Homologous expression of a mutated beta-tubulin gene does not confer benomyl resistance on Trichoderma virens

Authors


Benjamin A. Horwitz, Department of Biology, The Technion – Israel Institute of Technology, Haifa 32000, Israel (e-mail: horwitz@tx.technion.ac.il).

Abstract

Aims: To clone the beta-tubulins and to induce resistance to benzimidazoles in the biocontrol fungus Trichoderma virens through site-directed mutagenesis.

Methods and Results: Two beta-tubulin genes have been cloned using PCR amplification followed by the screening of a T. virens cDNA library. The full-length cDNA clones, coding for 445 and 446 amino acids, have been designated as T. virens tub1 and T. virens tub2. A sequence alignment of these two tubulins with tubulins from other filamentous fungi has shown the presence of some unique amino acid sequences not found in those positions in other beta-tubulins. Constitutive expression of the tub2 gene with a histidine to tyrosine substitution at position 6 (known to impart benomyl/methyl benzimadazol-2-yl carbamate resistance in other fungi), under the Pgpd promoter of Aspergillus nidulans, did not impart resistance to benomyl.

Conclusions: The homologous expression of tub2 gene with a histidine to tyrosine mutation at position +6, which is known to impart benomyl tolerance in other fungi, does not impart resistance in T. virens.

Significance and Impact of the Study: Unlike other Trichoderma spp., T. virens, has been difficult to mutate for benomyl tolerance. The present study, through site-directed mutagenesis, shows that a mutation known to impart benomyl tolerance in T. viride and other fungi does not impart resistance in this fungus. Understanding the mechanisms of this phenomenon will have a profound impact in plant-disease management, as many plant pathogenic fungi develop resistance to this group of fungicides forcing its withdrawal after a short period of use.

Introduction

In recent years, the focus on the management of plant diseases has shifted from chemical control to more eco-friendly methods of biological control. Several fungi have the potential to control crop loss caused by plant-pathogenic fungi. Among them, the common soil inhabiting Trichoderma spp. (perfect state, where known, belongs to Ascomycetes) have gained maximum popularity, resulting in several commercial formulations in many countries (Mukhopadhyay and Mukherjee 1996; Herrera-Estrella and Chet 1998; Mukherjee 1999a). A combination of the biofungicides with lower doses of chemical fungicides in the form of ‘integrated control’ has been useful in many instances.

The essence of such an approach is that the biofungicide must be insensitive to the chemical fungicide that is applied along with it. Trichoderma spp. are highly sensitive to a very important group of fungicides – the benzimidazoles (e.g. methyl benzimadazol-2-yl carbamate (MBC)/carbendazim, benomyl, thiabendazole). Thus, it was not possible to combine these fungi with this group of fungicides, until the pioneering work of Papavizas and his group at Beltsville who induced benomyl tolerance in T. harzianum through u.v.-mutagenesis (Papavizas et al. 1982). A series of papers soon followed on induction of mutations conferring benomyl tolerance in many other Trichoderma spp. (Papavizas and Lewis 1983; Ahmad and Baker 1988; Mukherjee et al. 1997, 1999). Strains which tolerated very high concentrations (100–200 mg l−1) of benomyl were easily obtained. However, attempts to mutagenize Trichoderma (Gliocladium) virens by several methods failed (Papavizas 1987; Mukherjee and Mukhopadhyay 1993; P.K. Mukherjee, unpublished data). Papavizas et al. (1990) nevertheless succeeded in inducing tolerance to benomyl in T. virens by a sequential treatment with ethyl methane sulphonate and UV radiation. The mutation frequency was extremely low, the mutants could tolerate only a low concentration (5–10 mg l−1) of benomyl and were ineffective as biocontrol agents.

Benomyl acts as an antimitotic agent, binding to the beta-tubulins (Davidse 1986). Jung et al. (1992) carried out a systematic study of the molecular mechanism of benomyl tolerance in Aspergillus nidulans and concluded that regions around amino acids 6, 165 and 198 of beta-tubulin are important for benzimidazole binding, and mutation in any of these residues leads to benomyl insensitivity. This is indeed evident from the crystallographic data on the structure of the αβ dimer of pig brain-tubulin, which indicates that these residues form a loop (Nogales et al. 1998). A single amino acid change from phenylalanine to tyrosine at position 167 results in benomyl tolerance in Neurospora crassa and Acremonium chrysogenum (Orbach et al. 1986; Nowak and Kuck 1994). In T. viride and Septoria nodorum, a tyrosine for histidine substitution at position 6 (H6Y) of tub2 confers resistance (Cooley and Caten 1993; Goldman et al. 1993). Although benomyl-tolerant T. virens could not be obtained through classical mutagenesis, heterologous expression of a benomyl-tolerant allele of beta-tubulin from N. crassa did result in tolerance (Ossanna and Mischke 1990). The failure to isolate resistant mutants might, therefore, result from specific structural properties of T. virens beta-tubulin. To examine this possibility, we generated a T. virens H6Y tubulin allele, and tested whether expression of this allele confers benomyl resistance in T. virens.

Materials and methods

Strains and growth conditions

Trichoderma virens (Gliocladium virens Miller, Giddens and Foster, IMI 304061; Mukherjee et al. 1993) was used in this investigation. The fungus was routinely grown on potato dextrose agar (PDA) (Difco) at 25°C.

Design of primers, amplification and cloning of the beta-tubulin genes

High molecular weight genomic DNA from Trichoderma mycelia was extracted as described by Mukherjee (1999b), treated with RNase A and purified by a single phenol extraction. Two primers designated as tubfor (TTC CCC GAC CGA ATG ATG G) and tubrev (CCT CGC CAG TGT ACC AAT GC) were designed from the sequences of tub2 of T. viride (Goldman et al. 1993), from the regions of high homology with beta-tubulins from other fungi, namely, Mycosphaerella pini, N. crassa, Colletotrichum graminicola, Erysiphe graminis and C. gloeosporioides f. sp. aeschynomene. These two regions correspond to the highly conserved amino acid sequences FPDRMM and HWYTGE, at amino acid positions 159 and 396, respectively. About 100 ng of genomic DNA was used as template for PCR amplification, which was performed with 2.5 mm each of dNTPs, 10 pmol each of the primers, 1.5 U of Pfu polymerase (Stratagene, La Jolla, CA, USA) in a total volume of 50 μl at 60°C annealing. The products were separated in 1% agarose gel, and the bands excised and extracted with the QIAEX II Gel Extraction Kit (Qiagen, Hilden, Germany), and cloned in the SrfI site of the pCR-ScriptAmp cloning vector (Stratagene) using the pCR-ScriptAmp (SK+) cloning kit. The plasmid DNA was digested with NotI and PstI and size-separated on a 1% agarose gel. Two clones (of different insert size, no. 1 and 7, designated as pGV1 and pGV7) were sequenced. Homology of these clones with other genes was determined by BLAST X search (http://www.ncbi.nlm.nih.gov/blast/blast.cgi) and the alignments were performed using the CLUSTAL W program (http://pbil.ibcp.fr/cgi-bin/npsa_automat.pl).

Screening of the cDNA library of T. virens

Using an EcoRV–BamHI fragment from the clone GV1 and an EcoRV–SacI fragment of GV7 as probe, a T. virens cDNA library (Mukherjee et al. 2003) was screened, and the positive clones (designated as GV11, GV12, etc. and GV71, GV72, etc.) were sequenced using T3, T7 and three internal primers (tubfor, tubrev, gvtub-CAT GTT CCG ACG CAA GGC).

Site-directed mutagenesis and transformation of T. virens

A primer (tubhy) was designed to introduce a BspHI site (compatible with the NcoI site of the Pgpd promoter in the plasmid pAN52) before the start codon, and to change histidine to tyrosine at position 6 in the tub2 gene (GV71). The sequence of tubhy was CC ATC ATGAGA GAG ATT GTT TAC ATC (the BspHI site is in italics; the bold nucleotides were originally C, C, T and C, respectively, in the cDNA). The codon CGT was changed to AGA in order to accommodate the BspHI site, without changing the amino acid arginine at position 2. CAC was changed to TAC for changing the amino acid at position 6 from histidine to tyrosine, which is known to impart carbendazim tolerance in T. viride (Goldman et al. 1993). Using the cDNA clone (pGV71) as template, and tubhy and tubrev as primers, the 5-end was amplified at 60°C annealing with Pfu polymerase (Stratagene) as described earlier. For expression of the mutated beta-tubulin in T. virens, five DNA fragments were ligated together at 16°C overnight. The fragments were: BspHI–EcoRI fragment of this amplified product, EcoRI–KpnI fragment of the cDNA clone (these two fragments complete the cDNA sequence from start codon till the beginning of polyA sequence); KpnI–SacI fragment of the plasmid pAT-BS (a XhoI–XbaI fragment containing the hph gene of E. coli, with TrpC promoter and TrpC terminator from pUC-ATPH cloned into XhoI–XbaI sites of pBlueScript KS(+), P.K. Mukherjee and B.A. Horwitz, unpublished data) for selection on hygromycin B; pBlueScript KS(+) digested with SacI–BamHI; and BglII–NcoI fragment (the promoter sequence, Pgpd) from pAN52. This places the mutated cDNA under the Pgpd promoter of A. nidulans. A clone, designated as pGVHY, was obtained by ligation of these five fragments. After sequencing the plasmid and confirming the desired change at amino acid 6 (and no other change in the DNA sequence), this plasmid was used for biolistic transformation (Lorito et al. 1993) of T. virens conidia and the transformants were maintained on PDA amended with 50 mg l−1 hygromycin B. The transformation was confirmed by Southern hybridization, and PCR using primers specific for the hph gene. The ability of the transformants to grow on benomyl/carbendazim was tested by growing on PDA amended with 50 mg l−1 hygromycin B and 2 mg l−1 benomyl. The expression of the mutated tubulin was studied by RT-PCR of total RNA. RNA was extracted with hot phenol (Ausubel et al. 1999), and treated with RNase-free DNase I (Roche, Indianapolis, IN, USA). cDNA from 1 μg of total RNA was synthesized using a cDNA synthesis kit (Roche). cDNA (5 of 20 μl) was used to amplify the mutated tubulin using the primer pair pgpd (3-end of the Pgpd promoter GCT TGA GCA GAC ATC ACC) and tubrev. As positive control, amplification was also carried out from the cDNA of the wild type and the transformants using tubfor and tubrev primer pair.

Results

Using the primer pair tubfor and tubrev, one clear band of about 700–800 bp was amplified from genomic DNA of T. virens. Of the 10 clones screened, nine were of the same size, and one clone (no. 7) was a little larger in size (which was not resolved in the initial run with the PCR products because of a short run). Clone no. 1, representing the smaller insert size, and no. 7, were sequenced and found to be highly homologous to the published tub1 and tub2 (respectively) genes of T. viride. The major difference between these two sequences is an intron in GV7. These two clones were used for the screening of the cDNA library. Several positive clones were obtained from screening of the cDNA library, which were designated as pGV11, 12, etc. (for those derived from pGV1), and pGV71, 72, etc. (for those derived from pGV7). Of the three independent clones sequenced for both the genes, pGV13 (coding for 445 amino acids, hereafter designated as GV tub1) and pGV 71 (coding for 446 amino acids, hereafter designated as GV tub2) were full-length cDNAs, highly homologous to T. viride tub1 and tub2 genes, respectively (the cDNA sequences of these genes have been deposited in the GenBank database with accession numbers AY158202 and AY158203). An alignment of GV tub1 and GV tub2 with beta-tubulins of other fungi (Fig. 1) reveals that GV tub1 is not so close to T. viride tub1 (differs by 22 amino acids), but GV tub2 is close to T. viride tub2 (differs by only eight amino acids). The two beta-tubulins of T. virens differ from each other by 46 amino acid residues. GV tub1 has unique (not found in any other fungal beta-tubulins) amino acids at positions 294 (leucine), 332 (serine) and 434 (aspartate); GV tub2 has unique amino acids at positions 55 (asparagine), 120 (isoleucine), 442 (alanine) and 444 (histidine).

Figure 1.

Figure 1.

Alignment of deduced amino acid sequences of Trichoderma virens tub1 (GVtub1) and tub2 (GVtub2) with the beta-tubulins of other fungi (TV, Trichoderma viride; EP, Erysiphe pisi; AC, Acremonium chrysogenum; CG, Colletotrichum graminicola; CGA, C. graminicola f. sp. aeschynomene; NC, Neurospora crassa; BC, Botrytis cinerea; EG, Erysiphe graminis; AN, Aspergillus nidulans)

Figure 1.

Figure 1.

Alignment of deduced amino acid sequences of Trichoderma virens tub1 (GVtub1) and tub2 (GVtub2) with the beta-tubulins of other fungi (TV, Trichoderma viride; EP, Erysiphe pisi; AC, Acremonium chrysogenum; CG, Colletotrichum graminicola; CGA, C. graminicola f. sp. aeschynomene; NC, Neurospora crassa; BC, Botrytis cinerea; EG, Erysiphe graminis; AN, Aspergillus nidulans)

Twelve colonies were obtained when conidia of T. virens were transformed with the plasmid pGVHY (with mutated tubulin of T. virens– a histidine to tyrosine change in position 6). Even after repeated subculturing on 50 mg l−1 hygromycin B, all the transformants were stable. The integration of the plasmid was confirmed by Southern hybridization (data not presented). None of the transformants, as the wild-type strain, were able to grow on 2 mg l−1 benomyl even after 6 days of incubation. RT-PCR with some selected transformants confirmed the expression of the mutated tub2 gene (Fig. 2).

Figure 2.

RT-PCR analysis of total RNA from the wild type (W) and transformants (2, 4, 6, 7) for mutated tubulin under the Pgpd promoter of Aspergillus nidulans. Note that Pgpd-tubrev amplifies only the mutated tubulin, while tubfor-tubrev amplifies all the tubulins

Discussion

Trichoderma virens is an important biocontrol fungus effective against a range of plant pathogens, with some commercial formulations already available in the market. However, its application in the framework of an integrated disease management has been limited as a result of the unavailability of mutants resistant to the widely used benzimidazole fungicides. This is the only biocontrol species of Trichoderma where several independent groups have failed to obtain any useful benomyl/carbendazim-tolerant mutants even after repeated attempts.

In this study, we cloned two beta-tubulin genes from T. virens. Through site-directed mutagenesis, we found that homologous expression of the tub2 histidine to tyrosine (H6Y) allele does not confer benomyl resistance in T. virens. This is in contrast to several other fungi, where the cognate of this mutation is known to impart benomyl/MBC tolerance (Jung et al. 1992; Cooley and Caten 1993; Goldman et al. 1993). Furthermore, in all cases, these mutated genes were used as dominant selectable markers when expressed in the wild-type strain. Thus, even in the presence of native tubulins, the same level of expression (with native promoter) of the mutated form imparted benomyl resistance. In the present study, however, even constitutive over-expression of the mutated gene did not result in tolerance to benomyl. The molecular mechanism of benomyl tolerance in the T. virens mutants obtained by Papavizas et al. (1990) was not studied. It is possible that, in that case, tolerance was mediated by a mechanism other than mutations in the beta-tubulin, for example decreased permeability of the cell to the fungicide (Nachmias and Barash 1976).

These observations indicate that T. virens is different from other species of Trichoderma and many other fungi with respect to induction of benomyl tolerance. It is, therefore, important to reach an understanding, at the molecular level, of why this is so. The results reported here suggest that T. virens Tub2 is intrinsically benomyl sensitive and cannot be converted to a tolerant allele by the H6Y point mutation. Very rapid development of benomyl/carbendazim tolerance in many plant pathogenic fungi has forced the withdrawal of these fungicides after a very short period of use. The understanding of why it is difficult to mutate T. virens for benomyl tolerance would therefore have an impact on chemical control of plant diseases, in addition to helping provide ways to generate benomyl-tolerant strains of T. virens for use in integrated control of plant diseases. A combination of further transgenic and biochemical approaches will be needed to determine whether there is a fundamental structural difference underlying the failure of T. virens Tub2 H6Y to confer resistance. As point mutations at beta-tubulins resulting in benomyl resistance in plant pathogens pose a serious threat to the plant protection industry, this understanding might be useful in prolonging the life of fungicides used for plant disease management.

Acknowledgements

MM thanks Prof. Gadi Schuster, Chairman of the Graduate Program Committee, Department of Biology, Technion, for the opportunity to work as a graduate project student. Supported in part by a grant from the Israel Academy of Sciences (ISF).

Ancillary