Evaluation of biocontrol properties of Streptomyces spp. isolates against phytopathogenic fungi Colletotrichum gloeosporioides and Microcyclus ulei

South American Leaf Blight (SALB) of the rubber tree, caused by Microcyclus ulei and foliar anthracnose caused by Colletotrichum gloeosporioides, are diseases that adversely affect rubber cultivation in America. Both diseases have a significant economic impact on this agricultural subsector. The aim of the present study was to evaluate the potential as biological control agents of three Streptomyces species strains, namely A20, 7.1 and 5.1, against M. ulei and C. gloeosporioides. The results of analysis of variance (ANOVA) and Tukey post-hoc test of the in vitro antifungal activity assays evidenced the potential of the three Streptomyces strains to inhibit C. gloeosporioides growth through the production of diffusible (A20 and 5.1) and volatile compounds (7.1). Furthermore, other results indicated that strain 5.1 had a high biocontrol activity against C. gloesporoides, and thus such strain was selected for further evaluations as a possible biocontrol agent against M. ulei. In vitro assays suggested that active compounds produced by 5.1 inhibited M. ulei growth by interfering with conidia germ tube and stroma formation. Bioassay-guided fractionation with organic solvents of 5.1 fermentation broths, suggested that antifungal compounds produced by this strain were nonionic compounds of medium-polarity. Currently, studies are ongoing to elucidate the chemical structure of these antifungal compounds. These approaches aim to generate a biological control agent to provide the Colombian rubber subsector with a preventive measure for controlling M. ulei and C. gloeosporioides.


INTRODUCTION
Hevea brasiliensis is a native plant of the Amazon basin, member of the Euphorbiaceae family used for rubber production (natural rubber). Approximately, 10 million tons/year of natural rubber are produced from this tree, Afr. J. Microbiol. Res.
93% of the production comes from South East Asian countries such as Thailand, Malaysia, and Indonesia, 4.5% from Africa and only 2.5% from Latin American countries Rivano et al., 2015). Even though, Colombia is not a natural rubber exporter, it possesses an ample natural rubber cultivation tradition located in the departments of Caquetá, Putumayo, Guaviare and Córdoba and more recently Santander and Meta with a cultivation area of 44,100 ha (Castellanos et al., 2009;Confederación Cauchera Colombiana, 2015). The difference observed in productivity between Asian and Latin American countries is generated in great part by the presence of an endemic disease known as the South American Leaf Blight (SALB) of the rubber tree caused by the Ascomycete Microcyclus ulei fungus (Gasparotto and Pereira, 2012). This phytopathogen infects mature fruits, stems, and young leaves (stage B) of the Hevea genus reducing the plant's growth, causing premature leaf falling, with reduction of its photosynthetic area. Moreover, it causes death of susceptible greenhouse clones and garden clones (Chee and Holliday, 1986;García et al., 2007;Gasparotto and Pereira, 2012). In addition to SALB, foliar anthracnose caused by Colletotrichum gloeosporioides is the second disease of fungal origin limiting H. brasiliensis natural rubber production (Guyot et al., 2005;Castro, 2011;Gasparotto and Pereira, 2012). This disease has generated an important impact on Colombian and Brazilian crops, with increasing incidence (Furtado and Trindade, 2005;García et al., 2007;Castro, 2011). Despite the efficiency obtained from traditional chemical treatments for M. ulei and C. gloeosporioides in H. brasiliensis, high costs and environmental impact from airplane spraying has promoted the search for new strategies of biological control. For most cases, these microbial agents are capable of self-sustained growth after an initial inoculation, with mid-and long-term pathogen suppression; and less biological impact compared with traditional chemical control (Quimby et al., 2002;Palaniyandi et al., 2013;Yuliar et al., 2015). To generate this effect, biocontrol agents can use one or more mechanisms including nutrient competition, niche exclusion (competitive exclusion), signal interference of quorum sensing (Quorum quenching), parasitism, diffusible or volatile secondary metabolite production with antimicrobial activity (antibiosis), and induced systemic resistance in plants (Bloemberg and Lugtenberg, 2001;Hibbing et al., 2010).
Among the microorganisms evaluated as potential biocontrol agents, members of the Streptomyces genus stand-out.
These gram-positive bacteria are characterized by mycelial growth similar to that of fungi. They are commonly isolated from terrestrial (soil, rhizosphere, and endosphere of plants) or marine environments, as free-living microorganisms or in association with other organisms (Coombs and Franco, 2003;Cao et al., 2004;Kinkel et al., 2012). Furthermore, they possess a diverse secondary metabolism that allows them to produce a great array of metabolites with antibacterial, antifungal, and antiviral activity. In addition, they are bioinsecticides, antitumoral and immune suppressors, among others (Omura et al., 2001;Hopwood, 2007;Kaur et al., 2014). In agriculture, numerous Streptomyces species have demonstrated capability of controlling diverse fungal phytopathogens of great agroindustrial impact (Samac and Kinkel, 2001;Taechowisan et al., 2003;Tian et al., 2004;Khamna et al., 2009;Zarandi et al., 2009;Gopalakrishnan et al., 2011;Li et al., 2012;Kanini et al., 2013). However, relatively few biological inoculants have been developed for crop use, mainly based on the poor association between the efficiency determined in the laboratory compared to the one observed in the greenhouse or the field (Bonaldi et al., 2015).
For the present study, three strains namely A20, 5.1 and 7.1, previously isolated by the Bioprocess and Bioprospecting Research Group from the National University of Colombia, were selected with the aim to verify its potential to act as biological control agents against C. gloeosporioides and M. ulei and to perform a preliminary characterization of the compound(s) associated with this activity. Suarez-Moreno et al. (2016, in press), taxonomically classified all three isolates within the Streptomyces genus, based on their biochemical profiles, colony macro-and microscopic characterization and sequencing of the rRNA 16S gene by Supplementary  Tables 1 and 2. Within the initial characterization of these three strains, isolated from symptomatic rice, carnation and yam plants, their high potentials for bacterial and fungal phytopathogen control were determined. For these reasons, this work screened all three isolates aiming to determine their potential to control rubber fungal pathogens, looking forward to develop biological inoculants for natural-rubber farmers, as an approach to integral control for SALB and foliar anthracnose.

M. ulei and C. gloeosporioides isolation and identification
M. ulei and C. gloeosporioides isolates were obtained from affected foliage with SALB and anthracnose symptoms respectively from Corpoica's clonal garden -La Libertad Section, located in Villavicencio in the Mavalle S.A. plantation (Department of Meta, Colombia). For M. ulei isolation, pure in vitro cultures were obtained in M4 growth media (Junqueira et al., 1984) and identified by conidial asexual morphology reported for Fusicladium heveae (Anamorphic form of M. ulei, currently known as Pseudocercospora ulei) (Schubert et al., 2003;Hora et al., 2014) . On the other hand, C. gloeosporioides isolation was performed by direct seeding on PDA of anthracnose symptomatic foliage that was previously treated with hypochlorite (5%) and ethanol (70%) (Agostini et al., 1992;Pinzón, 2014). Obtained monosporic cultures were morphologically characterized by macro-and microscopic description of the pathogen including aspect and color of the mycelium, form, size, segmentation and conidia sporulation according to previous reports in the literature (Gunnell and Gubler, 1992;Barnet and Hunter, 1998;Pérez et al., 2003). For both isolates, internal transcribed spacer (ITS) analysis was performed using ITS1/ITS4 universal primers (White et al., 1990). Obtained amplicons were sequenced in duplicate, assembled and analyzed by BLASTN against GenBank data base.

Streptomyces spp. strains
Streptomyces spp. A20, 5.1 and 7.1 strains were isolated from rice rhizosphere soil from La Pilar (Venadillo) and El Puente (Armero) farms in the Department of Tolima (Colombia) and were identified in previous works of the research group (Suarez-Moreno et al., 2016, in press).

Effect of Streptomyces spp. isolates on C. gloeosporioides mycelia growth
Streptomyces spp. A20, 5.1 and 7.1 strains were evaluated by dual culture plate assay to verify their ability to inhibit growth of C. gloeosporioides isolates C1, C2 and C3. Each Streptomyces strain was seeded on PDA in a straight line at 3 cm from the Petri dish periphery. 5 mm disc containing phytopathogen mycelia from each fungus (previously obtained from a 5 days-old solid PDA culture) was placed in the center of the petri dish. Media without Streptomyces was used as negative control. All media were incubated for eight days at 25°C and examined to verify inhibition areas between C. gloeosporioides and Streptomyces spp. A20, 5.1 or 7.1, respectively (Yuan and Crawford, 1995).

Production of diffusible and volatile compounds with antifungal activity
To evaluate whether the observed antifungal effect was due to diffusible or volatile compounds production, an agar well diffusion test and a volatile compound production assay was performed following CLSI 2011 guide, as well as recommendations suggested by Arrebola et al. (2010), respectively.
For agar well diffusion tests, 200 µl of a conidial suspension (10 5 conidia·mL -1 ) of C. gloeosporioides isolates C1, C2 or C3 was massively seeded in Petri dishes with 25 ml PDA. Wells of 7 mm in diameter were opened at a distance of 4 mm from the edges of the Petri dish. Each well was inoculated with 100 µl of A20, 5.1 or 7.1 liquid spent supernatants from liquid cultures obtained from each Streptomyces isolate, as described previously. 10 µl Clotrimazole (100 µg µl -1 ) and 100 µl M3.7 of sterile media were used as positive and negative controls, respectively (CLSI, 2011). Plates were incubated for 48 h at 25°C, and inhibition halo diameters were recorded in triplicates for each evaluated sample (Equation 1).
Assays were performed three times using three biological replicates for each Streptomyces culture.
Inhibition diameter = Total inhibition diameter -Well diameter (1) Villarraga et al. 143 To test if antifungal activity of the bacterial strains was due to the production of volatile compounds, a double-dish chamber assay was carried out. Briefly, bases of two Petri dishes containing 25 ml of PDA were used. For the first Petri dish, Streptomyces spp. isolates A20, 5.1 and 7.1 were seeded. In the second one, a 5 mm agar disc containing C. gloeosporioides C1, C2 or C3 mycelium was seeded. Both plates were then confronted and sealed with Parafilm ™ aiming to obtain a chamber with a shared atmosphere without direct contact between both microorganisms. As a negative control, one experiment was set without Streptomyces in the first Petri dish (Arrebola et al., 2010). After eight days of incubation at 25°C, radial growth of C1, C2 and C3 were measured and compared to the negative control. Inhibition percentage was calculated using Equation 2 as described by Taechowisan et al. (2012).
[ ] where GDU refers to the growth diameter in untreated control and GDT corresponds to the growth diameter in treatments.

Effect of Streptomyces A20 and 5.1 filtered extracts on C. gloeosporioides mycelial growth and conidial germination
To assess the effect of Streptomyces A20 and 5.1 filtered extracts on C. gloeosporioides mycelial growth, methods suggested by Anthony et al. (2004) were used. Briefly, Streptomyces A20 and 5.1 crude extracts were obtained by centrifuging at 6000 rpm for 10 min 500 ml from three independent liquid fermentation cultures of each strain in M3.7 medium. The obtained supernatant was filtered through 0.22 µm nitrocellulose membrane and subsequently used to supplement 25 ml of PDA. Increasing volumes of filtered spentsupernatants were used in order to obtain a medium with concentrations of 6, 4 or 2% (v/v) of extract per petri dish. Later, 7 mm discs from C1, C2 and C3 C. gloeosporioides isolates were placed in the center of each dish, and incubated at 25°C to evaluate the radial growth of each isolate every 24 h until day 17 postinoculation. C. gloeosporioides growth on each treatment was compared to a negative control seeded on PDA without extract, and the growth inhibition percentage was obtained 17 days after treatment from Equation 2 (Anthony et al., 2004;Taechowisan et al., 2012). Based on the results obtained from this assay, supernatant effect on conidia germination process of the three C. gloeosporioides isolates was further evaluated for strain Streptomyces 5.1. For this purpose, conidia germination counts were measured with treated and no-treated conidia by using lyophilized crude extracts from Streptomyces 5.1. To this end, 100 µl of a conidial suspension (1×10 5 conidia·mL -1 ) was supplemented with previously lyophilized 5.1 crude extract (obtained as described previously) and seeded onto PDA previously divided into 1 cm 2 squares. Four different concentrations of crude extract were evaluated (10, 25, 50 and 100 mg ml -1 ). Conidia germination count was performed after 10 h incubation at 25°C. Percentage of germination inhibition was calculated in relation to the total number of conidia and the present germinated conidia in the sample as defined by Equation 3 (Palaniyandi et al., 2011).
Streptomyces 5.1 antifungal activity evaluation against M. ulei To determine the effects of active compounds produced by Streptomyces 5.1 on M. ulei pathogenic isolates growth, two antifungal assays were performed evaluating the percentage of 144 Afr. J. Microbiol. Res. germinated M. ulei conidia in suspension (1.65 × 10 5 conidia ml -1 ) supplemented with four different concentrations of lyophilized Streptomyces 5.1 supernatant (10, 25, 50 and 100 mg ml -1 ). 300 µl of saline solution were added to M. ulei conidia suspension as negative control. Conidia germination was determined as described for C. gloeosporioides assays (Equation 3). Moreover, the effect of 5.1 extracts on M. ulei stroma formation was evaluated by seeding 100 µl of treated conidia suspension on PDA with 5.1 supernatants. Stroma formation count on solid media was performed after 15 day incubation at 25°C, reporting growth as CFU ml -1 (Rocha et al., 2011).

Preliminary characterization of Streptomyces 5.1 produced active metabolites
In order to isolate and characterize antifungal compounds produced by Streptomyces 5.1, three independent liquid fermentations were carried out in 1 L of M3.7 medium. Culture media was inoculated and grown at 30°C, under constant agitation at 150 rpm. Each culture was then centrifuged for 10 min at 5000 rpm, and the obtained supernatants were filter sterilized by a 0.22 µm membrane. Spent supernatants were analyzed by sequential fractionation with dichloromethane and butanol (ratio solventsupernatant 2:1 and 1:1 for each solvent, respectively) in a continuous liquid-liquid extraction system. The aqueous and organic extracts obtained were separated by decantation and dried through lyophilization or in a vacuum rotary evaporator (Labconco® Kansas City, MO USA), respectively. All extracts were subsequently evaluated for C. gloeosporioides (C3) antifungal activity by using the agar well diffusion method described previously. Aqueous extracts were dissolved in distilled water to evaluate concentrations 10, 20, 50 and 100 mg ml -1 . Likewise, obtained organic extracts at 0.5, 0.8 and 1mg ml -1 were dissolved in DMSO (8% v/v). One liter of M3.7 sterile media was subjected to the same extraction process and evaluated under the same conditions as 5.1 samples in order to be used as negative control. As mentioned, this methodology was performed with three 5.1 biological replicates and the negative control, respectively.
Butanolic extracts obtained from isolate 5.1 which maintained antifungal activity against C. gloeosporioides were analyzed by Liquid chromatography-mass spectrometry (LC/MS) and Matrixassisted Laser Desorption and Ionization Time of Flight mass spectrometry (MALDI-TOF). To this end, 2 mg of butanolic extract previously dissolved in ethanol (90% v/v) was injected into the HPLC VWR-LaChrom coupled to an Amazon × mass spectrometry (Bruker Daltonics, Bremen Germany) at the Universidad Industrial de Santander (Santander, Colombia). Chromatography was run in an × Terra® RP18 5 µm (4.6 × 250 mm) column with a nitrile acetate gradient with 0.075% formic acid-H2O as mobile phase. Data were collected and analyzed using the Compass Data analysis (Bruker Daltonics ® ) and Mzmine 2.14.2 ® programs.
For MALDI TOF, an Autoflex (Bruker Daltonics, Bremen, Germany) mass spectrometry with a positive ion reflection mode was used. Identification and spec allocation were carried out automatically using a Flexanalysis software version 2.2 (Bruker Daltonics) and Mzmine. All m/z (mass to charge ratio) obtained by both methodologies were compared to those reported in the Streptome DB database (Lucas et al., 2013) and Antimarin Database in order to find compounds previously reported with the m/z ratios found in this study (Blunt et al., 2007).

Statistical analysis
All assays of the present study (either qualitative or quantitative) were performed in triplicate. To test for normal distributions for the quantitative data a Shaphiro-Wilk test was carried-out. Additionally, it was evaluated if all replicas for each assay presented similar tendencies. Tukey test for each data group was performed to find atypical data. Analysis of variance (ANOVA) was used to evaluate the effect of different treatments on phytopathogen fungal growth. Significant differences among means were compared with a Tukey post-hoc test (p = 0.05). GraphPad Prism © (GraphPad Software, Inc. © 2012) was used for all analyses.

RESULTS
Streptomyces A20, 5.1 and 7.1 were capable of reducing different C. gloeosporioides isolates mycelial growth The potential of Streptomyces A20, 5.1 and 7.1 to inhibit the growth of C. gloeosporioides isolates was initially determined by a confrontation dual culture test. In this assay, it was evidenced that Streptomyces A20 and 5.1 significantly reduced C. gloeosporioides mycelial growth ( Figure 1). Furthermore, agar well diffusion assays suggested that A20 and 5.1 antifungal activity could be generated by a diffusible antifungal metabolite production ( Figure 1B). It was determined that Streptomyces 5.1, presented significant differences compared with strain Villarraga et al. 145  Data shown represents mean in mm ± SD of three biological replicates per treatment. Different letters in the same column represent significant differences among treatments (p < 0.05). A20 in regards to C. gloeosporioides mycelial growth inhibition. However, it was not comparable with the positive control (Table 1). Otherwise, strain 7.1 did not demonstrate any antifungal activity mediated by diffusible compounds. Thus to determine if Streptomyces 7.1 presented another means of control, it was decided to evaluate active volatile compounds against the three C. gloeosporioides isolates. Through the double-dish chamber assay, it was identified that only 7.1 strain was capable to reduce C. gloeosporioides mycelial growth through the production of volatile compounds (Figure 2), generating growth inhibition percentages of 71.62, 63.28 and 52.27% against C1, C2 and C3, respectively. In contrast, A20 was only capable of inhibiting the growth of isolate C1 with an inhibition percentage of 49.24%. These results indicated that Streptomyces A20 and 5.1 characteristic antifungal activity were mediated by the production of antifungal diffusible compounds, whereas for 7.1 such effect was the result of volatile nature compounds.

Streptomyces 5.1 diffusible compounds had a fungistatic effect on C. gloeosporioides
Considering the aforementioned results, it was desirable to confirm the efficiency and stability of the diffusible compounds produced by strain A20 and 5.1 based on its ability to reduce C. gloeosporioides mycelial growth through time. It was observed that diffusible compounds produced by 5.1, retarded mycelial growth for all three C. gloeosporioides isolates generating a maximum growth of 61.11, 51.45 and 38.71% for isolates C1, C2 and C3, respectively at the sixth day of incubation (6% treatment) ( Figure 3A to C). An inverse correlation was observed between the pathogen growth and the concentration of the compounds produced by 5.1 against C. gloeosporioides C3, being 6% concentration of the most active between the treatments. Interestingly, there were no statistical differences within 4 and 6% treatments in the antagonistic assays against C1 and C2, which could be possibly associated with susceptibility differences 146 Afr. J. Microbiol. Res. Figure 3. Effect of 5.1 and A20 produced compounds on Colletotrichum gloeosporioides C1, C2 and C3 mycelial growth. In the figure, sections A-C corresponds to 5.1 treatments and D-F to A20 treatments. A and D are the assays against to C1, B and E against to C2 and C and F against to C3, respectively. Data presented as mean (mm) ± Standard Deviation (SD) of six biological replicas for each treatment (2%v/v --, 4%v/v --, 6%v/v --, Negative control --).
among the three isolates of C. gloeosporioides to the secondary metabolites produced by 5.1. After ten days of incubation, the inhibitory effect for all 5.1 treatments was reduced to such extent that at the end of the evaluation (Day 17), only the 6% treatment presented inhibition percentages greater than 10% against the three C. gloeosporioides isolates (16.48, 14.61 and 13.7% for C1, C2 and C3 respectively) ( Table   2 and Figure 3A to C). These results suggest that the effect generated by Streptomyces 5.1 active compounds is fungistatic and its stability after a single application has a maximum of 10 days (last measurement where the inhibitory percentage was above 30%).
On the other hand, treatments performed with three Villarraga et al. 147 concentrations of the lyophilized crude extract from Streptomyces A20 showed no deleterious effect on the growth of the three Colletotrichum isolates at the end of the incubation time (Table 2 and Figure 3D to F). Suggesting that the active metabolite produced by A20 either is unstable and was degraded during the test or, it is not produced under the fermentation conditions evaluated at the concentration sufficient to inhibit phytopathogen growth. Due to these results, this isolate was excluded for the following analysis.
Treatment of C. gloeosporioides C1, C2 and C3 conidia with different concentrations of 5.1 lyophilized crude extracts, allowed to determine that the antifungal compounds produced by this strain are able to inhibit the phytopathogen's conidia germination ( Figure 4A). As it was observed for mycelial development, the inhibitory effect was inversely proportional to the concentration of the lyophilized extract, being 100 mg ml -1 of the most efficient treatment with an inhibition percentage of 79.5%.
In these experiments, light microscopy evidenced that conidia presented the characteristic swelling of the first stage of germination, in particular for 50 and 100 mg·ml -1 treatments, however, in these treatments conidia were unable to form the germ tube ( Figure 4B and C). This result suggests that active compound(s) produced by Streptomyces 5.1 possibly interfere with the cell wall membrane formation in C. gloeosporioides, which could impair the formation of germ tubes. Nevertheless, this 148 Afr. J. Microbiol. Res. hypothesis must be confirmed through cytochemical and microscopic complementary studies. In addition, it had to be established if treated conidia remained viable (fungistatic effect) or were incapable of forming new vegetative mycelium, once they lost contact with the antifungal compound (fungicidal effect).

Streptomyces 5.1 active compound production against M. ulei
To determine Streptomyces 5.1 biocontrolling potential against the phytopathogenic fungus M. ulei, two approaches were carried-out. First, compound capacity to inhibit M. ulei conidia germination process present in lyophilized supernatant from 5.1 cultures was determined. As can be observed from Figure 5, M. ulei conidia suspension treatment at different 5.1 lyophilized supernatant concentrations significantly reduced germination percentage at 25, 50 and 100 mg·ml -1 . On the contrary, treatment with 10 mg·ml -1 did not present any significant difference compared to negative control ( Figure 5A).
Furthermore, it was evidenced that for all 5.1 supernatant treatments (25, 50 or 100 mg·ml -1 ) conidial germ tube formation was not completely inhibited. Nonetheless, germ tubes did not present typical M. ulei characteristics, where germ tube formation develops at Villarraga et al. 149 Figure 6. Bioassay guided fractionation of Streptomyces 5.1 active compounds. A. Flow diagram with solvents of different polarity. B. Antifungal assay against C. gloeosporioides C3 for fractions obtained from strain 5.1. Numbers in the figure of the right correspond to the ones presented in Figure 6A. As a solvent control DMSO (8% v/v) was used.
both poles of the conidia ( Figure 5C). For the treated conidia, germination occurred only at one pole of the conidia, and it was not possible to observe the typical division of the germ tube ( Figure 5D).
To determine if the M. ulei deficient conidia (classified as germinated conidia during germination count) had a subsequent effect on stroma formation (M. ulei vegetative growing stage), 100 µl of conidia suspension treated with phytopathogen were seeded in PDA to evaluate mycelia growth of the fungus after being exposed to compounds produced by Streptomyces 5.1. It was observed that all treatments with different supernatant concentrations lead to a significant reduction in the number of formed stroma ( Figure 5B). This result suggested that conidia with deformed germination were incapable of generating a vegetative mycelium, thus reducing the number of stromas formed per dish. It can therefore be inferred that compounds produced by strain 5.1 have a fungicidal activity against M. ulei.

Streptomyces 5.1 produces different compounds with medium polarity of non-ionic nature
To preliminarily determine the nature of active compounds produced by Streptomyces 5.1, bioguided fractionation of Streptomyces 5.1 supernatant was performed with solvents of different polarity ( Figure 6A), followed by antifungal activity evaluation of all obtained fractions against C. gloeosporioides isolate C3. Our results indicated that, antifungal compounds produced by strain 5.1 were retained in the aqueous phase of the first fractioning when dichloromethane was used as a solvent. Subsequent treatment of this aqueous phase using butanol as a solvent favored organic phase (butanolic) compound extraction ( Figure 6B). The solubility profile identified for Streptomyces 5.1 antifungal compounds suggested that active compounds are characterized by being non-ionic, with medium polarity.
Analysis of Streptomyces 5.1 butanolic fractions through LC/MS and MALDI TOF evidenced the presence of six peaks with m/z values of 252,25, 309,17, 427,24, 610,06, 610,39 and 778,28. These peaks were not observed in blank media (Figure 7). Therefore, it was assumed that they were produced by Streptomyces 5.1. Out of these peaks, the major peak 610,39 was recognized by both techniques, thus becoming the most interesting for structure elucidation. Search of compounds produced by different Streptomyces spp. presenting an antifungal activity or any antimicrobial activity, in addition to similar m/z values to those reported in this study were not found, which suggests that these compounds have not been described yet. Therefore, it is necessary to continue their purification and structural elucidation by means of complementary techniques to those utilized in the present study.

DISCUSSION
To date, Integrated system crop management of natural rubber (H. brasiliensis), which is a crop of agricultural importance, constitutes a main focus in its production chain. Within these systems, management with phytopathogens has a major role given its direct effect on Streptomyces genus is recognized for its ample secondary metabolism, which allows them to produce diverse bioactive compounds of interest for biological control including antibiotics, lytic enzymes (chitinases and glucanases, proteases, among others) and/or siderophores (Qin et al., 2011). An initial evaluation of three Streptomyces strains againts three C. gloeosporioides isolates evidenced their potential to inhibit mycelial growth of the phytopathogenic agent ( Figure 1 and Table 1). Further assays permitted to establish that Streptomyces 7.1 was able to reduce C. gloeosporioides mycelial growth by producing volatile compounds with an inhibition percentage ranging from 52 to 71% (Figure 2). Antifungal activity by Streptomyces' genus volatile compound production for phytopathogen control has been previously evaluated by Wan et al. (2008), who evidenced Streptomyces platensis F1 capacity to inhibit Rhizoctonia solani, Sclerotinia sclerotiorum and Botrytis cinerea growth, and to reduce the incidence and severity of diseases caused by these pathogens in foliage tissues of rice plants, turnips, and strawberry fruits, respectively, under controlled atmosphere conditions (Wan et al., 2008).
Likewise, Li et al. (2010Li et al. ( , 2012 developed similar studies where Streptomyces globisporus JK-1 capability to inhibit B. cinerea and Penicillium italicum growth in tomato plants (Lycopersicon esculentum) and Shatang mandarin fruit (Citrus microcarpa) was verified, with promising results for growth control of these pathogens through in vivo assays under shared atmosphere conditions (Li et al., 2010(Li et al., , 2012. Nonetheless, use of these types of metabolites is recommended for controlling diseases in environments that favour the presence of a microatmospheres, where a higher concentration of a volatile compound is achieved, as it is the case for control of pathogens in soil, management of foliage diseases under controlled conditions in greenhouse and use of post-crop fruit storage containers (Wan et al., 2008;Li et al., 2012). For these reason, application of microorganisms able to produce volatile compounds for management of foliage phytopathogens in H. brasiliensis under field cultivation conditions would not be recommended; due to possible reduction in metabolite efficiency by compound dilution effect, since there would not be a controlled atmosphere environment. Therefore, strain 7.1 was removed for posterior assays in the present study, even though its potential as a biocontrol agent should be studied in future research aimed for controlling post-crop disease control.
On the other hand, Streptomyces A20 and 5.1 were capable of inhibiting C. gloeosporioides growth by Villarraga et al. 151 producing extracellular diffusible metabolites, as suggested by the antifungal activity retained in liquid culture filtered supernatants from both microorganisms ( Figure 1 and Table 1). Prapagdee et al. (2008) reported similar results to those obtained in our study. They described C. gloeosporioides 50% radial growth inhibition using filtered extract of Streptomyces hygroscopicus during its exponential growth. Furthremore, Shahbazi et al. (2014) demonstrated antifungal activity of Streptomyces strains P8 and P42 isolated from chili pepper (Capsicum annuum L. Kulai) rhizosphere soils evaluated against C. acutatum, C. capsici and C. gloeosporioides phytopathogens. Antifungal activity was due to compounds present in supernatants from liquid culture of Streptomyces strains (Shahbazi et al., 2014). In both studies, lysis of the phytopathogen's hyphae cell wall by chitinase was evidenced.
Results obtained in the present study indicate that strain 5.1 had the highest antifungal activity against C. gloeosporioides and M. ulei, which was statistically significant compared to results obtained with Streptomyces A20 (Table 1), being the first report (as far as we know) that describe the antifungal activity of a Streptomyces isolate against M. ulei. Furthermore, results obtained from the bioassay guided fractionation suggested that enzymatic lytic activity was not produced, since antifungal activity remained without alterations after supernatant fractionation with two different organic solvents of different polarity (dichloromethane and butanol). Since, these two solvents have the capability to degrade enzymes, it can be suggested that Streptomyces antifungal compounds can be classified as a nonenzymatic metabolite with antifungal activity. Additionally, analysis of 5.1 filtered supernatants suggested that active compounds of this microorganism suppress C. gloeosporioides and M. ulei conidia germination process, as well as stroma formation for the later one, with a percentage inhibition of 79.5, 40.88 and 95.56%.
Light microscopy analysis, demonstrated that contact between a Streptomyces extract (25 to 100 mg·ml -1 ) and C. gloeosporioides or M. ulei conidia, reduced conidia development and germ tube elongation for each pathogen, disrupting mycelia development. This type of activity could interrupt the penetration process of both phytopathogenic fungus in rubber leaves inhibiting or delaying the infection in the plants, as it has been previously reported for others Streptomyces species evaluated for biological control of C. gloeosporioides, Fusarium oxysporum f. sp. lycopersici, Verticillium alboatrum and Alternaria solani. An inversely proportional association was observed between conidia percentage germination and extract concentration, containing the active compound evaluated (El-Abyad et al., 1993;Palaniyandi et al., 2011), as it was observed in the present study (Figures 4 and 6).
However, mechanism of action against the phytopathogens for antifungal compounds isolated from 152 Afr. J. Microbiol. Res.
Streptomyces 5.1 is still un-known. To elucidate such activity, first of all it is necessary to establish the type of metabolite produced. The present study evidenced Streptomyces 5.1 produced at least two non-ionic compounds. Analysis through LC/MS and MALDI TOF of 5.1 fermentation butanol extractions revealed six unique strain peaks; with the highest peak at a molecular mass of 610.2 m/z. Data obtained from these experiments did not correlate with compounds reported in StreptomeDB or Antimarin Database, therefore, these compounds could be assumed as new. However, it is necessary to perform a discrimination process among the six peaks to determine which of them is responsible for the observed antifungal activity. Moreover, complementary assays must be carried out to reveal the compounds and verify the new molecule hypothesis. Last, despite our results it is necessary to establish additional in vitro and in vivo experiments focused on (i) evaluating Streptomyces 5.1 and C. gloeosporioides and M. ulei population behavioral dynamics in association with rubber plants (H. brasiliensis), taking into account the plant's defense response to Streptomyces 5.1. inoculation, (ii) comparisons between incidence and severity of the disease treated with 5.1, and (iii) determining its efficiency with respect to the traditional chemical control.
This study evidenced that Streptomyces 5.1 strain competence to produce extracellular metabolites with antifungal activity to inhibit mycelial growth in addition to impede C. gloeosporioides and M. ulei germination process under in vitro conditions. Obtained results demonstrated Streptomyces 5.1 potential to be utilized as a biological control agent destined for rubber plant (H. brasiliensis) protection against foliar anthracnose and SALB. To the best of our knowledge this is the first report establishing Streptomyces genus for important phytopathogen fungi control of the natural rubber.