Molecular Identification of Trichoderma spp . in Garlic and Onion Fields and In Vitro Antagonism Trials on Sclerotium cepivorum

Trichoderma species are non-pathogenic microorganisms that protect against fungal diseases and contribute to increased crop yields. However, not all Trichoderma species have the same effects on crop or a pathogen, whereby the characterization and identification of strains at the species level is the first step in the use of a microorganism. The aim of this study was the identification – at species level – of five strains of Trichoderma isolated from soil samples obtained from garlic and onion fields located in Costa Rica, through the analysis of the ITS1, 5.8S, and ITS2 ribosomal RNA regions; as well as the determination of their individual antagonistic ability over S. cepivorum Berkeley. In order to distinguish the strains, the amplified products were analyzed using MEGA v6.0 software, calculating the genetic distances through the Tamura-Nei model and building the phylogenetic tree using the Maximum Likelihood method. We established that the evaluated strains belonged to the species T. harzianum and T. asperellum; however it was not possible to identify one of the analyzed strains based on the species criterion. To evaluate their antagonistic ability, the dual culture technique, Bell’s scale, and the percentage inhibition of radial growth (PIRG) were used, evidencing that one of the T. asperellum isolates presented the best yields under standard, solid fermentation conditions.


INTRODUCTION
Species from the genus Trichoderma include a vast number of strains with economic importance based on their biological control potential, being commercialized as biopesticides, biofertilizers, and soil enhancers.They are non-pathogenic microorganisms that provide protection against fungal diseases caused by Phytophthora, Rhizoctonia, Sclerotium, Pythium, and Fusarium genera (Peteira et al., 2001); additionally, they promote high yields in crops (Ezziyyani et al., 2004a).These traits derive from their ability to produce antifungal metabolites, release hydrolytic enzymes, and their mycoparasitic behavior, as well as the production of other substances that enhance plant growth (Suárez-Mesa et al., 2008).
In garlic (Allium sativum) crops, positive results of biological control against the pathogen Sclerotium cepivorum -responsible for white rot -have been observed through the use of native isolates of Trichoderma obtained from the region of Llano Grande, in Cartago, Costa Rica (Astorga Quirós et al., 2014a).However, not all Trichoderma species have the same effects over a pathogen or a specific crop.Hence the strain characterization and identification at a species level is the first step required to select the correct microorganism, since it determines which isolate presents a higher antagonistic effect over another particular organism (Rubio et al., 2005).
Studies on Trichoderma diversity and speciation have determined that the ribosomal RNA regions ITS1, 5.8S, and ITS2 are among the most informative sequences, in taxonomical and phylogenetic terms (Guigón-López et al., 2010).
The hypotheses was that Trichoderma is a genus present in the onion and garlic's rhizosphere in soil in Costa Rica, and the isolates obtained had different levels of antagonistic activity against S. cepivorum.The objective of this study was the identification, at a species level, of five different strains of Trichoderma isolated from soil samples gathered at garlic (Allium sativum) and onion (Allium cepa) production fields from the region of Llano Grande in Cartago, where microorganisms had not been implemented, through the study of the ribosomal RNA (rRNA) regions ITS1, 5.8S and ITS2; as well as the in vitro determination of the antagonistic effect of each strain on Sclerotium cepivorum Berkeley.

MATERIAL AND METHODS
Total DNA extraction and amplification of ITS1, 5.8S and ITS2 rRNA regions Five strains of Trichoderma spp., previously purified and identified as A, B, C, B 4 and D 4, were cultured in Petri dishes with Potato Dextrose Agar (PDA) medium, acidified with lactic acid.These strains were maintained at room temperature during three days, in order to allow growth of sufficient mycelium to perform total DNA extraction.
The DNA extractions were achieved through the use of Wizard Genomic DNA Purification Kit for A, B 4 and D 4 strains, while for B and C strains, the DNeasy® Plant Handbook kit was used.In both cases, the extraction was performed as described per the manufacturer's specifications, with a modification in the initial stage, with samples being exposed to liquid nitrogen in lysis buffer for 2 min, for the latter disruption of the material using a Retsch® macerator.DNA concentration and purity were quantified through spectrophotometry, using 260 and 280 nm absorbance values.
The amplification of the ITS1, 5.8S and ITS2 regions generated from the total DNA extractions was developed according to the method proposed by Guigón-López et al. (2010), using SR6R (5'-AAGTAGAAGTCGTAACAAGG-3') and LR1 (5'-GGTTGGTTTCTTTTCCT-3') oligos.The thermocycling program used was: 1 cycle at 94 °C for 2 min, 30 cycles at 94 °C for 30 s, 30 cycles at 50 °C cycles for 45 s, 30 cycles at 72 °C for 1 min and a final 72 °C cycle for 15 min.The PCR products were run using 1 % agarose gel during 30 min.The MassRuler TM DNA Ladder Mix molecular marker was used.Sequencing of the amplified products was performed sending 20 μL of each sample to the Macrogen Inc company (USA).

Sequence editing and alignment for phylogenetic reconstruction
The sequences were first edited with the BioEdit software (Hall, 1999), then assembled through the CAP3 tool (Huang and Madan, 1999); finally they were aligned to the homologous sequences through the tool MUSCLE, from EMBL-EBI (Edgar, 2004).The multiple alignment file was examined with the MEGA v6.0 software (Tamura et al., 2013) in order to calculate the genetic distance between the strains based on the number of base pair substitutions between the sequences.The positions with missing data were eliminated using the Tamura-Nei model.The phylogenetic trees were built through the Maximum Likelihood (ML) method, with a 500 repetition bootstrap.

Antagonistic activity of the isolated Trichoderma strains over S. cepivorum
The dual culture technique was used as described by Astorga Quirós et al. (2014b).For this study, five repetitions were performed for each strain, which were incubated under the same conditions for a 10-days period.
The antagonistic effect of the Trichoderma isolates was verified according to the scale proposed by Bell et al. (1982).Additionally, as a second indicator to test their antagonistic activity, the percentage inhibition of radial growth (PIRG) was determined employing the formula presented by Ezziyyani et al. (2004b), PIRG = (R 1 -R 2 )/R 1 × 100, where R 1 corresponds to the pathogen's growth radius (control) and R 2 is the growth radius of the pathogen once it is inoculated with Trichoderma.In order to establish the value for R 1 , five Petri dishes with PDA were inoculated with a disk 5 mm in diameter with 10-days of age active mycelium of S. cepivorum, which allowed us to determine the average radius of these cultures.A previous assay had been performed to corroborate the feasibility of reproducing these microorganisms in solid state fermentation.Considering these three characteristics (viz.Bell's scale, PIRG and reproduction feasibility), the potential application of each isolate, to be used in S. cepivorum biological control programs, was evaluated.
Aside from the previously mentioned factors, the ease of reproducing the different stains under a standardized protocol for solid state fermentation was also determined.

Reproduction Tests for every isolate
The ease of reproduction strains was assessed by the conidial concentration produced and viability of conidia.To do this, sterile trays with 500 g of hydrated rice were inoculated with 30 mL of a solution of spores at a concentration of 1.0 × 10 6 conidia/mL.Ten trays per strain were inoculated and incubated for 12 days at 28 °C, 80 % RH.Subsequently, for measuring the conidial concentration, serial dilutions in sterile distilled water were performed with 10 g of every tray.Counts in Neubauer's Chamber were performed using 10 4 dilution.The average of the samples was calculated for each strain.Viability was obtained by inoculating 0.2 mL of the same dilution on PDA medium.Ten plates were inoculated for each strain, and were incubated seven days at 28 °C.The colonies were counted and the average for each strain was obtained.

RESULTS
Total DNA extraction and amplification of ITS1, 5.8S and ITS2 rRNA regions An average concentration of 39.21 μg/mL with an average A 260 /A 280 absorbance ratio of 1.74 of the total DNA from the five strains of Trichoderma was obtained.The amplification of the segment of interest was approximately 600 pb (Figure 1).

Sequence editing and alignment for phylogenetic reconstruction
The amplified sequences from each strain were compared to the GenBank database through the "Nucleotide Blast" search, where the A, B and C strains (GenBank: KM456214, KM456216 and KM456217, respectively) evidenced a 99 % similarity with the KF723005.1 T. asperellum accession number.The D 4 strain (GenBank: KM456218) presented a 99 % similarity with the accession number EU280079.1, for T. harzianum.Regarding the B 4 strain (GenBank: KM456215), the oligonucleotides used were only able to identify the strain at a genus level, since there was not a significant percentage of similarity among the Trichoderma species to which it was compared in the database.
The development of the phylogenetic tree through the Maximum Likelihood (ML) method evidenced the existing differences between taxa, and presented a similar behavior to that observed when the sequenced PCR products were compared to the GenBank database, where the A, B and C strains were grouped together, while the strains B 4 and D 4 were more phylogenetically distant, being separated from the previously mentioned group.

Antagonistic activity of the isolated Trichoderma strains over S. cepivorum
The isolates presented divergent behaviors with regard to their antagonistic effects.According to Bell's scale, the strains ranged between classes 1 and 3; with B 4 , C and A being most aggressive (in that order), while B and D 4 showed a poor antagonistic activity.On the other hand, by analyzing the PIRG determinations, it was established that B 4 and C presented the best results (>50 %), being similar to the tendency observed in Bell's scale (Table 1).

Reproduction Tests for every isolate
Concentration measurements and viable conidia were different for each of the isolates (Table 2).Strain C had the highest production of conidia and as many viable conidia per ml of the tested solution.By contrast, B4 and D4 produced lower amounts of conidia and viable conidia.For the assessment of reproduction capacity, it was determined that viable conidia concentrations less than 1.0 × 10 8 were not suitable for controlled solid substrate fermentation processes; these values are indicative of poor reproducibility.

DISCUSSION
Total DNA extraction and amplification of the rRNA ITS1, 5.8S, and ITS2 regions The total DNA extraction procedure was executed using mycelium, in order to obtain the largest amount of genetic material, since various workers have mentioned the difficulties encountered when extracting total DNA from spores (e.g.Cassago et al., 2002;Lurá et al., 2003).The application of liquid nitrogen and the maceration along with the lysis buffer allowed a more efficient disruption, which solubilized a larger amount of nucleic acids, and ensured the deactivation of DNA degrading enzymes (Rønsted et al., 2007).The selected method allowed an average purity value of 1.74, which benefited the subsequent amplification and sequencing process of the bands of interest (Glasel, 1995).
The average 642 bp amplified region from ITS1, 5.8S, and ITS2 (Figure 1) presented a considerable similarity to that reported by Guigón-López et al. ( 2010), who obtained a 605 bp sequence using the same oligos for the identification of T. asperellum and T. longibrachiatum.In this context, it is possible to confirm the reports of Beiggi and Piercey (2007), as well as Besse (2014), who suggested the efficiency of this sequence for the determination of phylogenetic and evolutionary relationships, since it contained the highly conserved 5.8S rRNA region, which is surrounded by two highly variable regions (ITS1 and ITS2) that differ even between species (Deng et al., 2008).

Sequence editing and alignment for phylogenetic reconstruction
The use of Blast demonstrated that, through the amplified rRNA region of the studied strains, it was possible to determine homologies up to a 99 % similarity with species identified in the database.Strain B 4 was the only exception, where the identification reached the genus level.This outcome resembles Martin and Rygiewcz's (2005) report, where they suggest that obtaining low percentage results for identity through a Blast search can be due to the absence of reference sequences in the database.
On the other hand, the phylogenetic analysis of the strains in this study presented a similar tendency to the results gathered through the initial Blast search (Figure 2).Strains A, B and C were grouped in one same node, while D 4 was the most distantly related taxon.B 4 is more closely related to B and C; according to Mendoza-Revilla (2012), when more common ancestors are shared by two taxa, they are most related to each other.This is evidenced through the genetic distance matrix of the Trichoderma strains (Table 3), where the divergence of the taxa A, B and C in regards to B 4 is 0.06, while the divergence with D 4 is 0.70.
These results match those gathered by Sanz et al. (2004) in a study where they indicated that, due to the existing polymorphisms in each taxon, the rRNA coding sequence contributed to the distinction between Trichoderma genotypes; allowing exploration of the genetic diversity, as well as establishing associations among genotypes.

Antagonistic activity of the isolated Trichoderma strains over S. cepivorum
Trichoderma's biocontrolling ability is related to its development of different mechanisms that allow it to compete directly for space and nutrients.In fungi, these mechanisms are associated with the production of antifungal metabolites, the secretion of hydrolytic enzymes, and their mycoparasitic ability, being behaviors that are expressed at different intensities depending on the strain used (Michel Aceves et al., 2013).
The biocontrolling ability is shown in table 1, indicate the different antagonistic behaviors found in the same species.From Bell's scale as well as the PIRG determination, it was shown that B 4 , C and A strains, respectively, had the best antagonistic activity yields, even though strains A, B and C are the same species based on the phylogenetic study (Figure 1).A similar trend was found by Hoyos-Carvajal et al. ( 2008) when studying T. harzianum species, where it was concluded that each isolate could have a specific mycoparasitic activity.Therefore, it is not possible to generalize and indicate that a specific Trichoderma species can be classified as a good antagonistic fungi, because that characteristic is specific to each isolate.3. Genetic distance matrix between the isolated strains of Trichoderma

Figure 1 .
Figure 1.Electrophoretic bands generated during the optimized PCR reaction.

Table 1 .
Assessment of the Trichoderma spp.isolates in relation to their antagonistic activity and their reproduction ability in standard conditions for a solid state fermentation

Table 2 .
Measurement of the reproduction capacity in solid substrate fermentation