In vitro antimicrobial activity of methanolic extracts from cutaneous secretions of Amazonian amphibians against phytopathogens of agricultural interest

The biochemical defense mechanisms of amphibians involve cutaneous secretions of bioactive molecules with antimicrobial activity. This study evaluated the in vitro activity of methanolic extracts from cutaneous secretions of two amphibian species of the Bufonidae family, Rhaebo guttatus and Rhinella marina, in the control of the phytopathogens Fusarium udum, Fusarium solani, Colletotrichum truncatum, Aspergillus flavus, Rhizoctonia solani, Macrophomina phaseolina, and Calonectria pseudometrosideri. The R. guttatus extract decreased the mycelial growth of F. udum, F. solani, A. flavus, and M. phaseolina at some tested concentrations. The R. marina extract decreased the mycelial growth of C. truncatum at the concentration of 0.5 mg mL-1, and inhibited the mycelial growth of A. flavus at concentrations of 0.1 and 0.5 mg mL-1, which was similar to the inhibition by the positive control. The R. marina extract also decreased the microsclerotia production by R. solani at concentrations of 0.2 and 0.3 mg mL-1. In addition, the extracts inhibited conidial sporulation and germination at varying degrees. The inhibition of appressoria formation in C. truncatum by the R. guttatus and R. marina extracts was 85–99% and 63–100%, respectively. Our results demonstrated that treatment with extracts from R. guttatus and R. marina cutaneous secretions showed antifungal activity against the studied phytopathogens.


INTRODUCTION
The biodiversity of Amazonia is a potential source of chemical compounds of biological and industrial interest (Skirycz et al. 2016). Natural chemotypes have biological activities which result in combinatorial chemistry which, in addition to generating stereo chemically complex structures with many functional groups, also produce compounds which specifically interact with biological targets (Cragg and Newman 2013;Loiseleur 2017).
Natural products have a long history as a source of and inspiration for novel agrochemicals (Sparks et al. 2017). For example, strobilurins are synthetic pyrethroids which are reconfigurations of the core structure isolated from the mycelium of Strobilurus tenacellus (Pers.) Singer (Loiseleur 2017;Sparks et al. 2017). Certain phytochemicals act on various types of diseases and can be applied in the food and agro-industry (Dileep et al. 2013). Compounds obtained from natural products have shown direct fungitoxic action by inhibiting spore germination and mycelial growth of phytopathogens (Stangarlin et al. 2011;Oliveira et al. 2014). In this context, the prospection for new active compounds is relevant to overcome pathogen resistance to existing fungicides, which results from the indiscriminate and excessive use of these substances, leading to a feedback cycle of ever higher concentrations of antifungals and a consequent increase in toxic residues in food products (Barros-Velazquez 2016).
Some animals produce structurally unique secondary metabolites which can be useful as new chemical templates for drug discovery (Cunha-Filho et al. 2010;Banfi et al. 2016). An example is the development of Captopril, a medication derived from small peptides isolated from the venom of a South American snake (Bothrops jararaca Wied-Neuwied), known to potentiate the action of bradykinin (Opie and Kowolik 1995;Harvey 2014).
In the present study, we evaluated the in vitro activity of the methanolic extract of the cutaneous secretion of two amphibian species of the Bufonidae family, Rhaebo guttatus (Schneider) and Rhinella marina (Linnaeus), in the control of seven phytopathogenic fungi.

Cutaneous secretion extracts and phytopathogen isolates
The cutaneous secretion was obtained from six individuals of R. guttatus and nine of R. marina collected in  in the state of Mato Grosso, Brazil (collection license no. 30034-1 issued to D.J. Rodrigues by Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais -IBAMA/SISBIO). The procedure for collecting the cutaneous secretion of the animals is non invasive and does not cause harm to ther animals. All sampled animals were returned to nature after the procedure. Voucher specimens were deposited in the Biologic Collection of Meridional Amazonia -ABAM/UFMT/Sinop (R. guttatus -ABAM-H 1538 and R. marina -ABAM-H 1262) in the context of another study (Oliveira et al. 2019). The secretion was extruded by manual compression of the parotoid macrogland. The collected material was dried in silica gel, extracted three times in methanol, and evaporated on a rotary evaporator. The methanolic extracts were weighed, diluted in sterile water, and filtered through a Millipore membrane (0.22 µm) (Ferreira et al. 2013). These procedures were carried out in the Laboratory of Chemical Research (LiPeQ) of UFMT/Sinop.

Antimicrobial activity assays
The antifungal activity of the methanolic extracts was evaluated by determining the mycelial growth using the disc-diffusion test in PDA medium (Bauer et al. 1966). Four sterilized filter-paper discs with a 5-mm diameter were soaked in the treatment solution and arranged equidistant on a 9-cm petri dish, and a 5-mm mycelial disc was deposited on the center of each petri dish. Five replicates were used for each treatment and each control. Each petri dish constituted one replicate. The petri dishes were sealed and incubated in a BOD incubator at 25 °C in the dark. Evaluations were performed at the same time every day by measuring colony diameter along two orthogonal axes (average of two diametrically opposed measurements) starting 24 h after dish preparation.
The mycelial growth (MG) and the mycelial growth rate index (MGRI) were evaluated until one of the treatments reached two thirds of the surface of the petri dish in all replicates (Costa and Carvalho et al. 2013;Andrade et al. 2018). The mycelial growth rate index was subsequently calculated from the mycelial growth data: where MGRI = mycelial growth rate index, d = current average diameter of the colony; da = average diameter of the colony from the previous day and, n = number of days after inoculation (Dias et al. 2005;Maia et al. 2011).
The mycelial growth inhibition percentage (MGI%) was determined according to the methodology proposed by Edgington et al. (1971).
The effects of the extracts on the sporulation of F. udum, F. solani, C. truncatum, A. flavus, and C. pseudometrosideri were analyzed after completing the mycelial growth assay (covering two thirds of the petri dish). After this period, 10 mL of sterile water were added to each petri dish, and then each fungal colony was scraped. The resulting suspension was filtered through a gauze and the number of conidia was counted in a Neubauer chamber under an optical microscope. The obtained data were converted into sporulation inhibition percentage (SI%) for each treatment relative to the negative control (Fernandes et al. 2015).
The effect of the extracts on the production of R. solani microsclerotia was determined after completing the mycelial growth assay (21 days of incubation of the mycelial disc). Each petri dish was divided into four equal quadrants starting at the central mycelial disc. A scale score was used for the visual quantification of the number of microsclerotia as follows: ++++ indicates a large number, +++ a moderate number, ++ a small number, + a very small number, and − the absence of microsclerotia. The sum of the scores was converted to a numerical scale from 0 to 4 for statistical analysis.
The effect of the extracts on the conidial germination of F. udum, C. truncatum, and C. pseudometrosideri was evaluated according to the methodology proposed by Stangarlin et al. (1999). Colonies of the fungal pathogens cultured in PDA medium for approximately 10 days were used to obtain the conidial suspension. Sterile water (10 mL) was added to the fungal cultures, and the cultures were scraped with a Drigalski loop and then filtered through a sterile gauze. After filtration, the suspension was adjusted to the following concentrations using a Neubauer chamber: 1 × 10 4 , 1 × 10 5 , and 5 × 10 4 conidia mL -¹ for C. pseudometrosideri, C. truncatum, and F. udum, respectively. Conidial germination was evaluated under an optical microscope (400x magnification) using 100 conidia per repetition. Germination was deemed to have occurred when the germ tube emerged.
Sterile 96-well microplates were used to evaluate conidial germination (Almeida et al. 2009). In this study, an aliquot of 40 µL of the suspension of the evaluated phytopathogens was deposited in each well. A 40-µL aliquot of the treatments was subsequently added to each well. For the control, 40 µL sterile water were added instead of the treatments. The conidial germination assays were performed with eight replications for each treatment. The plates were incubated at 25 °C for 3, 6, and 24 h for C. pseudometrosideri, C. truncatum, and F. udum, respectively. After the incubation period, germination was interrupted with 20 µL of blue cotton and lactophenol. This period was determined with preliminary germination tests of the conidia of each phytopathogen in sterile water.
A Drigalski-loop scattering method was used to evaluate the conidial germination of A. flavus. This method was used because the conidia of A. flavus did not germinate when using microplates in preliminary tests. Petri dishes of 9-cm diameter were used, where each plate was considered one replicate, with eight replicates per treatment level (5) or control (2) and two treatments (R. guttatus and R. marina extracts). The plates were filled with agar-water medium and 100 µL of the test solution was distributed on the medium after solidification using a sterile Drigalski loop. After 2 h, 100 µL of conidial suspension (2.6 × 10 6 conidia mL -¹) of A. flavus were added and spread using a sterile Drigalski loop. Plates were incubated at 25 °C in the dark and evaluated after 48 h, when germination was interrupted with 20 µL of blue cotton and lactophenol. The VOL. 51(2) 2021: 145 -155 ACTA AMAZONICA results were expressed as conidial germination inhibition percentage (CGI%) for each treatment level compared to the negative control (Fernandes et al. 2015). The germination of F. solani was not evaluated because there was no conidial production for this isolate.
For C. truncatum, we evaluated the emergence of the appressoria in conidial germination assays (Bonaldo et al. 2004) and expressed this variable as the appressorial formation inhibition percentage (AFI%) relative to the negative control. A total of 100 randomly selected germinated conidia per repetition were examined for number of appressoria, totaling 800 conidia per treatment. Post-germination appressoria development was averaged using an optical microscope.

Statistical analysis
The assays were conducted in a completely randomized design and the data were analyzed using the SISVAR version 4.3 program (Ferreira 2008). The results of each extract were subjected to analysis of variance (ANOVA) using the F-test, and the means were compared by the Scott-Knott test (p < 0.05). The original SI% data did not meet the requirements of normality and homogeneity and were therefore transformed to (x + 1) 0,5 .

RESULTS
The R. guttatus extract significantly decreased MG and MGRI of F. udum, F. solani, A. flavus and M. phaseolina (Table 1; Figure 1). MG and MGRI of F. udum were significantly lower for the 0.5-mg mL -¹ concentration relative to the negative control. MG was lower in the concentration of 0.5 mg mL -¹ (5.21 cm) in relation to the negative control (6.60 cm). This concentration also showed reduced MGRI (0.74 cm day -1 ) in relation to the negative control (0.94 cm day -1 ). MG and MGRI of F. solani were significantly lower at all R. guttatusextract concentrations compared with the negative control, with MG values of 5.71, 5.96, 5.48, 5.62, and 5.84 cm for 0.1, 0.2, 0.3, 0.4 and 0.5 mg mL -¹, respectively (negative control =7.18 cm), and corresponding MGRI values of 0.71, 0.75, 0.69, 0.70, and 0.73 cm day -1 , respectively (negative control average = 0.90 cm day -1 ). This suggests that the R. guttatus extract has properties which reduce in vitro development of F. solani at the tested concentrations.
MG and MGRI of A. flavus were significantly lower than the negative control at 0.1 and 0.5 mg mL -¹ (MG = 4.20 and 4.13 cm, respectively, and negative control =6.78 cm). The MGRI was lower at 0.1 mg mL -¹ (0.47 cm) and 0.5 mg mL -¹ (0.46 cm) relative to the negative control (0.75 cm). MG of M. phaseolina was significantly lower at 0.1, 0.2 and 0.3 mg mL -¹, with mean values of 6.88, 7.53, and 7.76 cm, respectively, relative to the negative control (7.94 cm). MG was lower at 0.1 mg mL -¹, differing from the other concentrations. The MGRI of M. phaseolina was significantly lower than the negative control (2.65 cm) at 0.1 mg mL -¹ (2.29 cm).
MG and MGRI of C. pseudometrosideri indicated colony development promotion at 0.1 and 0.2 mg mL -¹ compared   with the negative control. MG mean values were 6.64 and 7.36 cm at 0.1 and 0.2 mg mL -¹, respectively, while the negative control showed 5.88 cm. MGRI was higher at 0.1 and 0.2 mg mL -¹, with mean values of 0.47 and 0.53 cm day -1 , respectively, relative to the negative control (0.4 cm day -1 ). These data suggest that lower concentrations showed no beneficial effects in the mycelial growth reduction of C. pseudometrosideri. The R. marina extract did not decrease the mycelial growth relative to the negative control (  Figure 2). The MGRI of C. truncatum was reduced at 0.5 mg mL -¹ (0.47 cm day -1 ) when compared with the negative control (0.54 cm day -1 ), and did not differ significantly from the positive control (0.50 cm day -1 ).
The R. guttatus extract inhibited the MGI% of F. udum, A. flavus and C. pseudometrosideri ( Table 2). MGI% of F. udum increased by 21.3% at 0.5 mg mL -¹ and also increased significantly in A. flavus at 0.1 mg mL -¹ (38.0%) and 0.5 mg mL -¹ (40.6%), thus showing fungitoxic effect of the extract at these concentrations. The MGI% of C. pseudometrosideri at 0.3, 0.4 and 0.5 mg mL -¹ did not differ significantly from the positive control. The R. marina extract increased the MGI% of C. truncatum at 0.5 mg mL -¹ to a similar level to the positive control ( Table 2).
The R. guttatus extract significantly affected the sporulation inhibition (SI%) in A. flavus at 0.1 and 0.5 mg mL -¹ (93.1 and 83.6%, respectively), showing similar toxic effect to the positive control (95.9%) ( Table 3). The SI% of A. flavus for the R. marina extract at 0.4 mg mL -¹ (54.9%) showed similar results to the positive control (85.7%).
Microsclerotia production in R. solani was significantly reduced at 0.2 and 0.3 mg mL -¹ of the R. marina extract when compared with the negative and positive control ( Table 4).

DISCUSSION
The reduction in MG, MGRI, MGI% and CGI% by the R. guttatus extract at 0.5 mg mL -¹ suggests antifungal activity against F. udum. The R. guttatus extract also had a fungitoxic effect on the development of F. solani, as all concentrations inhibited MG, even at the lowest concentration.
There is no information in the literature on the fungitoxic effect of secondary metabolites in extracts originating from animals against Fusarium Link ex Grey, 1821. Antifungal activity on MG of F. solani had been reported for the crude aqueous extracts of the floral buds of clove ( The in vitro response of the MG, MGRI and MGI% of A. flavus to the R. guttatus extract is of particular importance as this fungus is of major concern in agriculture. It produces highly carcinogenic toxins called aflatoxins in the seeds of a number of crops both before and after harvest which are a health hazard to animals (Adeyeye 2016), causing negative impact on the nutritional value of plants and considerable yield losses (Klich 2007;Htoon et al. 2019). The R. guttatus extract showed similar MG inhibition of A. flavus as the positive control agent in all tests, which makes it a candidate for future prospection against this phytopathogen.
The R. guttatus extract presented fungitoxic effect on M. phaseolina only in the lower concentrations. The action mechanisms involved in antifungal activities in phytopathogens by bioprospecting secondary metabolites of animals are still unknown. However, a higher inhibition rate at a lower concentration (50 µg mL -¹) was observed using crude aqueous extracts of ginger (Zinziber officinalis Roscoe) on banana fibers infected with Helminthosporium sp. Link as compared to higher concentrations (100, 200, and 400 µg mL -¹) (Rodrigues et al. 2006).
The antifungal effect of lower concentrations of R. guttatus extract on M. phaseolina, as well as its effect on the promotion of MG in C. pseudometrosideri, may be related to the interaction among chemical constituents of the extract. An extract can present compounds that stimulate pathogen growth (Venturoso et al. 2011). Thus, the relative amount of stimulating compounds and their half-lives, as well as the interaction between them, determines the action on the phytopathogen.
The R. marina extract reduced the MGRI of C. truncatum. This reduction is promising because the fungicide used as positive control was not very effective in controlling this pathogen. This response may be related to constant applications of the same fungicide over time. Constantly exposing the pathogen population to high rates of a fungicide may add undue selection pressure on the population, which in Values are expressed as the mean ± standard deviation. Means followed by the same letter in each column did not differ significantly by the Scott-Knott test (p < 0.05).
turn may lead to resistance or tolerance (Deising et al. 2008). Thus, a change in the pathogen at this point can render the fungicide less effective or ineffective. The sporulation inhibition of A. flavus by both extracts may be an important result, as this pathogen produces large quantities of asexual spores (conidia) which promote fungal dissemination (Mah and Yu 2006). Pimentel et al. (2010) also observed a significant reduction in sporulation of A. flavus with the essential oil from fresh leaves of Tanaecium nocturnum (Barb. Rodr.) Bureau & K.Schum using the contact and fumigation techniques. Likewise, the microsclerotia production inhibition observed in the treatment with the R. marina extract may be relevant, because these resistance structures ensure pathogen survival in adverse conditions, even in the absence of the host and in unfavorable climate conditions (Ritchie et al. 2013). Microsclerotia can remain in the soil and in organic matter after plant infection and serve as inocula for subsequent cultures (Chawla et al. 2012), with potential for more severe and frequent epidemics (Ma et al. 2015).
The inhibition of appressoria formation by both R. guttatus and R. marina extracts is also relevant, because pathogens with these structures can directly penetrate the intact surface of the host and remain at this point of contact, dissolve the tissue, and form a small hole in the tissue (Bergamin Filho et al. 1995). Consequently, inhibiting the emergence of this structure can contribute to controlling the disease caused by C. truncatum because its appressoria act as specialized infection structures formed from a hypha or germ tube and are important for fungal penetration through the leaf surface of the host (Khan and Hsiang 2003).
The difference in the antifungal activity of the R. guttatus and R. marina extracts observed in our study could be attributed to factors such as the existence of many receptors of the extract which select phytochemicals with different affinities in the receptors of the phytopathogens (Ahuja et al. 2012;Fesel and Zucca 2016), or the insufficient amount of active compound in the test concentrations . Also, the active principle may be present in the test concentrations, but other constituents are exerting antagonistic effects on the bioactive compounds (Caesar and Cech 2019). Although the bufadienolides are the main active components in the cutaneous secretion of R. marina and R. guttatus, there are interspecific differences in the number and type of constituents (Ferreira et al. 2013). Four bufadienolides are found in the R. marina cutaneous secretion (telocinobufagin, marinobufagin, bufalin and resibufogenin) and one in the secretion of R. guttatus (marinobufagin) (Banfi et al. 2016).
We used crude extracts containing several active compounds which have similar or synergistic activities that inhibit or promote pathogen growth. The advantage of using such extracts as antimicrobial agents is the lower risk of microbial resistance, because their high complexity limits microbial adaptability (Daferera et al. 2003). A crude (untreated) extract from any source of natural products typically contains novel, structurally diverse chemical compounds (Lahlou 2013), as well as active, partially active, or many inactive compounds (Cragg and Newman 2013). Therefore, it is necessary to study the application of these extracts through isolation processes, to evaluate more precisely the potential of the isolated compounds and understand their effective action in controlling the phytopathogens tested.
Our results present a potentially novel source of fungitoxic substances. The R. guttatus extract had a beneficial effect on reducing mycelial growth, and also lower concentrations of the extract showed antifungal activity. Our study provides a basis for further studies examining the antifungal activity of secondary metabolites produced by amphibians on phytopathogens of agricultural interest.

CONCLUSIONS
The methanolic extract of the cutaneous secretion of Rhaebo guttatus caused mycelial growth inhibition of Fusarium udum, Fusarium solani, Aspergillus flavus and Macrophomina phaseolina at a range of concentrations. Furthermore, all extract concentrations inhibited mycelial growth of Fusarium solani. The extract also inhibited Aspergillus flavus sporulation and conidia germination of Fusarium udum similarly to the positive control fungicide. The Rhinella marina extract inhibited the mycelial growth of Colletotrichum truncatum and microsclerotia production of Rhizoctonia solani. Both extracts decreased appressoria formation of Colletotrichum truncatum to the same level of the positive control. The results suggest that the R. guttatus extract has higher antifungal activity against the analyzed phytopathogens, but that both extracts represent a potential alternative for inhibiting appressoria formation and the consequent penetration through the leaf surface by phytopathogenic fungi. Further studies are needed to determine the bioactive compounds responsible for the antifungal activity of the extracts. The Rhaebo guttatus and Rhinella marina extracts should be tested in other phytopathogen fungi which affect the agricultural crops in the region. Our results provided relevant information for assays with components derived from secondary metabolites of amphibian species on the growth of phytopathogens.