol, 3-HP, 1,3-PDO and ace-
tate were determined by a high performance liquid chromatogra-
phy (1200 series, Agilent, Santa Clara, CA, USA) equipped with a
REZEX ROA-organic acid column (Phenomenex, Torrance, USA)
heated at 60 °C. A mobile phase of 5 mM H2SO4 was used at a flow
rate of 0.6 mL/min. Detection was made with a reflective index
detector and an UV detector at 210 nm (Kwak et al., 2013). The
3-HPA concentration was measured by the modified tryptophan-
HCl method (Kwak et al., 2013).
2.5. Enzyme activity assay
Activity assay of aldehyde dehydrogenase for 3-HPA followed a
previous report (Kwak et al., 2013) with some modification. For
preparation of crude enzyme extract, the recombinant cells were
cultured in 100 mL of R medium with 20 g/L glucose at 25 °C and
200 rpm. At the mid-exponential phase, IPTG was added into the
culture broth at the final concentration of 0.2 mM. After 6 h induc-
tion, an optical density of the culture broth was adjusted at 10 by
appropriate dilution and concentration. After cell harvesting, the
cell pellets were suspended in 100 mM potassium phosphate buf-
fer (pH 7.0). After cell disruption by an ultrasonic processor (Cole-
Parmer, IL, USA) and centrifugation at 12,000 rpm and 4 °C, the
supernatant was collected and used as the crude enzyme extract.
The reaction mixture was composed of 50 mM potassium phos-
phate buffer (pH 7.0), 0.4 mM NAD+, 5 mM MgCl2, 0.5 mM dithio-
threitol and 50 mM 3-HPA. The 3-HPA standard solution was
synthesized chemically and kindly donated by Prof. Sunghoon Park
at Pusan National University, Korea. After addition of the crude ex-
tract, reduction of NAD+ was monitored at 340 nm of wavelength
and 30 °C. One unit of aldehyde dehydrogenase was defined as
the amount of enzyme able to reduce 1 lmol NAD+ at the reaction
conditions. Protein concentration was determined by the protein
assay kit (Bio-Rad, Richmond, CA, USA) using bovine serum albu-
min as the standard. Specific enzyme activity (U/mg protein) was
obtained by dividing the enzyme activity by the total protein con-
centration of the crude enzyme solution. The assay was repeated
independently in triplicate.
3. Results and discussion
3.1. Effects of glpK deletion on 3-HP production from glycerol
For the endogenous glycerol metabolism of E. coli, glycerol is
mainly metabolized by glycerol kinase encoded by glpK in aerobic
conditions (Fig. 1). To direct the glycerol flux from glycerol-3-phos-
phate to 3-HP via 3-hydroxypropionaldehyde (3-HPA), the chromo-
somal glpK gene was disrupted by a homologous recombination
strategy. The effect of glpK disruption was evaluated in fed-batch
cultures of E. coli BL21star(DE3) and BL21star(DE3) DglpK strains
both transformed with plasmids pELDRR and pCEa. As shown in
Fig. 2(A), the control strain was grown exponentially by feeding a
concentrated glucose solution at the pH-stat mode. Finally,
90.3 g/L dry cell mass was obtained. Production of 3-HP was initi-
ated by IPTG induction as well as addition of glycerol and coenzyme
B12. About 58 g/L glycerol was converted to 3-HP. Finally, 18.3 g/L
3-HP concentration with two by-products of 2.94 g/L 3-HPA and
3.88 g/L 1,3-PDO was obtained in 50 h of fed-batch culture. The
glpK-disrupted strain showed the similar growth pattern to the con-
trol (Fig. 2(B)). However, the glpK mutant strain produced 3-HP
more rapidly than the control strain. Finally, 29.1 g/L 3-HP concen-
tration, 1.04 g/L-h 3-HP productivity and 0.47 g/g glycerol 3-HP
yield (equivalent to 48% based on the theoretical yield) were ob-
tained in 50 h of fed-batch culture. 7.89 g/L 3-HPA and 13.1 g/L
1,3-PDO also accumulated as byproducts during the fed-batch
operation. The same trend of the glycerol flux change to the 3-HP
biosynthesis could be observed in a glpK deletion mutant of K. pneu-
moniae (Ashok et al., 2013a).
3.2. Double deletion of glycerol metabolic enzymes for efficient
production of 3-HP
Glycerol is generally utilized into the central carbon metabo-
lism by glycerol kinase (GlpK) and/or glycerol dehydrogenase
 K. Kim et al. / Bioresource Technology 156 (2014) 170–175
Fig. 2. Fed-batch fermentations of E. coli BL21star(DE3) wild type strain (A) and its
mutant deficient in the glpK gene (4glpK) (B). The two strains harbored plasmids
pELDRR and pCEa. After depletion of 20 g/L glucose, feeding of 800 g/L glucose
solution at the pH-stat mode was started (thick arrow). At the mid-exponential
phase, IPTG, coenzyme B12 and glycerol were added to final concentrations of
0.2 mM, 20 lM and 60 g/L, respectively, and the culture temperature was changed
to 25 °C (thin arrow). Symbols: d, dry cell mass; s, glucose; h, glycerol; j, 3-HP;
N, 3-HPA; ., 1,3-PDO.
(GldA) depending upon the supply of oxygen (Durnin et al., 2009).
Additionally, recombinant E. coli able to produce 3-HP can convert
glycerol to 1,3-PDO via 3-HPA by the action of an endogenous
propanediol oxidoreductase encoded by yqhD (Jarboe, 2011). To
minimize the flux of glycerol to the central carbon metabolism
and 1,3-PDO production, and thereby to increase 3-HP yield from
glycerol, the gldA or yqhD gene in the glpK-deficient strain was
disrupted to examine the effects of the dual disruption on 3-HP
production in fed-batch cultures. As shown in Fig. 3(A), double
deletion of glpK and gldA showed similar growth rates in the glu-
cose consumption stage (0–18 h), compared with the control
strain (Fig. 2(A)) and glpK deletion mutant (Fig. 2(B)). After IPTG
and glycerol addition at 18 h, the cells grew to reach 100.4 g/L
dry cell mass. In the glycerol conversion to 3-HP, double disrup-
tion of glpK and gldA showed the similar results of 3-HP produc-
tion to the single deletion of glpK, which gave 30.0 g/L 3-HP
concentration, 0.94 g/L-h productivity and 0.50 g/g glycerol yield.
Byproducts of 6.92 g/L 3-HPA and 13.2 g/L 1,3-PDO were also pro-
duced. Single deletion of glpK and double disruption of glpK and
gldA showed 1.2–1.6 times higher 3-HP production parameters
than the control strain without modification, indicating that shift-
ing of the glycerol flux away from the central carbon metabolism
was effective for efficient 3-HP production. The gldA deletion did
not influence the glycerol consumption and 3-HP production in
this aerobic fed-batch culture, which might be due to the fact that
173
Fig. 3. Fed-batch fermentations of E. coli BL21star(DE3) mutants deficient in the
glpK and gldA genes (4glpK4gldA) (A), and glpK and yqhD genes (4glpK4yqhD) (B),
harboring plasmids pELDRR and pCEa. After depletion of 20 g/L glucose, feeding of
800 g/L glucose solution at the pH-stat mode was started (thick arrow). At the mid-
exponential phase, IPTG, coenzyme B12 and glycerol were added to final concen-
trations of 0.2 mM, 20 lM and 60 g/L, respectively, and the culture temperature was
changed to 25 °C (thin arrow). Symbols: d, dry cell mass; s, glucose; h, glycerol; j,
3-HP; N, 3-HPA; ., 1,3-PDO.
the GldA-DhaLKM pathway is responsible for the anaerobic catab-
olism of glycerol while the GlpK-GlpD/GlpABC pathway contrib-
utes to the aerobic utilization of glycerol (Durnin et al., 2009).
Interestingly, the change of the glycerol flux to the 3-HP pathway
by the glpK deletion significantly increased the accumulation of 3-
HPA and 1,3-PDO, requiring additional modifications of the 3-HP
pathway.
YqhD, an NADP+-dependent oxidoreductase with a broad-sub-
strate range, is known to be responsible for 1,3-PDO accumulation
in 3-HP producing E. coli strains (Jarboe, 2011; Ko et al., 2012;
Kwak et al., 2013). A fed-batch culture was performed for the re-
combinant E. coli strain deficient in both glpK and yqhD
(Fig. 3(B)). The specific growth rates during both the glucose con-
sumption stage (0–18 h) and the glycerol metabolizing stage
(18–50 h) were lower than other fed-batch cultures described
above and 82.3 g/L of maximum dry cell mass was obtained. After
glycerol addition, 3-HP concentration increased linearly up to
44.1 g/L in 50 h culture, which was 52% higher than that of the
recombinant E. coli DglpK strain. As expected, the yqhD deletion
decreased 1,3-PDO production considerably to 1.80 g/L, but 3-
HPA still accumulated at 5.66 g/L. Similarly, single deletion of the
1,3-PDO dehydrogenase (DhaT) gene or double deletion of dhaT
and yqhD in recombinant K. pneumoniae expressing puuC or kgsadH
also reduced 1,3-PDO production and hence improved 3-HP pro-
duction from glycerol (Valdehuesa et al., 2013).
174 K. Kim et al. / Bioresource Technology 156 (2014) 170–175

3.3. Replacement of new aldehyde dehydrogenase active on 3-HPA

Through the modulation of the glycerol metabolic pathway and

fed-batch cultivations, glycerol could be converted to 3-HP at a
high titer. But 3-HPA, the precursor of 3-HP, was detected over
5 g/L in the culture medium, which is known to be a broad-spec-
trum antimicrobial agent (Chen and Chen, 2013). To assess the
toxic effect of 3-HPA on cell growth, the host strain of E. coli
BL21star(DE3) was cultured in the presence of various concentra-
tions of 3-HPA (Fig. 4). The cell growth decreased significantly in
proportion to 3-HPA concentrations. Addition of only 2 g/L 3-HPA
reduced specific growth rate by 60% and maximum dry cell mass
by 37% compared with the control case without 3-HPA addition,
indicating that 3-HPA should be minimized in order to improve
3-HP production. In some reports for reduction of the 3-HPA toxic-
ity and acceleration of the conversion of 3-HPA to 3-HP, E. coli
AldH, K. pneumoniae PuuC and A. brasilense KGSADH with different
activities on 3-HPA were overexpressed in E. coli or K. pneumoniae
(Ko et al., 2012; Kwak et al., 2013; Rathnasingh et al., 2009). To re-
duce 3-HPA accumulation, aldehyde dehydrogenases with high Fig. 5. Activity assay of various putative aldehyde dehydrogenase on 3-HPA as a
activities on 3-HPA were identified. Six genes encoding putative substrate. The plasmids contained the dehydrogenase genes as follows: pCEa, E. coli
aldehyde dehydrogenases from P. aeruginosa ATCC10145 and K. K12 aldH; pCKa19, K. pneumoniae ACI11654; pCKa47, K. pneumoniae ACI07089;
pCKa57, K. pneumoniae ACI11169; pCKa99, K. pneumoniae ACI10731; pCPa53, P.
pneumonia 342 were PCR-amplified. Each gene was cloned into
aeruginosa PA1253; pCPa72, P. aeruginosa PSPA7_3072.
plasmid pCDFDuet-1 (Supplement 1), of which resulting plasmid
was transformed into E. coli BL21star(DE3). By simple batch fer-
with plasmids pELDRR and pCEa (Fig. 3(B)), and its final dry cell
mentations using R medium with 20 g/L glucose and 6 h of IPTG
mass reached 91.4 g/L. After addition of IPTG and coenzyme B12,
induction, the crude cell extracts were collected and subjected to
57 g/L glycerol was depleted rapidly in about 20 h and more
aldehyde dehydrogenase activity assay on 3-HPA (Fig. 5). Among
glycerol was added. According to the rapid consumption of glyc-
the six enzymes, a gamma-aminobutyraldehyde dehydrogenase
erol, 3-HP was produced efficiently and its maximum concentra-
(ACI11169) from K. pneumoniae and a putative semialdehyde dehy-
tion was achieved at 57.3 g/L in 46 h of fed-batch fermentation.
drogenase (PSALDH, PSPA7_3072) from P. aeruginosa possessed
During the glycerol metabolizing stage after 18 h culture, 3-HP
about 1.4 times higher specific activity than E. coli AldH. Finally,
yield of 0.88 g/g glycerol and 3-HP productivity of 1.59 g/L-h were
plasmid pCEa containing the E. coli aldH gene was replaced to plas-
obtained finally. In by-product formation, the change of E. coli AldH
mid pCPa72 containing the P. aeruginosa semialdehyde dehydroge-
to P. aeruginosa PSLADH reduced 3-HPA concentration to 1.22 g/L.
nase gene (PSALDH) for 3-HP production.
As a result, modulation of the glycerol metabolic pathway by the
glpK and yqhD disruption and overexpression of a highly active
3.4. Effects of 3-HPA reduction on 3-HP production aldehyde dehydrogenase from P. aeruginosa led to a considerable
reduction of by-product formation and hence a remarkable
To evaluate the effect of PSALDH expression on 3-HP produc- enhancement in 3-HP production from glycerol.
tion, a fed-batch culture of recombinant E. coli BL21star(DE3) defi- The two-step fed-batch cultivation of the optimal E. coli strain
cient in the glpK and yqhD genes (4glpK4yqhD) and harboring using glucose and glycerol as the carbon sources (Fig. 6 and Table 2)
plasmid pELDRR with the L. brevis DhaB-DhaR genes and plasmid
pCPa72 with the P. aeruginosa PSALDH gene was carried out in R
medium with 20 g/L glucose (Fig. 6). The cell growth of this strain
showed a similar pattern to the recombinant E. coli 4glpK4yqhD

Fig. 6. Fed-batch fermentation of E. coli BL21star(DE3) deficient in the glpK and

Fig. 4. Effect of 3-HPA concentration on cell growth of E. coli BL21star(DE3). At the
early exponential growth phase, 3-HPA at the final concentrations between 2 and
8 g/L was added into the culture broth of R medium with 20 g/L glucose.
yqhD genes (4glpK4yqhD) and harboring plasmids pELDRR and pCPa72. After
depletion of 20 g/L glucose, feeding of 800 g/L glucose solution at the pH-stat mode
was started (thick arrow). At the mid-exponential phase, IPTG, coenzyme B12 and
glycerol were added to final concentrations of 0.2 mM, 20 lM and 57 g/L,
respectively, and the culture temperature was changed to 25 °C (thin arrow).
Symbols: d, dry cell mass; s, glucose; h, glycerol; j, 3-HP; N, 3-HPA; ., 1,3-PDO.
 K. Kim et al. / Bioresource Technology 156 (2014) 170–175 175

Table 2
Summarized results of fed-batch fermentations of recombinant E. coli BL21star(DE3) strains producing 3-HP from glycerol.

E. coli BL21star(DE3) strain Dry cell 3-HP 3-HP 3-HP 3-HPA 1,3-PDO Glycerol
mass (g/L) concentration productivity1 yield2 concentration concentration recovery3
(g/L) (g/L-h) (g/g) (g/L) (g/L) (mol/mol)
pELDRR + pCEa 90.3 18.3 0.76 0.32 2.94 3.88 0.41
pELDRR + pCEa, 4glpK 88.3 29.1 0.90 0.47 7.89 13.1 0.74
pELDRR + pCEa, 4glpK4gldA 100.4 30.0 0.94 0.50 6.92 13.2 0.76
pELDRR + pCEa, 4glpK4yqhD 82.3 44.1 1.38 0.66 5.66 1.80 0.74
pELDRR + pCPa72, 4glpK4yqhD 91.4 57.3 1.59 0.88 1.22 3.21 0.92
E. coli BL21(DE3) expressing K. pneumoniae dhaB123, 6.7 38.7 0.54 0.35 N.D.5 9.68 0.46
gdrAB and KGSADH under the T7 promoter4
1
3-HP productivity was calculated in the period of 3-HP production stage after IPTG induction and glycerol addition.
2
3-HP yield was calculated on the basis of the amount of consumed glycerol.
3
Glycerol recovery was calculated by the division of total molar concentrations of 3-HP, 3-HPA and 1,3-PDO by that of consumed glycerol.
4
These data were referred in a previous report for fed-batch production of 3-HP from glycerol only (Rathnasingh et al., 2009).
5
The concentration of 3-HPA was not determined.

resulted in 2–3 times higher 3-HP production, relative to the same
type of a fed-batch culture of the recombinant E. coli expressing L.
brevis DhaB/DhaR cluster and E. coli AldH (Fig. 2(A)). The 3-HP yield
of 0.88 g/g glycerol was correspondent to 90% of the theoretical
yield (0.978 g/g glycerol). In comparison to other reports, our data
of 3-HP concentration, productivity and yield were 1.5, 2.9 and 2.5
times higher than the corresponding parameters, respectively,
when an E. coli strain expressing K. pneumoniae DhaB, GdrAB and
KGSADH was cultured by a fed-batch operation using glycerol as
a sole carbon source (Table 2) (Rathnasingh et al., 2009). Similar
3-HP titer and productivity could be found in a microaerobic fed-
batch culture of K. pneumoniae expressing E. coli AldH, but
25.3 g/L 1,3-PDO was also co-produced (Huang et al., 2013).
4. Conclusion
To develop an efficient microbial process for 3-HP production
from glycerol, the endogenous glycerol metabolism of E. coli was
modulated by the specific deletion of the genes encoding glycerol
metabolic enzymes. Deactivation of both glycerol kinase (GlpK)
and propanediol oxidoreductase (YqhD) improved 3-HP yield
significantly. Introduction of new aldehyde dehydrogenase with
high activity on 3-HPA enhanced 3-HP production more. In the
fed-batch process, glucose was fed at the pH-stat mode in order
to regenerate the reducing power and supply cellular energy. To
improve the 3-HP yield based on total sugars consumed, further
research is in progress to engineer E. coli host cells able to convert
glucose to 3-HP via the glycerol synthetic pathway.
Acknowledgements
This work was supported by the National Research Foundation
of Korea (NRF) Grant (2013R1A2A2A01006590) funded by the
Ministry of Science, ICT & Future Planning, and also by the R&D
Program of MOTIE/KEIT (10044647).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.biortech.2014.
01.009.
References
Ashok, S., Mohan Raj, S., Ko, Y., Sankaranarayanan, M., Zhou, S., Kumar, V., Park, S.,
2013a. Effect of puuC overexpression and nitrate addition on glycerol
metabolism and anaerobic 3-hydroxypropionic acid production in
recombinant Klebsiella pneumoniae DglpKDdhaT. Metab. Eng. 15, 10–24.
Ashok, S., Sankaranarayanan, M., Ko, Y., Jae, K.-E., Ainala, S.K., Kumar, V., Park, S.,
2013b. Production of 3-hydroxypropionic acid from glycerol by recombinant
Klebsiella pneumoniae DdhaTDyqhD which can produce vitamin B12 naturally.
Biotechnol. Bioeng. 110, 511–524.
Bozell, J.J., Petersen, G.R., 2010. Technology development for the production of
biobased products from biorefinery carbohydrates-the US Department of
Energy’s ‘‘Top 10’’ revisited. Green Chem. 12, 539–554.
Chen, G., Chen, J., 2013. A novel cell modification method used in biotransformation
of glycerol to 3-HPA by Lactobacillus reuteri. Appl. Microbiol. Biotechnol. 97,
4325–4332.
Datsenko, K.A., Wanner, B.L., 2000. One-step inactivation of chromosomal genes in
Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97, 6640–
6645.
Durnin, G., Clomburg, J., Yeates, Z., Alvarez, P.J.J., Zygourakis, K., Campbell, P.,
Gonzalez, R., 2009. Understanding and harnessing the microaerobic metabolism
of glycerol in Escherichia coli. Biotechnol. Bioeng. 103, 148–161.
Huang, Y., Li, Z., Shimizu, K., Ye, Q., 2013. Co-production of 3-hydroxypropionic acid
and 1,3-propanediol by Klebseilla pneumoniae expressing aldH under
microaerobic conditions. Bioresour. Technol. 128, 505–512.
Jarboe, L., 2011. YqhD: a broad-substrate range aldehyde reductase with various
applications in production of biorenewable fuels and chemicals. Appl.
Microbiol. Biotechnol. 89, 249–257.
Kim, S.G., Shin, S.Y., Park, Y.C., Shin, C.S., Seo, J.H., 2011. Production and solid-phase
refolding of human glucagon-like peptide-1 using recombinant Escherichia coli.
Protein Expr. Purif. 78, 197–203.
Ko, Y., Ashok, S., Zhou, S., Kumar, V., Park, S., 2012. Aldehyde dehydrogenase activity
is important to the production of 3-hydroxypropionic acid from glycerol by
recombinant Klebsiella pneumoniae. Process Biochem. 47, 1135–1143.
Kumar, V., Ashok, S., Park, S., 2013. Recent advances in biological production of 3-
hydroxypropionic acid. Biotechnol. Adv. 31, 945–961.
Kwak, S., Park, Y.C., Seo, J.H., 2013. Biosynthesis of 3-hydroxypropionic acid from
glycerol in recombinant Escherichia coli expressing Lactobacillus brevis dhaB and
dhaR gene clusters and E. coli K-12 aldH. Bioresour. Technol. 135, 432–439.
Mori, K., Tobimatsu, T., Hara, T., Toraya, T., 1997. Characterization, sequencing, and
expression of the genes encoding a reactivating factor for glycerol-inactivated
adenosylcobalamin-dependent diol dehydratase. J. Biol. Chem. 272, 32034–
32041.
Rathnasingh, C., Raj, S.M., Jo, J.E., Park, S., 2009. Development and evaluation of
efficient recombinant Escherichia coli strains for the production of 3-
hydroxypropionic acid from glycerol. Biotechnol. Bioeng. 104, 729–739.
Toraya, T., Shirakashi, T., Kosuga, T., Fukui, S., 1976. Substrate specificity of
coenzyme B12-dependent diol dehydrase: glycerol as both a good substrate and
a potent inactivator. Biochem. Biophys. Res. Commun. 69, 475–480.
Valdehuesa, K., Liu, H., Nisola, G., Chung, W.-J., Lee, S., Park, S., 2013. Recent
advances in the metabolic engineering of microorganisms for the production of
3-hydroxypropionic acid as C3 platform chemical. Appl. Microbiol. Biotechnol.
97, 3309–3321.
Zhang, X., Li, Y., Zhuge, B., Tang, X., Shen, W., Rao, Z., Fang, H., Zhuge, J., 2006.
Construction of a novel recombinant Escherichia coli strain capable of producing
1,3-propanediol and optimization of fermentation parameters by statistical
design. World J. Microbiol. Biotechnol. 22, 945–952.