Healthy leaves of P. edulis were randomly collected in the municipality of Planaltina do Paraná, Paraná State, southern Brazil (latitude: 23º00’24’’S, longitude: 52º54’50’’W) and transferred to hermetic sealing plastic bags and stored at 4ºC for 24 hours. The endophytic fungi isolation was performed as previously described by Bongiorno et al. (2016), with slight modifications. Briefly, leaves were rinsed with 70% ethanol for 1 min., surface-disinfected with sodium hypochlorite solution (1% available Cl^-) for 2 min., and rinsed once in 70% ethanol (for 30 s) and twice in sterile distilled water. In a laminar airflow chamber, leaves were aseptically cut into 500 fragments (5-mm² area). There were five foliar fragments placed on each PDA dish (Potato Dextrose Agar, pH 6.6) (Himedia, Mumbai, India) supplemented with 50 µg.mL^-1 tetracycline (Sigma, St. Louis, MO) and they remained incubated at 28ºC for 7 days. The determination of colonization frequency (%) and fungal purification followed the protocols described by Ribeiro et al. (2018).

For the molecular identification, genomic DNA was extracted as described by Pamphile and Azevedo (2002) except that endophytes were previously grown for seven days in PDB dishes (Potato Dextrose Broth, pH 6.8) (Himedia, Mumbai, India). Amplification of the ITS1-5.8S-ITS2 (V9G 5’-TTACGTCCCTGCCCTTTGTA-3’ and ITS4 5’-TCCGTAGGTGAACCTGCGG-3’) (Invitrogen, Carlsbad, CA, USA) was carried out as described by White, Marrow, and Jones (1990) and Van Den Ende and De Hoog (1999), with initial denaturation at 94ºC for 5 min.; followed by 35 cycles of denaturation at 94ºC for 30 s, annealing at 48ºC for 50 s, and extension at 72ºC for 2 min.; followed by a final extension at 72ºC for 5 min. To amplify the partial gene of β-tubulin (T1 5’-AACATGCGTGAGATTGTAAGT-3’ and Bt-2b 5’-ACCTCAGTGTAGTGACCCTTGGC-3’) (Invitrogen, Carlsbad, CA, USA) were used the protocol described by Glass and Donaldson (1995) and O’ Donnell and Cigelnik (1997) with initial denaturation at 94ºC for 4 min.; followed by 40 cycles of denaturation at 94ºC for 30 s, annealing at 55ºC for 1 min., and extension at 72ºC for 1 min.; followed by a final extension at 72ºC for 3 min. Amplification of the glyceraldehyde-3-phosphate dehydrogenase (GPD1F 5’-CAACGGCTTCGGTCGCATTG-3’ and GPD2R 5’-GCCAAGCAGTTGGTTGTGC-3’) (Invitrogen, Carlsbad, CA, USA) was amplified according to Berbee, Pirseyedi, & Hubbard (1997) with initial denaturation at 96ºC for 2 min.; followed by 25 cycles of denaturation at 96ºC for 1 min., annealing at 52ºC for 45 s, and extension at 72ºC for 45 s; followed by a final extension at 72ºC for 10 min.

PCR products were sequenced by ACTGene Análises Moleculares Ltd. (Ludwigbiotec). Sequences were compared to those deposited in the GenBank (Altschul, Gish, Miller, Myers, & Lipman, 1990; http://www.ncbi.nlm.nih.gov/genbank) using the BLASTn tool with a restricted search to type strains. Based on data available in the TreeBase database (Piel et al., 2009; http://www.treebase.org), closely related species were also selected (Studies S17061 and S14881). Sequence alignment and phylogenetic analysis were performed according to Polonio et al. (2016). The sequences obtained were deposited in GenBank corresponding to the ITS1-5.8S-ITS2 (accession numbers: MK804479 to MK804492), β-tubulin (MK848597 to MK848609) and GPDH (MK861119).

There were 43 endophytic isolates randomly selected and tested for their biocontrol capacity against the pathogen Colletotrichum sp. CNPUV378 (provided by EMBRAPA Grape and Wine, Bento Gonçalves, RS, Brazil). For this, the dual culture method of Campanile, Ruscelli, and Luisi (2007) modified by Polonio et al. (2015) was used. All assay was performed in triplicate.

The percentage inhibition rate of mycelial growth (IP%) was calculated following Polonio et al. (2015): IP% = (1 - MT MC^-1) x 100, where IP% = inhibition index percentage of mycelial growth, MT = mean of triplicate area measured for treatment in cm² , and MC = average of triplicate control area in cm².

Competitive interactions (CI) between endophytes and phytopathogens were determined according to the rating scale proposed by Innocenti, Garibyan, and Badalyan (2002) and slightly modified by Ribeiro et al. (2018), which considers four main types of interactions (A, B, C and D) and their subtypes. The types are: A = inhibition of mycelial growth with contact; B = inhibition from a distance; C = endophytic growth on the pathogen without initial inhibition; and C= partial (CA1) and complete (CA2) endophytic growth on the pathogen after initial inhibition with mycelial contact; C = partial (CB1) and complete (CB2) endophytic growth on the pathogen after initial inhibition from a distance; D = partial (DA1) and complete (DA2) pathogen growth on the endophyte after initial inhibition with mycelial contact.

Ten top endophytes (named by codes PE-08, PE-14, PE-17, PE-22, PE-27, PE-34, PE-36, PE-39, PE-40 and PE-41) ranked in the dual culture experiment were randomly selected for the obtainment of crude ethyl acetate extracts containing secondary metabolites (CEAEs). There were 3 mycelia plugs (6-mm diameter) of seven-day-old cultures of each endophyte inoculated into 500-mL Erlenmeyer flasks containing 250 mL of PDB (Potato Dextrose Broth; pH 6.8) (Himedia, Mumbai, India) and incubated at 28ºC for 21 days under stationary condition. The metabolic extraction from the fermented culture broth, by using ethyl acetate P.A (Synth, Diadema, SP, Brazil) as solvent, followed the protocol described by Polonio et al. (2015).

To evaluate the bioactivity of CEAEs against Colletotrichum sp. CNPUV378, triplicates of PDA dishes received 5 mm autoclaved filter paper plugs inoculated with 10 μL of CEAEs (10 mg.mL^-1 in methanol) in an opposite position of phytopathogen plugs and remained incubated at 28 ºC for 7 days. For controls, the paper plugs received (instead of CEAEs) the fungicide Benlate® (30 mg.mL^-1) (C1) or methanol P.A (Synth, Diadema, SP, Brazil) (C2); C3 consisted of a phytopathogen plug inoculated in only one point of the PDA dishes. The percentage inhibition rate of mycelial growth (IP%) was calculated according to Polonio et al. (2015).

For the preparation of fungal inoculum, 6 mm diameter plugs of endophyte PE-36 and pathogen were macerated into microtubes containing 1.5 mL of autoclaved distilled water and then vortexed until completely homogenized. P. vulgaris seeds were treated by immersion in 70% ethanol for 1 min., surface-disinfected with sodium hypochlorite solution (2.5% available Cl-) for 3 min. and 70% ethanol for 30 s, and finally rinsed three times in sterile distilled water. To evaluate the efficiency of surface-disinfection, aliquots (100 μL) of water from the last rinsing was spread on Petri dishes containing PDA and TSA (Tryptone Soy Agar, pH 7.3) (Himedia, Mumbai, India). Then, seeds were transferred to dishes containing cotton soaked with 20 mL of water (both of them were previously autoclaved) and incubated at 28ºC without photoperiod for 3 days. On the 3^rd day, germinated seeds were transferred to 250 mL polypropylene containers with a mixture of Mecplant® substrate + vermiculite (3:1 v v^-1) and soil (1:1 v v^-1). The soil was collected at Universidade Estadual de Maringá (23º24’12.18” S and 51º56’30.54” W), autoclaved at 121ºC for 120 min.

There was one germinated seed placed in each container (with prepared soil and previously moistened with water) at approximately 3 cm depth, with the radicle facing down. The treatments were: control with water (C), plants only inoculated with the endophyte (T1), plants only inoculated with the phytopathogen (T2) and plants simultaneously inoculated with the endophyte + pathogen (T3). A total of 800 μL of each fungal suspension was used with 10 plants for each treatment (one plant per container). The presence or absence of symptoms and measurement of the biometric parameters (height, number of leaves, and root length) was performed 25 days after inoculations.

Data from the in vitro antagonism and antifungal activity the IP% values were analyzed using analysis of variance and the means were compared by the Scott-Knott test (p < 0.05 was taken to indicate a statistically significant difference) using the statistical program SISVAR 5.6 (Ferreira, 2011). To greenhouse experiment, the means of biometric parameters were compared by the Tukey’s test also taking p < 0.005 to indicate a statistical significant difference (Ferreira, 2011).

From the 500 leaf fragments sampled from P. edulis, we obtained a colonization frequency of 89%. The efficiency of the surface-disinfection process was attested by the absence of microbial growth in the control dishes. In order to isolate strains from the endophytic community, the surface disinfection of host plant material is necessary to ensure the elimination of epiphytes. In agreement with this current study, sodium hypochlorite and alcohol solutions have been employed in endophytic isolation protocols (Rani, Sharma, Chaturvedi, & Yadav, 2017; Salazar-Cerezo, Martinez-Montienel, Cruz-Lopez, & Martinez-Contreras, 2018; Lateef et al., 2019). For a look into the properties of endophytic microorganisms, the fragmentation of plant material represents a successful methodology that demonstrates satisfactory results for isolating the endophytic community, as already reported for different host plants (Wang et al., 2015).

The results of the BLASTn analysis are shown in Table 1. Phylogenetic analysis confirmed the molecular identification of endophytes PE-08, PE-14, PE-22, PE-27, PE-34, PE-36, PE-40, and PE-41 at the genus level as Colletotrichum sp. The taxonomy of PE-10, PE-17, PE-19, PE-26, and PE-38 was confirmed at the species level as Colletotrichum brevisporum [ITS BLASTn 98%; β-tubulin BLASTn 100%; 100% Bayesian probability (BP)], while PE-39 was confirmed as Curvularia senegalensis (ITS BLASTn 97%; 100% PB) (Figure 1).

Table 1
Endophytic fungi isolated from passion fruit and sequences with the greatest identity when aligned to the GenBank database (NCBI).

Figure 1
Cladogram results of the Bayesian analysis of endophytic fungi from Passiflora edulis with the alignment of three genes combined (ITS, β-Tubulin and GPDH*). The Bayesian probability was demonstrated on the nodes between each individual. *Only for PE-39

Literature data had already reported the abundant occurrence of Colletotrichum in the endophytic community (Vieira, Michereff, Morais Jr, Hyde, & Câmara, 2014; Nascimento et al., 2015; Gonzaga, Costa, Santos, Araújo, & Queiroz, 2015; Correia, Lira, Assis, & Rodrigues, 2018) Herein, Curvularia endophytes were less frequent in the P. edulis community; this genus was also isolated as an endophytes from other vegetal species (Bezerra et al., 2013; Nascimento et al., 2015; Correia et al., 2018).

Dual culture results showed differences among the antagonistic actions, ranging from 28.8% to 48.8% of IP% against Colletotrichum sp. CNPU378. Based on the rating scale proposed by Innocenti et al. (2002) and modified by Ribeiro et al. (2018), the types of CI observed were: A = deadlock (inhibition of phytopathogen) with mycelial contact (77.2%); DA1 = partial pathogen growth on the endophyte after initial inhibition with mycelial contact (13.6%); CA1 = partial endophyte growth on the pathogen after initial deadlock with mycelial contact (4.5%), and DB1 = partial pathogen growth on the endophyte after initial deadlock at a distance (4.5%). The Scott-Knott test highlighted 15 endophytic strains as promising antagonists in comparison with the control 2 (phytopathogen plug inoculated in only one point of Petri dish). More details regarding the antagonism index and competitive interaction are shown in Table 2 and Figure 2.

Table 2
Inhibition index of pathogen mycelial growth and competitive interaction between Passiflora edulis endophytes and Colletotrichum sp. CNPU378 in dual culture assay.

*Means of triplicates. Means followed by the same letter in column do not differ by the Scott-Knott test (p < 0.05). **Rating scale based on Innocenti et al. (2002) and Ribeiro et al. (2018). The absence of competitive interaction was indicated by (---).

Figure 2.
In vitro antagonism of endophytic fungi associated with Passiflora edulis against pathogen Colletotrichum sp. CNPUV38. A. Negative control with pathogen; B. Positive control with commercial fungicide; C. PE-17 (right) against pathogen (left); D. PE-34 (right) against pathogen (left); E. PE-36 (right) against pathogen (left); F. PE-39 (right) against pathogen (left); G. PE-40 (right) against pathogen (left); H. PE-08 (right) against pathogen (left); I. PE-14 (right) against pathogen (left); J. PE-22 (right) against pathogen (left); K. PE-27 (right) against pathogen (left); L. PE-41 (right) against pathogen (left).

Many studies have investigated the in vitro activity of endophytes against Colletotrichum pathogens. Rabha, Naglot, Sharma, Gogoi, and Veer (2014) showed that the indexes of antagonistic activity of fungal endophytes against Colletotrichum camelliae ranged from 18.33 to 36.73. The inhibition indexes (17.52 to 49.71%) obtained by Bongiorno et al. (2016.) were similar to those obtained in this study (28.8 to 48.8%). Supporting our results, literature data has already been reported on the deadlock (mutual inhibition) at mycelial contact (type A) as the most frequent in the pathogen-endophytes competitive interaction (Campanile et al., 2007; Polonio et al., 2015; Wang et al., 2015; Bongiorno et al., 2016), in which mutual fungal inhibition occurs and neither microorganism is able to overgrow the other (Innocenti et al., 2002; Oliveira et al., 2019).

Table 3 describes the results obtained for the in vitro antifungal activity of CEAEs of secondary metabolites. The higher IP% (=20.8%) against the Colletotrichum sp. CNPU378 was reported for the extract (named as code CEAE_PE-36) obtained from the endophyte PE-36.

Table 3
In vitro antifungal activity of crude ethyl acetate extracts (CEAEs) from Passiflora edulis endophytes, based on the inhibition index of mycelial growth (%) of Colletotrichum sp. CNPU378.

* Means of triplicates. Means followed by the same letter in column do not differ by the Scott-Knott test (p < 0.05). Control 1 = paper plug with 10 µL of fungicide Benlate® (30 mg.mL^-1); Control 2 = paper plug with 10 µL of methanol; C3 = phytopathogen plug inoculated in only one point of Petri dish. Concentration of CEAEs = 10 µL of 10 mg.mL^-1 solution of crude extracts.

Landum et al. (2016) reported that extracts with non-volatile compounds, produced by endophytes from olive plants, reached up to 26.8% of inhibition rates in dual culture tests against C. acutatum, similar to rates found in this study (up to 20.8% of mycelial inhibition).

Table 4 details the biometric parameters related to the vegetative growth of bean plants. Plants inoculated only with the endophyte Colletotrichum sp. PE-36 (T1) was symptomless and showed the difference for number of leaves and root length when compared to control plants (C) and other treatments (T2 and T3). Our findings suggest that the endophyte strengthened the growth promotion of the treated plants (T1). The T2 treatment presented inferior results in all biometric parameters analyzed. In addition, the T2 treatment presented whitish lesions in the root system apparently caused by the mycelial growth of Colletotrichum sp. CNPUV378, which were not found in the other treatments, suggesting that plants inoculated with endophyte + pathogen (T3), the endophyte was able to soften the propagules of the pathogen Figure 3.

Table 4.
Biometric parameters related to the vegetative growth of bean plants with and without inoculation of the endophyte Colletotrichum sp. (PE-36) and phytopathogen Colletotrichum sp. CNPU378.

Means of triplicates followed by lower-case letters indicate that values are not significantly different according to the Tukey test (p < 0.05)

Figure 3
Evaluation of the endophytic strain PE-36 against the pathogen Colletotrichum sp. CNPUV38 under greenhouse conditions. A. Control plant (C); B. plant inoculated with pathogen Colletotrichum sp. CNPUV38 and endophyte Colletotrichum sp. PE-36 (T3); C. Plant only with the pathogen Colletotrichum sp. CNPUV38 (T2).

Different mechanisms of in vivo biocontrol can be employed alone or together with other strategies to trigger direct or indirect responses to overcome the pathogenicity of the causal agent (Ghorbanpour, Omidvari, Dahaji, Omidavar, & Kariman, 2018). Consistent with our in vivo assay, Rodríguez et al. (2016) reporting that the application of a suspension of an endophytic fungus on necrotic leaves reduced the spore production of the pathogen after 21 days. In addition, the authors evaluated the endophytic colonization through plant height, root length, biomass, and the number of leaves. Li, Hwang, Huang and Huang (2018) described an endophyte able to carry out the suppression of Fusarium wilt of tomato plants and beefing up their growth, mainly through the absorption of nutrients.