Phytophthora parasitica and Fusarium oxysporum infect a wide range of host plants and cause significant damage on important crops around the world (Leslie and Summerell, 2006; Martin et al., 2014). To date, most of the phytosanitary management has been done through chemical control strategies; nevertheless, some strains of P. parasitica and F. oxysporum have developed resistance to several fungicides (Lucas et al., 2008). Due to this, searching for alternative biological strategies to control plant diseases has been growing in recent years (Anitha and Rabeeth, 2010). Amongst these, the use of Trichoderma, has received huge attention, because of its adaptability and diverse action mechanisms (mycoparasitism, antibiosis, competence and induction of endogenous plant defense) (Vinale et al., 2014). Trichoderma species are widely used in conventional agriculture and for biopesticide industry as biological agents or as enzymes sources. Trichoderma spp. produce several enzymes involved in their biological processes, including chitinases, glucanases, cellulases, proteases and amylases (Gajera et al., 2012; González et al., 2012); among these, chintinases and β-glucanases are the more relevant in the mycoparasitism. Chitinase 46-KDa (CHIT 46) affects the growth and cell wall of Sclerotium rolfsii and Rhizoctonia solani (Lima et al., 1999); β-1,6 glucanase, chitinase and protease produced by Trichoderma harzianum and Trichoderma viride also showed the ability to degrade cell walls of S. rolfsii (Parmar et al., 2015); furthermore, an extracellular alkaline-protease produced by T. harzianum of 18.8 kDa and active at pH 7 and 8 hydrolyzed cell-wall proteins from Crinipellis perniciosa, a pathogen with high resistance to antagonist microorganisms. P. parasitica and F. oxysporum also are sensible to lytic enzymes, as demonstrated by Budi et al. (2000), where the confrontation of several hydrolytic enzymes (chitinases, cellulases, proteases and pectinases) from Paenibacillus species promoted cell-wall and cell disorganization in some phytopathogens. The ability of Trichoderma spp. to parasite fungal phytopathogens is well known, and it is associated with formation of rolling, hooks and apresoria used to penetrate cell walls combined with enzymes, including chitinases and glucanases (Zeilinger and Omann, 2007). These enzymes attack cell walls structures formed almost exclusively of polysaccharides as chitin and glucanes (Küçük and Kivanc, 2004). Based on this, the main goal of this work was the quantification of chitinase and glucanase activities produced by four Trichoderma spp. and the evaluation of the inhibitory effect on the mycelial growth of P. parasitica and F. oxysporum.

Biological material

P. parasitica and F. oxysporum were isolated from Roselle plants (Hibiscus sabdariffa L.) with specific symptoms in the Laboratory of Biochemistry from the Colegio de Postgraduados, Campus Montecillo, Estado de Mexico, Mexico. P. parasitica was identified based on taxonomic keys from Erwin and Ribeiro (1996) and F. oxysporum was identified through taxonomic keys of Leslie and Summerell (2006).

To evaluate several Trichoderma species, Trichoderma virens (T6), Trichoderma asperellum (T9), Trichoderma gamsii (T13) and Trichoderma longibrachiatum (T19) were used. Trichoderma spp. were obtained from strain collection at the Instituto of Fitosanidad, Colegio de Postgraduados, Campus Montecillo, Estado de Mexico, Mexico. Species were identified with morphology and ITS sequencing by Esparza-Luna (2009) (unpublished data).

Chitinases and glucanases production

Trichoderma spp. were grown in liquid media to produce chitinases and glucanases according to (Almeida et al., 2007). Erlemeyer flasks were filled with 200 ml of culture media (1 L was prepared by mixing 10 g of KNO_3, 5 g of KH₂PO₄, 2.5 g of MgSO₄/H₂O; 2 mg of FeCl₃, 150 ml of V8 commercial juice, 10 g de PVP and distilled water to 1 L). Salts for media preparation were purchased from J. T. Baker (PA, USA), PVP was purchased from Sigma-Aldrich, MO, USA and the commercial juice was obtained from JUMEX (MO, Mexico). Two culture media were prepared by addition of 1% of chitin from shrimp shells (Sigma-Aldrich, MO, USA) and 0.1% of laminarin from Laminaria digitata (Sigma-Aldrich, MO, USA) to induce chitinases and glucanases production, respectively. The final pH of media was 6 and then was autoclaved at 121°C for 20 min. Separately, each species was inoculated in a flask with the addition of 1 ml of 1 × 10⁶ conidia/ml and incubated at 25°C under 12h:12 h (L:D) in continuous stirring (Thermo Scientific MaxQ 4000 Incubated Shake, MA, USA) for 5 days for chitinase and eight days for glucanase production. After incubation, media were filtered through paper Whatman No. 1 (Maidstone, UK) and centrifuged at 5,000 rpm for 15 min, supernatant was adjusted to 20 ml and used as source for protein quantification, chitinases and glucanases assays and for the inhibition assays (sterilized through a 0.22 µm filter (Millex® Millipore, MA, USA)).

Chitinase activity detection

Five tubes were marked for the experiment, in the first, 1 ml of water was added as a reaction blank, in the remaining tubes 1 ml of filtered media was added and then 1 ml of substrate solution (3.8 mg/ml of chitin in Citrate Mcllvaine buffer at pH 5 (prepared with 49 ml of 0.1 M solution of citric acid plus 51 ml of 0.2 M solution of sodium phosphate)) was incorporated. One tube was heated at 95°C in 10 min, after this, heated tube and remaining tubes were incubated at 37°C for 24 h. After incubation, 500 µl of each tube were placed in new tubes and heated at 95°C for 10 min and followed by addition of 100 µl of 0.8 M potassium tetraborate and then tubes were heated again during 5 min. Tubes were incubated at 37°C during 20 min and 3 ml of p-dimethylamino benzaldehyde solution (Sigma-Aldrich, MO, USA) (1 g was dissolved in 100 ml of 1.25% HCl - 8.75% aqueous solution) were added. Absorbance was recorded at 544 nm (spectrophotometer JENWAY 6305, OSA, UK). All reactions were done with four replicates and absorbance was compared with a standard curve, prepared with 0, 25, 50, 75, 100, 125, 150, 175 and 200 µg/ml of N-acetylglucosamine in Mcllvaine buffer. Chitinase activity was related to release of N-acetylglucosamine by comparison of the recorded absorbance from the initial heated tube minus average absorbance of the remaining tubes. Total chitinase activity (TA) and specific chitinase activity (SA) were calculated according to González et al. (2012) and expressed as µmol of N-acetylglucosamine/h and SA was expressed as µmol of N-acetylglucosamine/µg of protein/h. Protein quantification was according to Bradford (BIORAD, CA, USA) (Bradford et al., 1976).

Glucanase activity detection

Glucanase production was determined according to four filtered Trichoderma spp. media. Five tubes were marked as “X” and five as “Y”. One milliliter of 10 mg/ml lamminarin dissolved in 0.1 M citrate buffer pH 5 (0.1 M C₆H₅Na₃O₇ and 0.1 M citric acid, in 21:29 ratio) was added in each tube. Then, 250 µl of filtered media were added only in tubes marked as “X” before incubation (4 h at 40°C in a warm water bath). After incubation, 250 µl of filtered media were added in tubes marked as “Y” and then 2 ml of copper reactive (CR) was added to all tubes and immediately heated by immersion in boiling water for 10 min followed by addition of 2 ml of arsenomolybdate. Final volume was adjusted at of 25 ml with water. Solution was stirred in 1 min and finally absorbance was recorded at 595 nm (spectrophotometer JENWAY 6305). Calibration blank was prepared with water (Bruce et al., 1995).

CR solution was prepared previously and consisted of Cu I solution (16.1 g KNaC₄H₄O₆.4H₂O, 24 g of Na₂CO₃, 16 g of NaHCO₃, 80.44 g of Na₂SO₄ dissolved in distilled water to a final volume of 800 ml) and Cu II solution (4 g of CuSO₄.5H₂O, 36 g of Na₂SO₄ dissolved in a final volume of 200 ml), both solutions were mixed before using. The arsenomolybdate solution was prepared by mixing 25 g of NH₄Mo₇O₂₄.4H₂O in 450 ml of water and 21 ml of concentrated H₂SO₄ and 3 g of Na₂HAs₅O₄.7H₂O in 25 ml of water incubated at 37°C for two days. Salts for solutions preparation were purchased from J. T. Baker (PA, USA). Standard curve was made with 25, 50, 75, 100, 125 and 150 µg glucose/ml. Absorbance differences between X and Y were used to detect glucanase activity, detection was done with four replicates. Glucanase activity was calculated according to González et al. (2012), where total glucanase activity was expressed as µmol glucose/h, and glucanase specific activity was expressed as µmol glucose/h/µg of protein.

Chitinase and glucanase activity against fungal phytopathogens

Aqueous fraction of Trichoderma spp. cultures with chitinase or glucanase activity were sterilized through 0.22 µm filter (Millex®, Millipore) and mixed separately with warm PDA (Bioxon, Estado de Mexico, Mexico) (ratio 1:1) before filling Petri dishes (8 cm). In each plate, a 5 mm-piece of active mycelium of P. parasitica (5 days) or F. oxysporum (7 days) was placed and incubated at 25°C. Controls were established using PDA with distilled water (ratio1:1). Radial growth in cm was recorded every two days until the media were completely covered by mycelium (five days for F. oxysporum and seven days for P. parasitica). To estimate conidial production of F. oxysporum, 10 pieces 5 mm-diameter were stirred in 10 ml of distilled water to obtain conidial suspensions which were quantified in a Neubauer chamber (Wertheim, GE). Results from three dishes were recorded and expressed as number of conidia in 1 cm of PDA. Number of colony-forming units (CFU) was evaluated by inoculation of conidia in PDA and counted 48 h later.

Experimental design and data analyses

A completely randomized design was used with 9 treatments (4 glucanase activities and 4 chitinase activities and a control) with 5 replicates for two pathogens (P. parasitica and F. oxysporum). Response variables were mycelial radial growth, the inhibition percentage (calculated with the formula: [I (%) = (1 – T /C) × 100]; where, I= Inhibition percentage of pathogen growth by antagonists, C=radial growth in control, T=radial growth in the treatments), number of conidia and number of CFU. Analyses of variance (ANOVA) was used to detect differences between treatments and comparison means were done by Tukey test (P≤0.05) with SAS program version 9 (2002).

Chitinases and glucanases quantitation

The four Trichoderma spp. produced chitinases and glucanases at different levels; in general, chitinase TA (Total chitinase activity) values were between rates of 0.107 and 0.228 µmol of N-acetylglucosamine/h^-1 and SA (Specific Chitinase activity) rates were from 0.102 to 0.255 µmol h^-1 µg^-1 de N-acetylglucosamine/µg of protein. Statistical analyses detect differences (P≤0.0001) between species in TA and AE when chitinase activity was analyzed (Table 1). Total and specific activities were found higher in T. longibrachiatum, but this kind of activity was wide spread in all species, which seems typical in Trichoderma as reported for strains of T. harzianum (Michel-Aceves et al., 2005); however, differences in activities levels between species was evident and coincide with results from González et al. (2012) and Lee et al. (2012), who related such differences to endogenous characteristics, media composition and to culture age; such differences are considered useful by several authors to select Trichoderma spp. for biological control. Several studies on Trichoderma spp. have demonstrated the secretion of hydrolytic enzymes, including chitinases and glucanases, which have the ability to breakdown cell wall and penetrate the mycelium of other species (González et al., 2011).

It is well known that fungal β-glucanases have roles in glucan mobilization in the cell wall and also during nutrient depletion (Parmar et al., 2015), callose degradation, pathogenesis mechanism of antagonist fungi and nutrition of mycoparasites, for example, β-1,3-glucanases was associated with the lysis and wall degradation of Sclerotium spp. Glucanase activities obtained from four Trichoderma spp. fluctuated from 0.22 to 0.47 µmol of released glucose, while specific activity varied from 1.24 to 2.78 (Table 2). T. longibrachiatum showed the highest activity (0.47 µmol); however, T. gamsii and T. virens showed the higher specific activity, supporting that production depends on the species, which is similar to that reported by Parmar et al. (2015). Glucanase activity has been reported in several species of Trichoderma by Michel-Aceves et al. (2005), where enzyme production was T. koningii (1747.2 µmol), T. longibrachiatum (1400 µmol), T. virens (1045 µmol) and T. harziamun (1232 µmol), being higher than levels detected in this work, probably due to the culture conditions or genetic background of each species, similar to results reported by González et al. (2011), who indicated that the β-1,3 glucanase specific activity of Trichoderma spp., showed variations associated with variation in the substrate carbon source. On the other hand, Bruce et al. (1995) reported variations of total activity from 50 to 241 µmol of released glucose; nevertheless, specific activity fluctuated between 0.9 and 4.44 µmol h^-1 µg^-1 of glucose/mg protein and was in the range of results reported here, which is indicative that enzyme production is also influenced by culture media composition and the Trichoderma spp. evaluated.

Evaluation of chitinase and glucanase activity against P. parasitica and F. oxysporum

Production of enzymes by Trichoderma as part of their mycoparasitic process, where chitinases and glucanases are the most relevant enzymes has been reported (Küçük and Kivanç, 2008; Parmar et al., 2015). Gajera et al. (2012) and Parmar et al. (2015) indicated that chitinases and glucanases produce the lysis of fungal cell walls in some phytopathogens. Based on this, chitinase effects on P. parasitica mycelium growth was different between the T. gamsii, T. asperellum and T. longibrachiatum (Tables 3 and 4). On the other hand, enzyme activity obtained from T. asperellum, T. virens and T. gamsii induced abundant mycelial growth of P. parasitica, which could be attributed to enzyme inhibitors possibly produced.

Antagonistic activity could be also influenced by specific recognition at the molecular level, which determines the success or failure for the control; for example, the glucanase AGN13.1 recognized the cell wall of Aspergillus niger, Botrytis cinerea, Colletotrichun acutatum, F. oxysporum, Penicillium aurantiogriseum and R. solani but did not affect other evaluated fungi (Ait-Lahsen et al., 2001). This suggests that no evident inhibitory effects observed here by the four Trichoderma spp. could be related to absence of specific recognition on P. parasitica and F. oxysporum.

Some reports indicated that P. parasitica could induce glucanase production to defend itself against antagonists, which in addition to the possible enzyme inhibitor production could be part of factors that influenced the null inhibitory effects; however, this remains for confirmation. Chitinase and glucanase antibiotic effect on mycelial growth of F. oxysporum showed significant differences between treatments (P≤0.0001) (Figure 1). T. longibrachiatum showed the highest inhibitory effect with a reduction of 40% (4.4 cm) when chitinase activity was evaluated and 45% (4.8 cm) when glucanase were used (Table 3). Michel-Aceves et al. (2005) indicated that glucanases and chitinases produced by T. koningii and T. harzianum presented variations in their accumulation; nevertheless, also inhibited growth of F. oxysporum. Nevertheless, Gajera et al. (2012) and Parmar et al. (2015) reported that a significant positive correlation between percentage growth inhibition (Macrophomina phaseolina and S. rolfsii) of test fungus and lytic enzymes (chitinase, β-1,3-glucanase and protease) in the culture medium of antagonist treatment established a relationship to inhibit growth of fungal pathogen by increasing the levels of these enzymes.

Similarly to P. parasitica results, it was also observed an abundant growth of F. oxysporum when challenged with enzymes activities of T. asperellum, T. virens and T. gamsii (Figure 2); however, F. oxysporum inoculated in media containing enzyme activities did not produce violet pigments, typical of this species, which accumulation was observed in control media, indicating that extracts did not affect the growth rate but could have effects at other levels. Such inhibition of pigment accumulation could diminish pathogenic potential, because such pigments usually are associated with pathogenic roles and are important in the interaction phytopathogen-plant (Medentsev et al., 2005). Nelson et al. (1993) reported that F. oxysporum produce secondary metabolites like micotoxins, trichothecenes, pigments, antibiotics and phytotoxins that could have influence on the null inhibition of F. oxysporum growth. Number of conidia and CFU were different (P≤0.0001) between treatments and it demonstrated a relation with the higher mycelium growth that presented the higher number of conidia and CFU, demonstrating that an important parameter for reproduction as the conidial viability was not affected, which contrasted with report of Michel-Aceves et al. (2005) who found a 95% reduction in conidial production and around 70% of loss of viability when T. harzianum and T. koningii were used, rising the arguments to consider that physiological conditions, genetic background and nutritional availability are directly related with antagonistic effects of each Trichoderma spp., which should be considered as part of the production management and for control strategies that guarantees successfully plant protection.

All the Trichoderma spp. produced hydrolytic activities depending on the carbon source and activity levels were independent of the species. T. longibrachaitum was the species with higher total and specific enzyme activities of both activities; it also showed antagonistic effect on both phytopathogens.

The authors have not declared any conflict of interests.

The authors wish to thank Colegio de Postgraduados and Consejo Nacional de Ciencia y Tecnologia (CONACYT) for the providing scholarship to the first author (Scholarship number: 50928) and to the grant SAGARPA-163972.”