SciELO - Scientific Electronic Library Online

vol.21 issue3Assessing residual leukemia through fluorescence in situ hybridizationGenetic control of the fungus Colletotrichum lindemuthianum (Sacc. & Magn.) Scrib. reaction and corona color in the common bean (Phaseolus vulgaris L.) author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Genetics and Molecular Biology

Print version ISSN 1415-4757On-line version ISSN 1678-4685

Genet. Mol. Biol. vol. 21 n. 3 São Paulo Sept. 1998 

Trichoderma harzianum transformant has high extracellular alkaline proteinase expression during specific mycoparasitic interactions


Maria Helena S. Goldman1 and Gustavo H. Goldman2
1Departamento de Biologia,Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto and 2Departamento de Ciências Farmacêuticas, Faculdade de Ciências Farmacêuticas de Ribeirão Preto, USP, Av. do Café s/n, 14040-903 Ribeirão Preto, SP, Brasil. Fax: 55-016-6331092. E-mail:  Send correspondence to G.H.G.




The mycoparasite Trichoderma harzianum produces an alkaline proteinase that may be specifically involved in mycoparasitism. We have constructed transformant strains of this fungus that overexpress this alkaline proteinase. Some of the transformants were assessed for alkaline proteinase activity, and those with higher activity than the wild type were selected for further studies. One of these transformant strains produced an elevated and constitutive pbr1 mRNA level during mycoparasitic interactions with Rhizoctonia solani.




The mycoparasite Trichoderma harzianum may prove to be an effective control agent for many phytopathogenic fungi. It has been proposed that its mycoparasitic interaction proceeds in three major steps (Goldman et al., 1994a,b; Haran et al., 1996). Initially, the mycoparasite hyphae grow toward the host hyphae (Chet et al., 1981). Then the parasite attaches to the target hyphae, presumably mediated by a host lectin, and appressoria-like structures coil around the attacked cells (Elad et al., 1983a,b; Barak et al., 1985). Concurrently, degradation of b(1,3)-glucans and chitin from the host cell wall has been observed (Elad et al., 1983b). Probably both mechanical pressure and cell wall degradation by hydrolytic enzymes are involved. Finally, the mycoparasite penetrates and/or lyses the host hyphae (Chet et al., 1981) and releases cellular contents, which provides nutrients to sustain growth. Extracellular enzymes corresponding to the main chemical constituents of the fungal cell wall, i.e., chitin, glucans, and proteins, have been detected when T. harzianum is grown on Rhizoctonia solani mycelia or cell walls as the sole carbon source (Ridout et al., 1988; Geremia et al., 1993). The enzymes appeared sequentially; an alkaline proteinase was produced first, followed by glucanases and chitinases (Geremia et al., 1991, 1993). Recently, we purified and cloned this alkaline protease gene (pbr1) specifically induced by Rhizoctonia solani cell walls and chitin (Geremia et al., 1993). Although the results support the hypothesis that this gene is specifically involved in mycoparasitism, there is no direct evidence of its role in this interaction.

The availability of overproducing alkaline protease strains of T. harzianum would be of help to establish this enzyme's role in mycoparasitism. Additionally, antagonistic activity in the soil could be improved by increased activity of one or more enzymes involved in cell wall degradation of the pathogens. This study demonstrated the feasibility of overexpressing pbr1 gene in T. harzianum by increasing its copy number through transformation. One of the transformants showed elevated and constitutive pbr1 mRNA expression during mycoparasitic interactions.



Strains and growth conditions

T. harzianum strain IMI206040 was used throughout this work. This strain was grown as described by Geremia et al. (1993). R. solani cell walls and colloidal chitin were prepared according to Geremia et al. (1993). Basic protease activity was determined by incubation with Succ-Ala-Ala-Pro-Phe-pNA (Sigma Chemical Co., St. Louis, MO) (Geremia et al., 1993). Protein concentration was determined using a Bio-Rad assay (Bio-Rad, USA). In the antagonistic dual interactions, simple mycelial discs (1.0 cm in diameter from five-day-old cultures of R. solani in potato-dextrose-agar (PDA) plates) were inoculated on cellophane discs, near the edge of Petri dishes containing minimal medium (Geremia et al., 1991) plus 0.5% glucose. After two days, a mycelial disc (1.0 cm) of a four-day-old culture of T. harzianum from a PDA plate was placed opposite to the growing R. solani colony, and incubated at 28oC for four to 10 days.

DNA/RNA manipulations

Restriction enzyme digests were performed according to manufacturer's recommendations. Isolation of plasmid DNA from Escherichia coli and Southern blots was performed using standard procedures (Sambrook et al., 1989). DNA probes were made using a random primer system (Boehringer). Northern analysis was made as described previously by Goldman et al. (1992). A cellophane area representing different steps of the interaction was cut to obtain RNA from the dual antagonistic interactions. The mycelia were carefully washed with TES buffer, disrupted and total RNA extracted (Goldman et al., 1994b; Vasseur et al., 1995). Cotransformation of the basic protease gene was made in combination with plasmids pHAT-a and pBP2.2 (Herrera-Estrella et al., 1990; Goldman et al., 1990; Geremia et al., 1993). A molar ratio of approximately 1 to 4 was used; afterwards, selection was made in PDA plates with 1.2 M Sorbitol and 100 µg/ml hygromycin. Plasmid pBP2.2 was made of pT3T7.lac (Boehringer) with a 5.5-kb EcoRI fragment of the pbr1 gene (Geremia et al., 1993). This insert has three HindIII and four PstI sites. Furthermore, it has a fragment of about 1.5 kb, downstream from the start codon. Densitometry of the fungal transformants was performed using an LKB-Pharmacia 2202 Ultrascan Laser Densitometer.



Hydrolytic enzymes secreted by T. harzianum are believed to play an important role in the parasitism of phytopathogenic fungi. The alkaline protease (PBR1) can be induced in simulated mycoparasitic conditions; furthermore, colloidal chitin can also induce this enzyme (Geremia et al., 1991). The gene encoding this alkaline proteinase has been cloned, and Northern analysis showed that induction of this enzyme was due to an increase in mRNA level (Geremia et al., 1993). To determine if the pbr1 gene was induced during dual interaction assays, total RNA was prepared from different stages of antagonism between T. harzianum and R. solani (Figure 1A). As the mycoparasitic interaction progressed pbr1 mRNA levels increased (Figure 1B). Goldman et al. (1992) reported repression of the pgk gene expression during a simulated mycoparasitic state. In fact, expression dropped markedly during the dual interaction assays (Figure 1C). There was constant b-tubulin RNA expression under all test conditions (Figure 1B and C). These results indicate that pbr1 mRNA accumulates progressively during antagonistic interactions between T. harzianum and R. solani. Furthermore, simulated mycoparasitic conditions, where growth takes place in minimal medium with the cell walls of a phytopathogenic fungus as the only carbon source, mimic antagonistic dual interaction assays, at least insofar as pbr1 and pgk gene expression are concerned (Figure 1B and C). This is the first molecular genetic demonstration of simulated mycoparasitic conditions that reflects what occurs during antagonistic dual interactions. Due to the complexity of the in vivo interaction between T. harzianum and R. solani on the plant rhizosphere or in the soil, dual interaction assays provide the closest and most plausible methodological approach for studying the molecular biological aspects of mycoparasitism.

21n31926f1.GIF (14229 bytes)

Figure 1 - Northern blot of the mRNA obtained during specific mycoparasitic interactions between Trichoderma harzianum and Rhizoctonia solani. A: Different stages of the interaction. B: Northern analysis of the pbr1 mRNA obtained during dual interactions. The left panel shows hybridization with pbr1 gene, whereas the right panel shows hybridization with the tub1 gene. C: Northern analysis of the pgk mRNA obtained during dual interactions. The left panel shows hybridization with pgk gene, whereas the right panel shows hybridization with the tub1 gene. In panel AII, T. harzianum has grown over R. solani mycelia. R: R. solani; T: T. harzianum; G: T. harzianum grown in MM plates with 0.5% glucose; CW: T. harzianum grown in MM broth with 0.1% cell walls of R. solani.


There are many examples of gene overexpression in fungus caused by an increase in copy number (Fowler and Berka, 1991). Consequently, we decided to overexpress pbr1 gene by co-transforming T. harzianum with its genomic clone, plasmid pBP2.2, and a plasmid containing a selected hygromycin B resistant marker, pHAT-a (Herrera-Estrella et al., 1990). Protoplasts (107-108 ml-1) were transformed with both plasmids and selected for hygromycin B resistance. Electroporation of T. harzianum protoplasts in the presence of these two plasmids yielded the expected high transformation frequency (Goldman et al., 1990). Fifty-three transformants were selected, purified, and stabilized through four consecutive rounds of transfer in selective medium. These transformants were assessed for alkaline protease activity and those with activity higher than the wild type were selected for further studies.

DNA from four transformants and the wild-type strain was isolated, digested with PstI or HindIII, and hybridized with a 5.5-kb EcoRI fragment containing the pbr1 gene. Digestion with HindIII produced hybridizing fragments of about 5.0 and 0.9 kb, in the wild type and transformants (Figure 2A). This digestion also showed an additional fragment of about 3.0 kb in all transformants (Figure 2A). Digestion with PstI produced hybridizing fragments of about 2.3, 1.7 and 0.6 kb in the wild type and transformants (Figure 2B). Additional fragments of about 5.5, 2.8, and 1.5 kb could also be seen in the transformants (Figure 2B). The additional fragments present in both digestions are presumably due to rearrangements of plasmid pBP2.2 after transformation. Hybridization of undigested DNA from T. harzianum transformants with labelled pPBR1 gave a high molecular weight signal. This showed that the plasmid had integrated into the fungus' genome (data not shown). Together, these results indicate that multiple integrated copies of the pbr1 gene were arranged in tandem repeats, which were stable in the absence of selective pressure. One striking feature of these transformants is that all of them showed the same hybridization pattern. This suggests that pBP2.2 had integrated at the same genomic site (see below). In previous studies, using hygromycin resistance as a marker, we showed that Trichoderma transformants contained several plasmid copies integrated into their genome. Integration apparently occurs at the same site in the different transformants (Herrera-Estrella et al., 1990; Goldman et al., 1990, 1993). Transformation of T. viride with a homologous marker, tub2 gene, that confers resistance to benzimidazoles, has shown that at least 73% of the transformants integrated the transforming DNA at the same site (Goldman et al., 1993). These integration events could be explained by the existence of a highly recombinogenic site that could provide a preferential site for integration (Herrera-Estrella et al., 1990). However, there is no experimental evidence supporting this hypothesis since the DNA sequence of the junction between integrated plasmids and the genome has not been determined. Such sequence determinations should clarify the mechanism(s) underlying plasmid integration.

21n31926f2.GIF (11408 bytes)

Figure 2 - Southern blot analysis of transformants of Trichoderma harzianum DNA. Genomic DNA (10 µg) from five transformants (pbr1-2A, 16A, 36A, 55A, and 57A) and from the wild type (WT) was digested with PstI (A) or HindIII (B), separated by agarose gel electrophoresis, and analyzed by the Southern blot technique. Blots were probed with a 32P-labelled 5.0 kb EcoRI fragment containing the pbr1 gene from T. harzianum.


Transformants pbr1-2A, pbr1-16A, and pbr1-36A were chosen for analysis of PBR1 production. These transformants have additional copies of the pbr1 gene, which was verified by densitometry: four copies for transformant pbr1-2A, and two copies each respectively for transformants pbr1-16A and 36A (Figure 2). When these transformants were grown in MM with colloidal chitin as the sole carbon source, they had 97, 87, and 86% more activity than the wild type, respectively (Figure 3). Transformants' alkaline protease activity did not seem to be related to pbr1 gene copy number. An inefficiency of some other aspect of pbr1 gene expression, such as mRNA stability, mRNA processing, translation, or secretion could explain this occurrence. Similar results have been reported in Aspergillus spp. transformed with a chymosin gene (Fowler and Berka, 1991). When T. harzianum was grown in the presence of glucose as single carbon source, the alkaline protease activity was very low (Geremia et al., 1991). The alkaline protease activity of these transformants, when grown in MM with colloidal chitin and glucose as carbon sources, was 91, 54, and 64% higher than the wild type, respectively (Figure 3). These results showed that higher alkaline protease activity was obtained by increasing the pbr1 gene copy number by transformation. Furthermore, the alkaline protease activity in the transformants was not completely abolished by glucose.

21n31926f3.GIF (10886 bytes)

Figure 3 - Alkaline protease activity of Trichodema harzianum transformants. T. harzianum wild type and transformants were grown for 72 h in MM + glucose 2% at 28oC. Afterwards, mycelia were transferred to MM + colloidal chitin 0.1% (open bars) or colloidal chitin 0.1% combined with glucose 2% (grid bars). These cultures were grown for 48 h at 28oC, and supernatants were assayed for alkaline protease activity. Activity was expressed as a percentage of wild type strain activity. These results are the means of three experiments.


The pbr1 gene was induced during dual interaction assays (see above). To determine whether pbr1 gene expression in the transformants increases similarly during dual interaction assays, Northern blot analyses were carried out with total RNA isolated from dual interaction assays between one of the transformants, pbr1-2A, and R. solani (Figure 4). As shown before (Figure 1), there was an increase of the pbr1 mRNA level following the different stages of the mycoparasitic interaction between the wild type T. harzianum and R. solani (Figure 4B). The transformant pbr1-2A showed a higher pbr1 mRNA expression than the wild type, but instead of a progressive accumulation of pbr1 mRNA transcript as in the wild type, there was an almost constant expression (Figure 4B). This was not due to differences in mRNA loading, because tub1 mRNA expression remained constant during all the stages of interaction between pbr1-2A and R. solani (Figure 4C). Thus, the higher alkaline protease expression that was seen in the transformant pbr1-2A during simulated mycoparasitism also occurred at the mRNA level during specific mycoparasitic interactions.

21n31926f4.GIF (9410 bytes)

Figure 4 - Northern blot was performed on the mRNA obtained during specific mycoparasitic interactions using Trichoderma harzianum wild type and the transformant strain pbr1-2A against Rhizoctonia solani. A: Different stages of the interaction. B: Northern analysis of the pbr1 mRNA obtained during dual interactions. The left panel shows dual interactions between the wild type and R. solani, whereas the right panel shows dual interactions between pbr1-2A and R. solani. C: Northern analysis of the tub1 mRNA obtained during dual interactions. The left panel shows dual interactions between the wild type and R. solani, whereas the right panel shows dual interactions between pbr1-2A and R. solani. In panel AIII, T. harzianum has grown over R. solani mycelia.


It was shown earlier that the induction of pbr1 gene by cell walls (or colloidal chitin) was due to an increase in the corresponding mRNA level: no pbr1 mRNA was detected when the fungus was grown using a combination of cell walls and glucose as carbon sources (Geremia et al., 1993). In the pbr1 transformants, alkaline protease production was not completely repressed by glucose. One possible explanation could be limiting amounts of putative trans-acting factor(s). A similar observation about titration of trans-acting factors by the introduction of additional copies of endogenous genes has already been made for A. nidulans transformants (Andrianopoulos and Hynes, 1988). An alternative explanation is that cis-acting element(s) responsible for the glucose repression is (are) not present in the introduced genomic clone. The transformant pbr1-2A produced an elevated and apparently constant pbr1 mRNA expression during dual interaction assays. This suggests that in in vivo conditions transformant pbr1-2A would be able to produce higher amounts of the alkaline protease.

Our results indicate that pbr1 gene expression is repressed in saprophytic conditions and it is only specifically induced during mycoparasitism. Possible roles for PBR1 alkaline protease during mycoparasitism are: i) it can aid in the penetration of the host; ii) it can directly provide nutrients in the form of amino acids and indirectly by exposing other host materials to enzymatic digestion; iii) it might be important for the release or activation of Trichoderma toxins during infection, and iv) it can activate other enzymes which are in the zymogenic state. Additional experiments have to be performed to verify these hypotheses. Furthermore, the effect of extra copies of the pbr1 gene on the overall biocontrol activity can now be assessed through bioassays using the pbr1 transformants against phytopathogenic fungi. The availability of pbr1 overexpressing transformants will provide a powerful tool for these studies.



G.H.G. thanks Dr. Enno Krebbers for critically reading the manuscript, Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq-Brazil) for financial support, and Departamento de Genética, Faculdade de Medicina de Ribeirão Preto, USP, for providing laboratory space. This work was partially carried out with a grant from the International Foundation for Science (C/2450). Publication supported by FAPESP.




O micoparasita Trichoderma harzianum produz uma protease alcalina que pode estar especificamente envolvida em micoparasitismo. Foram construídas linhagens transgênicas deste fungo que super-expressam esta protease alcalina. A atividade de protease alcalina foi verificada em alguns destes transformantes e aqueles com maior atividade do que o tipo selvagem foram selecionados para estudos posteriores. Uma destas linhagens produziu um nível elevado e constitutivo de mRNA do gene que codifica a protease alcalina, prb1, durante interações micoparasíticas com o fitopatógeno Rhizoctonia solani.




Andrianopoulos, A. and Hynes, M.J. (1988). Cloning and analysis of the positively acting regulatory gene amdR from Aspergillus nidulans. Mol. Cell. Biol. 8: 3532-3541.         [ Links ]

Barak, R., Elad, Y., Mirelman, D. and Chet, I. (1985). Lectins: a possible basis for specific recognition in the interaction of Trichoderma and Sclerotium rolfsii. Phytopathology 75: 458-462.         [ Links ]

Chet, I., Harman, G.E. and Baker, R. (1981). Trichoderma hamatum: its hyphal interactions with Rhizoctonia solani and Pythium spp. Microbiol. Ecol. 7: 29-38.         [ Links ]

Elad, Y., Barak, R. and Chet, I. (1983a). Possible role of lectins in mycoparasitism. J. Bacteriol. 154: 1431-1435.         [ Links ]

Elad, Y., Chet, I., Boyle, P. and Henis, Y. (1983b). Parasitism of Trichoderma spp. on Rhizoctonia solani and Sclerotium rolfsii - Scanning electron microscopy and fluorescence microscopy. Phytopathology 73: 85-88.         [ Links ]

Fowler, T. and Berka, R.M. (1991). Gene expression systems for filamentous fungi. Curr. Opin. Biotechnol. 2: 691-697.         [ Links ]

Geremia, R., Jacobs, D., Goldman, G.H., Van Montagu, M. and Herrera-Estrella, A. (1991). Induction and secretion of hydrolytic enzymes by the biocontrol agent Trichoderma harzianum. In: Biotic Interactions in Soil-Borne Diseases, Developments in Agricultural and Managed-Forest Ecology Series (Beemster, A.B.R., Bollen, G.J., Gerlagh, M., Ruissen, M.A., Schippers, B. and Tempel, A., eds.). Vol. 23. Elsevier, Amsterdam, pp. 181-186.         [ Links ]

Geremia, R.A., Goldman, G.H., Jacobs, D., Ardiles, W., Vila, S.B., Van Montagu, M. and Herrera-Estrella, A. (1993). prb1, a gene specifically expressed during simulated mycoparasitism by Trichoderma harzianum. Mol. Microbiol. 8: 603-613.         [ Links ]

Goldman, G.H., Van Montagu, M. and Herrera-Estrella, A. (1990). Transformation of Trichoderma harzianum by high-voltage electric pulse. Curr. Genet. 17: 169-174.         [ Links ]

Goldman, G.H., Geremia, R., Caplan, A., Vila, S., Vilarroel, R., Van Montagu, M. and Herrera-Estrella, A. (1992). Molecular characterization and regulation of the phosphoglycerate kinase gene from Trichoderma viride. Mol. Microbiol. 6: 1231-1242.         [ Links ]

Goldman, G.H., Temmerman, W., Jacobs, D., Contreras, R., Van Montagu, M. and Herrera-Estrella, A. (1993). A nucleotide substitution in one of the b-tubulin genes of Trichoderma viride confers resistance to the antimitotic drug methyl benzimidazole-2-yl-carbamate. Mol. Gen. Genet. 240: 73-80.         [ Links ]

Goldman, G.H., Hayes, C.H. and Harman, G.E. (1994a) Molecular and cellular biology of biocontrol by Trichoderma spp. Trends Biotechnol. 12: 478-482         [ Links ]

Goldman, G.H., Vasseur, V., Contreras, R. and Van Montagu, M. (1994b). Sequence analysis and expression studies of a gene encoding a novel serine + alanine-rich protein in Trichoderma harzianum. Gene 144: 113-117.         [ Links ]

Haran, S., Schickler, H. and Chet, I. (1996). Molecular mechanisms of lytic enzymes involved in the biocontrol activity of Trichoderma harzianum. Microbiology 142: 2321-2331.         [ Links ]

Herrera-Estrella, A., Goldman, G.H. and Van Montagu, M. (1990). High-efficiency transformation system for the biocontrol agents, Trichoderma spp. Mol. Microbiol. 4: 839-843.         [ Links ]

Ridout, C.J., Coley-Smith, J.R. and Lynch, J.M. (1988). Fractionation of extracellular enzymes from a mycoparasitic strain of Trichoderma harzianum. Enzyme Microbiol. Technol. 10: 180-187.         [ Links ]

Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning, A Laboratory Manual. 2nd end. CSHL Press, New York.         [ Links ]

Vasseur, V., Van Montagu, M. and Goldman, G.H. (1995). Trichoderma harzianum genes induced during growth on Rhizoctonia solani cell walls. Microbiology 141: 767-774.         [ Links ]


(Received June 26, 1997)

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License