Cloning, expression and characterization of an aryl-alcohol dehydrogenase from the white-rot fungus Phanerochaete chrysosporium strain BKM-F-1767

Background The white-rot fungus Phanerochaete chrysosporium is among the small group of fungi that can degrade lignin to carbon dioxide while leaving the crystalline cellulose untouched. The efficient lignin oxidation system of this fungus requires cyclic redox reactions involving the reduction of aryl-aldehydes to the corresponding alcohols by aryl-alcohol dehydrogenase. However, the biochemical properties of this enzyme have not been extensively studied. These are of most interest for the design of metabolic engineering/synthetic biology strategies in the field of biotechnological applications of this enzyme. Results We report here the cloning of an aryl-alcohol dehydrogenase cDNA from the white-rot fungus Phanerochaete chrysosporium, its expression in Escherichia coli and the biochemical characterization of the encoded GST and His6 tagged protein. The purified recombinant enzyme showed optimal activity at 37°C and at pH 6.4 for the reduction of aryl- and linear aldehydes with NADPH as coenzyme. NADH could also be the electron donor, while having a higher Km (220 μM) compared to that of NADPH (39 μM). The purified recombinant enzyme was found to be active in the reduction of more than 20 different aryl- and linear aldehydes showing highest specificity for mono- and dimethoxylated Benzaldehyde at positions 3, 4, 3,4 and 3,5. The enzyme was also capable of oxidizing aryl-alcohols with NADP + at 30°C and an optimum pH of 10.3 but with 15 to 100-fold lower catalytic efficiency than for the reduction reaction. Conclusions In this work, we have characterized the biochemical properties of an aryl-alcohol dehydrogenase from the white-rot fungus Phanerochaete chrysosporium. We show that this enzyme functions in the reductive sense under physiological conditions and that it displays relatively large substrate specificity with highest activity towards the natural compound Veratraldehyde.


Background
Lignin is, after cellulose, the second most abundant terrestrial biopolymer, accounting for approximately 30% of the organic carbon in the biosphere [1]. The biodegradation of lignin plays a crucial role in the earth's carbon cycle. Unlike cellulose and hemicellulose, this amorphous and insoluble aromatic material lacks stereoregularity and is not susceptible to hydrolytic attack. In nature, the white-rot fungus Phanerochaete chrysosporium is among the small group of fungi that can completely degrade lignin to carbon dioxide while leaving the crystalline cellulose untouched [2].
By following the reduction of 3,4-Dimethoxybenzaldehyde (Veratraldehyde) in Nitrogen-limited cultures of P. chrysosporium, Muheim et al. [19] purified an intracellular aryl-alcohol dehydrogenase (EC 1.1.1.91) from this lignin-degrading fungus. A cDNA coding for this protein was later isolated and characterized [20]. However, the biochemical properties of the Aadp enzyme were not extensively studied.
Due to its high efficiency in lignin degradation, and to its potential applications in the textile, fuel and paper industries, the 35-Mb haploid genome of P. chrysosporium strain RP78 has been sequenced [2]. The current draft release, version 2.0, includes a total of 10,048 gene models [21] and reveals that the secreted oxidases, peroxidases and hydrolytic enzymes that cooperate in wood decay exist as large multi-gene families. Taking advantage of this genome sequence, this work describes the cloning of an AAD cDNA and the comprehensive biochemical characterization of the encoded enzyme in order to get deeper insight into its biological relevance and biotechnological applications potential such as the degradation of aromatic inhibitors in lignocellulosic hydrolysates that strongly impair ethanol fermentation by yeast [22], as well as for the microbial production of natural flavour and fragrance molecules like 2-Phenylethanol.

Results and discussion
Cloning of a cDNA from Phanerochaete chrysosporium encoding an aryl-alcohol dehydrogenase Using the amino acid sequence coded by a previously cloned AAD ORF from Phanerochaete chrysosporium (Pc) strain OGC101 [20] as query, a BLAST alignment was performed against the translated predicted ORFs of the genome sequence of P. chrysosporium strain RP78 [2,21]. The results showed the existence of 8 AAD homologues that consist of six to nine exons and encode proteins from 240 to 398 amino acids. The presence of multiple AAD genes in the Pc genome is in accordance with strong multiple bands observed in a Southern blot by Reiser et al. [20]. Interestingly, in scaffold_1, two tandem AAD homologues (scaffold_1:1025231 to 1023962, and scaffold_1:1027063 to 1025827) were found adjacent to each other. The distance between these two adjacent ORFs is only 596 base-pairs. This extensive genetic diversity was also observed for other lignin-biodegradation related genes encoding peroxidases, oxidases, glycosydases and cytochrome P450s [2]. The existence of multiple AAD genes might suggest multiple specificities required to reduce various aryl-aldehydes arising from the catabolism of complex wood polymers.
Among the 8 predicted homologous ORFs in the genome of Pc strain RP78, the one in scaffold_3:2235704-2237287 (JGI Transcript Id: 11055) has only 37 base pairs differences with the cDNA previously cloned by Reiser et al. [20] and encodes a 100% identical amino acid sequence. Considering that the remaining 7 AAD homologues show 72.1, 66.7, 64.6, 55, 54.1, 49.9 and 45.7% amino acid identity with this cDNA sequence, we designed specific primers on the coding region from scaffold_3:2235704-2237287 (hereafter termed AAD1) to clone the full length cDNA using RACE (rapid amplification of cDNA ends, [23,24]) and PCR techniques. The method was adopted because of the presence of 5 introns in the genomic sequence of this Pc AAD1 gene. The RNA used for this cloning was obtained from a six days Nitrogen-limited culture of Pc strain BKM-F-1767. As shown in Figure 1, qPCR assays under this growth condition showed that the AAD1 transcript began to accumulate at day 2 and continued over 6 days. This result nicely correlated with an increase of aryl-alcohol dehydrogenase activity acting on Veratraldehyde during N-limited culture and reaching a maximum after 6 days of growth [19]. The RACE-PCR method on the 6-days purified RNA allowed us to isolate a 1.4 kilobase fulllength cDNA containing a 1155 bp ORF that encoded a protein 100% identical with the translated genomic sequence from Pc RP78 strain [2,21] as well as with that of Reiser et al. [20]. The sequencing results of the cloned Pc AAD1 cDNA also showed the presence of a 5′ untranslated region (UTR) and of a 3′ poly(A) tail, confirming the integrity of the mRNA template. Comparison of the 5′UTR (159 nucleotides in total) with that of the cDNA by Reiser et al. [20] revealed 94.3% nucleotide identity, suggesting they are the same gene in the two strains.

Heterologous expression in E. Coli and purification of recombinant Pc Aad1p
In order to obtain large amounts of purified recombinant enzyme for biochemical characterization, the Pc AAD1 ORF was cloned in pGS-21a and pGEX-6p-1 vectors and expressed in E. coli to produce GST and/or His 6 tagged proteins. The expression conditions were optimized using different E. coli strains, cultivation temperatures, IPTG concentrations and induction times. The highest accumulation of recombinant Pc Aad1p was obtained with E. coli BL21 Star ™ (DE3) strain harbouring the pGS-21a-AAD1 expression vector after overnight induction with 0.1 mM IPTG at 16°C allowing the production of up to 1.8 ± 0.1 gÁL −1 of recombinant protein after purification. After cell disruption, the recombinant Aad1p was purified by Glutathione affinity chromatography to yield a single protein band as shown on SDS-Polyacrylamide gel electrophoresis ( Figure 2, lane 3). This SDS-PAGE also showed that the recombinant protein was the major band in the cell lysate ( Figure 2, lane 1) and that the purified protein migrates at an apparent molecular mass of 70 kDa in our conditions of electrophoresis. Taking into account the presence of the GST and His 6 tags in the fusion protein, which correspond to~30 kDa, the molecular mass of our purified Pc Aad1p is in accordance with the theoretical molecular mass calculated from its amino acid composition (43 kDa) and very close to the apparent 47 kDa of the Aad enzyme purified from P. chrysosporium by Muheim et al. [19].
Biochemical characterization of the purified recombinant Pc Aad1p Structure analysis of Pc Aad1p We searched for functional domains of the Pc Aad1 protein using the Pfam database server [25,26]. This in silico analysis identified the protein as belonging to subfamily AKR9A of the aldo-keto reductase (AKR) superfamily with residues D71, Y76 and K103 as predicted activesites. The AKR superfamily is one of the three enzyme superfamilies that perform oxidoreduction on a wide variety of natural and foreign substrates [27]. The large AKR superfamily includes presently 15 families, with more than 170 proteins identified in mammals, plants, fungi and bacteria. AKR structures share a highly conserved (α/β) 8 -barrel motif, a conserved cofactor (mostly NADPH) binding site and catalytic tetrad, and a variable loop structure  which usually defines broad substrate specificity. The majority of AKRs are monomeric proteins of about 320 amino acids in length, although several members from families AKR2, AKR6 and AKR7 were found to form multimers [28]. The closest AKR protein 'relatives' of Pc Aad1p (AKR9A3) are the fungal norsolorinic acid reductase from Aspergillus flavus (AKR9A2) and sterogmatocystin dehydrogenase from Aspergillus nidulans (AKR9A1) and the putative yeast proteins Aad14p, Aad3p, Aad4p and Aad10p from Saccharomyces cerevisiae. According to the family tree structure, the nearest AKR with 3D structure characterized is AKR11C1 from the bacterium Bacillus halodurans [27,29]. Aldo-keto reductases catalyze oxidation and reduction reactions on a range of substrates using NAD(P)(H) as cofactor. An ordered Bi Bi kinetic mechanism, in which cofactor binds first and leaves last, has been demonstrated for pig kidney aldehyde reductase (ALR) [30], bovine kidney aldose reductase ADR [31], rat liver 3-alpha-hydroxysteroid dehydrogenase (3α-HSD) [32] and 3-oxo-5b-steroid 4-dehydrogenase [33], and may be a characteristic feature of other AKRs [34]. The reduction reaction involves 4-pro-R hydride transfer from NAD(P)H to the substrate carbonyl and protonation of the Oxygen by a residue of the enzyme acting as a general acid [34]. The rate of this reaction is increased with substrates harbouring chemical structures that facilitate their nucleophilic attack by the hydride ion. It is also influenced by the orientation and/or relative mobility of the carbonyl function with respect of the rest of the molecule that would affect its protonation by one or more possibly acid residues of the active site.

Temperature-and pH-dependence of Pc Aad1p activity
To determine pH and temperature optimum of the recombinant purified Pc Aad1p, we used Veratraldehyde as substrate for the reductive sense, and the corresponding alcohol for the oxidative sense of the reaction, while NADP(H) was used as the cofactor. As shown in Figure 3A, the activity of this enzyme was optimal at about pH 6.4 in the reductive sense whereas oxidation rates could only be measured in basic conditions with an optimum at pH 10.4. At this pH, the oxidation activity was 7-fold lower than at the optimal pH for the reductive reaction. These results strongly support the fact that the Pc Aad1p works in the cells predominantly as an aldehyde reductase. The optimal temperature for activity was only determined in the reductive sense and was found to be close to 37°C ( Figure 3B).

Substrate specificity and kinetic properties of Pc Aad1p
The substrate specificity of the purified recombinant Pc Aad1p protein was determined with a large spectrum of chemical molecules including linear aliphatic and arylaldehydes and alcohols, and ethyl-, ramified and aryl acetate esters (Table 1), keeping in mind that the presence of a GST tag at the amino terminus could modify the enzyme properties. Figure 4 shows some of the aldehyde and alcohol substrates analyzed in this study ordered by chemical function and substitution. For comparative analysis, we carried out our assays at pH 6.1 in 50 mM MES and at 30°C using the same concentration of substrate molecules and NADPH and compared the measured activity to that obtained with Veratraldehyde, which was used as the reference. The activity value with this substrate was set to 100%. As indicated in Table 1, Pc Aad1p activity with mono-methoxylated Benzaldehyde at positions 3 (meta) or 4 (para), or dimethoxylated at positions 3,5 was very close or even slightly higher than with Veratraldehyde (3,4-Dimethoxybenzaldehyde). Activity was reduced by two when the methoxy radical was on carbon 2 (ortho). The presence of a hydroxyl group on Benzaldehyde or on methoxy-substituted Benzaldehyde resulted in a dramatic drop of the activity of Pc Aad1p. Likewise, the enzyme was 3 to 5-fold less active on other types of substitutions of the Benzaldehyde molecule such as with Chlorine, Fluorine or Nitro functional groups. Furthermore, the Pc Aad1p activity on Phenylacetaldehyde was comparable to that of Veratraldehyde. Linear aldehydes of 3 to 11 carbon atoms were also assayed for substrate specificity of Pc Aad1p. The highest activity was observed on C6 to C8 aldehydes, with reaction rates about 2-fold lower than on Veratraldehyde but comparable to that on Benzaldehyde. No activity was detected for Propanal (C3) and Butanal (C4) and very low activity for C9 to C11 aldehydes.  Among the substrates assayed for the oxidation reaction by Pc Aad1p with NADP + as cofactor, the highest activity was by far that on Veratryl alcohol (3,4-Dimethoxybenzyl alcohol), whereas other mono-, di-or tri-substituted methoxybenzyl alcohols showed poor reactivity with this enzyme. Interestingly, the Pc Aad1p showed 46% activity on 4-Hydroxy-3-Methoxybenzyl alcohol (Vanillyl alcohol) as compared to that on Veratryl alcohol. No activity could be detected on many other linear aliphatic, ramified aliphatic or aryl alcohol substrates as well as on some acetate esterified aryl and ramified alcohols. Altogether, these   Table 1 and 2.
results suggest that a specific size, structure and conformation of the substrate are necessary to allow concurrent interactions of the carbonyl group of the substrate molecule with the cofactor and with key amino acids of the active site. Other parameters like the relative hydrophilic/hydrophobic character of the substrates and of the active site as well as the possibility of resonance delocalization within a conjugated π system of the substrate might also account for relative specificity of the Aad1p enzyme to its substrate.
We then obtained precise kinetic parameters of Pc Aad1p with respect to cofactor dependency and affinity to several substrates like Veratraldehyde or Veratryl alcohol ( Table 2). In the reductive sense, using 0.2 mM Veratraldehyde, the activity of Pc Aad1p for NADPH oxidation followed a Michaelis-Menten curve with an apparent K M = 39 μM. NADH could also be used as electron donor though exhibiting a lower affinity (K M = 220 μM). The enzyme was only active with NADP + in the oxidation sense of the reaction, with a K M of 38 μM. Moreover, the activity of this enzyme determined against Veratraldehyde or Veratryl alcohol using NADPH or NADP + as cofactor showed a slight inhibition at elevated concentration of substrate ( Figure 5). However, the apparent K M for Veratraldehyde was 30-fold that for Veratryl alcohol. This explained also that the catalytic efficiency k cat /K M of Pc Aad1p was about 100-fold higher in the reductive than in oxidative sense of the reaction. Reduction activity towards Veratraldehyde has also been described for the enzymes Adh6p and Adh7p from the yeast Saccharomyces cerevisiae [35][36][37].

Conclusion
This study describes the cloning and biochemical properties of an aryl-alcohol dehydrogenase of the white-rot fungus Phanerochaete chrysosporium. It also shows its wide spectrum of activity on various chemicals (natural and non-natural) such as linear aliphatic and aryl-aldehydes, as well as its preference to function in the reductive sense under physiological conditions. This enzyme can be considered in the design of metabolic engineering strategies/synthetic biology systems for biotechnological applications such as the degradation of aromatic inhibitors present in lignocellulosic hydrolysates that impair yeast fermentation, or the microbial production of natural flavours and fragrances like the rose-like flavour compound 2-Phenylethanol. Further studies on the crystal structure of the protein and the determination of the key amino acids in its active site would be extremely

Strain and growth conditions
The white-rot basidiomycete Phanerochaete chrysosporium BKM-F-1767 strain used in this study (CBS 481.73) was purchased from Centraalbureau voor Schimmelcultures (Utrecht, Netherlands) in the form of a freeze-dried fungal culture. The mycelium was inoculated on freshly prepared Difco ™ Potato Dextrose Agar (PDA) plates and incubated at 37°C for four days before storage and maintenance at 4°C on PDA plates or at −80°C in 30% glycerol for longterm preservation. Spore suspensions were prepared after 4-days propagation at 37°C on PDA plates by washing the agar surface with 10 mL of 50 mM sodium acetate buffer at pH 4.5. Spore counts were determined with a counting chamber Thoma double cell.
To induce AAD1 expression in P. chrysosporium, 600 mL of Nitrogen-limited liquid medium was inoculated at 10 4 spores.mL -1 in a 1 L Erlenmeyer flask and cultivated at 37°C and 150 rpm on a TR-225 rotary shaker (Infors AG, Bottmingen, Switzerland) for 1 week. The medium was composed of basal elements, trace elements and vitamins according to [39][40][41] Table 2. Results are the mean ± SEM from at least three separate experiments.
NaOH. Trace elements and vitamins were prepared in 10000-fold concentrated stock solutions and added to the basal solution after autoclaving at 120°C for 20 min.
Analysis by qPCR of Phanerochaete chrysosporium AAD1 gene expression The expression of Pc AAD1 during Nitrogen-limited cultivation was analyzed by real-time PCR (qPCR). The frozen mycelia were disrupted with TissueLyser II grinder for 2  were used to inoculate 150 mL of the same medium in 1 L Erlenmeyer flasks at an initial OD 600 of 0.1. The bacterial biomass was grown at 37°C and 100 rpm until OD 600 0.7-0.9. The production of the recombinant protein was induced by addition of Isopropyl β-D-1-thiogalactopyranoside (IPTG) at 0.1 mM final concentration followed by incubation at 16°C and 120 rpm for 12 h. Bacterial cells were collected by centrifugation (4°C, 10000 g, 1 min), resuspended in PBS buffer at pH 7.3 containing 200 μgÁmL -1 Lysozyme and disrupted by sonication (ten 30 s pulses with a Vibra Cell ™ 72434 ultrasonicator operating at 35% power in 25 W scale). After addition of Triton W X-100 at 1% (v/v) final concentration, the cell lysate was left on ice for 20 min and centrifuged (4°C, 10000 g, 20 min) to remove cell debris.
The recombinant Pc Aad1p fusion protein was purified by a single-step batch affinity chromatography process on Glutathione Sepharose ™ 4B previously equilibrated with PBS buffer at pH 7.3 according to the manufacturer's instructions. The Glutathione Sepharose ™ 4B beads (0.75 mL) were added to the cell lysate supernatant (15 mL) and incubated 2 h at 4°C under gentle agitation (end-over-end rotation) in 50 mL Falcon ™ Conical Tubes (BD Biosciences, NJ, USA). Non-adsorbed proteins were removed by washing the beads with PBS buffer at pH 7.3 several times until the Bradford assay for protein did not react any more. The recombinant protein was eluted with 50 mM Tris-HCl, pH 8.0, containing 10 mM reduced L-Glutathione and stored at 4°C.

Enzyme assays
Enzymatic activity of Pc Aad1p was determined spectrophotometrically using an Agilent HP 8453 UV-visible spectrophotometer (Agilent Technologies, Massy, France). Unless otherwise specified, all assays were carried out at 30°C in 1 mL reaction mixtures using 1 cm optical path length microcuvettes. Reactions were initiated by substrate addition and were monitored by recording the absorption at 355 nm. At this wavelength, the molar extinction coefficients of the substrate compounds could be considered as negligible (less than 4%) compared to that of NAD(P)H (E 355 = 5.12 mM -1 .cm -1 ). The effect of pH was studied at 30°C, using 25 mM MES (pH 5.5 − 6.4), 50 mM HEPES (pH 6.9 − 8.2), 25 mM Tris-HCl (pH 8.8) or 100 mM Glycine-KOH (pH 9.0 − 10.7) as buffers. The temperature dependence was evaluated in 50 mM MES buffer (pH 6.1) in the presence of 0.2 mM 3,4-Dimethoxybenzaldehyde and 0.2 mM NADPH and the reaction was started by adding 9.0 μg of the enzyme. The substrate specificity towards a range of substrates (Table 1) and the kinetic parameters determinations ( Table 2) were determined in 50 mM MES buffer (pH 6.1) using 0.3 mM NADPH and 1 mM substrate in the reduction sense, or in 100 mM Glycine-KOH buffer (pH 10.3) using 0.3 mM NADP + and 10 mM substrate (except for Octanol where 1 mM was used, and for 2-Chlorobenzyl alcohol and 4-Chlorobenzyl alcohol where 3 mM were used) for the oxidation sense. The specific activity towards 3,4-Dimethoxybenzaldehyde (5.1 μmolÁmin -1 Ámg -1 ) and to 3,4-Dimethoxybenzyl alcohol (2.0 μmolÁmin -1 Ámg -1 ) were taken as 100% for the reduction and oxidation reactions, respectively ( Table 1).
The kinetic parameters K M , k cat and K i for aldehyde and alcohol substrates ( Table 2) were computed by fitting initial reaction rates, measured as a function of substrate concentration, to the Michaelis-Menten equation (Equation 1) or, when substrate inhibition was observed, to the uncompetitive substrate inhibition equation (Equation 2) with the non-linear regression Enzyme Kinetics 1.3 module of the SigmaPlot 11.0 package (Systat Software, IL, USA): where V represents the reaction rate, V max is the limiting reaction rate, S is the substrate concentration, K M is the Michaelis constant and K i is the substrate inhibition constant. The catalytic constant k cat of the enzyme for the different substrates was derived from k cat ¼ V max = E ½ . The total enzyme concentration [E] was evaluated using a protein molecular mass of 74.2 kDa. The enzyme kinetic parameters for NAD(P)H and NAD(P) + + were determined with 0.2 mM 3,4-Dimethoxybenzaldehyde and 10 mM 3,4-Dimethoxybenzyl alcohol, respectively. Results are the mean ± SEM from at least three separate experiments.