OUP user menu

Antagonism of Pythium blight of zucchini by Hypocrea jecorina does not require cellulase gene expression but is improved by carbon catabolite derepression

Verena Seidl , Monika Schmoll , Barbara Scherm , Virgilio Balmas , Bernhard Seiboth , Quirico Migheli , Christian P. Kubicek
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00157.x 145-151 First published online: 1 April 2006


Toward a better understanding of the biochemical events that lead to biocontrol of plant pathogenic fungi by Hypocrea/Trichoderma spp., we investigated the importance of carbon catabolite (de)repression and cellulase formation in the antagonization of Pythium ultimum by Hypocrea jecorina (Trichoderma reesei) on agar plates and in planta. Hypocrea jecorina QM9414 could antagonize and overgrow P. ultimum but not Rhizoctonia solani in plate confrontation tests, and provided significant protection of zucchini plants against P. ultimum blight in planta. A carbon catabolite derepressed cre1 mutant of H. jecorina antagonized P. ultimum on plates more actively and increased the survival rates of P. ultimum-inoculated zucchini plants in comparison with strain QM9414. A H. jecorina mutant impaired in cellulase induction could also antagonize P. ultimum on plates and provided the same level of protection of zucchini plants against P. ultimum as strain QM9414 did. We conclude that cellulase formation is dispensable for biocontrol of P. ultimum, whereas carbon catabolite derepression increases the antagonistic ability by apparently acting on other target genes.

  • biocontrol
  • Hypocrea jecorina
  • Trichoderma
  • cellulases
  • Pythium
  • carbon catabolite derepression


Biocontrol of plant pathogens is an attractive alternative to the strong dependence of modern agriculture on fungicides, which results in environmental pollution and the selection of resistant strains. Replacement or reduction of chemical application has been achieved through use of biologically based pesticides, a concept included in the definition of biocontrol proposed by Cook & Baker, 1983: ‘Biological control is the reduction of the amount of inoculum or disease-producing activity of a pathogen accomplished by or through one or more organisms other than man.’ Some mycoparasitic Hypocrea/Trichoderma species, primarily Hypocrea lixii (Trichoderma harzianum) and Hypocrea atroviridis (Trichoderma atroviride), have been used as biocontrol agents against several economically important plant pathogenic fungi (Harman et al., 2004). Unfortunately, the successful application of biocontrol strains is not easy to predict, and these strains are usually not economically competitive with chemical fungicides. A better understanding of the biochemical events that lead to mycoparasitism and their regulation could identify methods to improve the reliability of Hypocrea/Trichoderma strains as biocontrol agents (Roderick & Navajas, 2003).

The majority of the molecular research on biocontrol has so far been focused on the role of hydrolytic enzymes in antagonism of plant pathogenic fungi (for reviews see Chet et al., 1998; Kubicek et al., 2001; Howell, 2003; Benitez et al., 2004). However, these studies dealt only with the effect of single enzymes/genes, which – in view of the high number of hydrolytic enzymes known to be encoded in the genome of Hypocrea jecorina (Trichoderma reesei; http://gsphere.lanl.gov/trire1/trire1.home.html) and other filamentous fungi (Magnaporthe grisea, Fusarium graminearum, etc.; http://www.broad.mit.edu/annotation/fungi/) – only leads to limited insights into the general role of these enzymes. In addition, many of these hydrolases are controlled by carbon catabolite repression. In vitro studies with H. atroviridis suggested that the onset of mycoparasitism is indeed accompanied by carbon catabolite derepression (Lorito et al., 1996), but this has so far not been proven in vivo. One of the obstacles toward investigating these points is that the Hypocrea/Trichoderma species and strains used in biocontrol studies are difficult to manipulate genetically (Rocha-Ramirez et al., 2002; Delgado-Jarana et al., 2003). In addition, creating mutants deleted in 10–20 different hydrolase genes is a demanding task. In contrast, well characterized mutants defective in the formation of all cellulases (Torigoi et al., 1996; Zeilinger et al., 2000) and characterized mutants defective in carbon catabolite repression (Strauss et al., 1995; Ilmen et al., 1996) are already available for H. jecorina.

Hypocrea jecorina is widely used for the production of cellulolytic and hemicellulolytic enzymes and recombinant proteins (Archer & Peberdy, 1997; Mach & Zeilinger, 2003). While H. jecorina has not been evaluated for its ability to antagonize other fungi, Trichoderma longibrachiatum, a species closely related to H. jecorina, can successfully protect cucumber plants against Pythium ultimum (Migheli et al., 1998).

The objective of this study was (i) to test whether H. jecorina can be used to antagonize plant pathogenic fungi and protect plants against their attack; and (ii) to make use of available H. jecorina mutants to address issues that have been reported as being potentially relevant for antagonism of other fungi (Lorito et al., 1996; Kubicek et al., 2001; Howell, 2003; Roderick & Navajas, 2003; Harman et al., 2004): how important is the formation of hydrolytic enzymes for antagonism of plant pathogenic fungi? And would carbon catabolite derepression be beneficial for it? Answers to these questions will expand our understanding of the mechanism of biocontrol and eventually provide new strategies toward improvement of existing biocontrol strains of Hypocrea/Trichoderma.

Materials and methods

Strains and culture conditions

Hypocrea jecorina QM9414 (ATCC 26921) is a moderate cellulase producing second-generation mutant of the wild-type strain QM6a; H. jecorina QM9978 (obtained from K. O'Donnell, US Department of Agriculture, Peoria, IL) is a mutant strain unable to produce cellulases (Zeilinger et al., 2000); H. jecorina RUT-C30 (ATCC 56765) is a carbon catabolite-derepressed strain because of a truncation of its cre1 gene (Ilmen et al., 1996); Hypocrea atroviridis P1 (ATCC 74058) is a strain used in biocontrol (Lorito et al., 1996); Pythium ultimum (Woo et al., 1999) and Rhizoctonia solani (strain RT-10, belonging to anastomosis group AG-4, kindly provided by Dr E. Lahoz). All strains were maintained on malt extract agar (MEX; 3% weight in volume).

Plate confrontation tests

Strips of 3 mm × 3 cm were cut from the growing front of H. jecorina and H. atroviridis, and P. ultimum and R. solani on MEX plates, respectively, and placed on fresh plates at a distance of 4 cm from each other. Both MEX and minimal medium (MM; g L−1: agar, 10; glucose, 3; MgSO4·7H2O, 0.2; K2HPO4, 0.9; KCl, 0.2; NH4NO3, 1; FeSO4·7H2O, 0.005; ZnSO4·7H2O, 0.002; MnSO4·H2O, 0.002; and CoCl2·2H2O, 0.002) were used; plate confrontation assays were performed in the dark at 28°C.

RNA isolation and northern analysis

For RNA extraction from confrontation experiments, plates were covered with a dialysis membrane (cut-off size 12 kDa, Sigma, Vienna, Austria) and mycelia in the area of interaction or close to it harvested with a spatula, immersed in liquid N2 and ground to a fine powder. Equivalent zones were collected from control plates, inoculated with either H. jecorina or P. ultimum.

Total RNA was isolated by the guanidinium thiocyanate method (Chomczynski & Sacchi, 1987). Following electrophoretic separation on a 1.2% agarose-gel containing 2.2 M formaldehyde in 1 × MOPS buffer (40 mM MOPS, 1 mM EDTA, pH 7.0), RNA was blotted onto nylon membranes (Biodyne B, Pall Corporation, VWR International, Vienna, Austria) and hybridized in 50% formamide, 10% dextransulfate, 0.5% SDS, 5 × Denhardt's solution and 125 μg mL−1 denatured fish sperm DNA at 42°C for 20 h. Washing was performed with 2 × SSC+0.1% SDS at 42°C (2 × 10 min).

A 1264 bp PCR fragment of cbh1 (Shoemaker et al., 1983) was used as probe for northern hybridizations, and a 297 bp PCR fragment of 18S rRNA gene (Accession number: Z48932) was used as the hybridization control.

Greenhouse experiments

Three experiments were carried out on zucchini (Cucurbita pepo L.) seedlings (F1 hybrid cultivar Greyzini; Zorzi Seeds, San Pietro Viminario, Italy). Using a sterile cork borer, plugs of 15 mm in diameter were cut from 7-day-old cultures of P. ultimum or R. solani grown at 25°C on Schmitthenner's Agar (Atlas, 1997) and potato dextrose agar (PDA, Difco, Becton Dickinson, Franklin Lakes, NJ), respectively. They were then placed in the center of plastic sowing pots (4.5 cm in diameter, 55 mL capacity, one plug per pot), and covered by a 2 cm layer of sterilized (121°C for 60 min on 2 successive days) potting mix (Humin-Substrat N17, Neuhaus, Germany). Agar plugs of H. jecorina or H. atroviridis were cut from 7-day-old cultures grown at 25°C on PDA. One plug of H. jecorina or H. atroviridis was paired with one plug of P. ultimum or R. solani, by placing the mycelia in direct contact. Plug pairs were incubated in the dark at 25°C for 24 h and then transferred to the center of plastic pots and covered by sterilized substrate as described. For each treatment, five replicates (10 seeds for each replicate) were incubated on a bench in a glasshouse for 7 days before zucchini seeds (one seed per pot) were added. Pots were watered daily and the average temperature was 25–20°C (min 10–15°C, max 25–35°C). Seedling emergence was checked weekly after 7–21 days. After the last survey, the healthy plant stand was assessed.

Statistical analysis

Data from greenhouse experiments were analyzed using the analysis of variance (anova). anova was conducted after transforming the original data (expressed as percentage, %) using the arcsinx function, where x is the relative proportion. This transformation is appropriate to percentage and useful when original data do not fall between 30% and 70% (Sokal & Rohlf, 1995). Means separation was done by the Tukey–Kramer honestly significant difference (HSD; Sokal & Rohlf, 1995) test. All the analyses were performed by using J.M.P. ver. 3.1.5 software (SAS Institute Inc., Minneapolis, MN).


Plate confrontation experiments between Hypocrea jecorina and Pythium ultimum or Rhizoctonia solani

Hypocrea jecorina QM9414 antagonized Pythium ultimum (Fig. 1a). Although P. ultimum initially grew faster than H. jecorina, its growth stopped immediately upon physical contact with H. jecorina, which began to overgrow P. ultimum after approximately 2 days of incubation. Overgrowth was completed 7–9 days after inoculation and resulted in complete degradation of the host and sporulation of H. jecorina over the entire plate. The characteristic stages of the plate confrontation experiments were observed with fungi cultured on both rich (MEX) and minimal medium (MM). Hypocrea jecorina also exhibited the characteristic coiling around the host hyphae (Fig. 2) that has been implicated in the recognition and subsequent penetration of the host (Inbar & Chet, 1996).


(a) Plate confrontation tests of Hypocrea jecorina QM9414 (T) against Pythium ultimum (P) performed on malt extract agar medium. (b) Plate confrontation tests of H. jecorina QM9414 (T) against Rhizoctonia solani (R).


Light microscope-micrographs of Hypocrea jecorina QM9414 coiling around Pythium ultimum. The pictures shown were taken from plate confrontation assays on MM medium after 3 days. Arrows point to interaction zones between Hypocrea jecorina and P. ultimum hyphae. The bar marker indicates 10 μm (same scale on all pictures).

Hypocrea jecorina did not antagonize Rhizoctonia solani, as no overgrowth occurred after physical contact between H. jecorina and R. solani (Fig. 1b). Hypocrea jecorina sporulated only on its half of the plate and the characteristic brown pigments of R. solani appeared only in the other half.

Protection of zucchini plants against Pythium ultimum and Rhizoctonia solani blight

Although H. jecorina has an effective cellulase system, it causes no negative effect on the plant in the absence of pathogens (Table 1). In the presence of P. ultimum, H. jecorina QM9414 significantly antagonized the pathogen P. ultimum, the healthy plant stand being increased from 39% to 55% (Table 1). Plant damage caused by R. solani was not reduced significantly by H. jecorina, and the recovery of healthy plants in the presence of H. jecorina was only 13% in the presence of a high level of disease pressure (healthy plant stand in the inoculated control: 6%; Table 1).

View this table:

Effect of Hypocrea jecorina QM9414, RUT-C30 and QM9978 and Hypocrea atroviridis P1 on the development of zucchini seedlings (cv Greyzini) sown in greenhouse in the absence or in the presence of Pythium ultimum or Rhizoctonia solani

The biocontrol fungus Hypocrea atroviridis P1 was included as an internal control. This strain was able to effectively control both P. ultimum and R. solani, allowing a significant improvement of the healthy plant stand from 39% and 6% to 84% and 78%, respectively (Table 1).

Interaction of Pythium ultimum and Rhizoctonia solani with a cre1 mutant of Hypocrea jecorina

The H. jecorina mutant RUT-C30 contains a truncated copy of cre1 (Strauss et al., 1995; Ilmen et al., 1996) and is consequently carbon catabolite derepressed. This strain was much more active in antagonizing P. ultimum on plates (Fig. 3) than H. jecorina QM9414. Upon physical contact of P. ultimum with RUT-C30, P. ultimum growth stopped completely and RUT-C30 overgrew its host much faster than QM9414 did. As with H. jecorina QM9414, no overgrowth occurred in confrontation experiments with R. solani.


Plate confrontation assays of the carbon catabolite derepressed mutant Hypocrea jecorina RUT-C30 (T) against Pythium ultimum (P), performed on malt extract agar medium.

The increased antagonistic ability of RUT-C30 against P. ultimum can also be seen in the results from the two greenhouse experiments, which show that RUT-C30 protected zucchini plants more effectively than QM9414 (Table 1), thereby equaling H. atroviridis (Table 1).

Cellulase gene expression and biocontrol of Pythium ultimum

Hypocrea jecorina QM9978 is a mutant that is unable to induce the formation of cellulases due to a defect in cellulose to cellulase signaling (Torigoi et al., 1996; Zeilinger et al., 2000). Despite this fact, this strain also antagonized and overgrew P. ultimum, but not R. solani, on plates (Fig. 4). Protection of zucchini plants against P. ultimum was slightly, although statistically not significant, enhanced over that achieved with strain QM9414 (Table 1).


Plate confrontation assays of the cellulose-negative mutant Hypocrea jecorina QM9978 (T) against Pythium ultimum (P), performed on malt extract agar medium.

Strain QM9978 does not produce cellulases during growth on cellulose or sophorose (Zeilinger et al., 2000) but its lack of cellulase formation has not been proven under plate confrontation conditions. We compared therefore cellulase gene expression (using the gene encoding the major secreted cellulase Cel7A, cbh1) during confrontation assays of H. jecorina QM9414, QM9978 and RUT-C30 with P. ultimum. The results obtained confirm that QM9978 does not express cbh1 (Fig. 5). Interestingly, cbh1 expression in the cellulase-producing strain QM9414 only occurred before contact and was not detectable after contact with P. ultimum. In contrast, the catabolite-derepressed strain RUT-C30 showed significant cellulase gene expression before and also after contact with P. ultimum.


Abundance of the cbh1 transcript in Hypocrea jecorina during incubation with Pythium ultimum under plate confrontation conditions after 4 days of incubation. Plates with H. jecorina alone were used as control and 18S rRNA gene was used as loading control.


We found that Hypocrea jecorina, a fungus used for the production of industrial enzymes, can act as an antagonist against Pythium ultimum. However, despite the strong antagonism of H. jecorina QM9414 in plate confrontation experiments, protection of zucchini seedlings in in planta assays was only moderate in comparison with the mycoparasitic strain Hypocrea atroviridis P1. This could be owing to soil preference of H. jecorina: this fungus is known only from a narrow belt around the equator (Turner et al., 1996), and is particularly abundant in tropical soils (Lieckfeldt et al., 2000; Kubicek et al., 2003). It is possible that H. jecorina would display better plant protection when the respective soil characteristics of these regions are taken into account. However, a deficiency in other factors required for competence in the rhizosphere of the plant cannot be excluded at this stage. Nevertheless, the protection of zucchini plants against Pythium by H. jecorina QM9414 was significant enough to justify subsequent studies on the role of cellulolytic enzymes and carbon catabolite derepression in this process. The moderate biocontrol potential of QM9414 might even be advantageous if it enables us to identify both genes that are essential for biocontrol and those whose manipulation could improve biocontrol properties.

Production of hydrolytic enzymes has frequently been emphasized as one of the major factors contributing to the biocontrol activity of Hypocrea/Trichoderma spp. (Migheli et al., 1998; Zeilinger et al., 2000; Kubicek et al., 2001; Howell, 2003; Roderick & Navajas, 2003). As H. jecorina was able to antagonize P. ultimum, which as an oomycete has cellulose as a major component of its cell wall, we tested the potential role of cellulases in the antagonism of P. ultimum. We used a mutant (QM9978) which can not hydrolyze amorphous cellulose under conditions that induce cellulases in other strains (Torigoi et al., 1996; Zeilinger et al., 2000; M. Mandels and C.P. Kubicek, unpublished results), and thus probably forms none of the 11 cellulases which are present in the H. jecorina genome (Foreman et al., 2003; http://gsphere.lanl.gov/trire1/trire1.home.html). The results show that the cellulase-negative mutant QM9978 can also overgrow P. ultimum and that this mutant protects zucchini plants against Pythium blight, suggesting that cellulases are not essential for the antagonism. One could argue that the cell wall of P. ultimum may contain still unidentified inducers of the cellulases of H. jecorina, whose signaling is not impaired in QM9978. To rule out this possibility, we examined the expression of cellulases by H. jecorina during confrontation with P. ultimum, using the major cellulase cbh1 as a model gene. All studies so far showed that the expression of the various cellulase genes is coregulated (Torigoi et al., 1996; Archer & Peberdy, 1997; Foreman et al., 2003; Mach & Zeilinger, 2003), and we therefore assume that the expression of cbh1 also reflects that of the other cellulases. Our data indicate that QM9978 indeed does not form cellulases during interaction with P. ultimum on plates. Further, the fact that QM9414 expresses some cbh1 before contact but shuts it off during overgrowth further adds to the conclusion that cellulases are dispensable for H. jecorina during antagonization of P. ultimum. The fact that the catabolite derepressed strain RUT-C30 shows enhanced cbh1 gene expression during confrontation and maintains this expression during overgrowth of P. ultimum indicates that carbon catabolite repression is one reason for the low cbh1 expression and its turn-off in the cellulase producer QM9414.

The finding that cellulose hydrolysis is dispensable for antagonism of P. ultimum by H. jecorina does not rule out that other hydrolytic enzymes, such as β-glucanases or proteases (Kim et al., 2002; Delgado-Jarana et al., 2003; Pozo et al., 2004; Suarez et al., 2004), may still be important for it; however, it clearly indicates that hydrolysis of the major structural polymer of the pathogens cell wall is less significant.

In the light of these findings it is unclear, why H. jecorina antagonized P. ultimum and not Rhizoctonia solani. It would be tempting to speculate that the reason could be related to a general difficulty to attack fungi with chitin as major structural cell-wall polysaccharide. However, the H. jecorina genome contains 18 genes encoding putative chitinases (Seidl et al., 2005), including orthologues of all of the chitinases which have so far been characterized from H. atroviridis, Hypocrea lixii, Hypocrea virens and Trichoderma asperellum (Kubicek et al., 2001; Viterbo et al., 2001) and has therefore at least the necessary genes available. The inability of H. jecorina to overgrow R. solani must therefore be more complex, and probably related to other factors such as signaling of the presence of the host or regulation of enzyme induction.

Lorito (1996) demonstrated that in H. atroviridis P1, binding of the Cre1 carbon catabolite repressor to the promoter of the ech42 endochitinase is abolished during contact with the host and is replaced by a mycoparasitism-specific protein. Relief from carbon catabolite repression could thus accelerate the induction of the mycoparasitic response and improve antagonism. The results shown here with the carbon catabolite derepressed RUT-C30 demonstrated in fact improved biocontrol of P. ultimum in planta. While this confirmed the hypothesis, our findings that cellulases are of little relevance to antagonism of P. ultimum raises the question: What are the targets of mycoparasitism-related carbon catabolite repression? β-glucanases or proteases, as discussed above, could be such targets; but other physiological responses of H. jecorina, such as the formation of antifungal compounds (Sivasithamparam & Ghisaberti, 1998), penetration structures (Goodwin & Chen, 2002), or competition for fungal germination elicitors (Howell, 2003) might also be subject to carbon catabolite repression and be responsible for the increased antagonistic capability of RUT-C30. In any case, our results justify a closer look at the role of carbon catabolite repression in the capabilities of agriculturally important biocontrol strains.


This work was supported by grants from the Austrian Science Foundation (FWF-P13672) and the European Commission (QLK3-CT-2002-02032). Q.M. and C.P.K. also acknowledge support from an Austrian–Italian bilateral grant (Project 10/2004). The authors wish to thank Dr Domenico Rau for helpful suggestions concerning the statistical analysis of data.


View Abstract