Current Genetics

, Volume 55, Issue 5, pp 551–560

Endochitinase CHI2 of the biocontrol fungus Metarhizium anisopliae affects its virulence toward the cotton stainer bug Dysdercus peruvianus

Authors

  • Juliano Tomazzoni Boldo
    • Programa de Pós-graduação em Biologia Celular e Molecular (PPGBCM)Centro de Biotecnologia da Universidade Federal do Rio Grande do Sul (CBiot/UFRGS)
  • Angela Junges
    • Programa de Pós-graduação em Biologia Celular e Molecular (PPGBCM)Centro de Biotecnologia da Universidade Federal do Rio Grande do Sul (CBiot/UFRGS)
  • Karina Bohrer do Amaral
    • Programa de Pós-graduação em Biologia Celular e Molecular (PPGBCM)Centro de Biotecnologia da Universidade Federal do Rio Grande do Sul (CBiot/UFRGS)
  • Charley Christian Staats
    • Programa de Pós-graduação em Biologia Celular e Molecular (PPGBCM)Centro de Biotecnologia da Universidade Federal do Rio Grande do Sul (CBiot/UFRGS)
  • Marilene Henning Vainstein
    • Programa de Pós-graduação em Biologia Celular e Molecular (PPGBCM)Centro de Biotecnologia da Universidade Federal do Rio Grande do Sul (CBiot/UFRGS)
    • Programa de Pós-graduação em Biologia Celular e Molecular (PPGBCM)Centro de Biotecnologia da Universidade Federal do Rio Grande do Sul (CBiot/UFRGS)
Research Article

DOI: 10.1007/s00294-009-0267-5

Cite this article as:
Boldo, J.T., Junges, A., do Amaral, K.B. et al. Curr Genet (2009) 55: 551. doi:10.1007/s00294-009-0267-5

Abstract

Chitinases have been implicated in fungal cell wall remodeling and play a role in exogenous chitin degradation for nutrition and competition. Due to the diversity of these enzymes, assigning particular functions to each chitinase is still ongoing. The entomopathogenic fungus Metarhizium anisopliae produces several chitinases, and here, we evaluate whether endochitinase CHI2 is involved in the pathogenicity of this fungus. We constructed strains either overexpressing or lacking the CHI2 chitinase. These constructs were validated by Southern, Northern and Western blot analysis, and chitinase production. To access the effects of CHI2 chitinase in virulence, the cotton stainer bug Dysdercus peruvianus was used as a host. CHI2 overexpression constructs showed higher efficiency in host killing suggesting that the production of this chitinase by a constitutive promoter reduces the time necessary to kill the insect. More significantly, the knock out constructs showed decreased virulence to the insects as compared to the wild type strain. The lack of this single CHI2 chitinase diminished fungal infection efficiency, but not any other detectable trait, showing that the M. anisopliae family 18, subgroup B endochitinase CHI2 plays a role in insect infection.

Keywords

ChitinaseOverexpressionGene knock outAgrotransformationMetarhizium anisopliaeDysdercus peruvianus

Introduction

Chitin is the major structural component of the invertebrate exoskeleton, and is an important component of the cell wall of filamentous fungi. This polysaccharide is an abundant linear insoluble polymer and is produced by a number of organisms. It is composed of several units of the acetylated amino sugar N-acetylglucosamine linked by β-1,4 bonds and is generated to provide stiffness to structures where it is deposited due to the rigidity of the polymer and by forming crosslinks in chitin fibers (Seidl 2008). In addition, it provides protective functions, acting as an initial barrier to hostile environments. In filamentous fungi, the rigid structure formed by chitin and other polysaccharides, which together represent almost 90% of the cell wall, is also useful for penetration of the hyphae into insoluble substrates for nutrition (Latge 2007). Fungi and other microorganisms can also use chitin as an exogenous carbon and nitrogen source and are, therefore, the main recyclers of this abundant polymer, second only to cellulose (Duo-Chuan 2006).

The enzymes that can fully degrade chitin into N-acetylglucosamine monomers are divided into N-acetylglucosaminidases [EC3.2.1.52, Glycoside hydrolase (GH) family 20] and chitinases (EC3.2.1.14, GH family 18 and 19). N-acetylglucosaminidases catalyze the release of GlcNAc monomers from GlcNAc dimers or from the non-reducing terminal end of GlcNAc multimers. Chitinases can be classified into two major classes: endochitinases that cleave the chitin polymer at any point inside the fiber and exochitinases that cleave from the non-reducing end of the polymer and release N-acetylglucosamine (GlcNAc) dimers. All fungal chitinases analyzed so far (including all of the available fungal genomes) belong to GH family 18. Current classification divides GH family 18 chitinases into three subgroups. Subgroup A (former class V) chitinases contain catalytic domain, but no carbohydrate-binding module (CBM), and have molecular mass ranging from 40 to 50 kDa. Most of these chitinases possesses N-terminal signal peptide and are targeted to the secretory pathway, with some exceptions that have intracellular localization. Chitinases from subgroup B (former class III) also possesses N-terminal signal peptide and vary greatly in size, varying from 30 to 90 kDa. Smaller subgroup B chitinases contain a CBM, traditionally described as cellulose-binding domain (CBD), which has a broader polysaccharide binding spectrum and can also bind chitin. Larger subgroup B chitinases usually have serine/threonine rich domains and/or GPI-anchoring signal so that the mature proteins are bound to the plasma membrane. Subgroup C is composed by large chitinases that have a molecular mass of 140–170 kDa. They have N-terminal signal peptide and several features that distinguish them from other subgroups, like the peptidoglycan-binding regions. Filamentous fungi are prolific chitinase producers in terms of both quantity and variety. These enzymes have been implicated in many aspects of fungal biology such as conidial germination, hyphal growth and morphogenesis; in defense against niche competitors; and in nutrition by solubilizing exogenous chitin fibers (Seidl 2008).

Considering the participation of chitinases in nutrition, it is important to emphasize the lifestyle acquired by some filamentous fungi through evolution. Several filamentous fungi such as Metarhizium anisopliae have specialized in the acquisition of energy by penetrating and colonizing terrestrial arthropods, whose cuticle is composed of almost 30% chitin fibers (Duo-Chuan 2006). To acquire nutrients in nature, the entomopathogenic fungi have developed a complex infection process, involving mechanical pressure exerted by appressorium formation and the secretion of hydrolases, including chitinases (da Silva et al. 2005; Pedrini et al. 2007; Qazi and Khachatourians 2007; Wang and St Leger 2007). Until now, however, only one M. anisopliae chitinase has been experimentally shown to participate in the infection process (da Silva et al. 2005). M. anisopliae chitinases are induced in vitro when the fungus is grown in chitin-containing medium (Krieger de Moraes et al. 2003; Barreto et al. 2004; Baratto et al. 2006; da Silva et al. 2005). At least six chitinases have been detected, although only three genes have been isolated (Bogo et al. 1998; Freimoser et al. 2003; Baratto et al. 2006). The chit1 gene encodes a 42-kDa endochitinase (Bogo et al. 1998; Baratto et al. 2003); the chi2 gene also encodes a 42-kDa endochitinase (Baratto et al. 2006); and the chi3 gene encodes a 30-kDa exo/endo-acting chitinase (Pinto et al. 1997; Freimoser et al. 2003; da Silva et al. 2005). Although chitinases are induced by chitin and host cuticle-containing media, the importance of these enzymes in the pathogenesis process is still unclear. The evaluation of constructs carrying deletions or overexpressing chitinase in bioassays could help the understanding of their function.

Few studies have been conducted to analyze whether chitinases participate in the infection process of entomopathogenic fungi. The overexpression of the M. anisopliaechit1 gene did not alter fungal virulence toward Manduca sexta (Screen et al. 2001). In contrast, M. anisopliae CHIT30 chitinase (chi3 gene) was found to be highly expressed during Rhipicephalus (Boophilus) microplus infection (da Silva et al. 2005), and the overexpression of the Beauveria bassiana endogenous Bbchit1 gene enhanced fungal virulence toward aphids (Fan et al. 2007).

The M. anisopliaechi2 gene was cloned from strain E6 (used as the wild-type strain in this work) and characterized (Baratto et al. 2006). The chi2 gene comprises 1,524 bp, including two introns of 210 and 72 bp. The endochitinase encoded by this gene has a predicted size of 423 aa and a predicted mass of 42 kDa. CHI2 possesses a catalytic domain containing a substrate binding site and a fungal CBD at the C-terminus, as well a signal-peptide at the N-terminus, which is characteristic of GH family 18 subgroup B (Fig. 1a). Phylogenetically, the CHI2 chitinase clusters with the CHIT30 chitinase and other family 18 plant-like chitinases, subgroup B (Seidl et al. 2005). As suggested by Seidl et al. (2005), the chi2 and chi3 genes are related to mycoparasite chitinases since those chitinases clustered as a separated clade within subgroup B chitinases and potentially have a role in myco- and entomopathogenicity processes. Consistently, both chi2 and chi3 genes from M. anisopliae, as well as chi18-13 from the mycopathogen Trichoderma harzianum, are upregulated in the presence of host cuticle (da Silva et al. 2005; Seidl et al. 2005; Baratto et al. 2006).
https://static-content.springer.com/image/art%3A10.1007%2Fs00294-009-0267-5/MediaObjects/294_2009_267_Fig1_HTML.gif
Fig. 1

Overexpression and knock out constructs. a CHI2 domain prediction. SP signal peptide; CD glycosyl hydrolases family 18 catalytic domain; SBS substrate binding site; Ce-BD fungal cellulose binding domain. b Overexpression cassette. c Overexpression transformants screened using PS1 primers: 1 T5; 2 T12; 3 T33; 4 E6; 5 positive control (pPZP::upchi2::bar plasmid); 6 no DNA template control. d Overexpression transformants screened using PS2 primers: 1 T5; 2 T12; 3 T33; 4 positive control (plasmid pPZP::upchi2::bar); 5 E6; 6 no DNA template control. e Southern blot analysis of the overexpression transformants: 1 E6; 2 T5; 3 T12; 4 T33. f Knock out cassette and schematic genome integration. g Knock out transformants screened using PS3 primers: 1 Δ22; 2 Δ25; 3 E6; 4 positive control (pPZP:: Δchi2::bar plasmid); 5 no DNA template control. h Southern blot analysis of the knock out cassettes: 1 E6; 2 Δ22; 3 Δ25. PS primer sets used for transformants screening PCR; LB binary vector left border; RB binary vector right border

In order to assign a function to the M. anisopliae CHI2 chitinase, we constructed chi2 overexpressing and knock out constructs and evaluated their virulence to insects.

Materials and methods

Plasmid constructs

For the construction of the CHI2 overexpression plasmid, a cDNA of the chi2 gene from M. anisopliae (Baratto et al. 2006) was amplified using specific primers [chi2cdnaF: 5′ GGTACCAATGCATCATCTACGCGCTC 3′ (KpnI site underlined); chi2cndaR: 5′ GGATCCCACCGGCATGACGACCACCA 3′ (BamHI site underlined)] and blunt cloned into the pUC18 SmaI site, generating the pUC::upchi2 plasmid. This plasmid was then digested with KpnI and BamHI restriction enzymes, and the chi2 ORF was subcloned into the pTEF/TRPC plasmid digested with KpnI and BamHI (Nakazato et al. 2006), which contained the M. anisopliae tef1-α gene constitutive promoter and the Aspergillus nidulanstrpc gene terminator. The pPZP201BK binary plasmid (Covert et al. 2004) was digested with EcoRI and HindIII, and the overexpression cassette of the chi2 gene was inserted, generating the pPZP::upchi2 plasmid. The pPZP::upchi2 plasmid was digested with PstI, and the resistance cassette carrying the bar gene, which confers resistance to the fungicide ammonium glufosinate (Nakazato et al. 2006), was inserted, generating the pPZP::upchi2::bar plasmid. The plasmid used to knock out the M. anisopliae chi2 gene was constructed using fusion PCR methodology (Yu et al. 2004). The 5′ portion of the chi2 gene, including 520 bp of the promoter region and 140 bp of the coding region, was amplified using primers 5CHI2F (5′ GCAAGACATCAGATCCTCGTGCCC 3′) and 5CHI2R (5′ AGTCACCGGTCACTGTACAGCAGTAGACGACATTTT 3′). The 3′ portion of the chi2 gene, comprising 632 bp of the coding region, was amplified using primers 3CHI2F (5′ CATGCTCCTCTTCTTTACTCTCGACATTGAAGTCAA 3′) and 3CHI2R (5′ AAGATGGCATCGTGGTCGCG 3′). The 5CHI2R and 3CHI2F primers span the 5′ and the 3′ ends of a resistance cassette expressing the bar gene. Both PCR products were purified and used for PCR in a mix containing the resistance cassette and primers 5CHI2F and 3CHI2R. The expected product of 3.3 kb was purified and blunt cloned into the pPZP201BK vector, generating the pPZP::Δchi2::bar plasmid.

Agrobacterium tumefaciens-mediated transformation

Metarhizium anisopliae transformation mediated by A. tumefaciens was carried out as previously reported (Staats et al. 2007). Positive transformants were screened by PCR.

Southern blot analysis

According to PCR results, both overexpressing and knock out transformants were selected for Southern blot assays to further characterize cassette integration into the M. anisopliae genome. The selected transformants were grown on MCc media (Cove 1966) to obtain mycelia. The mycelia were grounded, and genomic DNA was extracted using standard protocols (Sambrook and Russel 2001). DNA samples were digested with the appropriate restriction endonucleases and subjected to electrophoresis in a 1.2% agarose gel for 16 h at 1 V cm−1. After electrophoresis, the samples were electroblotted onto nylon membranes using a Trans-blot Electrophoretic Transfer Cell (Bio-Rad) and baked at 80°C for 2 h. A chi2 gene probe (comprising nt 683–1185 (503 bp) of the chi2 gene ORF—accession number DQ011663.2; Fig. 1b) was labeled with 32P and hybridized for 16 h at 60°C. Gradual stringency washes were performed, and the membrane was exposed to the Kodak-K Screen. Digital images of the blots were obtained using the Pharos system (Bio-Rad).

Anti-serum production

To raise the anti-CHI2 polyclonal anti-serum, the coding region of the chi2 gene was amplified using primers chi2_S_EC (5′ CCATGGATCATCTACGCGCTCTCGTCG 3′) and chi2_A_EC (5′ CTCGAGCCGGCATGACGACCACCAAT 3′). The amplified product was blunt cloned into the pUC18 SmaI site. The pET-23d(+) vector was digested using NcoI and XhoI restriction enzymes, and the chi2 coding region was subcloned. The recombinant plasmid was transformed into Escherichia coli BL21(DE3)pLysS (Stratagene) cells for heterologous expression. Cells were cultivated on LB media containing 1% glucose for 2 h, and expression was induced by 0.1 M IPTG for 3 h. Cells were harvested by centrifugation, and expression was analyzed by SDS-PAGE. The recombinant protein was partially purified using a Hi-Trap chromatography column (GE Healthcare) following solubilization in 6 M urea. The eluted fraction with the recombinant chitinase was submitted to SDS-PAGE, and the corresponding band was excised from the gel and used to immunize rabbits. Antiserum against CHI2 was produced in New Zealand white rabbits, and reactivity to CHI2 chitinase was tested using standard procedures.

Northern blot analysis

The wt and constructs of M. anisopliae were cultured in media with 1% crystalline chitin or 1% glucose as the sole carbon sources for 96 h at 28°C and 150 rpm. The mycelia were filtered, and total RNA was extracted as previously described (Dutra et al. 2004). Total RNA samples were electrophoretically separated, and sample normalization was carried out using the Qubit system (Invitrogen). Northern blot analysis was carried out using standard procedures (Sambrook and Russel 2001) using radiolabeled chi2 (same as used in the Southern analyses; Fig. 1b) and tub (comprising nt 347–864 (517 bp) of the α-tubulin ORF—accession number DQ393576.1) probes. After hybridization, the Kodak-K Screen (Kodak) was exposed to the membranes, and digital images of the hybridization signals were obtained using the Pharos system (Bio-Rad).

Western blot analysis

The wt and constructs from M. anisopliae were grown in media containing 1% crystalline chitin or 1% glucose as the sole carbon sources in a bench incubator for 96 h at 28°C and 150 rpm. The mycelia mat was filtered, and the supernatant was lyophilized. The pellets were suspended in one-tenth of the original volume, and approximately 300 μg of each sample was subjected to SDS-PAGE. The proteins were electroblotted onto PVDF membranes for 16 h at 100 V. The membranes were then blocked in blocking buffer for 2 h at room temperature, incubated with primary antibodies (1:1,000) in PBS/0.1% Tween-20 for 1 h at room temperature and washed three times for 10 min each with PBS/0.1% Tween-20. The membranes were subsequently incubated in the presence of secondary antibodies (anti-rabbit IgG conjugated to peroxidase—Sigma Chemicals; 1:2,000) for 1 h at room temperature and washed three times in PBS/0.1% Tween-20. The hybridization signal was generated using an ECL Plus kit (GE Healthcare), and a digital image was obtained using the VersaDoc system (Bio-Rad).

Total chitinase activity assay

A total of 0.25 mL of M. anisopliae culture supernatant of liquid media containing 1% crystalline chitin or 1% glucose as the sole carbon source was incubated in 0.5 mL of PBS (pH 7.0) buffer containing 0.5 g of colloidal chitin for 5 h at 37°C. The production of N-acetyl-d-glucosamine (GlcNAc) monomers was evaluated, and the total chitinase activity present in each assay was determined in triplicate. One nkat unit of chitinase was defined as the enzyme amount necessary to catalyze the release of 1 nmol s−1 of GlcNAc. A standard curve was generated using different concentrations of GlcNAc (1–500 μM). Statistical analyses of the results, which represent data from three different experiments, were done using Analyses of Variance (ANOVA) and Duncan’s test (P < 0.001) performed by SPSS 13.0 software.

Endochitinase specific activity assay

The endochitinase specific activity assay was carried out using the 4-methylumbelliferyl-β-d-N, N′, N″-Tetraacetylchitotriose oligomer according to standard procedures (McCreath and Gooday 1992), with some modifications. Using a 96-well fluorimeter plate, 95 μL of McIlvain Buffer (0.1 M citric acid; 0.1 M Na2HPO4; pH 6.0) and 5 μL of 0.8 mM 4-methylumbelliferyl-β-d-N, N′, N″-Tetraacetylchitotriose were added and incubated for 5 min at 37°C. Then, 10 μL of M. anisopliae culture supernatant in liquid media containing 1% crystalline chitin or 1% glucose as the sole carbon sources was added and incubated for 30 min at 37°C. The reaction was stopped using 100 μL of Stop Buffer (1 M glycine/NaOH; pH 10.6) for 5 min at 37°C. Water was used in the control sample. The reading was performed using a fluorimeter set for 355 and 460 nm as the excitement and emission wavelengths, respectively. One nkat unit of chitinase was defined as the enzyme amount necessary to catalyze the release of 1 nmol s−1 of 4-MU. A standard curve was generated using different concentrations of 4-MU (0.1–10 μM). Statistical analyses of the results, which data from three independent experiments, were done using analyses of variance (ANOVA) and Duncan’s test (P < 0.001) performed by SPSS 13.0 software.

Bioassays

The bioassays were conducted according to Lubeck et al. (2008) with modifications as follows. Ten male adults of the cotton stainer bug Dysdercus peruvianus (kindly provided by Dr. Célia Carlini, Centro de Biotecnologia, CBiot/UFRGS) were placed inside plastic containers with a water supply and cottonseeds for feeding (Staniscuaski et al. 2005). A fresh conidia suspension at 108 conidia/mL was produced, and each bug was immersed for 10 s in the suspension, except those used for the negative control, which were immersed in sterile distilled water. Each of the three biological replicates had three experimental replicates. Host mortality was recorded everyday for 10 days. The obtained data were analyzed by ANOVA, and groups were clustered by Duncan’s test (P < 0.001) using SPSS 13.0 software.

Results

In order to elucidate the role of the M. anisopliae CHI2 endochitinase, we generated overexpression and knock out constructs for the chi2 gene using the previously reported capability of A. tumefaciens to transfer T-DNA to M. anisopliae (Staats et al. 2007). The overexpression and knock out cassettes are shown in Fig. 1b, f, respectively. Fungal transformations using the pPZP::upchi2::bar plasmid yielded approximately 60 ammonium glufosinate resistant transformants per cellophane disk in five independent experiments. A total of 200 transformants were screened by PCR using primer set PS1 (chi2CDNAF—5′ GGTACCAATGCATCATCTACGCGCTC 3′; PANDOWN—5′ TCCAGATTCGTCAAGCTGTTTGA 3′), which confirms the correct cassette assembly and insertion into M. anisopliae genome, and primer set PS2 (BARF—5′ CCATGGCCATGAGCCCAGAACGAC 3′; BARR—5′ GGATCCTCAGATCTCGGTGACGGG 3′), which amplifies a portion of the bar resistance gene. We detected 152 positive transformants with PS1 and PS2 PCRs, and 3 (T5, T12 and T33) clones were used for further analysis (Fig. 1c, d). Fungal transformation using the pPZP::Δchi2::bar plasmid was less efficient and yielded an average of 40 transformants per cellophane disk in five independent experiments. Primer set PS3 (chi2CDNAF—5′ GGTACCAATGCATCATCTACGCGCTC 3′; chi2CDNAR—5′ GGATCCTCACCGGCATGACGACCACC 3′) was used to screen for positive transformants (Fig. 1g). Two positive transformants (Δ22 and Δ25) were identified out of 150 screened.

In order to analyze genomic integration of the overexpression cassette, DNA from M. anisopliae strain E6 and from the selected transformants, based on the PS1 and PS2 PCR experiments, was extracted and digested using EcoRV. The blotted DNA was hybridized with a probe comprising 0.5 kb of the coding region of the chi2 gene. A single hybridization signal of 6.5 kb was detected in strain E6, indicating the presence of a single copy of the chi2 gene (Fig. 1e). Transformants T5 and T12 showed distinct hybridization bands compared to the wt strain. The wt chi2 locus was replaced with the overexpression cassette since no 6.5-kb signal was detected. In transformant T33 there were hybridization signals originating from both the wt chi2 locus and the overexpression cassette, suggesting ectopic integration. Therefore, this transformant was chosen for further experiments. Transformant T33 also displayed the highest chitinase activity. Transformants from the knock out experiment selected by PS3 PCR experiments were grown, and DNA was extracted from mycelia. The DNA was digested with HindIII, separated by electrophoresis and blotted. The same probe mentioned above was used. We observed a 4.7-kb hybridization signal in the wt strain E6 (Fig. 1h). Transformants Δ22 and Δ25 did not show the wt chi2 locus hybridization signal; instead, a 5.9-kb signal was detected, confirming that the wt locus was replaced by the knock out cassette.

Expression analyses were performed to confirm the overexpression and deletion of the chi2 gene in the constructs generated. Chitinase activity, transcripts and CHI2 protein were assayed in the overexpression and knock out constructs and compared to wt levels. Total chitinase and endochitinase activities from constructs T5, T12, T33, Δ22 and Δ25 were assayed in culture filtrates from both glucose and crystalline chitin liquid media. All overexpression constructs showed chitinase activity that was significantly higher (P < 0.001) in both experiments compared to wt. The chitinase and specific endochitinase activities in the knock out constructs were slightly diminished in both media with different carbon sources when compared to the wt strain E6 (Fig. 2a, b).
https://static-content.springer.com/image/art%3A10.1007%2Fs00294-009-0267-5/MediaObjects/294_2009_267_Fig2_HTML.gif
Fig. 2

Total chitinase and endochitinase activity of wild type and constructs. Total chitinase (a) and endochitinase activity (b) were detected in culture supernatants of M. anisopliae wild type strain (E6) and constructs overexpressing (T5; T12; T33) or knock out (Δ22; Δ25) for the chi2 gene. Supernatants were obtained from fungi cultivated on medium containing 1% glucose (dark gray bars) or 1% chitin (light gray bars). Bars represent the average of three independent experiments and the lines above bars represent the standard deviation. Bars with the same letter do not differ statistically according to the Duncan’s test (P < 0.001). One nkat unit of chitinase was defined as the enzyme amount necessary to catalyze the release of 1 nmol s−1 of GlcNAc for total chitinase activity experiments and 1 nmol s−1 of 4-MU for specific endochitinase activity experiments

Transcripts of the chi2 gene were analyzed by Northern blot for the constructs as shown in Fig. 3a from RNA extracted from cultures in glucose and in the presence of chitin. The overexpressing transformant T33 produced chi2 transcripts that were not suppressed by glucose. The knock out construct Δ22 did not produce chi2 transcripts. These results were further confirmed by Western blot analysis using anti-CHI2 serum. Expression of CHI2 in construct T33 was detected in both culture media. No CHI2 recognition was observed in the Δ22 construct in medium containing 1% glucose as the sole carbon source (Fig. 3, panel 1% Glucose, lane 3). A faint recognition band, however, was observed for construct Δ22 in medium containing 1% chitin as the sole carbon source (Fig. 3, panel 1% Chitin, lane 3). This was probably caused by non-specific recognition of another isoform of chitinase once the chitinase system was activated in such medium, and no chi2 transcript was observed in the Northern blot assays.
https://static-content.springer.com/image/art%3A10.1007%2Fs00294-009-0267-5/MediaObjects/294_2009_267_Fig3_HTML.gif
Fig. 3

Expression analysis of the chi2 gene and CHI2 chitinase. a Northern blot analysis of chi2 and tub gene expression in wild type and constructs: 1 E6; 2 T33; 3 Δ22. b Western blot analysis of CHI2 chitinase expression in wild type and constructs: 1 E6; 2 T33; 3 Δ22

To analyze the participation of CHI2 chitinase in the infection process, both the chi2 overexpression and knock out constructs were used in bioassays with the cotton stainer bug D. peruvianus as a host. A survival curve of the tested constructs is shown in Fig. 4. The LT50 and LT90 results demonstrated differences in the ability of the constructs to infect the host (Table 1). The wt strain had an LT50 of 156 h and an LT90 of 209 h. In contrast, the CHI2 overexpression construct T33 had an LT50 of 125 h and an LT90 of 154 h. This represents a drop of over 20% in the time required to kill the host, suggesting that the CHI2 chitinase enhances virulence. The knock out construct Δ22 had an LT50 of 186 h and an LT90 of 267 h, therefore requiring over 20% more time to kill the insect host. These results show that overexpression of CHI2 increased the efficiency of killing D. peruvianus, while the chi2 knock out construct had reduced efficiency compared to wt. The overexpression of the CHI2 chitinase could represent a synergistic effect on the infection process. More significant, however, is the fact that the blocking of a single chitinase gene (CHI2 chitinase) had an important effect in the insect infection efficiency.
https://static-content.springer.com/image/art%3A10.1007%2Fs00294-009-0267-5/MediaObjects/294_2009_267_Fig4_HTML.gif
Fig. 4

Mortality of Dysdercus peruvianus in bioassays. Virulence of M. anisopliae wild type strain (E6) and constructs overexpressing (T33) or knock out (Δ 22) for the chi2 gene was tested on D. peruvianus males. Groups of ten insects were treated with each M. anisopliae strain or construct (108 conidia mL−1) and checked everyday until 100% mortality was achieved. Bars represent the standard error

Table 1

M. anisopliae LT50 and LT90 values obtained in bioassays using D. peruvianus

Strain/construct

LT50 (h)#

LT90 (h)#

E6

156 (150–162)a

209 (199–222)a

T33

125 (121–129)b

154 (149–161)b

Δ22

186 (172–208)c

267 (238–318)c

Values in parentheses represent the confidence interval (95%)

#Means followed by the same letter do not differ statistically according to a t test (P < 0.001)

Discussion

Metarhizium anisopliae secretes several chitinases, including CHI2, an endochitinase of 42 kDa, which belongs to the subgroup B. Chitinases are thought to be key enzymes during the early stages of the infection process since chitin, associated with proteins, is one of the main compounds of the host cuticles and is highly insoluble. Another important factor is that chitinases are needed during the fungal morphogenesis process once chitin is present in the cell wall and must be remodeled in order to allow hyphae to grow and/or for cell differentiation into appressorium. Metarhizium has high chitinase diversity, and at least six chitinases have been observed in glycol-chitin zymogram assays after cultivation in chitin-containing medium (da Silva et al. 2005). In analyzing this complex enzyme family, the elucidation of each chitinase function will allow a better understanding of the chitinolytic system. Although little is known about the participation of chitinases in the infection process of entomopathogenic fungi, the endochitinases seem to be responsible for initiating the chitin cuticle solubilization process due to their cleavage characteristics.

Since the M. anisopliae chitinase encoded by the chi2 gene constitutes an endochitinase and is phylogenetically related to other endochitinases previously proven to be important in the infection process of other pathogenic fungi, specifically in mycopathogenic fungi (Seidl et al. 2005), it is probable that the CHI2 chitinase is also important for the M. anisopliae host infection process, as shown here. Considering the number of chitinases secreted by Metarhizium, it is expected that these enzymes act directly and synergistically along with other hydrolases throughout the infection process in order to solubilize the host cuticle, providing nutrition and allowing fungal penetration, to colonize the entire host. In this study, we demonstrated that overexpression and deletion of the chitinase encoded by the chi2 gene alter fungal virulence toward the model host D. peruvianus.

Chitinases are a long-term subject of study in filamentous fungi. Genome-wide analyses have suggested that the chitin degradation enzymes are much more complex in these organisms (Seidl et al. 2005). Assigning function for putative chitinases is the actual challenge to understand if most of such diversity is simply redundant or if each chitinase has specific roles in fungal biology. This is particularly important for the entomopathogenic and mycopathogenic fungi that use chitinases during their life cycle and in the course of infection to digest the protective chitin barrier of their hosts.

Subgroups A and B chitinases have been implicated in exogenous chitin degradation, like the subgroup A chitinases PjChi-1 from Paecilomyces javanicus (Chen et al. 2007), the Crchi1 from the mycoparasitic fungus Clonostachys rosea (Gan et al. 2007); and subgroup B chitinases chi2 and chi3 from M. anisopliae (Baratto et al. 2006; da Silva et al. 2005). Both subgroups are also involved in cell wall remodeling. A. nidulans chiA, a subgroup B chitinase, is a large GPI-anchored chitinase and its deletion decreased germination rates and hyphal growth (Takaya et al. 1998; Yamazaki et al. 2008). Moreover, a subgroup A chitinase from A. nidulans, chiB, is implicated in autolysis (Yamazaki et al. 2007). In this study, the deletion of the subgroup B chitinase chi2 did not show morphological alterations, but diminished M. anisopliae virulence. It seems that chitinases from all subgroups are involved in hyphal-related processes and/or infection processes (Seidl 2008).

Some chitinases also have been shown to be important in the infection process of pathogenic fungi. The mycoparasitic fungus Trichoderma secretes chitinases (specially subgroup B) in order to break down chitin polymers of host fungi cell walls such as Rhizoctonia solani for nutrition, consequently interrupting their development (Seidl et al. 2005). Mutants of the fungus C. rosea lacking each of the three different endochitinases displayed altered mycoparasitic activity during interaction with Fusarium culmorum in liquid culture. The knock out of the endochitinase genes cr-ech58 (possibly subgroup A), cr-ech42 (subgroup A) and cr-ech37 (subgroup B) each diminished virulence of C. rosea. Moreover, disruption of the cr-ech42 gene showed the stronger effect in reducing the mycoparasitism, indicating that this gene is the most important of the three deleted endochitinases (Mamarabadi et al. 2008).

An acidic mammalian chitinase (subgroup not determined) seems to be important as a defense mechanism against the pathogenic fungi Candida albicans and Aspergillus fumigatus (Chen et al. 2009). In addition, plant endophytic bacterial and viral chitinases are probably involved in defense against pathogenic fungus and insect plagues (Corrado et al. 2008; Quecine et al. 2008; Silvar et al. 2008). The altered virulence of the constructs overexpressing and lacking the chi2 gene demonstrated that the endochitinase CHI2 (subgroup B) participates in the infection process. This observation is in contrast to previous work in which the non-repressible expression of endochitinase CHIT1 (subgroup A) did not affect M. anisopliae virulence toward M. sexta (Screen et al. 2001). These authors hypothesized that the basal levels of endochitinase production were sufficient for the early period of infection once the chitin fibers were coated with proteins and somehow occluded, hampering the action of chitinases. As previously demonstrated (Limon et al. 1999; Bowman and Free 2006), however, the overexpression of a chitinase (subgroup B) in Trichoderma spp. enhanced mycoparasitism despite of the protein coat in the host cell wall (Bowman and Free 2006; Latge 2007). It is more likely that a specific chitinase, or group of chitinases, may act differently depending on the host cuticle. Recently, a gene expression analysis comparing strains of M. anisopliae var. anisopliae and M. anisopliae var. acridum, which are specific or generalist when considering the ability to infect a narrow or a wide range of hosts, showed that CHI2 is present in the generalist strain Ma2575 and not in the specific strain Ma443 (Wang et al. 2009). This suggests that CHI2 participates in the infection processes and that its absence reduces or impairs the fungal capacity to infect certain hosts. As demonstrated by Fan et al. (2007), overexpression of the Bbchit1 gene (subgroup B) induced high expression of the chitinase encoded by the aforementioned gene and enhanced B. bassiana virulence to aphids, diminishing both lethal conidial concentration and lethal time. In addition, genetically modified plants expressing different fungal/bacterial chitinases (subgroup A) have been found to be more resistant to phytopathogen fungi infection and insect feeding (Distefano et al. 2008; He et al. 2008; Kern et al. 2009; Shah et al. 2009).

Based on our findings, we propose that the endochitinase CHI2 (subgroup B) is important for pathogenicity by being secreted and directly acting in the penetration process of the host cuticle. Moreover, neither the overexpressing nor the knock out mutants presented detectable morphological alterations in spores, hyphae and appressorium formation (data not shown). These results strengthen the hypothesis of a specialization of the different chitinases in Metarhizium. However, other functions for the same chitinase cannot yet be ruled out at this point. The analysis of mutations in other chitinase genes as well as mutants lacking more than one chitinase should in the future allow a more comprehensive view of the chitin degradation system in Metarhizium and other fungi. Understanding fully the nature of chitinase functions could facilitate more effective utilization of Metarhizium in pest biocontrol strategies.

Acknowledgments

This work was supported by CNPq and FAPERGS. JTB and CCS are recipients of CAPES fellowships. AJ and KBA are recipients of CNPq fellowships. Authors acknowledge the time and many suggestions of the two anonymous referees that greatly improved the original manuscript.

Copyright information

© Springer-Verlag 2009