While lignocellulose is a promising source of renewable sugars for microbial fermentations, the presence of inhibitory compounds in typical lignocellulosic feedstocks, such as furfural, has hindered their utilisation. In Escherichia coli, a major route of furfural toxicity is the depletion of NADPH pools due to its use as a substrate by the YqhD enzyme that reduces furfural to its less toxic alcohol form. Here, we examine the potential of exploiting benzyl alcohol dehydrogenases as an alternative means to provide this same catalytic function but using the more abundant reductant NADH, as a strategy to increase the capacity for furfural removal. We determine the biochemical properties of three of these enzymes, from Pseudomonas putida, Acinetobacter calcoaceticus, and Burkholderia ambifaria, which all demonstrate furfural reductase activity. Furthermore, we show that the P. putida and B. ambifaria enzymes are able to provide substantial increases in furfural tolerance in vivo, by allowing more rapid conversion to furfuryl alcohol and resumption of growth. The study demonstrates that methods to seek alternative cofactor dependent enzymes can improve the intrinsic robustness of microbial chassis to feedstock inhibitors.

In recent years, lignocellulose has become increasingly apparent as a potential renewable feedstock for bioprocessing [1]. Lignocellulose is a complex material, found in plant cell walls, which consists primarily of three major classes of polymer: cellulose, a polymer of glucose joined by β-(1,4) glycosidic bonds, which forms a crystalline structure; hemicellulose, a heterogeneous group of polysaccharides, typically polymers of d-xylose, d-glucose or l-arabinose with various degrees of substitution; and lignin, a polymer of phenolic compounds [2]. Enzymatic hydrolysis of cellulose and hemicellulose yields sugars, which can be used as a feedstock for microbial growth [1]. However, lignocellulosic biomass is extremely recalcitrant to enzymatic saccharification [3], and must be pre-treated to remove lignin and hemicellulose and to disrupt the cellulose crystal structure [4]. This pre-treatment typically leads to the formation of toxic compounds. One of the more significant of these is furfural, an aromatic, heterocyclic aldehyde which is formed from pentose sugars such as D-xylose when heated in the presence of acids [5]. Thus, furfural is produced from the pentose fraction of hemicellulose when subjected to pre-treatments such as dilute acid [6] or steam explosion [7]; while more severe pre-treatments solubilise greater amounts of hemicellulose, they also result in the production of greater levels of furfural.

A number of routes have been implicated in the mechanism of furfural toxicity [8]. In Saccharomyces cerevisiae [9] and Escherichia coli [10], furfural has been associated with the accumulation of reactive oxygen species. Furfural is also capable of directly damaging DNA in vitro [11], and has been identified to react with DNA at AT sites [12]. This observation could be explained by a proposed reaction between furfural and adenine, forming kinetin [13], which has been shown to promote the mismatching of bases during polymerisation when incorporated into DNA [14]. Some engineering strategies, such as the overexpression of thyA [15] or of polyamine transporters [16], have been proposed to alleviate furfural toxicity by improving DNA repair or by protecting DNA from damage respectively.

In microorganisms, furfural is typically detoxified by conversion into less harmful compounds such as furfuryl alcohol and/or furoic acid (Figure 1) [17–20]. Exposure of different microbes to furfural results in decreases in intracellular NADH and NADPH levels as these are used as substrates for furfural reduction [17,21]. In E. coli, this is a significant cause of furfural-induced growth inhibition; furfural is converted to furfuryl alcohol primarily through the actions of NADPH-dependent reductases with a high affinity for NADPH, such as YqhD and DkgA [22], which can outcompete biosynthetic enzymes for reduced cofactor at low concentrations. One identified effect of this is the inhibition of sulphur uptake, as the assimilation of sulphate into organic sulphur requires 4 units of NADPH [23]. Correspondingly, for E. coli, many of the successful strategies for engineering of furfural resistance have focused on addressing the depletion of NADPH. These strategies include the identification and deletion of the furfural-active reductases that use NADPH [22], the expression of a transhydrogenase capable of replenishing the pool of NADPH [23], the rerouting of carbon flux through the pentose phosphate pathway [24,25], and the expression of reductases that are able to reduce furfural without using NADPH [26,27].

### Identification of novel furfural reductase candidates

To identify suitable reductase candidates, we examined the literature, including the BRENDA database [31], for enzymes with either known NADH-linked furfural reductase activity or alternatively enzymes known to catalyse NAD+-linked furfuryl alcohol dehydrogenase activity which can also function in the reverse reaction, for example the alcohol dehydrogenase ADH1 from S. cerevisiae [32]. However, most had features that might prevent their use in E. coli. For example, the FurX protein from Cupriavidus necator is able to reduce furfural efficiently, but requires ethanol as the source of reducing power [33], whereas the xylose reductase ZMO0976 from Zymomonas mobilis can use NADH but has a higher affinity for NADPH [34].

### Purification and cofactor specificity of candidate furfural reductases

Expression of the recombinant enzymes was assessed in E. coli using the pBADcLIC system, which adds a C-terminal deca-histidine tag to the protein to enable purification by nickel affinity chromatography. All three enzymes were readily purified via HPLC using a nickel affinity purification method (Figure 2). The expression levels of XylB and BaBAD appeared to be somewhat higher than that of AcBAD, although a sufficient quantity of AcBAD was purified for the assay.

Experimentally characterised examples of benzyl alcohol dehydrogenases can use NADH [43] or NADPH [44] as reductants when operating in the reverse direction. When functioning as a dehydrogenase, the P. putida MT-2 XylB protein is highly specific for NAD+ over NADP+ [35], and the known structure of AcBAD contains bound NADH; however, the preferred cofactor of BaBAD is not known. To examine the cofactor preference of our selected enzymes, we used a spectrophotometric assay utilising the ability of NADH and NADPH to absorb at 340 nm [45], which is reduced as the cofactor is oxidised during the conversion of the substrate to its alcohol form.

To understand this difference in selectivity between the enzymes, we examined the known and predicted structures, which suggest differences in the cofactor binding pocket. V227 in BaBAD is replaced with a lysine in XylB (Supplementary Data S2); in an SWISS-MODEL [46] prediction of the XylB structure (Supplementary Data S3), this residue projects into the binding pocket in the region where the phospho group of NADPH would be located, potentially restricting the binding of NADPH relative to NADH (Figure 3).

### Kinetic analysis of substrate specificity

Using NADH as the reductant, the three enzymes were then analysed in more detail for their activity against furfural and the related furanic compound hydroxymethylfurfural, as well as benzaldehyde as a positive control, and the aliphatic compound hexanal as a negative control. As expected, all three enzymes are active with benzaldehyde as a substrate (Figure 4, Table 1 and Supplementary Data S4). Conversely, only BaBAD was able to utilise hexanal, albeit with a low catalytic efficiency (kcat/KM). For the test compounds, furfural and HMF, we observed detectable activity for all three enzymes (Table 1), with furfural being reduced with greater catalytic efficiency than HMF. Interestingly, all three enzymes had comparable efficiencies for furfural and benzaldehyde. However, the kcat and KM values varied greatly. XylB showed much lower kcat and KM values towards benzaldehyde than the other enzymes, which may suggest that it functions more efficiently during alcohol oxidation than during aldehyde reduction. Conversely, AcBAD showed much greater kcat and KM values towards benzaldehyde than the other enzymes; it may be the case that the active site binds benzaldehyde more weakly, allowing faster turnover but requiring a higher substrate concentration to saturate the active site. Finally, BaBAD showed comparable kcat and KM values for furfural and benzaldehyde, suggesting that it interacts with both substrates in a similar manner. Perhaps the most promising from this in vitro data was the XylB protein, which demonstrated a higher rate constant against furfural compared with the other three substrates.

### Effect of furfural reductase expression on growth in furfural-containing medium

The results of the in vitro assay suggested that all three of the chosen enzymes might have the potential to confer furfural tolerance in vivo. Thus, the growth of all three expression strains was tested in modified M9-xylose medium containing 10 mM furfural (Figure 5). In this medium, the growth of the control strain is inhibited up until ∼18 h, at which point the furfural has presumably all been converted to furfuryl alcohol; this pattern is typically observed in E. coli when exposed to furfural where growth is completely inhibited until the toxin is removed, and hence the rate at which it is removed is a good indicator of activity [24]. The strains synthesising XylB and BaBAD showed substantial improvements in growth in the presence of furfural, entering exponential growth at ∼9 h, and reaching maximal OD600 at ∼22 h for XylB or 16 h for BaBAD. Conversely, the AcBAD-expressing strain showed a smaller increase in furfural tolerance, entering exponential phase at 12 h and showing much slower outgrowth. However, expression of AcBAD appears to be detrimental for growth, with a reduced maximal OD600 even in the absence of furfural. Nevertheless, the data suggests that the expression of heterologous NADH-dependent BAD enzymes is a promising route to confer furfural tolerance to E. coli.

To further investigate the mechanism of resistance, we undertook analysis on supernatants from the furfural removal stage of cultures grown in the presence of 10 mM furfural and followed both furfural removal and accumulation of furfuryl alcohol (Figure 6). In this assay, the strain expressing XylB was able to reduce furfural more rapidly, with furfural no longer detectable at 12 h. At this time point, ∼5 mM furfural remained in the control cultures, confirming that heterologously expressed XylB is able to increase the rate of furfural reduction in vivo. Interestingly, the conversion rates of both cultures is very similar for the first 6 h. However, between 6 and 12 h, the rate of conversion in the XylB-expressing culture proceeds at a faster rate than in the control, with a rate of 0.355 mM/hr (R2 = 0.9689) for the control culture and 1.07 mM/hr (R2 = 0.9998) for the XylB-expressing culture (Supplementary Figure S1). Thus, after 6 h, the XylB-expressing strain was able to reduce furfural at approximately three times the rate of the control strain. Curiously, the 6 h time point coincides with a decrease in growth rate in both strains, which is consistent with the biphasic growth curves observed in Figure 4. One potential explanation for these observations is that furfural reduction could be dominated by NADPH-utilising reductases in the beginning of the culture, but after 6 h, NADPH production becomes insufficient to sustain the initial rates of outgrowth and furfural conversion. This would lead to a drop in the rates of growth and NADPH-linked furfural reduction in both cultures, but the XylB-expressing cultures would continue to reduce furfural at a higher rate using NADH. Furfuryl alcohol is likely to be the only product; furoic acid was not detected in any of the cultures (data not shown), which is consistent with other reports of E. coli [26]. Furthermore, the concentration of furfuryl alcohol in the XylB culture remained roughly constant from 12 to 24 h (Supplementary Data S5), confirming that furfuryl alcohol was not converted further by the cells. However, the amount of furfuryl alcohol detected was less than the amount of furfural removed. This could be explained by evaporation of some of the furfural; alternatively, it is possible that some of the furfural is being retained by the biomass, either through partition into the cell membrane, or by formation of adducts with DNA or with cysteine residues as has been observed in vitro for furfural and hydroxymethylfurfural respectively [13,47]. A final sample taken at 24 h (Supplementary Data S5) confirmed that furfural was also undetectable in the control cultures at this time and that both strains were able to grow to a similar OD600 (mean OD600: control, 4.41; XylB, 5.25).

In this work we aimed to find alternative improved enzymes that would catalyse the aerobic reduction in furfural to furfuryl alcohol and that would function readily in microbial cells. Our choice of enzymes identified primarily for catalysing the reverse reactions was vindicated in detection of the desired reaction both in vitro and in vivo. Although the canonical role of XylB is the formation of benzaldehyde from benzyl alcohol, it has been shown that when the Xyl pathway is expressed in E. coli, XylMA is able to catalyse both the oxidation of toluene to benzyl alcohol and the subsequent oxidation of benzyl alcohol to benzaldehyde, whereas coexpression of XylB results in back-formation of benzyl alcohol. Thus, under physiological conditions in E. coli, the back reaction is heavily favoured [48]. It was suggested that the role of XylB may be to prevent the accumulation of benzaldehyde or to allow oxidation when the concentration of benzaldehyde is very low. This may not be optimal for toluene degradation in engineered strains, but it is ideal for our purpose of furfural reduction. Furthermore, there is some evidence that these enzymes may be involved in furfural tolerance in nature. For example, in Acinetobacter baylyi [49], the gene areB was recently shown to be up-regulated by over 20-fold after the addition of furfural; the corresponding protein is a close homologue of the A. calcoaceticus protein used in this work, sharing 82.75% identity. A. baylyi was shown to convert furfural into difurfuryl ether via furfuryl alcohol, and AreB was suggested as a candidate for catalysis of the initial reduction of furfural.

All three reductases were shown to give improvements in growth in the presence of furfural. However, the final OD600 of the AcBAD-expressing culture was markedly lower even under control conditions, and AcBAD was less well expressed than the other two proteins, which suggests that expression of AcBAD may be toxic to the cells.

While XylB and AcBAD have been previously shown to interact with furfuryl alcohol, BaBAD has not been previously characterised to our knowledge, and was selected purely on the basis of homology and the existence of a solved crystal structure. Examination of the gene neighbourhood of BaBAD shows that BaBAD is in an operon with benzaldehyde dehydrogenase; immediately upstream in the opposite direction is a gene cluster encoding genes homologous to benABCD and catBCA, which are involved in the conversion of benzoate to catechol and subsequent conversion of catechol to 3-oxoadipate-enol-lactone [50]. Thus, the role of BaBAD appears to be the degradation of benzyl alcohol and related compounds via the catechol ortho-cleavage pathway [51].

The expression of NADH-dependent furfural reductases is one route towards the improvement of bioproduction using lignocellulosic feedstocks. Most research has focused on the use of FucO, an iron-dependent alcohol dehydrogenase, for this purpose [28,29]. Additionally, the short-chain dehydrogenase enzymes UcpA [26] and, very recently, YghA have been shown to confer improved tolerance to furfural in E. coli [27]. However, this work is, to our knowledge, the first demonstration of improved furfural tolerance through heterologous expression of benzyl alcohol dehydrogenases. It is probable that the family of benzyl alcohol dehydrogenases will contain uncharacterised enzymes with even greater furfural reductase activity. This work is certainly not an exhaustive screening of benzyl alcohol dehydrogenases and examining a wider variety of genes could be an interesting route for future work to take. It may also be possible to further improve the activity or specificity of these enzymes for furfural; this could be accomplished by rational engineering based on the crystal structures, directed evolution based on tolerance to furfural as a method of selection, or the saturation mutagenesis approach applied to FucO by Zheng et al. [29]. Nevertheless, the benzyl alcohol dehydrogenases can be added to the library of known protective enzymes, providing more tools for the improvement of lignocellulosic bioproduction strategies.

### Strains, plasmids and culture conditions

Lysogeny broth (LB; 10 g/L tryptone, 5 g/L yeast extract, 10 g/L sodium chloride, pH ∼7.0) was used for routine culture of E. coli. For growth on solid medium, LB with 15 g/L bacteriological agar was used. Strains were stored at −80°C in LB with 20% (v/v) glycerol. Where relevant, ampicillin was added to a concentration of 100 µg/ml. Cultures were routinely incubated at 37°C and liquid cultures were shaken at 220 rpm. For a full list of strains and plasmids used in this work, see Table 2.

### Protein expression

For protein purification, a 20 ml overnight culture in LB was used to inoculate 1 L LB in a 2 L baffled flask, which was incubated at 37°C with shaking at 120 rpm until the culture reached an OD600 of ∼0.6. The culture was then induced by addition of arabinose to 0.01% (w/v). After 3 h, the culture was harvested by centrifugation at 5000g for 20 min.

Protein purification was carried out in potassium phosphate (KPi) buffers at pH 7.2. Centrifuged cell pellets were resuspended in 30 ml KPi wash buffer (50 mM potassium phosphate pH 7.2, 200 mM sodium chloride, 20% (v/v) glycerol, 40 mM imidazole) and lysed by sonication (3 s on and 7 s off for 10 min, total sonication time 3 min). Lysates were then centrifuged at 27 000g for 30 min and the lysates transferred into a fresh 50 ml Falcon tube. Proteins were purified via HPLC using a HisTrap HP 5 ml column (Cytiva); after loading, the column was washed with 10 column volumes of KPi wash buffer, followed by elution with KPi elution buffer (50 mM potassium phosphate pH 7.2, 200 mM sodium chloride, 20% (v/v) glycerol, 500 mM imidazole).

The IMAC purified proteins were buffer exchanged into the assay buffer (50 mM potassium phosphate pH 7.2, 200 mM sodium chloride) using a HiTrap Desalting column (GE healthcare). Protein concentrations were calculated by absorbance measured at 280 nm according to the respective extinction coefficients as determined using Expasy ProtParam [52].

### Enzyme activity assay

Reaction mixtures for the enzyme activity assay were prepared in assay buffer (50 mM potassium phosphate pH 7.2, 200 mM sodium chloride) with 0.2 mM of either NADH or NADPH and a range of concentrations of either furfural, hydroxymethylfurfural (HMF), benzaldehyde or hexanal. Purified enzymes in assay buffer were added to a final concentration of 0.25 μM. The reaction mixtures were incubated in a Nunc 2.0 microtitre plate with a plate seal at 37°C, shaking at 200 rpm in an Epoch 2 microplate spectrophotometer. The reaction was allowed to progress for an hour, measuring the reduction in absorbance at 340 nm every minute which corresponds to the oxidation of NADH and NADPH to NAD+ and NADP+, respectively. Utilisation of cofactor was calculated using the formula
$ε=cv,$
using an extinction coefficient of 6220 M−1 cm−1 and estimating a path length of 0.625 cm.
The kcat and KM values were calculated using GraphPad Prism v5. The fastest rate of cofactor utilisation was determined by plotting the converted NADH against time for each ligand at a specific concentration. The highest rates of conversion of NADH (V) from the first time point for each ligand concentration ([S]) were then plotted against the latter. A curve was fitted on each plot using non-linear regression for the estimation of Vmax. This analysis has been provided as ‘Supplementary Data S4.xlsx’. The following equations were used to determine the KM and kcat:
$V=Vmax∗[S]/(Km+[S])$
$Vmax=Et∗kcat,$
where Et = concentration of enzyme catalytic sites.

### Structural analysis

The structures for AcBAD and BaBAD were obtained from the Protein Data Bank [53] (1F8F [39] and 5TNX [40] respectively). The model of XylB was prepared on SWISS-MODEL (swissmodel.expasy.org [46]) using the known structure of AcBAD as the template for homology modelling. The QMEANDisCo global score [54] was calculated to be 0.82 ± 0.05 which suggests consistency with other models of a similar size. The derived GMQE (global model quality estimation) for the model [46] was determined to be 0.85, indicating high reliability. Structures were visualised using CCP4mg [55]. The SWISS-MODEL predicted structure of XylB is provided as ‘Supplementary Data S3.pdb’.

Growth assays were carried out in M9 medium [56] with 2% (w/v) xylose, 2 mM MgSO4, and 100 µM CaCl2. M9 medium was modified by addition of 1 ml L−1 1000× trace metals solution [57] (50 mM FeCl3, 20 mM CaCl2, 10 mM MnCl2, 10 mM ZnSO4, 2 mM CoCl2, 2 mM CuCl2, 2 mM NiCl2, 2 mM Na2MoO4, 2 mM Na2SeO3, 2 mM H3BO3, 0.1 M HCl). Additionally, citric acid was added to 0.05% (w/v) to prevent precipitation of iron, and casamino acids to 0.01% (w/v) in order to alleviate stress caused by recombinant protein expression [58]. This took the final pH of the medium to ∼4.6–4.7. Ampicillin was added to 100 µg/ml, and arabinose added to 0.01% (w/v) to induce expression of the target protein.

For the growth assays, strains were pre-cultured overnight in 5 ml modified M9-xylose medium. To ensure that the strains were at the same phase of growth at the beginning the experiment, the overnight cultures were used to inoculate over-day cultures of the same medium to an OD600 of 0.05. At an OD600 of ∼0.5, these cultures were diluted to an OD600 of 0.2 in the same medium. 96-well plates were prepared with 200 µl modified M9-xylose medium with furfural at the required concentration; wells were inoculated with 10 µl of the diluted cultures. The inoculated plates were incubated at 37°C in a Biotek Epoch 2 microplate spectrophotometer with double orbital shaking at 282 cpm for 48 h with OD600 being measured every 30 min. Data was plotted using OriginPro 2021b software.

### Supernatant analysis

For the supernatant analysis, cultures were grown in 250 ml flasks containing 30 ml modified M9-xylose medium with 10 mM furfural. Overnight cultures were prepared in 5 ml modified M9-xylose medium and used to inoculate the flasks to an OD600 of 0.05. Flasks were then grown at 37°C with shaking at 180 rpm; at 0, 2, 4, 6, 8, 10, and 12 h, 1 ml of the culture was harvested for OD600 measurement and analysis. Samples were centrifuged at 15 000g for 1 min and the supernatants transferred to a fresh 1.5 ml centrifuge tube. Samples were stored at −20°C until required.

Supernatant samples were diluted 20-fold in deionised water and analysed for furfural, furfuryl alcohol and furoic acid. Standards for these compounds were prepared to 500 µM in deionised water. Additionally, a calibration curve was prepared with a mixture of all three compounds diluted to 500 µM, 300 µM, 100 µM, 50 µM, 25 µM and 10 µM. Samples were analysed on an Acquity I-Class UPLC equipped with a TUV detector (Waters, Wilmslow, U.K.). The column was an Acquity UPLC BEH C18, 130 Å, 1.7 µm, 2.1 mm × 100 mm (Waters). The mobile phase A) was 5% (v/v) methanol (Rathburn, Walkerburn, U.K.) with 0.1% (v/v) acetic acid (Fisher Optima), B) methanol with 0.1% (v/v) acetic acid. The elution gradient was 16% (v/v) B) isocratic for 2.5 min, which was ramped up to 100% B) over 0.3 min, left there for 0.1 min, and returned to 16% (v/v) B) over 0.4 min; total run time was 4 min. Flow rate was 0.5 ml/min, injection volume 2 µl, column temperature 45°C, and sample temperature 10°C. The UV detection wavelength was 230 nm. Standard curves of furfural, furoic acid and furfuryl alcohol (Sigma/Merck, Gillingham, U.K.) were constructed in water. Data analysis was undertaken with Xcalibur 4.0 (Thermo Scientific, Loughborough, U.K.). OriginPro 2021b software was used to plot the concentration and OD600 data and to estimate the rate of furfural conversion.

The authors confirm that all relevant data has been provided either in the article or as Supplementary Material.

The authors declare that there are no competing interests associated with the manuscript.

B.J.W. and R.H. are funded by the BBSRC IB Catalyst Grant ‘DeTOX — Productive whole cell biocatalysis by engineering resistance to toxic products and substrates’ (Ref: BB/N01040X/1). S.L. is funded by the York Centre of Excellence in Mass Spectrometry, which was created with funding from Science City York, Yorkshire Forward and the Northern Way Initiative, and subsequent support from EPSRC (EP/K039660/1; EP/M028127/1) and BBSRC (BB/R013403/1).

Gavin Thomas: Conceptualization, Supervision, Funding acquisition, Visualization, Project administration, Writing — review and editing. Benjamin James Willson: Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing — original draft, Writing — review and editing. Reyme Herman: Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing — review and editing. Swen Langer: Data curation, Formal analysis, Validation, Investigation, Methodology, Writing — review and editing.

benzyl alcohol dehydrogenase

•
• HMF

hydroxymethyl furfural

•
• MIC

minimal inhibitory concentration

1
Saini
,
J.K.
,
Saini
,
R.
and
Tewari
,
L.
(
2015
)
Lignocellulosic agriculture wastes as biomass feedstocks for second-generation bioethanol production: concepts and recent developments
.
3 Biotech
5
,
337
353
2
Bajpai
,
P.
(
2016
) Structure of Lignocellulosic Biomass. In
Pretreatment of Lignocellulosic Biomass for Biofuel Production
(Sanjay, K.S., ed.), pp.
7
12
,
Springer
,
Singapore
3
Zhao
,
X.
,
Zhang
,
L.
and
Liu
,
D.
(
2012
)
Biomass recalcitrance. Part I: the chemical compositions and physical structures affecting the enzymatic hydrolysis of lignocellulose
.
Biofuels Bioprod. Biorefining
6
,
465
482
4
Jönsson
,
L.J.
and
Martín
,
C.
(
2016
)
Pretreatment of lignocellulose: formation of inhibitory by-products and strategies for minimizing their effects
.
Bioresour. Technol.
199
,
103
112
5
Hurd
,
C.D.
and
Isenhour
,
L.L.
(
1932
)
Pentose reactions. I. Furfural formation
.
J. Am. Chem. Soc.
54
,
317
330
6
Kabel
,
M.A.
,
Bos
,
G.
,
Zeevalking
,
J.
,
Voragen
,
A.G.J.
and
Schols
,
H.A.
(
2007
)
Effect of pretreatment severity on xylan solubility and enzymatic breakdown of the remaining cellulose from wheat straw
.
Bioresour. Technol.
98
,
2034
2042
7
Horn
,
S.J.
,
Nguyen
,
Q.D.
,
Westereng
,
B.
,
Nilsen
,
P.J.
and
Eijsink
,
V.G.H.
(
2011
)
Screening of steam explosion conditions for glucose production from non-impregnated wheat straw
.
Biomass Bioenerg.
35
,
4879
4886
8
Mills
,
T.Y.
,
Sandoval
,
N.R.
and
Gill
,
R.T.
(
2009
)
Cellulosic hydrolysate toxicity and tolerance mechanisms in Escherichia coli
.
Biotechnol. Biofuels
2
,
26
9
Allen
,
S.A.
,
Clark
,
W.
,
McCaffery
,
J.M.
,
Cai
,
Z.
,
Lanctot
,
A.
,
Slininger
,
P.J.
et al (
2010
)
Furfural induces reactive oxygen species accumulation and cellular damage in Saccharomyces cerevisiae
.
Biotechnol. Biofuels
3
,
2
10
Wang
,
J.
,
Zhang
,
Y.
,
Chen
,
Y.
,
Lin
,
M.
and
Lin
,
Z.
(
2012
)
Global regulator engineering significantly improved Escherichia coli tolerances toward inhibitors of lignocellulosic hydrolysates
.
Biotechnol. Bioeng.
109
,
3133
3142
11
Khan
,
Q.A.
,
Shamsi
,
F.A.
and
,
S.M.
(
1995
)
Mutagenicity of furfural in plasmid DNA
.
Cancer Lett.
89
,
95
99
12
,
S.
and
Rehman
,
A.
(
1989
)
Specificity of the interaction of furfural with DNA
.
Mutat. Res.
225
,
101
106
13
Barciszewski
,
J.
,
Siboska
,
G.E.
,
Pedersen
,
B.O.
,
Clark
,
B.F.C.
and
Rattan
,
S.I.S.
(
1997
)
A mechanism for the in vivo formation of N6-furfuryladenine, kinetin, as a secondary oxidative damage product of DNA
.
FEBS Lett.
414
,
457
460
14
Wyszko
,
E.
,
Barciszewska
,
M.Z.
,
Markiewicz
,
M.
,
Szymański
,
M.
,
Markiewicz
,
W.T.
,
Clark
,
B.F.C.
et al (
2003
)
“Action-at-a distance” of a new DNA oxidative damage product 6-furfuryl-adenine (kinetin) on template properties of modified DNA
.
Biochim. Biophys. Acta Gene Struct. Expr.
1625
,
239
245
15
Zheng
,
H.
,
Wang
,
X.
,
Yomano
,
L.P.
,
Shanmugam
,
K.T.
and
Ingram
,
L.O.
(
2012
)
Increase in furfural tolerance in ethanologenic Escherichia coli LY180 by plasmid-based expression of thyA
.
Appl. Environ. Microbiol.
78
,
4346
4352
16
Geddes
,
R.D.
,
Wang
,
X.
,
Yomano
,
L.P.
,
Miller
,
E.N.
,
Zheng
,
H.
,
Shanmugam
,
K.T.
et al (
2014
)
Polyamine transporters and polyamines increase furfural tolerance during xylose fermentation with ethanologenic Escherichia coli strain LY180
.
Appl. Environ. Microbiol.
80
,
5955
5964
17
Tsuge
,
Y.
,
Hori
,
Y.
,
Kudou
,
M.
,
Ishii
,
J.
,
Hasunuma
,
T.
and
Kondo
,
A.
(
2014
)
Detoxification of furfural in Corynebacterium glutamicum under aerobic and anaerobic conditions
.
Appl. Microbiol. Biotechnol.
98
,
8675
8683
18
Franden
,
M.A.
,
Pilath
,
H.M.
,
Mohagheghi
,
A.
,
Pienkos
,
P.T.
and
Zhang
,
M.
(
2013
)
Inhibition of growth of Zymomonas mobilis by model compounds found in lignocellulosic hydrolysates
.
Biotechnol. Biofuels
6
,
99
19
,
M.J.
,
Gustafsson
,
L.
,
Niklasson
,
C.
and
Lidén
,
G.
(
1999
)
Conversion of furfural in aerobic and anaerobic batch fermentation of glucose by Saccharomyces cerevisiae
.
J. Biosci. Bioeng.
87
,
169
174
20
Koopman
,
F.
,
Wierckx
,
N.
,
de Winde
,
J.H.
and
Ruijssenaars
,
H.J.
(
2010
)
Identification and characterization of the furfural and 5-(hydroxymethyl)furfural degradation pathways of Cupriavidus basilensis HMF14
.
107
,
4919
4924
21
,
M.
,
Bettiga
,
M.
,
Mapelli
,
V.
and
Olsson
,
L.
(
2013
)
The influence of HMF and furfural on redox-balance and energy-state of xylose-utilizing Saccharomyces cerevisiae
.
Biotechnol. Biofuels
6
,
22
22
Miller
,
E.N.
,
Jarboe
,
L.R.
,
Yomano
,
L.P.
,
York
,
S.W.
,
Shanmugam
,
K.T.
and
Ingram
,
L.O.
(
2009
)
Silencing of NADPH-dependent oxidoreductase genes (yqhD and dkgA) in furfural-resistant ethanologenic Escherichia coli
.
Appl. Environ. Microbiol.
75
,
4315
4323
23
Miller
,
E.N.
,
Jarboe
,
L.R.
,
Turner
,
P.C.
,
Pharkya
,
P.
,
Yomano
,
L.P.
,
York
,
S.W.
et al (
2009
)
Furfural inhibits growth by limiting sulfur assimilation in ethanologenic Escherichia coli strain LY180
.
Appl. Environ. Microbiol.
75
,
6132
6141
24
Glebes
,
T.Y.
,
Sandoval
,
N.R.
,
Reeder
,
P.J.
,
Schilling
,
K.D.
,
Zhang
,
M.
and
Gill
,
R.T.
(
2014
)
Genome-Wide mapping of furfural tolerance genes in Escherichia coli
.
PLoS ONE
9
,
e87540
25
Jilani
,
S.B.
,
Dev
,
C.
,
Eqbal
,
D.
,
Jawed
,
K.
,
,
R.
and
Yazdani
,
S.S.
(
2020
)
Deletion of pgi gene in E. coli increases tolerance to furfural and 5-hydroxymethyl furfural in media containing glucose–xylose mixture
.
Microb. Cell Fact.
19
,
1
13
26
Wang
,
X.
,
Miller
,
E.N.
,
Yomano
,
L.P.
,
Shanmugam
,
K.T.
and
Ingram
,
L.O.
(
2012
)
Increased furan tolerance in Escherichia coli due to a cryptic ucpA gene
.
Appl. Environ. Microbiol.
78
,
2452
2455
27
Jilani
,
S.B.
,
,
R.
and
Yazdani
,
S.S.
(
2021
)
Overexpression of an oxidoreductase YghA confers tolerance against furfural in ethanologenic Escherichia coli strain SSK42
.
Appl. Environ. Microbiol.
87
,
e01855
21
28
Wang
,
X.
,
Miller
,
E.N.
,
Yomano
,
L.P.
,
Zhang
,
X.
,
Shanmugam
,
K.T.
and
Ingram
,
L.O.
(
2011
)
Increased furfural tolerance due to overexpression of NADH-dependent oxidoreductase FucO in Escherichia coli strains engineered for the production of ethanol and lactate
.
Appl. Environ. Microbiol.
77
,
5132
5140
29
Zheng
,
H.
,
Wang
,
X.
,
Yomano
,
L.P.
,
Geddes
,
R.D.
,
Shanmugam
,
K.T.
and
Ingram
,
L.O.
(
2013
)
Improving Escherichia coli FucO for furfural tolerance by saturation mutagenesis of individual amino acid positions
.
Appl. Environ. Microbiol.
79
,
3202
3208
30
Cabiscol
,
E.
,
Hidalgo
,
E.
,
,
J.
,
Baldomá
,
L.
,
Ros
,
J.
and
Aguilar
,
J.
(
1990
)
Oxygen regulation of L-1,2-propanediol oxidoreductase activity in Escherichia coli
.
J. Bacteriol.
172
,
5514
5515
31
Chang
,
A.
,
Jeske
,
L.
,
Ulbrich
,
S.
,
Hofmann
,
J.
,
Koblitz
,
J.
,
Schomburg
,
I.
et al (
2021
)
BRENDA, the ELIXIR core data resource in 2021: new developments and updates
.
Nucleic Acids Res.
49
,
D498
D508
32
,
B.
,
Wallace-Salinas
,
V.
,
Carlsson
,
Å.J
,
Almeida
,
J.R.
,
Rådström
,
P.
and
Gorwa-Grauslund
,
M.F.
(
2014
)
Furaldehyde substrate specificity and kinetics of Saccharomyces cerevisiae alcohol dehydrogenase 1 variants
.
Microb. Cell Fact.
13
,
112
33
Li
,
Q.
,
Metthew Lam
,
L.K.
and
Xun
,
L.
(
2011
)
Cupriavidus necator JMP134 rapidly reduces furfural with a Zn-dependent alcohol dehydrogenase
.
22
,
1215
1225
34
Agrawal
,
M.
and
Chen
,
R.R.
(
2011
)
Discovery and characterization of a xylose reductase from Zymomonas mobilis ZM4
.
Biotechnol. Lett.
33
,
2127
2133
35
Shaw
,
J.P.
and
Harayama
,
S.
(
1990
)
Purification and characterisation of TOL plasmid-encoded benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase of Pseudomonas putida
.
Eur. J. Biochem.
191
,
705
714
36
Pick
,
A.
,
Rühmann
,
B.
,
Schmid
,
J.
and
Sieber
,
V.
(
2013
)
Novel CAD-like enzymes from Escherichia coli K-12 as additional tools in chemical production
.
Appl. Microbiol. Biotechnol.
97
,
5815
37
Gillooly
,
D.J.
,
Robertson
,
A.G.
and
Fewson
,
C.A.
(
1998
)
Molecular characterization of benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase II of Acinetobacter calcoaceticus
.
Biochem. J.
330
,
1375
38
MacKintosh
,
R.W.
and
Fewson
,
C.A.
(
1988
)
Benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase II from Acinetobacter calcoaceticus. Substrate specificities and inhibition studies
.
Biochem. J.
255
,
653
PMID:
[PubMed]
39
Beauchamp
,
J.C.
,
Gillooly
,
D.
,
Warwicker
,
J.
,
Fewson
,
C.A.
and
Lapthorn
,
A.J.
(
2000
)
Crystal Structure of Benzyl Alcohol Dehydrogenase From Acinetobacter calcoaceticus
40
Abendroth
,
J.
,
Mayclin
,
S.J.
,
Lorimer
,
D.
and
Edwards
,
T.E.
(
2016
)
Crystal Structure of Alcohol Dehydrogenase Zinc-Binding Domain Protein From Burkholderia ambifaria
41
Shrivastava
,
R.
,
Basu
,
A.
and
Phale
,
P.S.
(
2011
)
Purification and characterization of benzyl alcohol- and benzaldehyde- dehydrogenase from Pseudomonas putida CSV86
.
Arch. Microbiol.
193
,
553
563
42
Geertsma
,
E.R.
and
Poolman
,
B.
(
2007
)
High-throughput cloning and expression in recalcitrant bacteria
.
Nat. Methods
4
,
705
707
43
Uthoff
,
S.
and
Steinbüchel
,
A.
(
2012
)
Purification and characterization of an NAD+-Dependent XylB-Like aryl alcohol dehydrogenase identified in Acinetobacter baylyi ADP1
.
Appl. Environ. Microbiol.
78
,
8743
44
Ying
,
X.
,
Wang
,
Y.
,
Xiong
,
B.
,
Wu
,
T.
,
Xie
,
L.
,
Yu
,
M.
et al (
2014
)
Characterization of an allylic/benzyl alcohol dehydrogenase from Yokenella sp. strain WZY002, an organism potentially useful for the synthesis of α,β-unsaturated alcohols from allylic aldehydes and ketones
.
Appl. Environ. Microbiol.
80
,
2399
2409
45
Shaw
,
J.P.
,
Schwager
,
F.
and
Harayama
,
S.
(
1992
)
Substrate-specificity of benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase encoded by TOL plasmid pWW0. Metabolic and mechanistic implications
.
Biochem. J.
283
,
789
46
Waterhouse
,
A.
,
Bertoni
,
M.
,
Bienert
,
S.
,
Studer
,
G.
,
Tauriello
,
G.
,
Gumienny
,
R.
, et al (
2018
)
SWISS-MODEL: homology modelling of protein structures and complexes
.
Nucleic Acids Res.
46
,
W296
W303
47
Zou
,
Y.
,
Pei
,
K.
,
Peng
,
X.
,
Bai
,
W.
,
Huang
,
C.
and
Ou
,
S.
(
2016
)
Possible adducts formed between hydroxymethylfurfural and selected amino acids, and their release in simulated gastric model
.
Int. J. Food Sci. Technol.
51
,
1002
1009
48
Bühler
,
B.
,
Schmid
,
A.
,
Hauer
,
B.
and
Witholt
,
B.
(
2000
)
Xylene monooxygenase catalyzes the multistep oxygenation of toluene and pseudocumene to corresponding alcohols, aldehydes, and acids in Escherichia coli JM101
.
J. Biol. Chem.
275
,
10085
10092
49
Arteaga
,
J.E.
,
Cerros
,
K.
,
Rivera-Becerril
,
E.
,
Lara
,
A.R.
,
Le Borgne
,
S.
and
Sigala
,
J.-C.
(
2021
)
Furfural biotransformation in Acinetobacter baylyi ADP1 and Acinetobacter schindleri ACE
.
Biotechnol. Lett.
43
,
1043
1050
50
Pérez-Pantoja
,
D.
,
Donoso
,
R.
,
Agulló
,
L.
,
Córdova
,
M.
,
Seeger
,
M.
,
Pieper
,
D.H.
et al (
2012
)
Genomic analysis of the potential for aromatic compounds biodegradation in Burkholderiales
.
Environ. Microbiol.
14
,
1091
1117
51
Basu
,
A.
,
Dixit
,
S.S.
and
Phale
,
P.S.
(
2003
)
Metabolism of benzyl alcohol via catechol ortho-pathway in methylnaphthalene-degrading Pseudomonas putida CSV86
.
Appl. Microbiol. Biotechnol.
62
,
579
585
52
Gasteiger
,
E.
,
Hoogland
,
C.
,
Gattiker
,
A.
,
Duvaud
,
S.
,
Wilkins
,
M. R.
,
Appel
,
R. D.
et al (
2005
) Protein Identification and Analysis Tools on the ExPASy Server. In
The Proteomics Protocols Handbook
(John, M.W., ed.), pp.
571
607
, Humana Press
53
Berman
,
H.M.
,
Westbrook
,
J.
,
Feng
,
Z.
,
Gilliland
,
G.
,
Bhat
,
T.N.
,
Weissig
,
H.
et al (
2000
)
The protein data bank
.
Nucleic Acids Res.
28
,
235
242
54
Studer
,
G.
,
Rempfer
,
C.
,
Waterhouse
,
A.M.
,
Gumienny
,
R.
,
Haas
,
J.
and
Schwede
,
T.
(
2020
)
QMEANDisco—distance constraints applied on model quality estimation
.
Bioinformatics
36
,
1765
1771
55
McNicholas
,
S.
,
Potterton
,
E.
,
Wilson
,
K.S.
and
Noble
,
M.E.M.
(
2011
)
Presenting your structures: the CCP4mg molecular-graphics software
.
Acta Crystallogr. Sect. D Biol. Crystallogr.
67
,
386
394
56
Neidhardt
,
F.C.
,
Bloch
,
P.L.
and
Smith
,
D.F.
(
1974
)
Culture medium for enterobacteria
.
J. Bacteriol.
119
,
736
747
57
Studier
,
F.W.
(
2005
)
Protein production by auto-induction in high-density shaking cultures
.
Protein Expr. Purif.
41
,
207
234
58
Shachrai
,
I.
,
Zaslaver
,
A.
,
Alon
,
U.
and
Dekel
,
E.
(
2010
)
Cost of unneeded proteins in E. coli is reduced after several generations in exponential growth
.
Mol. Cell
38
,
758
767
59
Datsenko
,
K.A.
and
Wanner
,
B.L.
(
2000
)
One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products
.
97
,
6640
6645
60
Park
,
S.E.
,
Koo
,
H.M.
,
Park
,
Y.K.
,
Park
,
S.M.
,
Park
,
J.C.
,
Lee
,
O.K.
et al (
2011
)
Expression of aldehyde dehydrogenase 6 reduces inhibitory effect of furan derivatives on cell growth and ethanol production in Saccharomyces cerevisiae
.
Bioresour. Technol.
102
,
6033
6038
61
Donoso
,
R.A.
,
González-Toro
,
F.
and
Pérez-Pantoja
,
D.
(
2021
)
Widespread distribution of hmf genes in Proteobacteria reveals key enzymes for 5-hydroxymethylfurfural conversion
.
Comput. Struct. Biotechnol. J.
19
,
2160
2169