Materials
All buffering chemicals and substrates including glucose, xylose, mannose, galactose, cellobiose, lactose, mannobiose, β-(1→4)-D-galactobiose, isomaltose and maltose were purchased from Sigma-Aldrich (Germany). The other substrates including cellotriose, cellotetraose, xylobiose, xylotriose, xylotetraose, 32-α-L-arabinofuranosyl-xylobiose (A3X), 23-α-L-arabinofuranosyl-xylotriose (A2XX), 22-(4-O-methyl-α-D-glucuronyl)-xylobiose (U4m2X), 23-(4-O-methyl-α-D-glucuronyl)-xylotriose (U4m2XX), chitosanbiose, 32-β-D-glucosyl-cellobiose, β-(1→4)- D -glucosyl- D -mannose, gentiobiose, kojibiose, sophorose, laminaribiose, and nigerose were purchased from Megazyme (Ireland). The laccase from T. versicolor (38429, Sigma-Aldrich, Germany) was used in oxidation reactions to recycle the electron acceptor, 1,4-benzoquinone (BQ, PHR1028, Sigma-Aldrich, Germany). Methylated cellobiose and lactose was produced in house (details in Additional File 1).
Phylogenetic analysis
To detect proteins of T. myriococcoides CBS 389.93 from the Auxiliary Activity families of interest, an hmm-scan search with default cut-off values (if alignment > 80aa, use E-value < 1e-5, otherwise use E-value < 1e-3; covered fraction of HMM > 0.3) were performed against all proteins of this organism using HMMs from dbCANv9 database. The protein sequences were retrieved from the Centre of Structural and Functional Genomics at Concordia University [18]. The result from hmm-scan was combined with the best hits from blastp search against all sequences from CAZy database [38].
To build the phylogenetic tree of AA3_1 proteins, 206 protein sequences from this subfamily were retrieved from CAZy. Among those, 14 were reported as experimentally characterized. Four AA3_1 proteins of T. myriococcoides CBS 389.93 found from the hmmscan and blastp searches were added to this AA3_1 dataset. The multiple sequence alignment profile was created using MUSCLE [39] and the trimming was discribed by [13]. CBM1 and cytochrome domain information marked on the tree were obtained from the pfam scan against Pfam database v34. The tree was then constructed and edited using FastTree [40] and iTOL [41].
RNA extraction, cloning of xdhA and recombinant protein production
XdhA coding DNA sequence was obtained from the T. myriococcoides CBS 398.93 genome portal [18, 42]. The strain was grown on YpSs agar medium for three days at 45°C at a shaking speed of 220 rpm [43]. Ten plugs from the mycelium grown on agar were used to inoculate a primary culture of 50 mL liquid mycological broth (MB) (1% soytone, 0.4% D-glucose, trace elements, pH 5). Trametes defined medium (TDM) containing 2% of a mix of alfalfa and barley was used for the main culture with an inoculum of 10% volume from the primary culture [43]. Mycelia was harvested after 24 hours of incubation at 45°C at a shaking speed of 220 rpm and was grinded into powder as described by [44]. Total RNA was extracted using the RNeasy® Plant Maxi Kit (Qiagen) and complementary DNA (cDNA) was synthesized using the SuperscriptTM III reverse transcriptase (Invitrogen) according to instructions from manufacturers.
For cloning, xdhA gene was amplified from cDNA by PCR using Phusion® High-Fidelity DNA Polymerase (New England BioLabs Inc.) The following forward and reverse primers were used: 5’-CCCCAGCAACAAAACACCGGCTCAGCAATGCAAACTGCTTCGAAATTAGC-3’ and 5’- GAAGGACGGCGACGGACGGCTTCACGATTCCGCATCCTC-3’. Ligation-independent cloning (LIC) method [45] was used to clone the gene into the LIC-adapted vector pGBFIN49, in which the gene is flanked by 1972 bp of the Aspergillus niger glucoamylase A (glaA) promoter and 701 bp of the A. nidulans trpC terminator.
CRISPR-Cas9 technology was used to replace the glaA gene with the xdhA gene into the genome of the engineered A. niger strain CSFG_9057 (FGSC #A1513 ∆pyrG ∆kusA ∆[prtT amyC agdA] ∆bglA ∆laeA ΔglaATt::trpCTt). Geneious software was used to select the 20 bp guide RNA sequence targeting the glaA gene. The guide RNA sequence was cloned into ANEp8-Cas9 as described by [46]. The expression vector containing the cdh1 gene was co-transformed into CSFG_9057 strain with the ANEp8-Cas9 plasmid containing the glaA-targeting guide RNA sequence using protoplast-mediated transformation method [47]. Transformants were selected on minimal medium without uracil and uridine [48]. Supernatants from transformants were screened for recombinant protein production after growth in MMJ medium supplemented with 0.1% arginine and containing 15 g/L of maltose for induction of protein production [49].
Spores from transformant producing the XdhA recombinant protein were inoculated in 500 mL MMJ medium supplemented 0.1% arginine at a concentration of 2x106 conidia/mL. Supernatant was harvested after six days of stationary incubation at 30°C. Desalting and concentration of the supernatant was done using Vivaflow® cassette as described in the manufacturer’s protocol (Sartorius). Protein production and concentration were checked on SDS-PAGE gel using standard techniques [50].
Purification of the recombinant protein
The secreted recombinant protein was first concentrated to smaller volume using centrifuge filter with a cut-off of 10 kDa. Afterwards, the concentrated fraction was filtrated through 0.45 µm filter and purified with Size Exclusion Chromatography (SEC, HiLoad 16/600 Superdex 200 pg column, GE healtheare, USA) in 10 mM sodium citrate buffer (pH 5) with 0.15 M sodium chloride. The protein purity was checked with SDS-PAGE for each fraction. Semi-purified fractions of XdhA were pooled and loaded to the SEC again for a second round of purification. After that, the purified fractions of XdhA were pooled and exchanged to 10 mM sodium acetate buffer (pH 5) and concentrated using 30 kDA cut off Vivaspin 20 spin columns (Sartorius, Germany). The final protein concentration was measured using the Bradford method (Bio-Rad Laboratories, USA) and the purified protein was snap-freezed and stored in -80 °C in aliquots.
Confirmation of protein purity by deglycosylation, 2-D Electrophoresis and MALDI-TOF
The purified protein was treated with PNGaseF (New England Biolabs, USA) at both denaturing and native conditions at 37 °C for 4h. The samples before and after deglycosylation at both denaturing and native conditions were analyzed by SDS-PAGE, the specific activity of the purified enzyme that was deglycosylated under native condition was also analyzed to evaluate the impact of glycosylation. Subsequently, the purified enzyme was analyzed by 2-D electrophoresis using pH 3-10 strip and Criterion TGX- 4-20 % gel (Bio-Rad Laboratories, USA). The spots from 2D gel were cut, destained, digested with trypsin, and then subjected to MALDI-TOF. Proteins were identified by correlating the mass spectra to the A. niger protein database from Uniprot and the xdhA sequence.
Enzymatic assays
To select the optimal pH for TmXdhA oxidation reactions, the activity of TmXdhA was screened at 30 °C with 5 mM cellobiose and 1 mM 2,6-Dichloroindophenol (DCIP, D1878, Sigma-Aldrich) in McIlvaine’s buffer at pH values from 3.0 to 7.5. The activity of TmXdhA was also analyzed at 30 °C with 5 mM cellobiose and 1 mM BQ in 50 mM ammonium acetate buffer (pH 5.5) to investigate the influence of volatile buffer and e-acceptor. Reduction of DCIP (εabs520nm=7.8 mM-1cm-1) and BQ (εabs290nm=2.24 mM-1cm-1) was followed using an Eon plate reader (BioTek, USA). All reactions were performed in triplicates.
Spectrophotometric assays to determine the initial activity were carried out towards glucose and the di- and oligosaccharides that were oxidized after 24h incubation. Reactions (250 µl) were performed at 30 °C in 50 mM ammonium acetate buffer (pH 5.5) with 5 mM substrate and 1 mM BQ, and the reduction of the BQ (εabs290nm=2.24 mM-1cm-1) was measured for up to 40 minutes at 290 nm.
Steady-state kinetic constants for cellobiose, cellotriose, cellotetraose, xylobiose, xylotriose, and xylotetraose were measured using the conditions for initial activity determination. The reduction rate of BQ was plotted versus substrate concentration (8 points, from 0 to 5 mM). The Michaelis-Menten constant (Km) was estimated by fitting of the data to the Michaelis-Menten equation using GraphPad Prism 6.0 (GraphPad Software, USA). Te measurements were followed using an Eon plate reader (BioTek, USA). All reactions were performed in triplicates.
The hydrolytic activity was tested on five different pNP-sugars. Each reaction mixture contained 2.5 mM substrate in 25 mM ammonium acetate buffer (pH 5.5). 1 µg of TmXdhA was added in each reaction. Reactions were terminated after 20 mins incubation at 30 °C by adding equal volumn of 1 M Na2CO3. The activity was determined by measuring pNP release at 405 nm and calculated with pNP standards.
Laccase activity was measured using 5 mM hydroquinone (HQ, H9003, Sigma Aldrich, Germany), and oxidation of HQ (εabs249=17.25 mM-1cm-1) in 250 μl reaction was followed at 249 nm.
Substrate screening and identification of reaction products with ESI-Q-Tof
The substrate specificity of TmXdhA was determined by using 29 different carbohydrates (Table 2) including monosaccharides, glucosidic-disaccharides with different glucosidic linkages, β-(1→4)-linked disaccharides, cello-/xylooligosaccharides, and C-1 methylated cellobiose and lactose.. All reactions (125 µL total reaction volume) were performed at 30 °C with shaking (400 rpm) in 10 mM ammonium acetate buffer (pH 5.5) containing 25 mU TmXdhA, 5 mM sugar substrate, 1 mM BQ as e-acceptor, and 25 mU T.versicolor laccase for the regeneration of BQ. For the reactions with α-glucosidic-disaccharides, 0.1 mM castanospermine (532673, Sigma Aldrich, Germany) was included to inhibit the hydrolytic side activity [51]. Oxygen availability was not controlled during the reactions. The sampling (50 µL) was done at 3 h for the reaction mixtures containing A3X and A2XX and at 24 h for all other reactions.
Reactions were stopped after sampling by filter through 10 kDa cut off Vivaspin 500 spin columns (Sartorius, Germany). Mass spectrometric analysis was then done for checking if the substrates were oxidized by TmXdhA and to identify the reaction products using Quadruple Time-of-flight (Q-TOF) mass spectrometry with an ESI source (SYNAPT G2-Si, Waters, MA, USA). In practice, a 5 μl sample from each reaction was mixed with 5 μl of 10 mg/ml ammonium chloride and 500 μl 50% methanol in water prior introducing to the ESI-Q-Tof. The analysis was done in negative mode and the ions were collected in m/z range of 50 to 1200 with the parameters developed by [52]. Fragmentation analysis was done to ions generated from oxidized products to identify the reaction products.
Enzymatic conversion of cello- and xylooligosaccharide series
The efficiency of TmXdhA for oxidizing cello- and xylooligosaccharides was compared individually and in mixtures. The reactions (250 µL) were performed at 30 °C in 10 mM ammonium acetate buffer (pH 5.5) with shaking (400 rpm) using 0.2 mM BQ as an e-acceptor in triplicates. For the reactions on each substrate individually, 1.25 mU TmXdhA and 3.75 mU T.versicolor laccase were added to convert 5 mM substrate. For the conversions of oligosaccharide mixtures, the enzyme dose to molarity of reducing end ratio was kept the same, with 3.75 mU of TmXdhA and 11.25 mU of laccase added to convert a mixture of three oligosaccharides in equal molarity (5 mM each). Five series were studied, including cellooligosaccharide series with cellobiose, cellotriose, and cellotetraose and xylooligosaccharide series with xylobiose, xylotriose, and xylotetraose. The three other series mixed cello- and xylooligosaccharides with combinations of xylotriose, xylotetraose, and cellotetraose; xylotriose, cellotriose, and cellotetraose; and xylobiose, cellobiose, and xylotetraose.
Time course sampling (50 µL) was done at 1, 3, 7, and 24h. The oxygen level was not controlled. Reactions were stopped by adding 200 µL 0.1 M ammonia solution directly after sampling and filtering through 10 kDa centrifuge filter. The flow through were lyophilized and redissolved in 50% ACN. The samples were kept frozen at -80 °C prior to further analysis by mass spectrometry and HILIC-ELSD.
Semi-quantitative analysis by mass spectra
5 μL sample solution from each time point was mixed with 5 μl of 10 mg/ml ammonium chloride and 500 μl 50% methanol in water before introducing to ESI-Q-TOF. Ions were analyzed in negative mode and collected in m/z range of 50 to 1200. The oxidized substrates were deprotonized and the original substrates presented as chloride adducts. The following equation was used d to estimate the extent of oxidation completeness employing the relative ratio in peak height between the non-oxidized substrate and the formed product:
Quantification of oxidized products by HILIC-ELSD
The depletion of substrates and formation of the corresponding aldonic acids were followed using an Acquity UPLC coupled with evaporative light scattering detector (ELSD). A 1.7 μm, 2.1*150 mm Acquity UPLC BEH Amide column (HILIC amide, Waters, MA, USA) was used to separate the reaction products. The elution gradient and instrument settings were according to [52]. External standard series with cello- and xylooligosaccharides, and their corresponding aldonic acids were made by injecting 200 ng to 1500 ng of each compound. The injection volume for the samples varied so that the injection amount of each substrate and its reactions product fall in the quantification range. Each standard curve was fitted to a quadratic polynomial equation of f(x)=ax2+bx+c, where f(x) is the peak area, x is the sample amount.
The aldonic acids were made in-house using wild-type glucooligosaccharide oxidase (GOOX) from S. strictum [35]. The GOOX activity was determined according to [53]. For aldonic acids production, 13.5 mg/mL substrate was oxidized for 24h at 37 °C in 10 mM ammonium acetate buffer (pH 5.5) by 400 mU/mL GOOX. Catalase from bovine liver (1200 mU/mL, C30, Sigma, Germany) was also included to remove the formed hydrogen peroxide. The reaction mixtures after 24h incubation were filtered through 10 kDA centrifugal filters and lyophilized to dryness to recover the aldonic acids. The pure aldonic acids were redissolved to proper dilution series with 50 % ACN prior HILIC-ELSD injection.
Sequence alignment and structure prediction
The multiple sequence alignment with TmXdhA, cellobiose dehydrogenase from N. crassa (PDB code 4QI7), and MtCDH cellobiose dehydrogenase from T. myriococcoides (PDB code 4QI4, strain CBS 208.89) was created using MUSCLE [39]. A structural model of TmXdhA was built using a colab version of AlphaFold2 [26] and the model was displayed with FAD cofactor added using PyMOL v 2.1.0 (PyMOL Molecular Graphics Systems, Schrödinger, LLC). The PyMOL molecular graphic system (V 2.1.0, Schrödinger, LLC) was used for structure visualization and structural alignments.
Statistical Analyses
Averages and standard deviations were calculated over three replicate reactions (n = 3). One-way ANOVA (p < 0.05; Graphpad) with F-test and Tukey’s test were performed to ascertain the difference in specific activities towards different substrates and Km values.