Cloning and characterization of an endo-β1 , 4-xylanase gene from Colletotrichum lindemuthianum and phylogenetic analysis of similar genes from phytopathogenic fungus

Colletotrichum lindemuthianum is the etiological agent of anthracnose, one of the main diseases of bean (Phaseolus vulgaris). In this study, the complete cDNAs of two endo-β-1,4-xylanase genes (xyl1) from non-pathogenic (0) and pathogenic (1472) races of C. lindemuthianum were isolated and characterized. To get an insight into the role of endo-β-1,4-xylanases in their different lifestyles, xyl1 gene expression and enzyme activity in mycelia of both races grown in the presence of xylan or P. vulgaris cell walls were investigated. The xyl1 sequence analysis and Clustal alignment revealed the characteristic elements of genes coding for endo-β-1,4-xylanases of the GH11 family. The growth of the two races with glucose as the sole carbon source showed both basal transcription levels of xyl1 and endoxylanase activity. When glucose was substituted with xylan or plant cell walls, xyl1 transcription, and enzyme activity significantly increased in race 1472 as compared to race 0. The pathogenic race degraded xylan faster and grew better than the non-pathogenic counterpart. Seemingly, the regulation of xylanolytic gene expression, enzyme production and the nature of the assimilatory carbon substrates processed by these organisms play a determinant role in their lifestyle. Phylogenetic analyses of XYL1 and endo-β-1,4-xylanases from other fungi revealed a diversification process and separation of proteins from the same fungal species into different lineages.


INTRODUCTION
Colletotrichum lindemuthianum is an economically important phytopathogen and together with its host Phaseolus vulgaris, represents a convenient model for studying the physiological and molecular basis of plantpathogen interactions (Dean et al., 2012;Perfect et al., 1999).This species encompasses different strains or special forms known as races, physiological races or pathotypes identified through the interaction with a group of 12 different cultivars of P. vulgaris, a system used worldwide (Rodríguez-Guerra et al., 2006).A non-pathogenic race and more than 100 pathotypes with different virulence levels have been reported around the world.AFLP analyses of 10 out of the 54 C. lindemuthianum pathotypes identified in México have shown high genetic diversity with several lineages (Gonzalez et al., 1998;Sánchez-García et al., 2009).C. lindemuthianum is an intracellular hemibiotrophic whose physiological races invade the plant in a manner consistent with the gene-for-gene model interactions (Flor, 1971;Oblessuc et al., 2012).Monogenic dominant resistance in common bean cultivars leads to the appearance of localized necrotic spots that are typical of the hypersensitive response (HR) (O'Connell and Bailey, 1988).After penetrating a host epidermal cell in a susceptible cultivar, pathogenic races of the fungus develop an infection vesicle and extend into adjacent cells by means of large primary hyphae, which invaginate without penetrating the cell membrane and thus persisting as a biotrophic interaction (Münch et al., 2008;O'Connell and Bailey, 1988).Once a large area of the plant tissue has been colonized, necrotrophic hyphae develop.This step closely correlates with the production of a number of host cell wall degrading enzymes that are characteristic of phytopathogenic fungi (Dodds et al., 2009;King et al., 2011;Wijesundera et al., 1989).Currently, race 0 is one strain of C. lindemuthianum unable to infect 12 different cultivars of P. vulgaris, which contrasts with race 1472, one virulent isolated in México (Rodríguez-Guerra et al., 2006).This difference makes the comparison of the two races a convenient approach to investigate the role played by host cell wall degrading enzymes in the pathogenicity of C. lindemuthianum.
Analysis of genomic sequences from plant saprophytic and pathogenic fungi has led to the identification of putative genes encoding for carbohydrate-active enzymes (CAZymes) involved in the degradation of plant cell wall.Comparison of these genes has contributed to our understanding of their lifestyle and helped to create infection models (Zhao et al., 2013).For example, biotrophic fungi tend to have fewer CAZymes than necrotrophic and hemibiotrophic fungi.Saprophytic fungi have fewer CAZymes than plant pathogenic fungi, and dicot pathogens often contain more pectinases than monocot pathogens (Zhao et al., 2013).Nevertheless, there have been few studies examining the genetic expression and enzymatic activity of these CAZymes as compared to the vast diversity of substrates presented by hosts.
On this background, here, for the first time, the isolation and characterization of xyl1 cDNA, which encodes an extracellular endo-β-1,4-xylanase in non-pathogenic (0) and pathogenic (1472) races of C. lindemuthianum was reported.Moreover, to understand the role of β-1,4xylanase in the different fungal lifestyles, xyl1 gene expression and endoxylanase activity in mycelia of both races grown in the presence of xylan or P. vulgaris cell walls were investigated.
Finally, the results of Clustal alignment and phylogenetic analyses of XYL1 from C. lindemuthianum and similar enzymes reported in other species of fungi are also presented.

Strains and culture conditions
C. lindemuthianum races 1472 and 0 were kindly provided by Dr. June Simpson (CINVESTAV-IPN, Unidad Irapuato, México), which were reported and characterized by interaction with differential varieties of bean (Phaseolus vulgaris) and molecular strategies (RAPD and AFLP) as the pathotypes 1472 and 0 by González et al. (1998) and subsequently analyzed by Rodríguez-Guerra et al. (2006).C. lindemuthianum was maintained on potato dextrose agar (PDA) (Difco, México) at 20°C.For expression analysis, 1.6 mg (approximately 5 cm 2 ) of mycelia from both races was inoculated into 250 mL-Erlenmeyer flasks containing 50 mL of PD medium and shaken (150 rpm) at 20°C.After 9 days, mycelia was collected by filtration, washed with water and transferred to 125 mL-Erlenmeyer flasks containing 50 mL of modified Mathur's medium (Acosta-Rodriguez et al., 2005), supplemented with 2.5% of glucose, xylan (from beechwood; Sigma-Aldrich, St. Louis, MO, USA) or cell walls from P. vulgaris (cv.Flor de Mayo).Flasks were shaken (150 rpm) at 20°C and after various periods of time, mycelia were collected by filtration, washed with water and stored at -80°C until use.
Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License  Lee et al., 2005) supplemented with one of the carbon sources described above, were inoculated with 1.6 mg dry weight (approximately 5 cm 2 ) of a 9-day-old colony grown on PDA and incubated at 20°C with continuous shaking (150 rpm).After different periods of time, cultures were centrifuged at low speed and the mycelia and supernatants were saved.Fungal growth was measured as mg of wet mycelia except in those experiments where plant cell walls were used as carbon sources.In these cases, growth was measured as the amount of mycelial protein, as residual undegraded cell walls interfered with weight quantification.A 3-mL aliquot of the cell-free supernatant was filtered through a column (1.5 x 6 cm) of Bio-Gel P-6 (Bio-Rad, Hercules, CA, USA), equilibrated and eluted with 50 mM sodium acetate buffer, pH 5.0 (buffer A), at 4°C to prepare the enzymatic fraction.Fractions corresponding to the void volume (Vo) were pooled and the pool, labeled as the filtered extracellular medium (FEM), was used to determine protein and enzyme activity.

Preparation of plant cell walls
P. vulgaris seedlings (cv.Flor de Mayo) were grown for seven days, and cell walls were extracted and purified from hypocotyls by washing in organic solvents as described elsewhere (Fry, 2006).

RNA isolation
Total RNA was purified from mycelia using the Sokolovsky method (Sokolovsky et al., 1990).RNA samples were treated with DNAse I according to the manufacturer's instructions (Invitrogen, Grand Island, NY, USA) to eliminate DNA.The quality and concentration of total RNA were assessed using a Biophotometer Plus system (Eppendorf, Barkhausenweg, Hamburg, Germany).

cDNA isolation, sequencing and analysis
A cDNA fragment (223 bp) of the endo-β-1,4-xylanase gene (xyl1) from race 1472 of C. lindemuthianum was amplified using the TT-3´] according to Kimura (Kimura et al., 2000).Total RNA was isolated from mycelia induced with xylan for 24 h.This fragment was sequenced (data not shown), the specific primers Xyl-D110 (5'-GCGTGAACCAGCCCAGCATC-3') and Xyl-GSP1 (5'-CAGCGTTGAAGTGGTTG-3´) were designed.The cDNA of xyl1 was amplified by 3´and 5´RACE as specified by the manufacturer using the SuperScript III First-Strand Synthesis System (Invitrogen, Grand Island, NY, USA) using total RNA isolated from mycelia of race 1472 induced with xylan for 24 h.Finally, the complete cDNAs of xyl1 were amplified with the specific primers designed on the 5´ UTR, Clxil1-F (5´-ACTTATCATCGTCCGCTTCAACCA-3´) and 3´ UTR Clxil1-R (5´-GCAATCCTCGGAGTTCCAATCTGA-3´), using total RNA of mycelia from both races induced with xylan for 24 h.The PCR incubation mixture was heated at 96°C for 3 min in a Thermocycler (Eppendorf Master Cycler Gradient, Brinkmann, Westbury, NY), followed by 30 cycles of denaturation for 30 s at 96°C, annealing for 35 s at 60°C, extension for 1 min at 72°C and then by a final extension for 10 min at 72°C.All PCR products obtained from both races were ligated into the pCR 2.1 vector (Invitrogen).
The sequences of both strands of cDNA were determined by automatized sequencing using the dideoxy-chain termination method by the commercial service of Macrogen USA.Nucleotide sequences were analyzed using DNAsis (Hitachi), Mega6 (Tamura et al., 2013) and 4Peaks v 1.7.2 software (Griekspoor, 2012).The sequence of the N-terminal secretion signal was identified with SignalP 4.1 Server (Bendtsen et al., 2004).The protein molecular masses, pI values and N-glycosylation sites were calculated using ExPASy Proteomics Server (Wilkins et al., 1999).The nucleotide sequences of 12 endo-β-1,4-xylanases from fungal species were obtained from the NCBI GenBank (Table 1) and were numbered when more than one gene was present in a genome.Multiple sequence alignments were performed with Clustal X software (Larkin et al., 2007) using the default parameters.The signal peptide sequences and N-and C-terminal extensions were excluded.
Data were obtained from three independent experiments performed in triplicate, and analysis of variance (ANOVA) was carried out.The results are reported as the mean and standard errors (SE).P values <0.05 were considered significant.

Assay of endoxylanase activity
Endoxylanase catalyzes the hydrolysis of the 1,4-β-D-xylosidic linkages present in xylan, releasing β-D-xylopyranosyl oligomers and smaller molecules such as β-D-xylopyranosyl mono-, di-and tri-saccharides (Collins et al., 2005;Polizeli et al., 2005).These products can be quantified as reducing sugars using colorimetric methods.Accordingly, reaction mixtures containing 0.5 mL of 0.5% xylan (Sigma), 0.5 mL FEM and buffer A in a final volume of 1 mL were incubated at 30°C.After 10 min, the reaction was terminated by heating the samples in boiling water for 5 min and, after cooling, the amount of reducing sugars was measured using the Nelson-Somogyi method (Nelson, 1944).Activity was expressed, as µg of reducing sugars released in one min.Specific activity was referred to one milligram of protein.

Statistical analysis
Statistical analysis of data was performed using ANOVA with a GLM using a 2 x 3 factorial design (two races and three carbon sources) at each time point.The mean values of the enzyme activity of the fungus and its substrate and the amount of protein and mycelial protein were compared and grouped using the Tukey's test.All analyses were performed with STATISTICA v 10 software (Inc., 2013).

Phylogenetic analyses
Phylogenetic analyses were performed on the deduced amino acid sequence for C. lindemuthianum XYL1 and 11 endo-β-1,4xylanases characterized from other ascomycetes (Table 1) and one sequence from the basidiomycete Lentinula edodes as an outgroup.Deduced amino acid sequences were aligned with Clustal X software (Larkin et al., 2007) using the default parameters.Before phylogenetic analyses, the signal peptide sequences, and N-and C-terminal extensions were excluded.Phylogenetic analysis was performed under Neighbor-Joining and Maximum likelihood criteria using Mega6 (Tamura et al., 2013).The JTT substitution model and gamma correction were used, 1000 bootstrap replicates were performed.The amino WAG evolution model with gamma correction was utilized for maximum likelihood analysis, and the most parsimonious trees were estimated using the heuristic search option (Nearest-Neighbor-Interchange-NNI) with random sequence addition (five random replicates).A WAG+G substitution model was used, and 1000 bootstrap replicates were performed.

Isolation and sequence analysis of xyl1
cDNA encoding an endo-β-1,4-xylanase was isolated from each race and deposited in GenBank (Accessions: KF487129, KM587707).The C. lindemuthianum xyl1 cDNA of race 1472 has 905 bp, with a 5´ UTR of 25 bp and a 3´ UTR of 211 bp (Figure 1).The xyl1 cDNA of race 0 has 751 bp, with a 5´ UTR of 23 bp and a 3´ UTR of 59 bp (Figure 2).At nucleotide and amino acid levels, the sequence of both races showed 100% identity.
Comparison at amino acid level with corresponding sequences in GenBank showed 67, 64, 62 and 61% identity with a xylI of P. tritici-repentis, xil1 of Cochliobolus carbonum, htxyl1 of Helminthosporium turcicum, Xyn22 of Magnaporthe grisea, respectively.The putative protein has an open reading frame of 222 amino acids with a signal peptide cleavage site between Ala 19 and Ser 20 (Figures 1 and 2), according to the SignalP 4.1 web server (Bendtsen et al., 2004), which is consistent with previously reported sequences (Apel-Birkhold and Walton, 1996;Kimura et al., 2000).The putative mature protein (residues 20 to 222) has a calculated molecular mass of 21.71 kDa and a pI of 8.94.A potential N-glycosylation site at Asn 71 was found with the ExPASy Proteomics Server (Ellouze et al., 2011;Wilkins et al., 1999).
The multiple sequence alignment of the deduced amino acid sequences of C. lindemuthianum XYL1 with the endo-β-1,4-xylanases of other fungi revealed the conserved motif EYY where a residue corresponding to the catalytic site is found (Figure 3).In various bacteria and fungi, this motif was reported to be the key segment for enzyme catalytic activity or substrate binding (Apel et al., 1993;Degefu et al., 2001;Li and Ljungdahl, 1994).The deduced amino acid sequence of C. lindemuthianum XYL1 revealed two Glu residues (E 96 , E 187 ) that are highly conserved in xylanases of family GH11 and are likely to be involved in hydrolysis of the glycosidic bond (Figures 1, 2 and 3) (Degefu et al., 2001;Kimura et al., 2000;Tanaka et al., 2005), with one acting as an acid catalyst/base and the other as a nucleophilic residue (Davies and Henrissat, 1995;Sapag et al., 2002).Additionally, XYL1 has an Asp in position 55 that may be necessary for maintaining the optimum pH of these enzymes (Ellouze et al., 2011;Lübeck et al., 1997b).

Analysis of xyl1 expression and production of endoxylanase activity
When used as the principal carbon source, glucose sustained growth of both fungal races but the maximum growth of race 0 (501.67 mg after 10 days) was delayed by approximately two days and was 29% lower when compared with race 1472 (708 mg after 8 days) (P<0.0001)(Figure 4A).Extracellular endoxylanase (XYL) production by the two races was very low, with values of specific activity in the range of 6.4 and 10.5 reducing sugars/min/mg protein after 2 and 12 days of growth, respectively, for the race 1472 (Figure 4A).Corresponding values for the race 0 were 2.5 and 11.83 (Figure 4A).However, even under these conditions, significant differences in XYL production were observed with an increase for race 1472, particularly during early days of growth.For both races, a peak was observed after seven days (P<0.0001).The C. lindemuthianum xyl1 transcript exhibited basal levels of expression (1-fold) in both races (Figure 5A and B).
As described earlier (Hernández-Silva et al., 2007), both races utilized xylan as a carbon substrate, but growth of race 1472 (416.67 mg of mycelia after 8 days) was over 5.5-fold higher and substantially faster than that of race 0 (75 mg of mycelia after 8 days) (P<0.0001)(Figure 4B).Values of 416.67 and 75 mg represent 59 and 15% of the maxima obtained in glucosesupplemented cultures, respectively.Xylan stimulated the production of XYL activity, which was increased and peaked at 90 µg reducing sugars/min/mg protein after seven days of growth and was faster in the pathogenic race (P<0.0001)(Figure 4B).This result was by far the highest activity detected in this study and was 1.5-fold higher than that of race 0, which reached a maximum of 59.67 µg reducing sugars/min/mg protein one day later (Figure 4B).The pathogenic race strongly expressed the xyl1 transcript after 6 to 48 h and three days (16.5-fold) and decreased over the following 4-9 days (Figure 5A).In contrast, the non-pathogenic race weakly expressed the xyl1 transcript between 0 h and three days, increased and peaked after four days (6.6-fold) and decreased over the following 5-9 days (Figure 5A).
The ability of cell walls fractions from P. vulgaris to sustain the growth of C. lindemuthianum and induce endoxylanase was also tested.As observed with other carbon sources, the pathogenic race grew faster and to a greater extent than the non-pathogenic race.This difference was maintained up to approximately 8-9 days of incubation (P<0.0001)(Figure 4C).After that, both races grew with similar rates producing comparable amounts of mycelium protein after 12 days of growth (P>0.05).Due to technical restrictions to measure growth on cell walls, values obtained with this substrate could not be compared with those obtained in soluble carbon substrates.The pathogenic race exhibited low expression levels of the xyl1 transcript during early time points, then a weak peak was observed after 48 h (3.3-fold), and this increased over the following 5-9 days (9.3-fold) (Figure 5B).In contrast, the non-pathogenic race weakly expressed the xyl1 transcript between 0-48 h, then it increased, peaked after 3 and five days (~2-fold) and finally decreased over the following 7-9 days (Figure 5B).Interestingly, after eight days of incubation, the pathogenic race produced an amount of XYL activity (80 µg reducing sugars/min/mg protein) equivalent to 89% of the maximum observed after seven days of growth on xylan (Figure 4C and B).The non-pathogenic race produced a 1.6-fold lower amount of enzymatic activity(50 µg reducing sugars/min/mg protein) that corresponded to 67% of the maximum induced by xylan (Figure 4C and B).It should be noted that in this case, the maximum XYL activity was produced by the non-pathogenic race two days after the maximum produced by the pathogenic race.
The regulation of lytic enzymes in some fungi (St. Leger et al., 1988) and gene expression in several species of Colletotrichum can be modulated by ambient pH (Prusky et al., 2001) and this mechanism of regulation operates in a vast diversity of organisms (Denison, 2000).Therefore, we considered it necessary to determine whether the pH of the extracellular medium varied during the period of incubation under the different culture conditions.Throughout this study, the initial pH of all culture media was adjusted to 5.5.The growth of the pathogenic race in the presence of glucose, xylan or cell walls resulted in a rather irregular profile of pH variation.After 12 days, the pH reached values of 6.0, 6.5 and 7.2, respectively (Table 2).On cell walls, an abrupt acidification of the medium was observed after two days and then it became increasingly alkaline.The growth of the non-pathogenic race on glucose or xylan resulted in a steadier pattern of pH alkalinization to final corresponding values of 6.2 and 6.5.Alkalinization of the medium to pH 7.0 by race 0 grown on plant cell walls followed a more irregular profile.

Phylogenetic analyses
Comparison of amino acid sequences of endoxylanases showed 40 to 67% identity suggesting a diversification process that gave rise to proteins that shared identity mainly in sequence and the structure of the catalytic site.Clustal alignment identified the location of amino acids

DISCUSSION
No differences were found in the coding region of the endo-β-1,4-xylanase of the non-pathogen and pathogen races of C. lindemuthianum.The xyl1 sequence analysis and Clustal alignment with xylanases reported for other fungi revealed the characteristic elements of genes coding for endo-β-1,4-xylanases of family GH11.Induction of cell wall degrading enzymes by different carbon substrates has been studied in some phytopathogenic fungi such as Sclerotinia sclerotiorum (Riou et al., 1991), Sclerotium rolsfii (Sachslehner et al., 1998) and Penicillium sp.(Rahman et al., 2003), among others.The involvement of some of these enzymes in the development of bean anthracnose by C. lindemuthianum race ɣ was first described by Wijesundera et al. (1989).
Results presented here show a clear difference between the non-pathogenic (0) and pathogenic (1472) races of C. lindemuthianum regarding growth, induction of xyl1 transcript expression and production of extracellular endoxylanase activity when they are challenged with different carbon substrates.Accordingly, though maximum growth was obtained on glucose, a readily metabolizable nutrient, basal expression of xyl1 and only trace amounts of XYL were produced by both races; however, although low, a higher XYL production was observed for race 1472, particularly early during incubation.A similar basal production of xylanase and endoglucanase (Sachslehner et al., 1998;Tuncer et al., 2004), pectinases (Hernández-Silva et al., 2007;Oyeleke et al., 2012;Riou et al., 1991) and cellulases (Acosta-Rodriguez et al., 2005;Carle-Urioste et al., 1997;Sharada et al., 2013) has been described in the fungi S. rolsfii, S. sclerotiorum and C. lindemuthianum, respectively.
In fungi, the expression of extracellular hydrolytic enzymes is coordinately regulated by transcriptional activators and repressors (Aro et al., 2005;Tani et al., 2014).The expression of genes encoding xylanolytic enzymes is subject to catabolic repression through the action of CreA under a preferred carbon source and the activation through the action of XlnR under carbon limitation (Cho and Choi, 1999;de Vries and Visser, 2001;Tani et al., 2014).
In this study, significant levels of the xyl1 expression and endoxylanase were produced in pathogenic race 1472 only when the enzyme substrate was available in a medium lacking other nutrients of easier assimilation, such as glucose, indicating that expression of xyl1 can be regulated by the carbon source.It has been proposed that basal levels of endoxylanase commence degradation of xylan generating products that induce further enzymatic activity.Xylose, as the final product of xylan degradation, functions as a regulator of the expression of xylanases, acting as an inducer at low concentrations and as a repressor at high levels (de Vries et al., 1999;de Vries, 2003;Kulkarni et al., 1999;Mach-Aigner et al., 2010, 2012;Stulke and Hillen, 2000).These results support this idea as degradation of xylan occurred when basal levels of endoxylanase activity produced low levels of xylose, which then induced expression of higher levels of expression of xyl1 and enzyme production in the pathogenic race 1472.At later time points, a reduction of activity was observed most likely due to repression by the accumulation of xylose.
The non-pathogenic race of C. lindemuthianum used in this work is unable to infect P. vulgaris, and thus its lifestyle is closer to that of a saprophytic fungus.Therefore, it is possible that the differences found between the non-pathogenic and pathogenic races of C. lindemuthianum are related to the speed of activation of the lytic enzyme genes during the interaction with the host.Additionally, the pathogenic race degrades xylan faster and grows better than the non-pathogenic race, suggesting a different ability in the degradation of this polysaccharide and the use of oligo-or monomeric sugars.As previously described (Hernández-Silva et al., 2007;Lara-Marquez et al., 2011), the expression of Clpnl2 gene and activity of pectin lyase between the two races were similar to that observed in this study when 92% esterified pectin was utilized as the sole carbon source.In other words, the pathogen requires a rapid and higher level expression of endoxylanase activity and other related lyticases for successful interaction with the live plant tissue, which implies an energy cost and the non-pathogen does cannot invest because it feeds on dead plant tissue.
The response of the fungus to cell walls from P. vulgaris is interesting if it is considered that only a certain proportion of the provided substrate corresponds to xylan.A typical primary cell wall contains 9-25% cellulose, 25-50% hemicellulose (whose main structural polymer is xylan), 10-35% pectin and 10% proteins (Cosgrove, 1997).The authors previously demonstrated that P. vulgaris cell walls also induced pectin lyase (PNL) activity in the pathogenic race to levels that represent up to 46% of the maximum obtained with 92% esterified pectin, the best PNL inducer.This value was approximately 2.8-fold higher than that produced by race 0 (Hernández-Silva et al., 2007).This difference is close to that observed for endoxylanase activity in this study.These data indicate that polysaccharides present in the cell wall structure cooperate efficiently to induce a range of polysaccharidases specific for the types of glycosidic linkages present in the cell wall components, thus granting the fungus the ability to degrade the wall barrier efficiently.
A number of evidences indicate a role of ambient pH in the regulation of production of pectinolytic enzymes in fungi such as Penicillium paxilli (Szajer and Czajer, 1985), P. itallicum (Alaña et al., 1989), the avocadopathogen C. gloeosporioides (Drori et al., 2003;Kramer-Haimovich et al., 2006;Yakoby et al., 2000), Trametes trogii (Levin and Forchiassin, 1998) and Aspergillus oryzae (Fontana and Silveira, 2012).In the latter, alkalinization during fruit infection is necessary for the conversion of the biotrophic stage into the necrotrophic stage (Kramer-Haimovich et al., 2006).Also, ambient pH has been described as a regulatory factor related to the pathogenesis of S. sclerotiorum (Rollins and Dickman, 2001), C. gloesporoides (Alkan et al., 2013) and C. acutatum (You et al., 2007).Contrary to these findings, an effect of the pH on growth, expression of xyl1 gene and production of XYL activity by C. lindemuthianum was not observed, which is consistent with previous results on PNL activity (Hernández-Silva et al., 2007).
The results, allow us to hypothesize that the regulation of enzyme expression and nature of the assimilatory carbon substrates processed by these organisms play a determinant role in their lifestyle.The differences in growth, xyl1 expression and production of enzymatic activity between the two races of C. lindemuthianum suggest an adaptation of race 1472 that results in a rapid degradation of xylan, induction of increased activity and utilization of depolymerization products as carbon nutrients.Race 0 does not seem to prefer xylan as a carbon source but instead grows better with bean cell walls, suggesting that differences exist in the utilization of mono-or oligosaccharides on race 1472.The authors observed a similar behavior of other enzymes of the complex involved in the degradation of the cell wall suggesting that it may be a general phenomenon (Acosta-Rodriguez et al., 2005;Hernández-Silva et al., 2007;Lara-Marquez et al., 2011).The differences at this level can be part of the general response of fungi to host components.However, future studies comparing the enzymatic complexes of degradation of more fungal species with different lifestyles will be required to confirm this hypothesis.
Finally, phylogenetic analyses showed a diversification of endo-β-1,4-xylanases and separation of proteins from the same fungal species into various groups or lineages.Similar results were described after a phylogenetic analysis of the nucleotide sequences of the htxyl1 and htxl2 xylanase genes from the corn pathogen, H. turcicum.These genes showed differential expression related to the substrate type (xylan and/or xylose) or stages of infection of maize, suggesting a role in saprophytic or pathogenic phases (Degefu et al., 2004;Ellouze et al., 2011).Here, a phylogenetic separation of other xylanases with differential expression was found; xyl1 and xyl3 in C. carbonum (Apel-Birkhold and Walton, 1996) and xyl4 and xyl5 in F. oxysporum (Gomez-Gomez et al., 2001, 2002), suggesting diversifying selection (Brunner et al., 2013).In this context, the differential expression of cutinases, cellulases, hemicellulases and pectinases related to different stages of the life cycle, namely, biotrophic, necrotrophic and saprophytic, has been reported in the hemibiotrophic pathogen Zymoseptoria tritici (Brunner et al., 2013).Also, purified selection has been detected in many genes, which can be related to the optimization of enzymatic activity.In some of these genes, diversifying selection has also been detected, which is possibly related to the adaptation to the host and/or the life cycle of the fungus (Brunner et

Figure 1 .
Figure 1.Nucleotide and deduced amino acid sequences of the xyl1 gene of C. lindemuthianum race 1472.The signal peptide sequence is underlined, the catalytic residues (Glu-96 and 187) are boxed and an asterisk indicates the stop codon.

Figure 2 .
Figure 2. Nucleotide and deduced amino acid sequences of the xyl1 gene of C. lindemuthianum race 0. The signal peptide sequence is underlined, and the catalytic residues (Glu-96 and 187) are boxed and an asterisk indicates the stop codon.

Figure 3 .
Figure 3. Clustal alignment of endoxylanases.Identical residues and conserved catalytic residues are marked with an asterisk (*).Dots indicate the change of one amino acid for another of the same group.

Figure 4 .
Figure 4. Growth and production of extracellular endoxylanase activity (XYL) by race 1472 and race 0 cultivated in the presence of glucose (A), xylan (B), or plant cell walls (C) as the sole carbon sources.Diamonds, the growth of race 1472; circles, the growth of race 0; stripped bars, XYL activity of race 1472; gray bars, XYL activity of race 0.

Figure 5 .
Figure 5. Expression analysis of xyl1 by RT-qPCR induced with xylan (A) or bean cell walls (B).Stripped bars, show xyl1 expression in race 1472; gray bars, show xyl1 expression in race 0. Each bar indicates the mean of triplicates ± SE of three independent experiments.The symbol "*" indicates significant changes (P<0.05) about the control (glu, glucose).

Figure 6 .
Figure 6.Phylogenetic analyses of endoxylanases from C. lindemuthianum and other fungal species.The tree was constructed using the Neighbor-Joining (NJ) and Maximum likelihood methods and includes only the conserved region of the catalytic domain (190 aa) of the proteins used in the analysis.General topology obtained is represented by the 50% of majority rule consensus tree, in which the NJ posterior probabilities and ML bootstrap support are indicated on the branches.The numbers at the nodes indicate bootstrap values based on 1,000 bootstrap replications.

Table 1 .
Sequences of endoxylanases used for the analyses.

Table 2 .
Kinetic data of pH analysis in races 1472 and 0 of C. lindemuthianum, grown with 2.5% of glucose, xylan or P. vulgaris cell walls.
*Time expressed in days.