Whole-cell biocatalytic selective oxidation of 5-hydroxymethylfurfural to 5-hydroxymethyl-2-furancarboxylic acid

Article abstract of DOI:10.1039/c7gc01751k

Currently, the catalytic upgradation of 5-hydroxymethylfurfural (HMF), an important bio-based platform chemical, is of great interest. In this work, we reported the biocatalytic oxidation of HMF to 5-hydroxymethyl-2-furancarboxylic acid (HMFCA) by using a newly isolated strain Comamonas testosteroni SC1588. Compared to growing cells, resting cells were found to be a better biocatalyst for the synthesis of HMFCA. This strain exhibited high tolerance to HMF (up to 180 mM). However, its catalytic performances were highly sensitive to pH. The product exerted a significant negative effect on the catalytic performances and viability of the cells, partially due to the acidity of this compound. The product inhibition and toxicity toward this strain were reduced considerably after adjusting the pH of the reaction mixtures to 7.0. Excellent substrate conversions (approximately 100%) and good HMFCA yields (88-99%) were obtained when the substrate concentrations were less than 130 mM. In addition, the HMFCA synthesis could be improved significantly by the combination of histidine addition with pH tuning at higher substrate concentrations. Catalytic activities of the cells increased markedly when they were cultivated in the presence of a low concentration of furfural and furfuryl alcohol. HMFCA was afforded in a yield of approximately 98% after 36 h at a substrate concentration of up to 160 mM. Besides, this strain was capable of selectively transforming other furfurals to the target acids with good yields of 90-93%.

Full text of DOI:10.1039/c7gc01751k

Green Chemistry
View Article Online
View Journal
PAPER
Whole-cell biocatalytic selective oxidation of
5-hydroxymethylfurfural to 5-hydroxymethyl-
Cite this: DOI: 10.1039/c7gc01751k
2-furancarboxylic acid†
Xue-Ying Zhang, Min-Hua Zong and Ning Li
*
Currently, the catalytic upgradation of 5-hydroxymethylfurfural (HMF), an important bio-based platform
chemical, is of great interest. In this work, we reported the biocatalytic oxidation of HMF to 5-hydroxy-
methyl-2-furancarboxylic acid (HMFCA) by using a newly isolated strain Comamonas testosteroni
SC1588. Compared to growing cells, resting cells were found to be a better biocatalyst for the synthesis
of HMFCA. This strain exhibited high tolerance to HMF (up to 180 mM). However, its catalytic perform-
ances were highly sensitive to pH. The product exerted a significant negative effect on the catalytic per-
formances and viability of the cells, partially due to the acidity of this compound. The product inhibition
and toxicity toward this strain were reduced considerably after adjusting the pH of the reaction mixtures
to 7.0. Excellent substrate conversions (approximately 100%) and good HMFCA yields (88–99%) were
obtained when the substrate concentrations were less than 130 mM. In addition, the HMFCA synthesis
could be improved significantly by the combination of histidine addition with pH tuning at higher sub-
strate concentrations. Catalytic activities of the cells increased markedly when they were cultivated in the
presence of a low concentration of furfural and furfuryl alcohol. HMFCA was afforded in a yield of
approximately 98% after 36 h at a substrate concentration of up to 160 mM. Besides, this strain was
capable of selectively transforming other furfurals to the target acids with good yields of 90–93%.
Received 15th June 2017,
Accepted 4th August 2017
DOI: 10.1039/c7gc01751k
rsc.li/greenchem
1. Introduction
In the last decade, the production of bio-based fuels and plat-
form chemicals from biomass has attracted growing
interest.1–4 5-Hydroxymethylfurfural (HMF) that can be syn-
thesized via carbohydrate dehydration5 is one of the most
important bio-based chemicals.6 It has the essential reactivity
for its upgradation into a variety of useful intermediates,7,8
due to the presence of active groups such as primary hydroxyl
and formyl. For example, it could be transformed into 2,5-bis
(hydroxymethyl)furan (BHMF) and 2,5-dimethylfuran via selec-
tive hydrogenation,9 while the selective oxidation of HMF
could furnish 2,5-diformylfuran (DFF), 5-hydroxymethyl-2-
furancarboxylic acid (HMFCA), 5-formyl-2-furancarboxylic acid
(FFCA) and 2,5-furandicarboxylic acid (FDCA) (Fig. 1). HMFCA
that is the oxidation product of the formyl group in HMF is a
versatile building block for the synthesis of polyesters10 as well
as an interleukin inhibitor.11 In addition, it was reported that
HMFCA had an antitumor activity.12
Fig. 1 HMF and its value-added derivatives.
Presently, chemical approaches are playing a dominant role
in the oxidation of HMF to HMFCA.8 Two groups reported the
synthesis of HMFCA from HMF via the Cannizzaro reac-
tion;13,14 nonetheless, the theoretical selectivity of the desired
product was 50% because of the formation of an equimolar
State Key Laboratory of Pulp and Paper Engineering, School of Food Science and
Engineering, South China University of Technology, 381 Wushan Road,
Guangzhou 510640, China. E-mail: lining@scut.edu.cn; Fax: +86 20 2223 6669;
Tel: +86 20 2223 6669
†Electronic supplementary information (ESI) available. See DOI: 10.1039/c7gc01751k by-product (BHMF), although good yields (>80%) were
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
obtained. Various noble metal catalysts were developed for C. testosteroni SC1588 are available in the ESI.† Based on the
HMFCA synthesis.15–17 Zhang et al. reported that HMF was analysis of 16S ribosomal DNA (available in the ESI,† which
selectively oxidized to HMFCA in toluene within 3 h with an has been submitted to GenBank with accession no. KX961682)
immobilized molybdenum acetylacetonate complex, with a as well as its physiological and biochemical characteristics
yield of approximately 87%.18 Recently, Han and co-workers (Table S1, available in the ESI†), the strain was identified as
reported a selective and mild photocatalytic approach toward C. testosteroni. The phylogenetic tree of this strain is shown in
carboxyl compounds over Au/TiO2, where HMFCA was syn- Fig. S1 (available in the ESI†). C. testosteroni SC1588 (CCTCC
thesized from HMF under ultraviolet and visible light in no. M2016562) was maintained in the China Center for Type
aqueous Na2CO3 solution, with the yields of 90–95%.19
Culture Collection (CCTCC, Wuhan, China).
Biocatalytic oxidation represents a promising option to sup-
HMF (98%), FFCA (98%) and FDCA (97%) were purchased
plement and even replace chemical methods,20–22 because of from J&K Scientific Ltd (Guangzhou, China). HMFCA (98%)
many advantages such as excellent selectivity, mild reaction was bought from Adamas Reagent Ltd (Shanghai, China).
conditions, high efficiency and being environmentally friendly. BHMF (98%) and furfural (99%) were obtained from Macklin
Nevertheless, the biocatalytic oxidation of HMF received little Biochemical Co., Ltd (Shanghai, China). Furfuryl alcohol
attention previously,7,8,23 most of which involved the synthesis (98%), DFF (>98%), furoic acid (98%) and 5-methylfurfural
of FDCA.24–31 To our knowledge, there are few reports on the (97%) were purchased from TCI (Japan). 5-Methylfurfuryl
biocatalytic synthesis of HMFCA in the literature.28,32,33 Krystof alcohol (98%) was purchased from Apollo Scientific Ltd (UK).
et al. described the lipase-mediated oxidation of HMF using 5-Methyl-2-furoic acid (97%) was obtained from Sigma-Aldrich
H2O2 as the oxidant in a mixture of t-butanol and ethyl acetate, (USA). Propidium iodide (PI, 50 μg mL−1) staining solution was
which furnished HMFCA in a yield of approximately 80%, purchased from BD Biosciences (USA).
together with 20% of HMFCA acetate.32 Our group found that
xanthine oxidase was an efficient catalyst for the synthesis of
2.2. Cultivation of C. testosteroni SC1588 cells
HMFCA from HMF in buffer, and the desirable product was C. testosteroni SC1588 cells were pre-cultivated at 30 °C and
afforded with a yield of 94% and a selectivity of >99% after 150 rpm for 12 h in 1.8% nutrient broth. Then, 1% seed
7 h.28 To continue our interest in biocatalytic upgradation of culture was inoculated into the fresh nutrient broth (1.8%).
HMF,28,34,35 we reported the whole-cell biocatalytic oxidation of After incubation at 30 °C and 150 rpm for 5 h, the cells were
HMF to HMFCA in this work (Scheme 1). In fact, the whole-cell harvested by centrifugation (8000g, 5 min, 4 °C) and washed
biocatalytic transformation of HMF remains challenging, since twice with 0.85% NaCl solution, followed by dispersion in
the substrate HMF is a well-known potent inhibitor to micro- phosphate buffer to give a cell concentration of 30 mg (cell wet
organisms.36,37 In addition, the over-oxidation reaction(s) weight) per mL.
readily occurred in the whole-cell biotransformation of
2.3. General procedure for biocatalytic oxidation of HMF
HMF,29–31,38 due to the presence of complex enzyme systems in
the cells. Herein, we isolated a new bacterial strain (Comamonas Typically, 4 mL of phosphate buffer (200 mM, pH 7.0) contain-
testosteroni SC1588) from soil samples, which is not only highly ing 40 mM HMF and 30 mg (cell wet weight) per mL microbial
tolerant to HMF, but also shows good selectivity toward the cells was incubated at 30 °C and 150 rpm. Aliquots were with-
formyl group. The catalytic performances and substrate- and drawn from the reaction mixtures at specified time intervals
product-tolerance of this strain were evaluated in the synthesis and diluted with the corresponding mobile phase prior to
of HMFCA. And various strategies were applied for improving HPLC analysis. The initial reaction rate (V0) was calculated
HMFCA synthesis at high substrate concentrations. Besides, the based on the decrease in HMF concentrations in the initial
substrate spectrum of the new biocatalyst was studied.
reaction stage (in general 30 min). The conversion was defined
as the percentage of the consumed HMF amount in the initial
HMF amount. The yield was defined as the percentage of the
measured product amount in the theoretical product amount
based on the initial amount of HMF. All the experiments were
conducted at least in duplicate, and the values were expressed
as the means standard deviations.
2. Experimental
2.1. Biological and chemical materials
C. testosteroni SC1588 was isolated from soil samples obtained
from the grounds of University Town (Guangzhou, China). The
experimental details for the isolation and screening of
2.4. HPLC analysis
The reaction mixtures were analyzed on a Zorbax Eclipse
XDB-C18 column (4.6 mm × 250 mm, 5 μm, Agilent, USA) by
using a reversed phase HPLC equipped with a Waters 996
photodiode array detector (Waters, USA). Unless otherwise
specified, the mobile phase for HPLC analysis was a mixture of
acetonitrile/0.4% (NH4)2SO4 solution with pH 3.5 (10 : 90, v/v)
at a flow rate of 0.6 mL min−1. The retention times of FDCA
(maximum absorption wavelength: 264 nm), FFCA (286 nm),
Scheme 1 Biocatalytic synthesis of HMFCA from HMF.
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
HMFCA (252 nm), BHMF (223 nm), HMF (283 nm) and DFF
(288 nm) were 4.79, 5.96, 6.14, 8.08, 9.74 and 12.59 min,
respectively. The retention times of furoic acid (245 nm), fur-
furyl alcohol (216 nm) and furfural (278 nm) were 9.21, 14.29
and 16.29 min, respectively. The mobile phase for analyzing
the reaction mixtures in 5-methylfurfural biotransformation is
a mixture of acetonitrile/0.02% phosphoric acid solution
(40 : 60, v/v) at a flow rate of 0.3 mL min−1. The retention times
of 5-methyl-2-furoic acid (264 nm), 5-methylfurfuryl alcohol
(221 nm) and 5-methylfurfural (289 nm) were 12.08, 13.05 and
15.67 min, respectively. HPLC spectra are shown in Fig. S2
(available in the ESI†).
2.5. Cell viability assay
Cell viability was measured by PI uptake and flow cytometry.39
Briefly, the cells were harvested after incubation for 12 h under
the designated conditions and diluted to 105 colony-forming
units per mL with phosphate buffer salt (pH 7.2). The cells
were stained with 2.5 μg mL−1 PI for 15 min in the dark, and
the cell viability was assayed using a Gallios flow cytometer
(Beckman Coulter Inc., USA). The fluorescence emission was
observed when excited at 488 nm and was recorded at 630 nm.
Data were acquired and analyzed using the Kaluza software.
The cell viability was expressed as the percentage of the cells
unstained with PI in the total cells.
Fig. 2 Time courses of HMF transformation with growing (open
symbols) and resting cells (solid symbols). Reaction conditions: 40 mM
HMF, 30 mg mL−1 microbial cells, 4 mL phosphate buffer (200 mM, pH
7.0), 150 rpm, 35 °C. 0.9% nutrient broth was supplemented when
growing cells were used for biotransformation.
presence of nutrients (especially carbon sources),40 thus facili-
tating the synthesis of the reduced product. Then, BHMF was
gradually re-oxidized into HMFCA, along with the formation of
a minor amount of an over-oxidized product (FDCA). As shown
in Fig. 2, the yield of the desired product HMFCA using resting
cells was much higher after the reaction of 11 h than that
using growing cells (94% vs. 73%). So promoting the re-oxi-
dation of BHMF was critical to enhance the HMFCA synthesis
catalyzed by this strain. We attempted to supplement the
reduced co-substrate (acetone) to facilitate the oxidation reac-
tions (data not shown). Unfortunately, no positive results were
obtained, since HMFCA was afforded with a yield of 81%,
together with the production of 11% BHMF.
2.6. Scale-up synthesis of HMFCA
Whole cells were prepared as described in section 2.2., with
the exception of the cultivation process in which 1% seed
culture was inoculated into 1.8% nutrient broth containing
5 mM furfuryl alcohol. Fifty mL of phosphate buffer (200 mM,
pH 7.0) containing 160 mM HMF, 20 mM histidine and 30
mg mL−1 cells was incubated at 30 °C and 150 rpm. The pH
was tuned to approximately 7.0 every 24 h. After the reaction,
the reaction mixture was acidified to approximately pH 1.0 by
adding 50% H2SO4 solution, followed by extraction 4 times
with ethyl acetate. The organic solvent was removed through
evaporation, affording the crude product.
3.2. Effects of temperature and pH on HMFCA synthesis
The effects of temperature and pH on the biocatalytic oxi-
dation of HMF were studied (Fig. 3). Fig. 3A shows the influ-
ence of the reaction temperature on the HMF oxidation when
the temperature varies from 25 to 45 °C. It was found that
HMF was consumed completely after the reaction of 11 h at
25–40 °C. However, the maximal substrate conversion
decreased to 83% at 45 °C, possibly due to the thermal inacti-
vation of the biocatalyst at this temperature. Good HMFCA
3. Results and discussion
3.1. Comparing the catalytic performances of growing and
resting cells
The catalytic performances of growing and resting cells were yields (approximately 90%) were obtained within the tempera-
compared in the oxidation of HMF (Fig. 2). As shown in Fig. 2, ture range of 25–40 °C. Nevertheless, the yield (53%) was pretty
growing cells showed a much higher activity in the oxidation low at 45 °C. In addition to incomplete substrate transform-
of HMF, compared to resting cells. The substrate was almost ation, no occurrence of BHMF (around 17%) re-oxidation
used up within 5 h, regardless of the cell status. Two major might account for the poor HMFCA yield at this temperature.
products (HMFCA and BHMF) were simultaneously formed in Considering the cell stability, the temperature of 30 °C was
the initial stage. In addition, the reduced product BHMF used for the subsequent studies.
seemed to be synthesized more readily in the presence of
Fig. 3B shows the effect of pH on the HMF oxidation. It was
nutrients. For example, the maximal yield of BHMF was found that pH had a significant effect on the biocatalytic oxi-
approximately 41% with growing cells, while being 14% with dation of HMF. Poor results were obtained at pH 5.0. Both sub-
resting cells (Fig. 2). The reason might be that reduced cofac- strate conversions and product yields were improved signifi-
tors (e.g., NAD(P)H) could be regenerated efficiently in the cantly with the increment of buffer pH. The best results (the
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
3.3. Inhibitory and toxic effect of the substrate
The HMF-tolerance level of microbial cells is key for the whole-
cell biocatalytic upgradation of HMF, because HMF is a well-
known toxic and inhibitory substance.36 So, the tolerance of
this strain toward HMF was examined (Fig. 4). As shown in
Fig. 4A, the highest initial reaction rate (approximately
31 mM h−1) was observed when the substrate concentration
was 100 mM. This suggests that there exists no substrate in-
hibition toward the biocatalytic oxidation at this concen-
tration. Interestingly, further increases in the substrate concen-
trations only resulted in moderate substrate inhibition, since
the initial reaction rates (24–29 mM h−1) remained good at the
HMF concentrations of 110–150 mM. Also, it could be verified
by excellent substrate conversions (approximately 100%) in all
cases. HMFCA was furnished with the yields of 92–99% when
the substrate concentrations were less than 110 mM. In spite
Fig. 3 Effects of temperature (A) and pH (B) on HMFCA synthesis.
Reaction conditions for (A): 40 mM HMF, 30 mg mL−1 microbial cells,
4 mL phosphate buffer (200 mM, pH 7.0), 150 rpm, the designated
temperature, 11 h; reaction conditions for (B): 40 mM HMF, 30 mg mL−1
microbial cells, 4 mL buffer [200 mM, pH 5.0–8.0 phosphate buffer (PB),
pH 8.0–9.0 Tris-HCl buffer, pH 9.0 glycine-NaOH buffer (Gly-NaOH)],
150 rpm, 30 °C, reaction periods of 11 h for pH 5.0–8.0 and 24 h for
pH 9.0.
conversions of up to 100% and the yields of up to 94%) were
observed in phosphate buffer of pH 7.0 and 8.0. The buffer
types seemed to have a moderate effect on the synthesis of
HMFCA, because the yield was approximately 92% in phos-
phate buffer (pH 8.0), while being 82% in Tris-HCl buffer of
the same pH. However, the conversions (41–63%) as well as
the yields (28–31%) decreased markedly when the pH was
increased to 9.0, regardless of the buffer type. It was interest-
ingly found that the synthesis of HMFCA was improved
considerably (the yield of 82%) when the concentration of
Fig. 4 Effects of the substrate concentrations on HMFCA synthesis (A)
and cell viability (B). Reaction conditions for (A): HMF of the designated
concentration, 30 mg mL−1 microbial cells, 4 mL phosphate buffer
(200 mM, pH 7.0), 150 rpm, 30 °C, reaction periods (12, 12, 18, 24, 36, 42,
Tris-HCl was reduced from 200 mM to 50 mM (Fig. S3, avail- 48, 60, 60 and 60 h, respectively, with gradually increasing HMF concen-
trations from 20 to 200 mM); conditions for (B): after microbial cells of
able in the ESI†). And almost no substrate transformation
occurred at pH 10.0 (data not shown). Compared to
30 mg mL−1 were incubated for 12 h in phosphate buffer (200 mM, pH
7.0) containing HMF of the designated concentration at 150 rpm and
30 °C, cell viability was measured. Control 1: cells were incubated in
buffer without HMF at 150 rpm and 30 °C for 12 h; control 2: cells stood
Meyerozyma guilliermondii SC1103 that was capable of selec-
tively reducing HMF to BHMF,35 this strain exhibited a much
narrower pH-tolerance range.
in buffer without HMF at 4 °C for 12 h; control 3: fresh cells harvested.
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
of the presence of moderate substrate inhibition, good HMFCA
yields (88–96%) were achieved at the substrate concentrations
of 110–130 mM. However, its yield decreased significantly to
approximately 70% at the substrate concentration of 150 mM.
The major reason for the unsatisfactory yield is that BHMF
could not be effectively re-oxidized into HMFCA by the cells.
There was approximately 30% BHMF in the reaction mixture
after 60 h, and a significant decrease in its concentrations was
not observed with the prolongation of the reaction period.
Therefore, it may be critical to promote BHMF oxidation for
improving HMFCA synthesis at high substrate concentrations.
Poor results were observed when the substrate concentration
was 200 mM (Fig. 4A). And even, almost no transformation of
HMF took place at 300 mM (data not shown).
Fig. 4B shows the effect of the HMF concentrations on cell
viability. Unexpectedly, cell viability (approximately 13%) was
very low after the cells were incubated for 12 h in phosphate
buffer at 30 °C. However, cell viability was improved to approxi-
mately 75% when the cells were incubated at 4 °C for 12 h,
which is almost comparable to that of the cells harvested
freshly (80%). The reason might be that cellular metabolism
became pretty poor at low temperature, thus preventing the
bacterial decay. Interestingly, good cell viability (74–80%) was
observed after incubation in the presence of 20–90 mM HMF,
which is comparable to those in the controls (controls 2 and
3). This suggests that HMF almost exerts no toxic effect on the
Fig. 5 Effects of the product concentrations on HMFCA synthesis and
cell viability: (A) without pH tuning; (B) after pH tuning. Reaction con-
ditions for (A): 50 mM HMF, HMFCA of the designated concentration,
30 mg mL−1 microbial cells, 4 mL phosphate buffer (200 mM, pH 7.0),
150 rpm, 30 °C; after 30 mg mL−1 microbial cells were incubated for
12 h in phosphate buffer (200 mM, pH 7.0) containing HMFCA of the
designated concentration at 150 rpm and 30 °C, cell viability was
measured. Reaction conditions for (B) were similar to the above ones,
with the exception that the pH of the reaction mixture was tuned to 7.0
by adding NaOH prior to reaction or incubation.
cells at less than 90 mM. On the other hand, cell viability in cates that there is remarkable product inhibition when the
the presence of less than 90 mM HMF was much higher than product concentration is more than 150 mM. Interestingly, the
that in the absence of HMF (control 1), indicating that the product had no significant toxicity toward the cells at less than
toxic feature of HMF may stimulate the cells to significantly 150 mM, because of high cell viability (53–70%), which is com-
improve their resistance against environmental stress. And parable to or slightly lower than that in the control (Fig. 5A).
even, HMF of up to 130 mM only resulted in slight toxicity, However, the product of more than 200 mM resulted in great
because cell viability (64%) remained high at this concen- toxicity. The significantly detrimental effect of the product on
tration. When the substrate concentration was increased to the biocatalyst might be partially attributed to its acidity,
150 mM, moderate substrate toxicity was observed. And HMF because of the presence of the carboxylic acid group in this
of 300 mM showed significant toxicity. Based on the above molecule. In fact, decreases in pH values occurred after adding
studies on substrate inhibition and toxicity, this strain was tol- HMFCA into the reaction mixtures (Fig. S4A†), although phos-
erant to HMF of at least 150 mM. Compared to the microor- phate buffer of up to 200 mM was used. In particular, the pH
ganisms reported previously,33,41 C. testosteroni SC1588 reduced markedly to 3.3–3.9 in the presence of 200–300 mM
appeared to be much more advantageous for the synthesis of HMFCA. Fig. S4B† shows the effects of pH on the catalytic
HMFCA from HMF, because of its high tolerance toward the activities and viability of the cells. Slight to moderate decreases
substrate (<10 mM vs. >150 mM) as well as good catalytic in the catalytic activities were observed at pH 5.7–6.9, com-
performances.
pared to that at pH 7.0. However, the cells displayed low activi-
ties as well as poor viability at less than pH 4.0. These results
were in good agreement with those shown in Fig. 3B.
3.4. Inhibitory and toxic effect of the product
In addition to the substrate, the product might have a negative
A significant effect of pH on the catalytic performances of
effect on this biocatalytic oxidation, since HMFCA is an acidic the cells has been demonstrated above. So the base NaOH was
compound. It has been demonstrated that the catalytic per- added into the reaction mixtures to neutralize HMFCA, thus
formances of this strain closely depend on pH (Fig. 3B). avoiding pH changes due to the presence of this acidic sub-
Hence, the inhibition and toxicity of the product toward the stance. The pH values of the reaction mixtures were tuned to
whole-cell biocatalyst were studied (Fig. 5). As shown in 7.0 in all cases (Fig. 5B). Interestingly, the product showed sig-
Fig. 5A, the initial reaction rates decreased significantly with nificantly reduced inhibition toward the cells at less than
the increasing product concentrations. The biocatalyst 150 mM after pH tuning, compared to that without pH tuning.
retained approximately 30% of the original activity when the In addition, a considerable decrease in the toxicity of the
product concentration was 150 mM, and became almost com- product was observed when its concentration was up to
pletely inactivated at more than 200 mM HMFCA. This indi- 300 mM, although severe product inhibition remained
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
(Fig. 5B). This suggests that controlling pH may be an effective
measure to relieve the negative effect of the product on the
whole-cell biocatalytic oxidation of HMF, thus significantly
improving HMFCA synthesis.
3.5. Improved synthesis of HMFCA
As described above, the HMFCA yields decreased significantly
when the substrate concentrations were more than 150 mM
(Fig. 4A). Based on the physiological and biochemical pro-
perties of this strain (Table S1†), it could grow on some amino
acids such as histidine. Hence, the addition of amino acids
might enhance the HMF detoxification of microbial cells, thus
facilitating the oxidation reactions. Also, the pH of the reaction
mixture decreased with the accumulation of the oxidized pro-
ducts (e.g., HMFCA), which exerted a detrimental effect on the
synthesis of the desired product (Fig. 3B and S4†). It was
reported that whole-cell biocatalytic oxidation reactions would
be improved significantly by enhancing oxygen supply.42,43
Therefore, three strategies (amino acid addition, pH tuning
and air bubbling) were applied to enhance the HMFCA syn-
thesis (Fig. S5A and B†). pH tuning proved to be effective for
improving the yield of the desired product (94% vs. 85% in the
control). Compared to the control, most amino acids were
capable of enhancing HMFCA synthesis; in addition, the
HMFCA yields depended on amino acids. Of the amino acids
Fig. 6 Improved synthesis of HMFCA. Reaction conditions: HMF of the
designated concentration, 30 mg mL−1 microbial cells, 4 mL phosphate
buffer (200 mM, pH 7.0), 20 mM histidine, 150 rpm, 30 °C, tuning pH of
the reaction mixture to approximately 7.0 every 24 h.
attributed to strongly detrimental effects of both the substrate
and product on the biocatalyst.
3.6. Scale-up synthesis of HMFCA
tested, the highest yield (91%) was obtained in the case of his- Although good yields were obtained at high substrate concen-
tidine. OD600 and pH values were monitored (Fig. S5C and trations (130–160 mM), the reaction periods (60 h) remained
D†). It was found that OD600 values decreased in all cases, long. To facilitate the HMFCA synthesis, inducible expression
indicating the toxicity of the substrate and product. After 60 h, of the enzyme(s) responsible for BHMF and HMF oxidation
comparable OD600 values were found in the control and in the was attempted by supplementing furfural and furfuryl alcohol
case of histidine (Fig. S5C†), indicating that histidine seems of a low concentration (5 mM) during cultivation (Fig. S7†).
not to work as a nutrient for cell growth. Nonetheless, the pH Indeed, the induced cells showed higher catalytic activities,
in the case of histidine was higher than that in the control compared to the control (Fig. S7†). The reaction periods
(6.47 vs. 6.06) after 60 h (Fig. S5D†), which may contribute to a decreased significantly from 60 h to approximately 36 h with
higher yield. In addition, cell viability seemed to be higher in the induced cells as biocatalysts, in spite of the comparable
the presence of histidine (Fig. S6†), which may also partially maximal HMFCA yields in all cases (72–76%). It was worth
account for the better results. However, air bubbling appeared noting that the cell biomass reduced markedly during cultiva-
not to be effective for improving the HMFCA synthesis. The tion in the presence of furfural, likely due to the significant
reason for this might be that oxygen supply by the method toxicity of this inhibitor.36,37 So furfuryl alcohol-induced cells
applied in this work is inefficient in flasks.
were used for the subsequent scale-up synthesis of HMFCA.
Fig. 7 shows the scale-up synthesis of HMFCA. It was found
So the combination of histidine addition and pH tuning
was applied for improving the synthesis of HMFCA (Fig. 6). As that the substrate of 160 mM was transformed completely in
shown in Fig. 6, a high HMFCA yield (approximately 90%) was 12 h with furfuryl alcohol-induced cells as a biocatalyst, which
obtained when the substrate concentration was up to 160 mM, is much faster than that in the control (24 h). More interest-
indicating the effectiveness of the combined strategy. In ingly, the desired product HMFCA was obtained with an excel-
addition, the substrate was completely transformed after 120 h lent yield (approximately 98%) after 36 h, along with the pro-
when its concentration was up to 180 mM (data not shown). duction of around 2% BHMF and trace FDCA. Compared to
Nonetheless, a low substrate conversion (41%) was obtained at those in the control (without induction, Fig. 6), both the
200 mM. This suggests that the HMF-tolerance level of this HMFCA yield (90% vs. 98%) and the reaction rate (reaction
strain is up to 180 mM, which is much higher than that period, 60 h vs. 36 h) were much higher with the induced cells.
(110 mM) of M. guilliermondii SC1103.35 To our knowledge, it After the reaction, the isolation of the desired product was con-
is the highest HMF-tolerance level of microorganisms ever ducted. The pH of the reaction mixture was tuned to approxi-
reported. However, the HMFCA yield (52%) remained unsatis- mately 1.0, followed by extraction 4 times with an equal
factory at the substrate concentration of 180 mM, because of volume of ethyl acetate. The removal of the solvent via evapor-
the formation of almost equimolar BHMF. The low catalytic ation afforded the crude product with an isolated yield of 99%
activity of the cells in the presence of 200 mM HMF might be and the purity of 94%.
Green Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Green Chemistry
Paper
furoic acid that has applications in the pharmaceutical, agro-
chemical, flavour and fragrance industries44 was obtained with
a good yield (93%); however, the oxidation rate of furfural was
much lower than that of HMF (the periods for reaching the
maximal yields: 96 h vs. 18 h). The reason may be that furfural
is more toxic toward microbes than HMF.37,45 Also, 5-methyl-
furfural could be oxidized into 5-methyl-2-furoic acid with a
yield of 90% after 120 h. When DFF acted as the substrate,
FDCA was afforded as the major product, along with the pro-
duction of minor HMF. The FDCA yield of 61% was obtained
after 48 h. FDCA is a useful starting material for the pro-
duction of bio-based polymers such as polyamides, polyesters,
and polyurethanes.46 However, the product distribution
changed significantly when the concentration of DFF was
reduced to 20 mM; after 24 h, DFF was transformed completely
into a mixture of equimolar FDCA and HMFCA (data not
shown). Interestingly, FFCA was a good substrate for this
strain. As shown in Table 1, it could be transformed quickly
into FDCA in a good yield (92%).
Fig. 7 Scale-up synthesis of HMFCA. Reaction conditions: 160 mM
HMF, 30 mg mL−1 microbial cells (which were cultivated in the presence
of 5 mM furfuryl alcohol), 20 mM histidine, 50 mL phosphate buffer
(200 mM, pH 7.0), 150 rpm, 30 °C, tuning the pH of the reaction mixture
to approximately 7.0 every 24 h.
Compared to chemical routes,15–19 this biocatalytic
approach to HMFCA requires much longer reaction periods,
and thus the reaction rates need to be improved significantly,
although the latter is free of using volatile organic solvents
and environmentally unfriendly catalysts as well as harsh reac-
tion conditions. Xanthine oxidase proved to be an excellent
biocatalyst for the synthesis of HMFCA from HMF, but its sub-
strate spectrum appeared to be pretty narrow, since it could
almost not accept DFF as a substrate.28 Compared to the
lipase-catalyzed chemoenzymatic approach,32 this whole-cell
biocatalytic method showed a much higher selectivity (80% vs.
98%). In addition, the use of toxic H2O2 was avoided herein.
Nonetheless, isolated enzymes, especially those cofactor-inde-
pendent ones, are still promising candidates for the transform-
ations of the toxic substrate HMF, since the toxic effect of
HMF on biocatalysts could be relieved significantly. Currently,
the isolation and characterization of the enzyme(s) responsible
for HMF oxidation are underway in our laboratory.
4. Conclusions
In summary, we have successfully developed an efficient and
selective approach for the synthesis of HMFCA from HMF by
using highly HMF-tolerant C. testosteroni SC1588 cells. An
excellent yield of HMFCA was obtained when the substrate
concentration was up to 160 mM. In addition, scale-up syn-
thesis and simple isolation of HMFCA were realized, which
may be promising to envisage an industrially sound biocataly-
tic process. Medium engineering (e.g., biphasic solvent
systems)47,48 may be a promising approach for improving
HMFCA synthesis at higher substrate concentrations, where
the water-immiscible solvent works as a reservoir for the sub-
strate and product, thus significantly reducing their negative
effects on whole-cell biocatalytic oxidation. Also, the knockout
of the enzyme(s) responsible for the reduction of HMF by
metabolic engineering strategies30 may be an alternative route,
which would block the BHMF formation and thus significantly
improve the HMFCA synthesis. In addition, this strain may
have promising application potential in the biodetoxification
3.7. Oxidation of furfurals
The substrate spectrum of this new biocatalyst was examined of lignocellulosic hydrolysates,49,50 because of its high detoxifi-
(Table 1). It was found that this strain could accept all furfurals cation efficiency and HMF tolerance. More importantly, it
tested as the substrates. However, its catalytic activities were can’t utilize both glucose and xylose as carbon sources, thus
highly correlated with the substrate structures. For example, avoiding the sugar losses during biodetoxification.
Table 1 Whole-cell biocatalytic oxidation of furfurals
Substrate
Product
Time (h)
Conversion (%)
Yield (%)
Furfural
5-Methylfurfural
DFF
Furoic acid
5-Methyl-2-furoic acid
FDCA
96
120
48
100
100
81.6 2.5
100
93.0 2.7
90.2 1.8
61.3 3.0
92.1 0.1
FFCA
FDCA
12
Reaction conditions: 50 mM furfurals, 30 mg mL−1 microbial cells, 4 mL phosphate buffer (200 mM, pH 7.0), 20 mM histidine, 150 rpm, 30 °C.
This journal is © The Royal Society of Chemistry 2017
Green Chem.

View Article Online
Paper
Green Chemistry
23 M. van Deurzen, F. van Rantwijk and R. Sheldon,
Acknowledgements
J. Carbohydr. Chem., 1997, 16, 299–309.
This research was financially supported by the State Key 24 J. Carro, P. Ferreira, L. Rodriguez, A. Prieto, A. Serrano,
Laboratory of Pulp and Paper Engineering (2017C01), the
National Natural Science Foundation of China (21676103) and
B. Balcells, A. Arda, J. Jimenez-Barbero, A. Gutierrez and
R. Ullrich, FEBS J., 2015, 282, 3218–3229.
the Fundamental Research Funds for the Central Universities 25 W. P. Dijkman, D. E. Groothuis and M. W. Fraaije, Angew.
(2017ZD065).
Chem., Int. Ed., 2014, 53, 6515–6518.
26 S. M. McKenna, S. Leimkuhler, S. Herter, N. J. Turner and
A. J. Carnell, Green Chem., 2015, 17, 3271–3275.
27 W. P. Dijkman, C. Binda, M. W. Fraaije and A. Mattevi, ACS
Catal., 2015, 5, 1833–1839.
References
1 R. A. Sheldon, Catal. Today, 2011, 167, 3–13.
2 C. O. Tuck, E. Pérez, I. T. Horváth, R. A. Sheldon and
M. Poliakoff, Science, 2012, 337, 695–699.
3 R. A. Sheldon, Green Chem., 2014, 16, 950–963.
4 A. J. Straathof, Chem. Rev., 2014, 114, 1871–1908.
5 S. P. Teong, G. Yi and Y. Zhang, Green Chem., 2014, 16,
2015–2026.
6 J. J. Bozell and G. R. Petersen, Green Chem., 2010, 12, 539–
554.
7 R.-J. van Putten, J. C. van der Waals, E. de Jong,
28 Y. Z. Qin, Y. M. Li, M. H. Zong, H. Wu and N. Li, Green
Chem., 2015, 17, 3718–3722.
29 F. Koopman, N. Wierckx, J. H. de Winde and
H. J. Ruijssenaars, Bioresour. Technol., 2010, 101, 6291–6296.
30 G. S. Hossain, H. Yuan, J. Li, H.-d. Shin, M. Wang, G. Du,
J. Chen and L. Liu, Appl. Environ. Microbiol., 2017, 83, e02312-16.
31 C.-F. Yang and C.-R. Huang, Bioresour. Technol., 2016, 214,
311–318.
32 M. Krystof, M. Perez-Sanchez and P. Dominguez de Maria,
ChemSusChem, 2013, 6, 826–830.
C. B. Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev., 33 K. Mitsukura, Y. Sato, T. Yoshida and T. Nagasawa,
2013, 113, 1499–1597. Biotechnol. Lett., 2004, 26, 1643–1648.
8 A. A. Rosatella, S. P. Simeonov, R. F. Frade and 34 Y. Z. Qin, M. H. Zong, W. Y. Lou and N. Li, ACS Sustainable
C. A. Afonso, Green Chem., 2011, 13, 754–793. Chem. Eng., 2016, 4, 4050–4054.
9 M. J. Climent, A. Corma and S. Iborra, Green Chem., 2014, 35 Y. M. Li, X. Y. Zhang, N. Li, P. Xu, W. Y. Lou and
16, 516–547. M. H. Zong, ChemSusChem, 2017, 10, 372–378.
10 H. Hirai, J. Macromol. Sci., Part A: Pure Appl. Chem., 1984, 36 E. Palmqvist and B. Hahn-Hägerdal, Bioresour. Technol.,
21, 1165–1179. 2000, 74, 25–33.
11 A. C. Braisted, J. D. Oslob, W. L. Delano, J. Hyde, 37 S. I. Mussatto and I. C. Roberto, Bioresour. Technol., 2004,
R. S. McDowell, N. Waal, C. Yu, M. R. Arkin and
B. C. Raimundo, J. Am. Chem. Soc., 2003, 125, 3714–3715.
12 M. Munekata and G. Tamura, Agric. Biol. Chem., 1981, 45,
2149–2150.
13 S. Subbiah, S. P. Simeonov, J. M. S. S. Esperanca,
L. P. N. Rebelo and C. A. M. Afonso, Green Chem., 2013, 15,
2849–2853.
93, 1–10.
38 D. Feldman, D. Kowbel, N. Glass, O. Yarden and Y. Hadar,
Biotechnol. Biofuels, 2015, 8, 63.
39 L. C. Crowley, A. P. Scott, B. J. Marfell, J. A. Boughaba,
G. Chojnowski and N. J. Waterhouse, Cold Spring Harb.
Protoc., 2016, DOI: 10.1101/pdb.prot087163.
40 Y. Ni and J.-H. Xu, Biotechnol. Adv., 2012, 30, 1279–1288.
14 E.-S. Kang, D. W. Chae, B. Kim and Y. G. Kim, J. Ind. Eng. 41 N. Terasawa, A. Sugiyama, M. Murata and S. Homma, Food
Chem., 2012, 18, 174–177. Sci. Technol. Res., 2002, 8, 28–31.
15 S. E. Davis, A. D. Benavidez, R. W. Gosselink, J. H. Bitter, 42 X. Zhou, S. Lü, Y. Xu, Y. Mo and S. Yu, Biochem. Eng. J.,
K. P. De Jong, A. K. Datye and R. J. Davis, J. Mol. Catal. A:
Chem., 2014, 388–389, 123–132.
16 S. E. Davis, L. R. Houk, E. C. Tamargo, A. K. Datye and
R. J. Davis, Catal. Today, 2011, 160, 55–60.
17 O. Casanova, S. Iborra and A. Corma, ChemSusChem, 2009,
2, 1138–1144.
2015, 93, 196–199.
43 X. Zhou, X. Zhou, Y. Xu and R. R. Chen, J. Chem. Technol.
Biotechnol., 2017, 92, 1285–1289.
44 R. Mariscal, P. Maireles-Torres, M. Ojeda, I. Sadaba and
M. Lopez Granados, Energy Environ. Sci., 2016, 9, 1144–1189.
45 M. A. Franden, P. T. Pienkos and M. Zhang, J. Biotechnol.,
2009, 144, 259–267.
18 Z. Zhang, B. Liu, K. Lv, J. Sun and K. Deng, Green Chem.,
2014, 16, 2762–2770.
46 Z. Zhang and K. Deng, ACS Catal., 2015, 6529–6544.
19 B. Zhou, J. Song, Z. Zhang, Z. Jiang, P. Zhang and B. Han, 47 Z.-J. Zhang, J. Pan, J.-F. Liu, J.-H. Xu, Y.-C. He and Y.-Y. Liu,
Green Chem., 2017, 19, 1075–1081. J. Biotechnol., 2011, 152, 24–29.
20 W. Kroutil, H. Mang, K. Edegger and K. Faber, Adv. Synth. 48 R. J. Cornmell, C. L. Winder, S. Schuler, R. Goodacre and
Catal., 2004, 346, 125–142. G. Stephens, Green Chem., 2008, 10, 685–691.
21 J. Brummund, T. Sonke and M. Müller, Org. Process Res. 49 D. Zhang, Y. L. Ong, Z. Li and J. C. Wu, Biochem. Eng. J.,
Dev., 2015, 19, 1590–1595. 2013, 72, 77–82.
22 F. Hollmann, I. W. C. E. Arends, K. Buehler, A. Schallmey 50 H. Ran, J. Zhang, Q. Gao, Z. Lin and J. Bao, Biotechnol.
and B. Buhler, Green Chem., 2011, 13, 226–265. Biofuels, 2014, 7, 51.
Green Chem.
This journal is © The Royal Society of Chemistry 2017