Identification of differentially expressed genes from Trichoderma harzianum during growth on cell wall of Fusarium solani as a tool for biotechnological application

Background The species of T. harzianum are well known for their biocontrol activity against many plant pathogens. However, there is a lack of studies concerning its use as a biological control agent against F. solani, a pathogen involved in several crop diseases. In this study, we have used subtractive library hybridization (SSH) and quantitative real-time PCR (RT-qPCR) techniques in order to explore changes in T. harzianum genes expression during growth on cell wall of F. solani (FSCW) or glucose. RT-qPCR was also used to examine the regulation of 18 genes, potentially involved in biocontrol, during confrontation between T. harzianum and F. solani. Results Data obtained from two subtractive libraries were compared after annotation using the Blast2GO suite. A total of 417 and 78 readable EST sequence were annotated in the FSCW and glucose libraries, respectively. Functional annotation of these genes identified diverse biological processes and molecular functions required during T. harzianum growth on FSCW or glucose. We identified various genes of biotechnological value encoding to proteins which function such as transporters, hydrolytic activity, adherence, appressorium development and pathogenesis. Fifteen genes were up-regulated and sixteen were down-regulated at least at one-time point during growth of T. harzianum in FSCW. During the confrontation assay most of the genes were up-regulated, mainly after contact, when the interaction has been established. Conclusions This study demonstrates that T. harzianum expressed different genes when grown on FSCW compared to glucose. It provides insights into the mechanisms of gene expression involved in mycoparasitism of T. harzianum against F. solani. The identification and evaluation of these genes may contribute to the development of an efficient biological control agent.


Background
Trichoderma harzianum is a soil-borne filamentous fungus that protects crop plants from attack by a range of pathogenic fungi [1]. Species of the genus Trichoderma are widely known for their biotechnological interest, however their use as biocontrol agents requires a comprehensive analysis of the biological principles of their action. Their antagonistic abilities are described as a combination of several mechanisms, including nutrient competition and direct mycoparasitism, which involves the production of antifungal metabolites and cell wall-degrading enzymes [2][3][4][5]. The use of these species as biocontrol agents represents an environmentally friendly alternative to chemical fungicides, and furthermore, some of their genes have been used to improve plant resistance to pathogens and salt stress [6]. Recently, biologically important proteins from Trichoderma have been successfully produced for agricultural and industrial applications [1].
The genus Fusarium comprises a wide and heterogeneous group of fungi that causes economically harmful diseases in many crops, such as soybean (Glycine max), tobacco (Nicotiana tabacum) and common beans (Phaseolus vulgaris), reducing both the quality and the quantity of their products [7]. Fusarium solani occurs in practically all of the common bean-producing regions in Brazil, and has been controlled through the use of chemical fungicides [8]. Studies on the antagonistic capacity of T. harzianum have revealed that it represents an important alternative to the use of chemical fungicides [9]. Comprehensive analysis of the molecular mechanisms used by T. harzianum during interaction with F. solani is required in order to identify the molecular determinants of its role as a biological control agent [10,11]. The identification of the Trichoderma genes involved in these mechanisms and analysis of their expression profiles can provide researchers with biotechnological tools that exhibit anti-fungal activity and that could potentially be used as transgenes capable of inducing resistance to pathogens in economically valuable plants.
The aim of the present study was to provide helpful insights into the mechanism of T. harzianum in its action against F. solani. We used a suppression subtractive hybridization (SSH) approach to obtain genes that are differentially expressed during T. harzianum growth on F. solani cell wall or glucose. We analyzed the differentiallyexpressed genes for homology and classified them into functional categories. Finally, we discuss the possible functional roles of the genes identified in the interaction between T. harzianum and F. solani by using quantitative real-time RT-PCR.

Results and discussion
Identification of differentially expressed genes during growth of T. harzianum in FSCW or glucose In this study, a suppression subtractive hybridization (SSH) approach, which is an efficient method for the isolation of differentially expressed genes, was used to isolate and identify genes that are differentially expressed during T. harzianum growth on FSCW or glucose. Samples of mRNA from four incubation times (24, 36 and 48 h) were used to construct a unique cDNA library. In order to obtain a cDNA library enriched for sequences representative of those genes up-regulated in the presence of FSCW-library, cDNA from T. harzianum grown in FSCW was used as the tester and cDNA from T. harzianum grown in glucose medium as the driver (Glc-library). The Glc-library was enriched with possible genes down-regulated during T. harzianum growth on FSCW, a mycoparasitism-related condition. The approach based on the construction of cDNA libraries from a mixture of conditions was previously used successfully in T. harzianum [10,12].
In the FSCW-library, 417 reads were generated, which were grouped into 77 Unigene clusters (representing 39 contigs and 38 singletons), with an overall EST redundancy of 91%. The inspection of these Unigene clusters detected matches for 64% using blastx. Seventy eight reads were generated from the Glc-library, which represented 47 Unigene clusters, including 19 contigs and 28 singletons, with an overall EST redundancy of 64%. Among these clusters, 57% presented sequence similarity with GenBank entries using the blastx algorithm. The comparison between the sequences from both libraries indicated that there was no overlap between them, suggesting that the subtraction approach was successful and that the two libraries were in fact enriched for sequences from genes differentially expressed under each condition. All ESTs were submitted to GenBank (accession numbers from JK840901to JK841024).

Functional annotation of the differentially expressed genes
ESTs were annotated according to Gene Ontology (GO) guidelines (Ashburner et al. 2000) with Blast2GO, a universal web-based annotation application [13]. Genes from both libraries were allocated to the main GO categories and the distribution of ESTs in these categories was taken as a measure of gene concurrence across the two libraries ( Figure 1). For this purpose, the total number of unique sequences from the two libraries that possessed an assigned GO term within each of the three organizing principles of GO (Biological Process, Molecular Function and Cellular Component) was taken as 100%. Functional annotation of the genes from the FSCW-library indicated that the highest percentage of GO terms was seen in the categories from Biological Processes: localization (40%), cellular process (40%), and metabolic process (64%), as well as the categories related to Molecular Functions: binding (51%), transporter activity (26%), catalytic activity (62%), and hydrolase activity (29%). The GO terms related to regulation: biological regulation (4%), enzyme regulator activity (4%), and transcription regulator activity (4%) as well as proliferation: cell proliferation (4%), demonstrated the lowest percentage values. Genes involved in sphingolipid metabolism, N-glycan biosynthesis, glycosaminoglycan degradation, glycosphingolipid biosynthesis, other glycan degradation, oxidative phosphorylation, amino sugar and nucleotide sugar metabolism, amino acid and water transport, hydrolytic activity and energy related processes were identified during T. harzianum growth in FSCW. Moreover, proteins found to be associated with the response of T. harzianum to the presence of phytopathogens included: MAP kinases (serine-threonine protein kinase and chk1), enzymes that are essential for maintenance of the cell wall integrity, hyperosmotic stress tolerance proteins, pathogenicity factors, a serine protease and a QID 74 protein considered to be involved in the mycoparasitism [14][15][16]. Annotation and KEGG analysis of the identified sequences demonstrated a clear relationship with the biological processes and molecular functions required during mechanism of biocontrol [1,5].
Functional annotation of the sequences from the Glclibrary identified genes belonging to the following categories of Biological Processes: cellular process (94%), cellular metabolic process (63%), metabolic process (69%), catabolic process (31%), biosynthetic process (56%), and also to the following categories from Molecular Functions: catalytic activity (65%) and lyase activity (35%). KEGG analysis showed that the major metabolism pathways corresponded to biosynthetic pathways: glycerophospholipid metabolism, and glycolysis/gluconeogenesis. This fact is consistent with the extensive metabolic activity expected for a filamentous fungus growing on a rich medium with an easily assimilable substrate [17].
Expression analysis of genes from T. harzianum during growth in FSCW Amongst the differentially expressed genes identified in the SSH analysis, twenty-eight genes were selected based on their predicted function or involvement in Trichoderma development, metabolism and biocontrol activity ( Table 1). All of these genes were further analyzed by quantitative real-time RT-PCR (RT-qPCR) in order either to validate the results obtained by the SSH method or to understand the kinetics of their expression in the evaluated conditions. We then extended the studies on the expression of these genes by evaluating samples obtained at 24, 36 and 48 hours during growth of T. harzianum on FSCW (Table 2 and Additional file 1). The RT-qPCR results suggest that the genes identified as altered in the FSCW-library and annotated as hydrolases (acid sphingomielinase (asm), β-1,3-endoglucanase (bgn), chitinase 33 (chit), endochitinase 42 (endo), exo-rhamnogalacturonase (exo) and glycosyl hydrolase (glyc)) were up-regulated with the highest expression values at 24 hours with a statistically significant decrease on these values at 36 and 48 hours of growth. Two other genes identified in the FSCW-library which were annotated in the catalytic activity category (amine oxidase (aoc) and phospholipase d (pld)), were down-regulated in all times of growth (p < 0.01). Genes associated with binding activity such as checkpoint-like protein (chk1), serine threonine-protein kinase (sck1) and senescence-associated protein (sag) presented the highest expression values at 24 hours and a decrease on these values at 36 and 48 hours of growth (Table 2 and Additional file 1). The observed expression of mannose-binding lectin (mbl2) increases over time ( Table 2 and Additional file 1).
A serine protease (ser) transcript was also identified in the FSCW-library and RT-qPCR analysis indicated that, differently from hydrolases, the highest expression value was detected after 36 hours of growth (Table 2 and Additional file 1). Moreover, similar to serine protease expression, three genes annotated in the transport GO category (peptide transporter (ptr2), aquaporin (aqp) and a DUF895 domain membrane protein (duf)), showed the highest expression values after 36 hours of growth (Table 2 and Additional file 1). Finally, RT-qPCR expression analyses were performed with qid74 and cfem (eight cysteinecontaining domain), genes that correspond to receptor activity on GO. Their highest expression values were detected at 36 hours and 24 hours of growth on FSCW, respectively.
RT-qPCR expression analyses with 10 genes identified on the Glc-library were also conducted to observe their expression profile during the growth of T. harzianum in FSCW. Our data showed that all genes analyzed were down-regulated (p < 0.01) ( Table 2 and Additional file 1). Enolase (eno), glyceraldehyde 3-phosphate dehydrogenase (gapd) and pyruvate decarboxylase (pdc), enzymes that are involved in glycolysis and/or glyconeogenesis pathways, showed similar expression profiles at 24 and 36 hours of growth (Table 2 and Additional file 1). Also, the genes with binding activity, hsp98 and a c2h2 (zinc finger domain protein), showed the lowest expression values after 24 hours of growth in FSCW. Moreover, the results obtained with genes corresponding to transporter activity on GO (hexose transporter-like protein (ht) and zinc-regulated transporter (zt)) demonstrated the same time-course expression profiles as the genes that present binding activity (Table 2 and Additional file 1). RT-qPCR expression analysis of norsolorinic acid reductase (norA), phosphatidylserine decarboxylase (psd) and coproporphyrinogen oxidase (cpox) demonstrated that these genes are down-regulated during growth of T. harzianum in FSCW (Table 2 and Additional file 1). These enzymes are essential in some filamentous fungi in the biosynthetic pathways of aflatoxins, glycerophospholipid metabolism and heme groups [18,19].
Expression analysis of genes from T. harzianum during interaction with F. solani In order to identify genes potentially involved in biocontrol, RT-qPCR was performed using total RNA from dual cultures of T. harzianum and F. solani in three different interaction stages: before contact, during contact and after contact (Table 3 and Additional file 2). Only the genes identified in the FSCW library were analyzed by RT-qPCR. As a control, a confrontation assay was conducted where T. harzianum was challenged with itself. RT-qPCR analysis showed that the genes studied were not expressed when T. harzianum was challenged with itself. We have previously reported that direct confrontation assays are a powerful tool to study the phenomenon of mycoparasitism by T. asperellum and T. harzianum [10,20]. During the confrontation assay most of the genes studied were down-regulated before contact, and the expression of these genes occurs only after contact with the host. However, three genes (asm, mbl2 and aqp) presented high expression values before contact. Previously works showed that Trichoderma species are able to sense the presence of its host and specific genes are expressed already before contact [21]. Interestingly, with the exception of serine protease (ser) all genes are up-regulated aftercontact, when the interaction has been established (Table 3 and Additional file 2).
We analyzed the expression of three genes encoding cell wall degrading enzymes (CWDE): β-1,3-endoglucanase (bgn), chitinase 33 (chit) and endochitinase 42 (endo). In agreement with previous studies, transcripts encoding these enzymes were highly expressed mainly after contact indicating intense cell wall degradation in this stage [6,10]. The expression of the genes encoding these enzymes was induced by FSCW (Table 2 and Additional file 2). These enzymes could be also induced by metabolites secreted by the host, and is strongly repressed by glucose [22,23]. These genes play a major role in the mycoparasitic activity against the pathogens especially R. solani, F. oxysporum and S. sclerotiorum [5].
Exo-rhamnogalacturonase (exo) and glycosyl hydrolase (glyc) annotated as hydrolases were also up-regulated after contact (Table 3 and Additional file 2). Exo-rhamnoga lacturonases belong to the group of pectin-degrading enzymes (PDE) and could be involved in cell degradation of F. solani. For instance, an endopolygalacturonase (PDE) produced by T. harzianum are known to be involved in the cell wall degradation of Rhizoctonia solani and Pythium [24]. Glycosyl hydrolases are a widespread group of enzymes that hydrolyse the glycosidic bond between two or more carbohydrates, or between a carbohydrate and a non-carbohydrate moiety. A classification system for glycosyl hydrolases has led to the definition of 85 different families, including amylases, cellulases, β-glucanases and chitinases [25]. Thus, the characterization of the glycosyl hydrolase identified in this work needs to be made to show its role in mycoparasitism.
The expression of genes related with amino acids metabolism (ser, aoc and ptr2) was also analyzed. Serine protease (ser) was up-regulated before and during contact, but down-regulated after contact (Table 3 and Additional file 2). These data show that these enzymes play an important role in the early stages of the interaction between T. harzianum and F. solani. On the other hand, the expression of amino oxidases (aoc) and a peptide transporter (ptr2) are up-regulated after-contact when the expression of serine protease decreased. Some proteases from Trichoderma species have been identified as having biocontrol functions, including aspartyl protease, serine protease and subtilisin-like protease [5,26]. These proteases can take part in the host cell wall breakdown process or act as proteolytic inactivators of pathogen enzymes. Amine oxidases (AO) are a large group of enzymes catalyzing oxidative deamination of amines to form the corresponding aldehydes, hydrogen peroxide and ammonia. AO in filamentous fungi are involved in amino acid metabolism and are described in many fungi such as Aspergillus oryzae, Penicillinium chrysogenum and Fusarium oxysporum [27]. Peptide transport is a universally observed physiological phenomenon in both prokaryotes and eukaryotes cells [28]. This process is characterized by the ability of cells to transport peptides across membranes in an energy-dependent manner. Internalized peptides are rapidly hydrolyzed by peptidases and the resulting amino acids are used for protein synthesis or as alternative sources of nitrogen and carbon. Vizcaíno et al., reported the cloning and characterization of ThPTR2, di/tri-peptide transporter gene from T. harzianum related to the mycoparasitic process. This hypothesis was supported by the fact that expression of ThPTR2 was triggered when Trichoderma directly interacted with B. cinerea [12]. The transcript asm, encoding an acid sphingomielinase (ASM), showed the lowest expression value after contact to the phytopathogen (Table 3 and Additional file 2). ASM are a group of hydrolases that cleave sphingolipids, a common component of plasma membranes, and their products can regulate a variety of cellular functions such as proliferation and differentiation [29]. This enzyme could be involved in providing nutrition for T. harzianum from F. solani cell components. All results were compared to the respective control group (T. harzianum grown in glucose media) at the respective time. (±) Represent standard deviation. The data were analyzed by the 2 -ΔΔCt method.
The transcripts duf (DUF895 domain membrane protein), mbl2 (mbl2-like protein) and cfem (eight cysteine-containing domain) identified in the cDNA library were used in order to learn whether proteins located in membrane or outer surfaces of the cell walls are expressed during the interaction between T. harzianum and F. solani. These proteins are described as involved in recognition, attachment, adhesion and appressorium development key events mycoparasitism [30]. The transcript duf and cfem are down-regulated before contact and increased their expression values after-contact (Table 3 and Additional file 2). However, the transcript mbl2, encoding for a mannose biding lectin, is expressed in the three stages of interaction. This may be due the fact that these proteins are important in coiling and appressorium formation a key step in the mechanism of mycoparasitism.
The transcript qid74, encoding the QID74 cell wall protein, is also up-regulated and was strongly expressed after contact. Studies of QID74 in T. harzianum CETC 2413 showed that this protein is involved in resistance of hyphae to lytic enzymes and the ability to adhere to hydrophobic surfaces [31]. High expression values of genes of CWDE (bgn, chit and endo) were seen after contact, suggesting that QID74 was produced with the purpose of protecting the cell wall of T. harzianum. These proteins could be also be involved in recognition, attachment and formation of specialized structures such as appressoria.
Among all studied genes only one encoding for aquaporin (aqp) was highly expressed in the three stages of interaction to F. solani (Table 3 and Additional file 2). These proteins mediate rapid and selective flux of water across biological membranes and hence play important roles in the osmoregulation of cells and organisms. These proteins also facilitate transmembrane transporte of small uncharged molecules like polyols, urea, arsenite and many more, thereby playing roles in nutrient uptake [32].
Interestingly, we found a significant expression of the gene sag encoding for senescence-associated protein after contact between T. harzianum and F. solani (Table 3 and Additional file 2). Senescence is the progressive loss of growth potential of mycelium culminating in total cessation of growth when the culture is considered as dead [33]. Some naturally occurring strains of fungi cease growing through successive subculturing [33]. However, this biological mechanism has to be studied in detail to offer an interpretation about the senescence in Trichoderma species and its role in mycoparasitism.
Finally, three genes encoding for phospholipase d (pld), serine threonine-protein kinase (sck1) and checkpoint-like protein (chk1) involved in transduction cascades, protein modification and fungal morphogenesis were studied [34]. All results were compared to the respective control group (T. harzianum challenged with itself) at the respective stage. (±) Represent standard deviation. The data were analyzed by the 2 -ΔΔCt method.
The transcript chk1 showed the greatest level of expression followed by sck1 and pld, mainly after contact (Table 3 and Additional file 2). The role of these proteins in mycoparasitism must be studied in more detail, because it involves the participation of a wide range of other genes not described in this work.

Conclusion
Our results provided a step toward the understanding of the mycoparasitic process of T. harzianum during its interaction with F. solani. However, future studies, aimed at the functional characterization of genes reported here, will help to better define pathways involved in T. harzianum interaction with F. solani. A better understanding of the expression profiles of these genes could improve T. harzianum performance, either by predicting the regulation of the genes involved in the mycoparasitism or by improving their use in biotechnology processes such as transgenic expression in plants.

Methods
Fungal strains and culture conditions Isolation of RNA, cDNA synthesis and construction of the suppression subtractive hybridization (SSH) libraries Total RNA from T. harzianum mycelia was extracted using the TRIZOL reagent (Invitrogen), and the resulting quality and concentration were checked by formaldehyde/ agarose gel electrophoresis and a spectrophotometer. To obtain the "driver" and "tester" cDNA populations, 0.6 μg of mRNA purified at 24 h, 36 h and 48 h periods were pooled for each set of conditions (FSCW and glucose).
The PCR-Select cDNA Subtraction Kit (BD Biosciences Clontech, Mountain View, CA) was used to generate two subtracted cDNA libraries enriched for genes up-and down-regulated in T. harzianum grown in FSCW, referred to hereafter as the forward and reverse subtracted libraries. The first subtracted library (forward) was produced by the subtraction of the cDNA population from T. harzianum grown in FSCW, used as "tester", from the cDNA population from T. harzianum grown in glucose, used as "driver". The second library (reverse) was obtained by the subtraction of the cDNA population from T. harzianum grown in glucose, used as "tester", from cDNAs obtained from T. harzianum grown in FSCW, used as "driver". The final PCR products obtained corresponded to genes differently expressed during T. harzianum grown in FSCW. PCR products were cloned using the pGEM-T Easy vector system (Promega). Positive colonies were picked out and grown in microtiter plates. Plasmids DNAs were prepared from clones using standard protocols.
DNA sequencing, processing and EST database construction T. harzianum expressed sequence tags (ESTs) were obtained by single-pass 5'-end sequencing of the cDNA inserts using cycle-sequencing and dye-terminator standard protocols. The automated capillary electrophoresis sequencing runs were performed on an ABI Prism 3100 (Applied Biosystems). EST sequences were pre-processed using Phred [35]. Only sequences with at least 100 nucleotides and with a Phred quality value greater than or equal to 20 were kept for further analysis. ESTs were screened for vector sequences using the CrossMatch program (www.phrap.org). The resulting sequences were assembled into contigs using the CAP3 assembly program [36]. The filtered sequences were compared against the GenBank non-redundant (nr) database using the BLASTX algorithm from the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov). Database sequence matches were considered significant at E values ≤ 10 -14 . Transcripts were annotated using Gene Ontology (GO) terms and hierarchical structure (http://www.geneontology.org). Redundancy of the collections of ESTs was calculated as [1-(number of singletons/total number of ESTs)] × 100 [20].

Quantitative real-time RT-PCR analysis (RT-qPCR)
In this study, RT-qPCR experiments were carried out to check the SSH results and the reliability of our approaches. New independent RNA samples were obtained from the same two conditions: T. harzianum grown in FSCW, and T. harzianum grown in glucose medium. Both conditions were sampled at three different time points (24,36 and 48 hours). Statistical tests (Student-t test, ANOVA and linear regression analysis) were performed when appropriate. Eighteen genes potentially involved in biocontrol were selected from the FSCW-library and 10 genes from Glc-library (Table 1). Additional expression analyses were performed by using the dual confrontation plate assay [10]. Circular plaques of 5 mm diameter were cut from mycelium of 7-day-old cultures of T. harzianum and of F. solani grown on MYG