Production of Adipic Acid from Sugar Beet Residue by Combined Biological and Chemical Catalysis

Article abstract of DOI:10.1002/cctc.201600069

Adipic acid is one of the most important industrial dicarboxylic acids and is used mainly as a precursor to nylon-6,6. Currently, commercial adipic acid is produced primarily from benzene by a chemical route that is associated with environmental, health, and safety concerns. Herein, we report a new process to produce adipic acid from an inexpensive renewable feedstock, sugar beet residue by combining an engineered Escherichia coli strain and Re-based chemical catalysts. The engineered E.coli converted d-galacturonic acid to mucic acid, which was precipitated easily with acid, and the mucic acid was further converted to adipic acid by a deoxydehydration reaction catalyzed by an oxorhenium complex followed by a Pt/C-catalyzed hydrogenation reaction under mild conditions. A high selectivity to the free acid products was achieved by tuning the acidity of the Re-based catalysts. Finally, adipic acid was produced directly from sugar beet residue that was hydrolyzed enzymatically with engineered E.coli and two chemical catalysts in a yield of 8.4 %, which signifies a new route for the production of adipic acid.

Full text of DOI:10.1002/cctc.201600069

DOI: 10.1002/cctc.201600069
Communications
Production of Adipic Acid from Sugar Beet Residue by
Combined Biological and Chemical Catalysis
Hongfang Zhang+,[a] Xiukai Li+,[b] Xiaoyun Su+,[a, c] Ee Lui Ang,[a] Yugen Zhang,*[b] and
Huimin Zhao*[a, d]
Adipic acid is one of the most important industrial dicarboxylic
acids and is used mainly as a precursor to nylon-6,6. Currently,
commercial adipic acid is produced primarily from benzene by
a chemical route that is associated with environmental, health,
and safety concerns. Herein, we report a new process to pro-
duce adipic acid from an inexpensive renewable feedstock,
sugar beet residue by combining an engineered Escherichia
coli strain and Re-based chemical catalysts. The engineered
E. coli converted d-galacturonic acid to mucic acid, which was
precipitated easily with acid, and the mucic acid was further
converted to adipic acid by a deoxydehydration reaction cata-
lyzed by an oxorhenium complex followed by a Pt/C-catalyzed
hydrogenation reaction under mild conditions. A high selectivi-
ty to the free acid products was achieved by tuning the acidity
of the Re-based catalysts. Finally, adipic acid was produced di-
rectly from sugar beet residue that was hydrolyzed enzymati-
cally with engineered E. coli and two chemical catalysts in
a yield of 8.4%, which signifies a new route for the production
of adipic acid.
rate of 3–5% per year. Commercial adipic acid is produced
mainly from benzene-based petrochemical reactions, which
generate huge amounts of the greenhouse gas nitrous oxide.[2]
To overcome this key drawback, many efforts have been made
to produce adipic acid by more sustainable and environmen-
tally friendly approaches.[3] For example, researchers at Renno-
via developed a chemical route that comprised two heteroge-
neous catalytic steps for adipic acid production from glucose.
Specifically, glucose was aerobically oxidized to glucaric acid
that was in turn converted into adipic acid by hydrodeoxyge-
nation with maximum yields of 66 and 89%, respectively.[4]
However, this chemical route was limited by the need for high
temperature and pressure and the use of a supporting solvent.
Scientists at Verdezyne genetically engineered yeast strains iso-
lated from petroleum-contaminated soils to produce up to
50 gLÀ1 adipic acid from fatty acids derived from vegetable
oil.[5] In addition, Frost and co-workers designed and construct-
ed a three-step biosynthetic pathway in which an E. coli
mutant that lacked shikimate dehydrogenase converted 3-de-
hydroshikimic acid (DHS) to cis,cis-muconic acid.[6] Up to
36.8 gLÀ1 cis,cis-muconic acid was produced from batch-fed
fermentation in 22 mol% yield from glucose. Subsequently,
cis,cis-muconic acid was hydrogenated chemically to adipic
acid in a yield of 97%. Beckham and co-workers engineered
Pseudomonas putida KT2440 to produce 13.5 gLÀ1 cis,cis-mu-
conic acid from p-coumarate, which can be derived from
lignin, in a yield of 67% that was then converted chemically to
adipic acid with 97% conversion.[7] Qian and co-workers de-
signed a recombinant E. coli for a six-step pathway to reverse
adipic acid degradation, which produced 639 mgLÀ1 of adipic
acid directly from 10 gLÀ1 glucose.[8] These biological routes
used either glucose or fatty acids as substrates and in most
cases showed low titers and yields.
Adipic acid, the precursor of nylon-6,6, is one of the most im-
portant dicarboxylic acids in the chemical industry and has
a market volume of 2.8 million ton per year[1] with a growth
[a] Dr. H. Zhang,+ Dr. X. Su,+ Dr. E. L. Ang, Prof. H. Zhao
Metabolic Engineering Research Laboratory
Science and Engineering Institutes
31 Biopolis Way, The Nanos, 138669 (Singapore)
[b] Dr. X. Li,+ Dr. Y. Zhang
Institute of Bioengineering and Nanotechnology
31 Biopolis Way, The Nanos, 138669 (Singapore)
E-mail: ygzhang@ibn.a-star.edu.sg
[c] Dr. X. Su+
Present address:
Sugar beet residue is an abundant (~25 millionton/yr) and
low-value byproduct of the beet-processing industry, which
contains approximately 24% pectin, 24% cellulose, and 21%
arabinan.[9] d-Galacturonic acid, the principal monomer of
pectin, has been converted into potential platform chemicals,
which include meso-galactarate,[10] keto-deoxy-l-galactonate,[11]
and l-galactonic acid,[12] by using engineered filamentous
fungi. Recently, a consolidated bioprocess was developed to
generate l-galactonic acid directly from orange peel.[13] In this
work, we developed a new route to produce adipic acid from
sugar beet residue by combining biological catalysis and
chemical catalysis (Scheme 1). Inspired by recent studies on
the exploration of the synergy between biocatalysis and chem-
ical catalysis[14] and newly developed Re-based chemical cata-
Key Laboratory for Feed Biotechnology of the Ministry of Agriculture
Feed Research Institute
Chinese Academy of Agricultural Sciences
No. 12 South Zhongguancun Street, Haidian District, Beijing
100081 (P.R. China)
[d] Prof. H. Zhao
Departments of Chemical and Biomolecular Engineering
Chemistry, Biochemistry and Bioengineering
Institute for Genomic Biology
University of Illinois at Urbana-Champaign
Urbana, IL 61801 (USA)
E-mail: zhao5@illinois.edu
[+] These authors contributed equally.
Supporting Information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cctc.201600069.
ChemCatChem 2016, 8, 1500 – 1506
1500
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
Scheme 1. Schematic representation of the green route for adipic acid synthesis from sugar beet residue by combining biological catalysis and chemical catal-
ysis.
lysts able to transform mucic acid (also known as meso-galacta-
ric acid) to adipic acid,[15] we designed and constructed
a whole-cell biocatalyst to convert d-galacturonic acid released
from sugar beet residue into mucic acid. In addition, we devel-
oped and screened chemical catalysts that could catalyze the
production of free adipic acid from a highly selective deoxyde-
hydration (DODH) process using mucic acid as the substrate
via muconic acid as the intermediate. The combination of
these two types of catalysts yielded a green route to adipic
acid from renewable biomass.
a NAD+-dependent dehydrogenase, and the cellular NAD+
pool in the resting cells used might be insufficient to drive the
reaction to completion.
To further improve mucic acid production from d-galactur-
onic acid, two endogenous genes, uxaC, which encodes uro-
noate isomerase, and garD, which encodes d-galactarate dehy-
drogenase, were deleted from the E. coli BL21(DE3) genome by
the lambda red-mediated gene disruption method[21] to yield
E. coli BL21(DE3) DuxaC DgarD strain. The primary monomers
of sugar beet residue are glucose, l-arabinose, and d-galactur-
onic acid in almost equal dry weight ratio.[22] We mimicked this
composition by using a mixture of 10 gLÀ1 glucose, 10 gLÀ1 l-
arabinose, and 10 gLÀ1 d-galacturonic acid as the substrate to
investigate the effect of uxaC and garD deletion on mucic acid
production. E. coli BL21(DE3) udh (the E. coli BL21(DE3) strain
that overexpresses A. tumefaciens udh) and E. coli BL21(DE3)
DuxaC DgarD udh (the E. coli BL21(DE3) DuxaC DgarD strain
that overexpresses A. tumefaciens udh) were cultured to pre-
pare whole-cell biocatalysts. In 48 h, E. coli BL21(DE3) udh con-
verted 68.5% of the d-galacturonic acid to mucic acid with
a titer of 7.4 gLÀ1, whereas E. coli BL21(DE3) DuxaC DgarD udh
converted 95.4% of the d-galacturonic acid to mucic acid with
a titer of 10.3 gLÀ1 (Figure 1). If we take the presence of glu-
cose and l-arabinose into consideration, the yields of mucic
acid were 22.8 and 31.8% for E. coli BL21(DE3) udh and E. coli
BL21(DE3) DuxaC DgarD udh, respectively. This result demon-
strated clearly that blocking the native catabolism of d-galac-
turonic acid and mucic acid in E. coli improved mucic acid pro-
Development of recombinant E. coli-based biocatalyst
There are three distinct catabolic pathways for d-galacturonic
acid in microorganisms.[16] In the oxidation pathway, d-galac-
turonic acid is oxidized to mucic acid by uronate dehydrogen-
ase (Udh), which was identified in two Pseudomonas strains
and Agrobacterium tumefaciens.[17] In addition, Udh is able to
convert d-glucuronic acid to d-glucaric acid.[18] Recently, three
additional Udhs that could serve as alternative enzymes in our
whole-cell biocatalyst were characterized.[19] We chose the Udh
enzyme from A. tumefaciens because of its high activity to-
wards d-galacturonic acid (the Michaelis–Menten constants for
d-galacturonic acid are Km =0.5 mM, Vmax =124 UmgÀ1).[20] The
overexpression of A. tumefaciens udh directed by a T7-promot-
er-based pET system in E. coli BL21(DE3) was induced by
0.2 mm isopropyl b-d-1-thiogalactopyranoside (IPTG) and con-
firmed by using sodium dodecyl sulfate polyacrylamide gel
electrophoresis (SDS-PAGE; Figure S1). After induction, cells
were collected by centrifugation and used as a whole-cell cata-
lyst to convert 5, 10, and 20 gLÀ1 pure d-galacturonic acid into
4.3, 8.3, and 11.5 gLÀ1 mucic acid, respectively, in 48 h. These
yields are much higher than that of the fungal strain Hypocrea
jecorina QM6a Dgar1 udh engineered previously, which pro-
duced 3.8 gLÀ1 mucic acid from 17.4 gLÀ1 d-galacturonic acid
in 211 h at pH 5.0.[10] The titer of mucic acid in the engineered
H. jecorina QM6a Dgar1 udh was increased from 3.8 to 4.2 gLÀ1
after the addition of 2 gLÀ1 glucose to the reaction medium.
An antagonistic effect was observed in our engineered E. coli
strain. In theory, 1.08 gLÀ1 mucic acid should be produced
from 1 gLÀ1 d-galacturonic acid if d-galacturonic acid was oxi-
dized completely into mucic acid. We speculated that the
yields we obtained that were lower than the theoretical values
were likely because of the native catabolism of d-galacturonic
acid and mucic acid in E. coli in which d-galacturonic acid is
isomerized to d-tagaturonate by uronate isomerase and mucic
acid is catalyzed by d-galactarate dehydrogenase into 5-dehy-
dro-4-deoxy-d-glucarate. Another reason might be that Udh is
Figure 1. Effects of uxaC and garD deletions on mucic acid production. The
substrates are 10 gLÀ1 d-galacturonic acid, 10 gLÀ1 glucose, and 10 gLÀ1 l-
arabinose in 280 mM M9 medium. Resting cells with OD600=20 were used,
and the pH was adjusted to neutral after each sampling point. WT is E. coli
BL21(DE3) udh strain, and DuxaC DgarD is E. coli BL21(DE3) DuxaC DgarD
udh strain.
ChemCatChem 2016, 8, 1500 – 1506
www.chemcatchem.org
1501
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
duction and channeled almost all of the d-galacturonic acid to
mucic acid. d-Galacturonic acid was consumed completely in
both of the whole-cell reactions (Figure 1), which indicates
that d-galacturonic acid entered its isomerization pathway in
wild-type E. coli. In pH-controlled bioreactors, H. jecorina QM6a
Dgar1 udh produced 5.9 gLÀ1 of mucic acid at pH 6.5 and
2.6 gLÀ1 of mucic acid at pH 4.0 from 9 gLÀ1 d-galacturonic
acid and 4 gLÀ1 glucose.[10] We found a similar pH-dependent
production in E. coli BL21(DE3) DuxaC DgarD udh, in which the
mucic acid titer decreased if the reaction medium was acidified
(Figure S2). This phenomenon could be explained by the char-
acteristics of A. tumefaciens Udh, the optimal pH and tempera-
ture of which are 8.0 and 358C, respectively.[17] Different from
d-galacturonic acid, glucose and l-arabinose were both con-
sumed during the biotransformation reactions (Figure S3).
their different polarities, 4 and 2 were separated easily by ex-
traction to give isolated yields of 79 and 16%, respectively. Ki-
netic studies of the DODH of 1 catalyzed by NH4ReO4 revealed
that more than 65% of 1 was converted in the first 8 h and
the selectivity to 4 was as high as 98% (Figure 2). After 8 h,
Development of the Re-based chemical catalyst
In our DODH reaction of mucic acid (1) reported previously,
only muconic acid esters (monoester 2 and diester 3) were
produced with pure methyltrioxorhenium (MTO) catalyst (Sche-
me 2a),[15a] but from a chemical synthesis point-of-view, the
free acid product is more desirable and cost efficient than the
ester form. It was clear that the MTO-catalyzed DODH reaction
and acidic Re-catalyzed esterification reaction are parallel and
competitive, which indicates that the free acid selectivity of
the DODH reaction could be improved by tuning the acidity of
the Re-based catalysts. To identify cheaper and more conven-
ient catalysts for the DODH reaction, several less acidic Re
compounds were screened (Table S1).[20] If NH4ReO4 was em-
ployed for the DODH of 1 (Scheme 2), 82% yield of free mu-
conic acid (4) and 16% yield of muconic acid monopentyl
ester (2) were produced. The selectivity for the production of
free muconic acid was much higher than 43% selectivity with
MTO (5 mol%) at 1508C and no selectivity with HReO4 report-
ed previously.[15b] Our result also demonstrated that the selec-
tivity of the Re-catalyzed DODH of 1 to 4 or 2 was sensitive to
the format of catalyst and reaction conditions. As a result of
Figure 2. Time course of the NH4ReO4-catalyzed DODH reactions. Reaction
conditions: mucic acid (210.0 mg, 1.0 mmol), NH4ReO4 (13.4 mg, 0.05 mmol),
3-pentanol (20.0 mL), 1208C, flowing N2. Yield and conversion were deter-
1
mined by H NMR spectroscopy with an internal standard.
the reaction proceeded at a slower rate and the total yields of
4 and 2 increased gradually, but with more 2 in the products.
Thus, it is possible to obtain 4 in high selectivity by using
NH4ReO4 as the catalyst and controlling the reaction time. So
far, it has been demonstrated that by selection of the NH4ReO4
catalysts, free acid products could be achieved from sugar
acids at high selectivity
After we had achieved the high selectivity to 4 with
NH4ReO4, further hydrogenation reactions towards adipic acid
(5) was investigated. The NH4ReO4-catalyzed DODH reaction of
1 was terminated after 8 h. Compound 4 was isolated in 72%
yield, and 20% of unreacted 1 was recovered by filtration
(Scheme 2c). Thus, based on the converted 1, 4 was isolated in
Scheme 2. DODH of mucic acid over a) MTO and b) NH4ReO4 and c) the highly selective procedure for the transformation of mucic acid to adipic acid.
ChemCatChem 2016, 8, 1500 – 1506
www.chemcatchem.org
1502
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
90% yield. The hydrogenation of 4 was performed under mild
conditions (room temperature, H2, 7 bar), and 92% isolated
yield of 5 was obtained in 8 h. Notably, the hydrogenation re-
action was conducted in water instead of organic solvent as re-
ported previously.[23] Despite the poor solubility of 4 in water,
the hydrogenation reaction could still proceed as 5 dissolves
well in water.
Production of adipic acid from mock substrate
To develop an integrated process that combines the engi-
neered recombinant E. coli-based biocatalyst and the NH4ReO4
chemical catalyst, we sought to produce 5 from mock mono-
sugar mixtures that mimic the composition of sugar beet resi-
due hydrolysates, that is, glucose, l-arabinose, and d-galactur-
onic acid mixed equally. The E. coli BL21(DE3) DuxaC DgarD
udh strain was used to catalyze the conversion of d-galactur-
onic acid to 1 with 50 mL reaction volume in a 250 mL shake-
flask. In 48 h, more than 99% of d-galacturonic acid in the mix-
ture was converted to 1 (Figure S4), and unsurprisingly, glu-
cose and l-arabinose were both consumed. Compared to test
tubes, the relatively high yield of 1 here might be because of
the better oxygen transfer in the shake-flask. Compound 1 was
then recovered directly from the reaction supernatant by acid
precipitation with a recovery yield of 91%. The resulting 1 was
very pure (confirmed by NMR spectroscopy) and was then sub-
jected to the Re-catalyzed DODH reaction to give the same
high yield of 4 and 2 (99%). If we consider that we fed the
cells with an equal mixture of glucose, l-arabinose, and d-gal-
acturonic acid and the consecutive steps of the biological and
chemical transformations, the overall yield of 4 and 2 here was
29.7%.
Figure 3. Mucic acid production from hydrolyzed sugar beet residue using
the engineered E. coli BL21(DE3) DuxaC DgarD udh strain.
on sugar beet residue. We did not purify the d-galacturonic
acid and simply removed undigested carbohydrate fibers by fil-
tration, so unknown components may have adverse effects on
the whole-cell biocatalyst, which may explain the reduced d-
galacturonic acid conversion and 1 production here. The re-
sulting 1 was purified by acid precipitation (confirmed by
using NMR spectroscopy; Figure S6). Subsequently, the synthe-
sized bio-1 underwent the DODH reaction and hydrogenation
under standard conditions. The yield of 5 was 83% from our
homemade 1, which is almost the same as that from commer-
cial 1. In the reaction of the real substrate, the overall yield of
5 based on sugar beet residue was 8.4%. The complex compo-
sition of sugar beet residue, the inefficient release of d-galac-
turonic acid, the reduced d-galacturonic acid conversion, and
the lower purity of 1 accounted collectively for the decreased
yield of 5 compared to that from the reaction of the mock sub-
strate.
Production of adipic acid from sugar beet residue
Next we extended our studies to the real-world substrate,
sugar beet residue. The composition of the sugar beet residue
was first analyzed by diluted acid hydrolysis and found to com-
prise 22% glucose, 21% l-arabinose, and 21% d-galacturonic
acid by dry weight, which was close to reported values
(Table S2).[9] We then used the sugar beet residue directly as
the substrate to produce adipic acid by using both the engi-
neered whole-cell biocatalyst and the developed chemical cat-
alyst. Briefly, the sugar beet residue was hydrolyzed by using
a combination of cellulase, pectinase, and Viscozymeꢁ L in
50 mM citrate buffer at 508C, which produced a mixture of
22% glucose, 17.4% l-arabinose, and 16.5% d-galacturonic
acid from the total residue (Table S2). After enzymatic hydroly-
sis, the supernatant was collected by centrifugation and used
directly as the substrate for the whole-cell-catalyzed conver-
sion from d-galacturonic acid to 1. Although the system was
more complex than the above-mentioned mock substrate re-
action, it was still very efficient and the production reached
a plateau after around 24 h (Figure 3 and Figure S5). After
48 h, 86% of d-galacturonic acid was transformed into 1. In
other words, 6.9 gLÀ1 of 1 was produced from 7.8 gLÀ1 of d-
galacturonic acid, and the overall yield of 1 was 14.2% based
In conclusion, we developed a new route for the production
of adipic acid from sugar beet residue. This route is a promising
alternative for the valorization of beet residue and it could be
extended to other pectin-rich biomass such as citrus peel. In
this route, d-galacturonic acid was first released from beet resi-
due through an enzymatic hydrolysis reaction and then con-
verted to mucic acid by the engineered recombinant E. coli
strain. The yield of mucic acid could be further improved by
two approaches: the improvement of the enzymatic hydrolysis
efficiency to release more d-galacturonic acid from sugar beet
residue and the improvement of the specific activity of the ur-
onate dehydrogenase by either protein engineering[24] or bio-
prospecting. Moreover, to better valorize the sugar beet resi-
due, l-arabinose could also be converted to useful chemicals,
such as rare sugar l-ribulose. The feasibility has been demon-
strated by our preliminary results, but further optimizations are
needed to find compatible reaction conditions for both d-gal-
acturonic acid and l-arabinose conversion. The bioproduced
mucic acid was precipitated easily and used for chemical con-
version and its purity and reaction activity were comparable to
that obtained commercially. Mucic acid was converted to
adipic acid by a deoxydehydration reaction catalyzed by an ox-
ChemCatChem 2016, 8, 1500 – 1506
www.chemcatchem.org
1503
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
orhenium complex reaction followed by a Pt/C-catalyzed hy-
drogenation reaction under mild conditions. The acidity of the
oxorhenium catalysts determined the selectivity of the free
acid products in the deoxydehydration reactions. We used
modified Re catalysts to convert mucic acid to muconic acid
and then to adipic acid with a high selectivity.
Deletion of the uronoate isomerase gene uxaC and d-galac-
tarate dehydrogenase gene garD in E. coli BL21(DE3)
Gene knockout was conducted using the lambda red-mediated
gene disruption method as described elsewhere.[21] The cassette
for the deletion of uxaC was amplified from the pKD4 template
plasmid using
TAGCTCACTCGTTGAGA
a
pair of primers: 5’-TCGCACCATAAGCAAGC
GGAAGACGAAAGCGATTGTGTAGGCTG
GAGCTGCTTC-3’ and 5’-TATCCAGCGCATGGATCTTGATGTATTG
CATATCAACCCCAGACCATGGTCCATATGAATATCCTCCTTAG-3’
with 40–50 bp homologous sequence (italic) to its upstream and
downstream part. The resulting ~1.6 kb fragment that contained
the kanamycin resistance marker was purified on a gel. E. coli
BL21(DE3) was transformed with helper plasmid pKD46. E. coli
BL21(DE3)/pKD46 was grown in LB medium at 308C to the early
exponential phase, and lambda red recombinase was induced with
the addition of 0.2% l-arabinose. This strain was cultivated for an-
other 3 h before it was made electrocompetent. The ~1.6 kb dele-
tion cassette was electroporated into E. coli BL21(DE3)/pKD46 com-
petent cells, and positive deletion colonies were selected on the
kanamycin plate. The selected colonies were further verified by
colony PCR. The kanamycin cassette was removed with the help of
the pCP20 plasmid. Both pKD46 and pCP20 are temperature-sensi-
tive plasmids and could be lost by heating to 428C.
Experimental Section
Materials, strains, and media
All starting materials are commercially available and were used as
received, unless otherwise indicated. E. coli BL21(DE3) and the pET-
46 Ek/LIC Cloning Kit were purchased from Novagen (Germany).
E. coli DH5a obtained from Invitrogen (USA) was used for cloning
and propagation of plasmids. All strains were cultured at 378C and
250 rpm unless otherwise stated. All polymerase chain reactions
(PCR) were conducted on Phusion High-Fidelity DNA Polymerase
from Thermo Scientific (USA). QIAprep Spin Miniprep Kit, QIAquick
Gel Extraction Kit, and QIAquick PCR Purification Kit were from
Qiagen (Germany). d-Galacturonic acid was obtained from Sigma–
Aldrich (USA). Mucic acid (98%) and 3-pentanol (98%) were pur-
chased from Merck; MTO (98%), Re2O7 (99.99%), AgReO4 (99.99%),
KReO4 (99.99%), HReO4 (80% in water), NH4ReO4 (> 99%), Cs2CO3
(99.9%), and 5%Pt/C were purchased from Aldrich. Other reagents
involved were from Sigma or Merck. P-Bn62 and PMF63 were synthe-
sized according to literature methods. M9 medium contained
12.8 gLÀ1 Na2HPO4·7H2O, 3 gLÀ1 KH2PO4, 0.5 gLÀ1 NaCl, 1 gLÀ1
NH4Cl, 2 mM MgSO4, 0.1 mM CaCl2, 10 mgLÀ1 Thiamine, and
10 mgLÀ1 Biotin. M9-glucose medium, that is, M9 medium supple-
mented with 4 gLÀ1 glucose, was used for uronate dehydrogenase
overexpression.
The deletion cassette for garD was amplified using a pair of pri-
mers: 5’-ACATGGCACTAAGGCCATGTTTTGCGAGGGACGTTCCAAAA
GAAAATGGTCCATATGAATATCCTCCTTAG-3’
and
5’-AC
CAGGTCCTCATTTTAATAACCCCTGGCTGGAGAATATTGCACAGC
GATTGTGTAGGCTGGAGCTGCTTC-3’. All the other procedures
were the same as for the deletion of the uxaC gene except that
the targeting strain was E. coli BL21(DE3) DuxaC instead of wild
type. The double deletion mutant was named as E. coli BL21(DE3)
DuxaC DgarD.
Cloning of the uronate dehydrogenase gene
Production of mucic acid with E. coli BL21(DE3) DuxaC
DgarD udh
The Agrobacterium tumefaciens Udh gene udh (GenBank accession
number is BK006462) was cloned into a pET-46 vector by using the
Ek/LIC Cloning Kit. The 798 bp udh ORF with overhang sequences
at both the 5’ and 3’ ends was amplified from A. tumefaciens ge-
nomic DNA with the forward primer 5’-GACGACGACAAGAT
GAAACGGCTTCTTGTTACCGG-3’ and reverse primer 5’-GAGGA
GAAGCCCGGTCAGCTCTGTTTGAAGATCGGG-3’. The amplified DNA
fragment was treated with a T4 DNA polymerase and assembled
into the vector according to the manufacturer’s instructions to
generate the recombinant plasmid pET46-udh. The correct se-
quence of udh was verified by the DNA sequencing service provid-
ed by Axil Scientific Pte Ltd, Singapore.
E. coli BL21(DE3) DuxaC DgarD was transformed with pET46-udh by
heat shock to give the recombinant strain E. coli BL21(DE3) DuxaC
DgarD udh. The strain was cultured in the same way as mentioned
above to prepare a whole-cell biocatalyst. Resting cells of OD600=
20 were used to catalyze the reactions from d-galacturonic acid to
mucic acid in 280 mM modified M9 medium. Generally, the reac-
tion substrates were a mixture of equal amounts of d-galacturonic
acid, glucose, and l-arabinose, such as 10, 20, and 40 gLÀ1 of each
sugar. Reactions were conducted at 378C and 250 rpm for 2 days.
Samples were aliquoted every few hours, and the reaction medium
was adjusted to neutral pH with 3M NaOH after each sampling
point.
Expression of Udh in E. coli BL21(DE3)
E. coli BL21(DE3) was transformed with pET46-udh by heat shock to
give the recombinant strain E. coli BL21(DE3) udh. This strain was
cultured in M9-glucose medium for 24 h at 378C, and the expres-
sion of Udh was induced by adding IPTG to a final concentration
of 0.2 mM. This culture was continued for another 12 h before the
cells were collected by centrifugation. Resting cells of OD600=2
were used in the conversion reaction from d-galacturonic acid to
mucic acid, and reaction substrates of 5, 10, and 20 gLÀ1 d-galac-
turonic acid was dissolved in M9 medium. Reactions were conduct-
ed at 378C and 250 rpm for 2 days.
Homogenous DODH reactions
A
mixture of mucic acid (1.0 mmol, 210.0 mg), Re catalyst
(0.05 mmol), and 3-pentanol (20.0 mL) was heated to reflux
(1208C) in a 50 mL flask under a flow of air or N2. For the kinetic
study, 1.0 mL of reaction mixture was taken at different time inter-
vals and dried for NMR spectroscopy, and a known amount of me-
sitylene was added as an internal standard.
ChemCatChem 2016, 8, 1500 – 1506
www.chemcatchem.org
1504
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
as glucose and l-arabinose were detected by using a RID-10A re-
Conversion of mucic acid to adipic acid
fractive index detector (RID).
In the first step of the DODH reaction, a mixture of mucic acid
(2.0 mmol, 420.0 mg), NH4ReO4 (0.1 mmol, 26.8 mg), and 3-penta-
nol (40.0 mL) was heated to reflux (1208C) in a 100 mL flask under
N2 atmosphere for 8 h. The reaction mixture was cooled to RT. The
unconverted mucic acid was recovered by filtration and dried
under vacuum at 508C overnight (42 mg, 20% of the initial
amount). The filtrate was evaporated to dryness under reduced
pressure. The solid was extracted into n-hexane (20 mL”2) and
then vacuum dried at 508C overnight to obtain muconic acid
(204 mg, 72% yield). In the subsequent hydrogenation step, a mix-
ture of muconic acid (100 mg, 0.7 mmol), 5.0% Pt/C (10.0 mg), and
H2O (10.0 mL) was charged into a Parr reactor. The reactor was
sealed, purged with N2 three times, and then pressurized with H2
(7 bar). The reactor was stirred at RT (248C) for 8 h with overhead
stirring before it was depressurized. The catalyst was separated by
filtration, the water solvent was removed by evaporation, and
adipic acid was obtained as a white solid (92 mg, 92% yield).
Acknowledgements
We thank the Agency for Science, Technology and Research, Sin-
gapore for supporting various research projects in the Metabolic
Engineering Research Laboratory (MERL) through the Visiting In-
vestigator Programme to H.Z. This work was also supported by
the Institute of Bioengineering and Nanotechnology (Biomedical
Research Council, Agency for Science, Technology and Research
(A*STAR), Singapore) and Biomass-to-Chemicals Program (Science
and Engineering Research Council, A*STAR, Singapore) (Y. Z.).
Keywords: biomass
· biocatalysis · gene technology ·
homogeneous catalysis · rhenium
Production of mucic acid from sugar beet residue
[1] N. Z. Xie, H. Liang, R. B. Huang, P. Xu, Biotechnol. Adv. 2014, 32, 615–
622.
[2] S. Van de Vyver, Y. Roman-Leshkov, Catal. Sci. Technol. 2013, 3, 1465–
1479.
[3] a) T. Polen, M. Spelberg, M. Bott, J. Biotechnol. 2013, 167, 75–84; b) R.
Beerthuis, G. Rothenberg, N. R. Shiju, Green Chem. 2015, 17, 1341–
1361.
[4] T. R. Boussie, E. L. Dias, Z. M. Fresco, V. J. Murphy, J. Shoemaker, R.
Archer, H. Jiang, (Rennovia, Inc. Menlo Park, CA), US 9156766 B2, 2014.
[5] T. Beardslee, S. Picataggio, Lipid Technol. 2012, 24, 223–225.
[6] a) K. M. Draths, J. W. Frost, J. Am. Chem. Soc. 1994, 116, 399–400; b) W.
Niu, K. M. Draths, J. W. Frost, Biotechnol. Prog. 2002, 18, 201–211.
[7] D. R. Vardon, M. A. Franden, C. W. Johnson, E. M. Karp, M. T. Guarnieri,
J. G. Linger, M. J. Salm, T. J. Strathmann, G. T. Beckham, Energy Environ.
Sci. 2015, 8, 617–628.
[8] J.-L. Yu, X.-X. Xia, J.-J. Zhong, Z.-G. Qian, Biotechnol. Bioeng. 2014, 111,
2580–2586.
[9] M. D. Sutton, J. B. Doran Peterson, J. Am. Soc. Sugar Beet Technol. 2001,
38, 19–34.
Sugar beet residue was hydrolyzed by a combination of three en-
zymes, namely, cellulase from Trichoderma reesei ATCC 26921, pec-
tinase from Aspergillus aculeatus, and Viscozymeꢁ L (a multi-
enzyme complex that contains a wide range of carbohydrases,
which includes arabanase, cellulase, b-glucanase, hemicellulase,
and xylanase), all of which were obtained from Sigma–Aldrich,
USA. The hydrolysis reaction was performed in 100 mM sodium cit-
rate buffer (pH 5.0) at 508C and 200 rpm for 48 h in 250 mL baffled
flask. The reaction suspension was then filtrated to recover the
liquid solution and discard undigested debris. The collected super-
natant contained sugar monomers, such as d-galacturonic acid,
glucose, and l-arabinose, and was neutralized to pH 7.0 with 1.0m
Na2HPO4 to serve as the substrate for the subsequent transforma-
tion reaction. E. coli BL21(DE3) DuxaC DgarD udh was cultured as
mentioned above and used as a whole-cell biocatalyst. The reac-
tion volume was 50 mL, and the reactions were conducted at 378C
and 250 rpm for 2 days in 250 mL baffled flask. Samples were with-
drawn every few hours, and the reaction medium was adjusted to
neutral pH with 3M NaOH after each sampling point. Mucic acid
was purified by an acid precipitation method similar to that dem-
onstrated elsewhere.[10] Briefly, reaction suspension was centrifuged
to remove cell pellet, and 1M HCl was added to liquid broth until
the pH decreased to ~1.5 with stirring. Mucic acid precipitate was
recovered by centrifugation and vacuum dried at 608C for 4 h. The
resulting mucic acid was used for the subsequent chemical conver-
sion to adipic acid.
[10] D. Mojzita, M. Wiebe, S. Hilditch, H. Boer, M. Penttila, P. Richard, Appl.
Environ. Microbiol. 2010, 76, 169–175.
[11] M. G. Wiebe, D. Mojzita, S. Hilditch, L. Ruohonen, M. Penttila, BMC Bio-
technol. 2010, 10, 63.
[12] J. Kuivanen, D. Mojzita, Y. Wang, S. Hilditch, M. Penttila, P. Richard, M. G.
Wiebe, Appl. Environ. Microbiol. 2012, 78, 8676–8683.
[13] J. Kuivanen, H. Dantas, D. Mojzita, E. Mallmann, A. Biz, N. Krieger, D.
Mitchell, P. Richard, AMB Express 2014, 4, 33.
[14] a) C. A. Denard, H. Huang, M. J. Bartlett, L. Lu, Y. Tan, H. Zhao, J. F. Hart-
wig, Angew. Chem. Int. Ed. 2014, 53, 465–469; Angew. Chem. 2014, 126,
475–479; b) C. A. Denard, J. F. Hartwig, H. Zhao, ACS Catal. 2013, 3,
2856–2864; c) R. M. Lennen, D. J. Braden, R. A. West, J. A. Dumesic, B. F.
Pfleger, Biotechnol. Bioeng. 2010, 106, 193–202; d) P. Anbarasan, Z. C.
Baer, S. Sreekumar, E. Gross, J. B. Binder, H. W. Blanch, D. S. Clark, F. D.
Toste, Nature 2012, 491, 235–239.
HPLC analysis
[15] a) X. Li, D. Wu, T. Lu, G. Yi, H. Su, Y. Zhang, Angew. Chem. Int. Ed. 2014,
53, 4200–4204; Angew. Chem. 2014, 126, 4284–4288; b) M. Shiramizu,
F. D. Toste, Angew. Chem. Int. Ed. 2013, 52, 12905–12909; Angew. Chem.
2013, 125, 13143–13147.
[16] P. Richard, S. Hilditch, Appl. Microbiol. Biotechnol. 2009, 82, 597–604.
[17] S.-H. Yoon, T. S. Moon, P. Iranpour, A. M. Lanza, K. J. Prather, J. Bacteriol.
2009, 191, 1565–1573.
[18] E. Shiue, I. M. Brockman, K. L. Prather, Biotechnol. Bioeng. 2015, 112,
579–587.
[19] A. Pick, J. Schmid, V. Sieber, Microb. Biotechnol. 2015, 8, 633–643.
[20] H. Boer, H. Maaheimo, A. Koivula, M. Penttila, P. Richard, Appl. Microbiol.
Biotechnol. 2010, 86, 901–909.
The concentrations of d-galacturonic acid, mucic acid, glucose, and
l-arabinose in the reaction solution were analyzed by using an
HPLC system from Shimadzu, Japan. The HPLC was equipped with
a Nucleogelꢁ sugar 810H column (300 mm”7.8 mm, Macherey–
Nagel, Germany) linked to a Synergi 4 mL Hydro-RP 80 A column
(250 mm”4.6 mm, Phenomenex, USA), and these two analytical
columns were protected with a SecurityGuard guard cartridge
from Phenomenex. Samples (10 mL) were injected and eluted with
a mobile phase of 5 mM H2SO4 at an isocratic flow rate of
0.4 mLminÀ1. The column temperature was maintained at 308C. d-
Galacturonic acid and mucic acid were measured at 210 nm by
using a SPD-M20A photodiode array UV/Vis detector (PDA), where-
[21] K. A. Datsenko, B. L. Wanner, Proc. Natl. Acad. Sci. USA 2000, 97, 6640–
6645.
ChemCatChem 2016, 8, 1500 – 1506
www.chemcatchem.org
1505
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communications
[22] a) V. Micard, C. M. G. C. Renard, J. F. Thibault, Enzyme Microb. Technol.
1996, 19, 162–170; b) M. Spagnuolo, C. Crecchio, M. D. R. Pizzigallo, P.
Ruggiero, Bioresour. Technol. 1997, 60, 215–222.
[24] a) D. Bçttcher, U. T. Bornscheuer, Curr. Opin. Microbiol. 2010, 13, 274–
282; b) R. E. Cobb, R. Chao, H. Zhao, AIChE J. 2013, 59, 1432–1440.
[23] a) J. M. Thomas, R. Raja, B. F. G. Johnson, T. J. O’Connell, G. Sankar, T. Khi-
myak, Chem. Commun. 2003, 1126–1127; b) X. She, H. M. Brown, X.
Zhang, B. K. Ahring, Y. Wang, ChemSusChem 2011, 4, 1071–1073.
Received: January 20, 2016
Revised: February 13, 2016
Published online on March 15, 2016
ChemCatChem 2016, 8, 1500 – 1506
www.chemcatchem.org
1506
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim