Analysis of the ergosterol biosynthesis pathway cloning, molecular characterization and phylogeny of lanosterol 14 α-demethylase (ERG11) gene of Moniliophthora perniciosa

The phytopathogenic fungus Moniliophthora perniciosa (Stahel) Aime & Philips-Mora, causal agent of witches’ broom disease of cocoa, causes countless damage to cocoa production in Brazil. Molecular studies have attempted to identify genes that play important roles in fungal survival and virulence. In this study, sequences deposited in the M. perniciosa Genome Sequencing Project database were analyzed to identify potential biological targets. For the first time, the ergosterol biosynthetic pathway in M. perniciosa was studied and the lanosterol 14α-demethylase gene (ERG11) that encodes the main enzyme of this pathway and is a target for fungicides was cloned, characterized molecularly and its phylogeny analyzed. ERG11 genomic DNA and cDNA were characterized and sequence analysis of the ERG11 protein identified highly conserved domains typical of this enzyme, such as SRS1, SRS4, EXXR and the heme-binding region (HBR). Comparison of the protein sequences and phylogenetic analysis revealed that the M. perniciosa enzyme was most closely related to that of Coprinopsis cinerea.


Introduction
Cocoa (Theobroma cacao L.) cultivation has suffered significant production losses because of diseases that affect its crops (Pereira et al., 1989;Purdy and Schimidt, 1996;Evans, 2007). Among the main pathogens that cause dis-eases in cocoa, the basidiomycete Moniliophthora perniciosa (Stahel) Aime & Philips-Mora has received considerable attention because it is the causal agent of witches' broom disease (Griffith et al., 1994(Griffith et al., , 2003Aime and Phillips-Mora, 2005).
Moniliophthora perniciosa has biotrophic and saprophytic stages (Griffith et al., 2003;Meinhardt et al., 2006). The biotrophic stage is characterized by basidiospore infection in meristematic regions that leads to hypertrophy of infected tissues and the proliferation of abnormal axillary branches known as green brooms. In the saprophytic stage, basidiomata that sporulate on plant material are generated and dry brooms occur as a result of cell death in infected tissues (Ceita et al., 2007), one of the characteristic symptoms of this disease. Oxalate production and calcium oxalate crystal accumulation also play a role in the pathogenesis of witches' broom disease (Rio et al., 2008).
The action of azole-group antifungals has created new perspectives for controlling witches' broom in cocoa trees and has highlighted the importance of analyzing ergosterol biosynthesis in M. perniciosa in order to develop specific inhibitors of this pathway (McQuilken and Rudgard, 1988;Mota et al., 2010). To accomplish this, it is essential to study the development of the disease from molecular and biochemical perspectives. In this sense, the sequencing of the M. perniciosa genome has led to the discovery of genes that are essential for metabolism and development in this species and has resulted in several studies that have focused on gene expression analysis Mondego et al., 2008).
Analyses of the biotrophic and saprophytic stages of M. perniciosa revealed high gene expression levels of the cytochrome P450 superfamily, this altered expression included lanosterol 14a-demethylase, a key enzyme in the ergosterol biosynthetic pathway in fungi that is a target for antifungals and is encoded by the ERG11 gene Pires et al., 2009). Lanosterol 14a-demethylase belongs to the CYP51 family of the cytochrome P450 superfamily, which is notable for being the only cytochrome P450 family that is present in all biological kingdoms (Waterman and Lepesheva, 2005;Lepesheva and Waterman, 2007).
Antifungals that affect the ergosterol biosynthetic pathway have been used for decades. Azole-group antifungals are potentially useful disease-modulating agents because of their specific binding to lanosterol 14a-demethylase and their selective inhibition of the removal of the methyl group by this enzyme, which leads to the accumulation of unsaturated intermediates and to the depletion of ergosterol (Hof, 2001;Carrillo-Muñoz et al., 2006;Sheng et al., 2009). Ergosterol is an important component of the fungal cell membrane that regulates membrane fluidity and permeability (Barrett-Bee and Dixon, 1995;Lees et al., 1995;Veen and Lang, 2005).
In this report, we describe the first identification, molecular characterization, cloning and phylogenetic analysis of the ERG11 gene that encodes lanosterol 14a-demethylase, an enzyme that is essential for the survival and pathogenicity of M. perniciosa. The results presented here should be useful in identifying antifungal drugs active against this enzyme and, consequently, in controlling witches' broom disease in cocoa trees.

Ergosterol pathway in Moniliophthora perniciosa and sequence analyses
The sequences examined in this work were retrieved from the Moniliophthora perniciosa Genome Sequencing Project database (www.lge.ibi.unicamp.br/vassoura/) and were used as the primary source for identifying genes of the ergosterol biosynthetic pathway in M. perniciosa. With this database, it was possible to obtain the genomic consensus sequences that were predicted to encode the main enzymes of this pathway, such as lanosterol 14a-demethylase, and to compares these sequences with those from other organisms.
The genes were accurately analyzed using ab initio programs for gene prediction, such as AUGUSTUS (Stanke et al., 2006) and GeneMark (Ter-Hovhannisyan et al., 2008). These analyses were then combined with the Basic Local Alignment Search Tool (BLAST) algorithm (Altschul et al., 1990) to align the sequences with those of phylogenetically related species. This approach allowed the identification of expressed regions, the presence and number of introns, and of regions bordering the ERG11 gene. BioEdit software (v. 7.1.3) (Hall, 1999) was used to align the predicted genomic DNA and cDNA sequences with sequences determined by sequencing.
Signal peptide prediction was done using SignalIP 4.0 software (Petersen et al., 2011), and probable transmembrane domains were determined using the Phobius program (Kall et al., 2007). The ProtoParam tool was used to analyze the physical and chemical parameters of the proteins, such as the theoretical isoelectric points (pIs) and molecular masses (Gasteiger et al., 2005). Additionally, the conserved domains, functional sites and protein family were determined using PROSITE software (Sigrist et al., 2009).
The lanosterol 14a-demethylase sequences of other organisms were identified in GenBank and analyzed using the Protein BLAST (Blastp) tool. The M. perniciosa sequence was used as a reference (Table 1).

M. perniciosa strain and culture conditions
The 948F strain of M. perniciosa (Ilhéus, Bahia, Brazil), which was kindly provided by the Phytopathology Laboratory at the Cocoa Research Center (Centro de Pesquisas do Cacau da Comissão Executiva do Plano da Lavoura Cacaueira -CEPEC/CEPLAC, Ilhéus, Bahia, Brazil), was used in the experimental procedures. The strain was inoculated in Petri dishes containing Sabouraud agar solid medium (Sigma-Aldrich) to maintain mycelial discs and in YPD liquid medium (1% yeast extract, 1% peptone and 2% dextrose) (Sigma-Aldrich) and incubated statically at 27°C for 15 days to obtain an adequate mycelial mass for genomic DNA and total RNA extraction.

Genomic DNA extraction
Total DNA was extracted from M. perniciosa mycelia according to the protocol described by Raeder and Broda (1985). The DNA concentration was determined spectrophotometrically (Novaspec II, Pharmacia Biotech) and the quality and quantity of the extracted DNA were analyzed in 1% agarose gels.

RNA isolation and reverse transcription
Total mycelial RNA was obtained using a RiboPure kit (Ambion) according to the manufacturer's recommendations. The first cDNA strand was produced from 1-3 mg of total RNA using the RevertAid H Minus M-MuLV Reverse Transcriptase kit (Fermentas Life Sciences) according to the manufacturer's recommendations.

Primer design and amplification conditions
Specific primers for PCR and RT-PCR were designed based on the predicted ERG11 sequence in M. perniciosa ( Figure S1). We aimed to obtain complete fragments of DNA and cDNA (1,668 bp) using the primers LanFNcoIfull and LanRXhoI. A fragment (1,496 bp) that lacked the signal peptide was obtained using the primers LanFNheIfull and LanRNheI. A fragment (1,400 bp) that lacked the first transmembrane helix was obtained with the primers LanFNheIdel1ahel and LanRNheI. Lastly, a fragment (1,316 bp) that lacked the regions corresponding to the first and second transmembrane helices was constructed using the primers LanFNheIdel2ahel and LanRNheI (Table 2).
Amplifications were done using a Mastercycler Gradient thermocycler (Eppendorf) with variable annealing temperatures. The following program was used: 3 min at 94°C, 35 cycles that included 1 min at 95°C, 1 min at 60°C and 2 min at 72°C and a final extension of 5 min at 72°C. ERG11 of Moniliophthora perniciosa 685 The variable temperatures used to amplify the genomic DNA were: The PCR products were visualized on 1% agarose gels containing ethidium bromide (0.5 mg/mL) and the gel was photographed using an EDAS 290 photodocumentation system (Kodak).

Sequencing of PCR products
The DNA and cDNA sequences that corresponded to ERG11 were confirmed by sequencing the purified PCR products on a MegaBACE 1000 capillary sequencer (Amersham Biosciences -GE Healthcare) using a DYEnamic ET dye terminator kit (GE Healthcare) according to the manufacturer's instructions.
Cloning of cDNA fragments of ERG11 cDNA fragments of M. perniciosa ERG11 were cloned into the pET-28a vector (Novagen) and recombinant plasmids that contained the ERG11 gene were transformed in E. coli Rosetta (DE3) that were grown in LB medium (1% Tryptone, 1% NaCl and 0.5% yeast extract) that contained appropriate antibiotics for selection. The identities of positive clones were confirmed by PCR and sequencing and were subsequently used for expression studies and the purification of lanosterol 14a-demethylase.

Sequence alignment and phylogenetic analysis
Sequences obtained from the NCBI database (Table 1) and the predicted sequence from M. perniciosa were used to perform alignments with BioEdit (Hall, 1999) and ClustalW2 (Thompson et al., 1994) and a data matrix was used for the phylogenetic analyses. Twenty sequences were analyzed and Mycobacterium tuberculosis was used as an outgroup.
Maximum parsimony analysis (MP) was done with PAUP 4.0b10 software (Swofford, 2002) using Fitch parsimony (unordered characters and equal weighting) and branch-swapping heuristic search algorithms. The search 686 Ceita et al.   consisted of 1,000 replicates in which data were randomly added using the tree bisection and reconnection (TBR) algorithm; 15 trees per replicate were saved and all gaps were treated as missing data. A strict consensus tree was then obtained and the relative clade support was determined using the bootstrap -BS test (Felsenstein, 1985). Only values that were ³ 75% are shown and the retrieved trees were edited using TreeView v.1.6 software (Page, 1996).

Results
Isolation and characterization of the ERG11 gene in M. perniciosa The complete ERG11 gene that was amplified from genomic DNA contained 2,030 base pairs (bp) while the gene that was amplified from cDNA was 1,668 bp long. The gene product with no region corresponding to the signal peptide contained 1,858 bp ( Figure 1) while the product obtained by RT-PCR contained 1,496 bp (Figures 2 and  S2).
Removal of the regions that corresponded to transmembrane insertions 1 and 2 generated genomic PCR products of 1,712 bp and 1,575 bp, respectively ( Figure 1) and cDNA products of 1,400 and 1,316 bp, respectively ( Figure  2), as previously predicted by our bioinformatic analysis. The reliability of the DNA and cDNA base composition of ERG11 was confirmed by sequencing and demonstrated the accuracy of the gene prediction models that we used. Sequence analysis showed that M. perniciosa ERG11 has eight exons and seven introns (positions 166-215, 333-385, 509-561, 1027-1076, 1250-1301, 1771-1821 and 1915-1967).

Structural prediction and physicochemical parameters of lanosterol 14a-demethylase
Determination of the sequence of the ERG11 transcript allowed prediction of the primary structure of lanosterol 14a-demethylase, which consisted of 555 amino acids; the predicted molecular mass was 62 kDa and the theoretical pI was 6.34. The ERG11 ORF in M. perniciosa exhibited two putative, highly hydrophobic transmembrane domains located near the N-terminal region of the protein at amino acid positions 33-55 and 62-84, which suggested a probable insertion in the endoplasmic reticulum membrane (Lepesheva and Waterman, 2007). In addition, signal peptide prediction identified a cleavage site between amino acid positions 46 and 47 and suggested that this protein was directed to the endoplasmic reticulum through the secretory pathway.
Analysis of CYP51 in M. perniciosa identified an iron-binding domain at amino acid positions 493-502 that is characteristic of proteins belonging to the cytochrome P450 superfamily.

Sequence alignment and detection of conserved domains
The alignment of ERG11 sequences from different organisms allowed the identification of several conserved domains that are characteristic of P450 proteins and, more specifically, of CYP51 proteins (EC 1.14.13.70). The C-terminal region of these proteins contains a binding domain that is common to the heme group (FxxGxxxCxG). This domain was identified in M. perniciosa CYP51 and contained a conserved cysteine residue at amino acid position 500 that is responsible for the binding of iron to this protein (Figure 3). The SRS1 and SRS4 domains (substrate recognition regions) are relatively well conserved in CYP51.
The SRS1 (YxxF/L(I)xxPxFGxxVxF/YD/a) and SRS4 regions (GQ/hHT/sS) of M. perniciosa CYP51 were very similar to those of other organisms (Figure 3). The SRS1 region (B'helix/B'/loopC) forms the upper surface of the substrate binding cavity of P450 and the replacement of some conserved residues in this region can lead to partial or complete loss of CYP51 activity in Homo sapiens and Mycobacterium tuberculosis (Zhao et al., 2007). The SRS4 domain is located in the C-terminal region of the P450 I-helix, which forms the right wall of the distal surface of the substrate binding cavity (Lepesheva and Waterman, 2007). The EXXR motif, which is characteristic of the K-helix of CYP51 , is also highly conserved (Figure 3).

Phylogenetic analysis
Phylogenetic analysis based on protein sequence data allowed assessment of the evolutionary relationships among CYP51 proteins of 20 organisms (Figure 4). Maximum parsimony analysis yielded three equally parsimonious trees, each with 2,700 steps (TreeBase access TB2:S14729). The maximum parsimony result was very similar to the maximum likelihood analysis (not shown), with the branches that clustered bacteria, fungi and animals being clearly distinct. The formation of these clusters agreed with the accepted phylogeny for the major kingdoms of organisms.
Species of the phylum Ascomycota were grouped in one clade and those belonging to the Basidiomycota were grouped in another (Figure 4). This sister group relation-ERG11 of Moniliophthora perniciosa 687 ship for CYP51 phylogeny agreed with the general taxonomy of fungal species (James et al., 2006).

Discussion
Many of the genes that comprise the ergosterol biosynthetic pathway in M. perniciosa have been identified by genomic data mining. Our results indicate that the M. perniciosa genes that are part of this pathway show great similarity with those of the Basidiomycota, such as Coprinopsis cinerea (ERG1, ERG2, ERG9, ERG11, ERG24, ERG25 and ERG26), Laccaria bicolor (ERG5, ERG6 and ERG7) and Cryptococcus gatti (ERG4) (Martin et al., 2008;Stajich et al., 2010;D'Souza et al., 2011). These data demonstrate the conservation of genes and biosynthetic pathways that are essential for the survival of related organisms.
The biosynthetic pathways for ergosterol and steroids have been studied mostly in model organisms such as Kluveromyces lactis and Saccharomyces cerevisiae, and in those that cause public health problems, such as Aspergillus fumigatus, Candida albicans, C. dubliniensis, C. glabrata and C. tropicalis (Goffeau et al., 1996;Dujon et al., 2004;Jones et al., 2004;Nierman et al., 2005;Butler et al., 2009). However, there is little information on the steroid bio- 688 Ceita et al. synthetic pathways of phytopathogenic filamentous fungi (except for A. fumigatus). Gaps in our knowledge concerning the genes of this pathway in some organisms reflect a lack of genomic data or the need for further studies related to sequences that have already been deposited in genomic databases and EST libraries. The results of the present study fill a gap in our knowledge of this metabolic pathway. This pathway is particularly important for the development of M. perniciosa and may be useful in designing new antifungal drugs.
The detection of ERG11 gene expression during the mycelial stage of M. perniciosa, as shown here, suggests that mycelial development requires intense steroidal activity. ERG11 is highly expressed since ergosterol is an essential sterol. Ergosterol is the major component of the fungal cell membrane and is essential as a bioregulator of membrane fluidity, asymmetry and integrity (Barrett-Bee and Dixon, 1995;Carrilo-Muñoz et al., 2006). The membranes of eukaryotic cells act as barriers between the inside of the cell or the lumen of organelles and the corresponding exter-nal environment. These membranes also contain proteins that selectively transport molecules or act as enzymes in different metabolic activities (Veen and Lang, 2005).
Macroarray analysis of the differential expression of ERG11 during M. perniciosa development, which was part of effort to construct a cDNA library for M. perniciosa, indicated that this gene was hyperexpressed during the white mycelial stage and had a lower level of expression during the primordial stage, prior to the formation of basidiomata (Pires et al., 2009). This finding indicated that expression levels of this gene vary during fungal development.
The strong conservation of regions SRS1, SRS4 and the EXXR motif of lanosterol 14 a-demethylases among organisms belonging to different groups indicated their importance in the recognition of basal substrates. Despite the variability among CYP51 sequences, the basic structure of this protein needs to be maintained in order to retain its catalytic activity, and this structure is determined by the HBR and the more conserved substrate recognition sites (SRS1 and SRS4) ( (Table 1) were indicators of the high similarity between the CYP51 orthologs of these phylogenetically related organisms. As with M. perniciosa, introns have also be observed in the homologous sequences of other basidiomycetes such as C. cinerea, C. neoformans and P. chrysosporium, that contain six, eight and six introns, respectively (Mellado et al., 2001;Revankar et al., 2004;Warrilow et al., 2010). Ustilago maydis is the only Basidiomycete that lacks introns in the ERG11 gene (Revankar et al., 2004). Furthermore, introns are usually not found in Ascomycota such as C. albicans, P. carinni and S. cerevisiae. The average size of the ERG11 introns in M. perniciosa was 52 bp, as estimated for other genes analyzed by genome sequencing . In contrast to the positional variation of ERG11 introns among organisms that may reflect evolutionary events in the P450 superfamily, the HBR, which is highly conserved in this family of proteins, has not suffered recombination events or exon shuffling (Rozman et al., 1996).
In phylogenetic analysis, the formation of welldefined clusters confirmed the accepted phylogeny for the major kingdoms of organisms. Fungi formed two distinct clades. The organisms that showed higher identity values with the CYP51 sequences of M. perniciosa (C. cinerea, P. chrysosporium, C. gattii, C. neorformans and U. maydis) formed a distinct clade that included M. perniciosa and had a significant bootstrap value ( Figure 4); this finding indicates that Mper ERG11 is phylogenetically closer to C. cinerea ERG11 than to all homologs of the other Basidiomycota with well-characterized sequences.
The cloning of M. perniciosa CYP51 in expression vectors will enable the recombinant production of lanosterol 14a-demethylase and should provide sufficient protein for elucidation of the structure of this enzyme. A similar approach applied to other enzymes in this pathway should contribute to the design and optimization of more effective drugs against this phytopathogenic fungus.