TOXICITY OF β-(1→3,1→6)-ᴅ-GLUCANS PRODUCED BY Diaporthe sp. ENDOPHYTES ON Metarhizium anisopliae (METSCHNIKOFF) SOROKIN ASSESSED BY CONIDIA GERMINATION SPEED PARAMETER TOXICIDADE DE Β-(1→ 3,1→ 6)-ᴅ-GLUCANAS PRODUZIDAS POR ENDÓFITOS Diaporthe sp. SOBRE Metarhizium anisopliae (METSCHNIKOFF) SOROKIN AVALIADA PELO PARÂMETRO DE VELOCIDADE DE GERMINAÇÃO DOS CONÍDIOS

We have previously reported that β-(1→3,1→6)-ᴅ-glucans produced by endophytes Diaporthe sp. G27-60 and G65-65 (GenBank accession codes JF766998 and JF767007, respectively) are promising anti-proliferation agents against human breast carcinoma (MCF-7) and hepatocellular carcinoma (HepG2-C3A) cells. However, the literature fails to describe the effects of Diaporthe exopolysaccharides (EPS) on eukaryotic healthy cells. The fungus Metarhizium anisopliae has been employed as model-system to evaluate the toxicity of pharmaceutical and agricultural-interest substances, taking into account, among other parameters, the speed of conidia germination. Current study verified the effect of different concentrations of Diaporthe β-glucans on the germination speed of M. anisopliae. Conidia were incubated with β-glucans treatments (50, 200 and 400 μg/mL) at 28oC, sampled during 24 h and analyzed by light microscopy. At the end of a 24-h incubation, the amount of germinated conidia reached ≈99% for controls and ranged between 97.7 and 98.6% for treatments. Bayesian analysis indicated that Diaporthe glucans had no toxicity on M. anisopliae and the curve of germination occurred as expected for this fungal strain. Considering the validity of filamentous fungi as model-systems, results are important data on the toxicity of endophytic EPS on healthy cells and may be associated with our previous results obtained for these polymers against tumor cells.

Literature data report that Phomopsis (= Diaporthe) foeniculi, a phytopathogen isolated from fennel, secreted two EPS (a galactan and a mannan) that exhibited toxic effects (i.e. chlorosis, necrosis and/or wilting) on fennel, tobacco and tomato plants (Corsaro et al. 1998). Different biological systems could be evaluated to provide an in-depth investigation on the toxic effects of these polymers on healthy cells.
In 1974, Smith and Rosazza suggested that microbial systems could be defined as those capable of mimicking the biotransformations reported in mammals (Cerniglia 1997). Once the basic principles of many cellular processes are conserved between animals and fungi, these microorganisms may be used for studying fundamental cell biological issues, with the advantage of their amenability to classical and molecular techniques (Steinberg and Perez-Martin 2008). So that conidia germination may be employed as a parameter, fungal conidia are inoculated into a liquid medium; samples are periodically collected, and the number of germinated conidia is counted (Milner, Huppataz and Swaris 1991) to determine whether chemical variables influence fungal development and conidiogenesis (Rangel et al. 2004).
Metarhizium anisopliae cells may be used as model-system for toxicity assays of chemical products (Almeida et al. 2014). In this context, the germination speed of M. anisopliae conidia has been employed to evaluate the effects of nutritional and physical factors, and the toxicity or compatibility of pharmaceutical and agricultural-interest substances (Rangel, Alston and Roberts 2008, Alves et al. 2011, Akbar et al. 2012, Tonussi et al. 2012, Bulla et al. 2013, Fabrice et al. 2013, Almeida et al. 2014, Mochi et al. 2017, Sohrabi et al. 2019. Current study employs M. anisopliae conidia germination speed as a parameter for the evaluation of the toxicity of β-(1→3,1→6)-ᴅ-glucans secreted by Diaporthe sp. G27-60 and Diaporthe sp. G65-65, and expects to detect a possible toxic effect of the fungal polymers against eukaryotic healthy cells.

CONTENTS
The Mato Grosso (MT) strain of M. anisopliae var. anisopliae -retrieved from the Collection of Endophytic and Environmental Microorganisms, Laboratory of Microbial Biotechnology, Universidade Estadual de Maringá, Brazil (CMEA/LBIOMIC-UEM) -was used as a model-system to evaluate the possible toxic effects of different concentrations of β-glucans. The strain was grown on Petri dishes containing Complete medium (CM) (Pontecorvo et al. 1953), whilst conidia were obtained directly from seven-day-old sporulation cultures by scraping and suspended in a 0.01% Tween 80 aqueous solution (7 mL). The conidia solution was filtered through a glass funnel with autoclaved gauze and added to a saline solution (9 mL) to obtain a solution with a concentration of 3×10 7 conidia/mL. Further, 300 µL of conidia solution were inoculated into 10-mL glass flasks containing 400 µL of Liquid complete medium (LCM) (Pontecorvo et al. 1953). Each flask received 100 µL of β-glucan solution, as shown in Table 1. The β-(1→3,1→6)-ᴅ-glucans were previously obtained by Orlandelli et al. (2017b) from submerged cultures of Diaporthe sp. G27-60 and Diaporthe sp. G65-65 (GenBank codes JF766998 and JF767007, respectively), two endophytes isolated from the medicinal plant Piper hispidum Sw. (Orlandelli et al. 2012) and retrieved from the CMEA/LBIOMIC-UEM. Controls comprised (C1), namely, β-glucan solution replaced by same volume of water, or (C2) LCM plus dimethyl sulfoxide (DMSO; Sigma-Aldrich Co.) at concentration 1%. The three concentrations (50, 200 and 400 μg/mL) of β-glucans tested were prepared under same conditions previously described by (Orlandelli et al. 2017b) for anti-proliferation assay and, then, dissolved in DMSO and diluted with culture medium (LCM). The DMSO concentration used (≤1%) has already been reported as compatible with M. anisopliae conidia germination (Schumacher andPoehling 2012, Wenzel Rodrigues et al. 2017).
All flasks remained incubated at 28±2ºC for 24 h and triplicates of control and treatment samples were collected periodically (at 0, 6, 8, 10, 12 and 24 h). Conidia were counted using a Neubauer hemocytometer and light microscopy, and the percentage of germinated conidia was assessed by randomly observing 100 conidia per sample. A conidium was considered germinated when a germtube was projected (Milner et al. 1991).
So that possible differences could be verified in the conidia germination among treatments and control groups, incubation time and their interactions, the counting data were analyzed with statistical package BRugs for software R (R Development Core Team 2008) and the Poisson distribution was assumed, implemented by Bayesian methodology. The Monte CarloMarkov Chain (MCMC) was composed of 10,000 samples for each parameter, with a burn-in period of 1,000 initial values and thinning interval of 10, or rather, at every 10 values generated, one belonged to the sample, with 900 values generated. Significant differences were considered at 5% level between treatments if the zero value was not included in the credibility interval of the desired contrast. A non-informative Gamma distribution was considered a priori for germinated conidia averages, that is, θ n ~ G(10 -3 ;10 -3 ), where θ n is the mean for each n treatment considered.
The analysis of the behavior of conidia germination over time consisted of a model of logistic regression, whilst data were analyzed with the same package and software described above. The binomial distribution was considered for the data of germination percentage, and the following formula (1) was used: log it(θij) = β 0 + β 1 time + β 2 time 2 (1) where: log it is the logistic link function; θij is the germination percentage; β 0 is the intercept; β 1 is the linear logistic regression coefficient; β 2 is the quadratic logistic regression coefficient and time is the number of hours elapsed since the beginning of incubation time.
The regression fit was tested by the coefficient of determination (r²) and binomial distribution was taken into account for germination percentage data. Further, 10,000 values were generated in a MCMC process for each parameter, considering a sample discard period of 1,000 initial values. The final sample was taken with steps of 10. Logistic regression coefficients were significant at 5% level if the zero value was not contained in the credibility interval for the parameter. A noninformative normal distribution was considered a priori for parameters b 0 , b 1 and b 2 , or rather, b 0 , b 1 , b 2 ~ N(0;10 -6 ).
When a logistic link function is considered, the conidia germination percentage is generally given by the formula (2) for quadratic regressions: where θij is the percentage of germinated conidia.
Herein, the MT conidia germination speed represented a rapid and alternative toxicity assay which significantly expanded investigation on toxicity studies. Comparative analyses of ribosomal RNA and protein sequences proved that fungi are closely related to animal cells (Baldauf andPalmer 1993, Wainright et al. 1993). Furthermore, fungi may be easily grown and manipulated under laboratory conditions for studying cellular and genetics processes (Kibbler et al. 2018), justifying the choice of current fungal model-system.
Approximately 16% of conidia were germinated at 6 h (Table 2), and significantly increased at 8 h (~66%), corroborating the curve of germination speed (Figure 1), which shows that, for all controls and treatments, the conidia germination started close to 4 -6h of incubation. It was more apparent at 8 h, as expected for this fungal strain in the absence of toxic substances (Alves et al. 2011). At the end of 24 h of incubation, the amount of germinated conidia was ≥99.0% for controls, and ranged between 97.7% and 98.6% for treatments. Germination speed curve indicated that no delays on fungal germination occurred when MT conidia was treated with Diaporthe β-glucans at 50, 200 and 400 µg/mL concentrations.  The spore germination is a process that comprises a sequence of events that activate the resting spore (D'Enfert 1997). When triggered to germinate, the cell becomes hydrated and there is a marked increase in respiratory activity, followed by a progressive increase in the synthesis of protein and nucleic acids (Deacon 2006). The resting spore is subsequently converted into a rapidly growing germ-tube from which the mycelium is formed by elongation and branching (D'Enfert 1997, Deacon 2006. Water, oxygen, and carbon dioxide are universally required to activate the spore germination (D'Enfert 1997). Moreover, optimum conditions such as temperature, humidity, pH and nutrient sources are essential for conidia germination (Almeida et al. 2014). The process is directly affected by nutritional, environmental, physical and chemical factors (Rangel et al. 2004). Variations in the speed of conidia germination may be observed in response to different stress conditions St Leger 2004, Rangel, Alston andRoberts 2008).
The Bayesian analysis is an approach that works on datasets with true distribution, being reliable for small groups of data (Alves et al. 2011). It has been used to obtain precise estimates without needing any kind of transformation (Gomes et al. 2014). The statistical method revealed the absence of a statistically significant difference between the germination speed of controls and treatments. A Bayesian ICr of 95% is the interval in which 95% of the samples are contained, and the smaller the interval, the less dispersed is the parameter. Overall (Table 3), the means of germinated conidia in the interval 0-24 h ranged between 57.40 and 59.75%. According to these results, the β-glucans had no inhibitory effect on the M. anisopliae germination when compared to controls. The means of germinated conidia significantly increased through the 0 -12 h interval, while the incubation time of 12 and 24 h was statistically equal, showing higher germination percentages that reached a mean of 97.93% of conidia germinated at 24 h (Table 4). The germination speed remained increasing between 8 and 14 h of incubation, after which it became stable. The logistic regression adjusted efficiently the conidia germination percentage over time (Table 5), showing that the germination percentage over time showed a quadratic behavior.   Conidial germination may be employed to determine whether the substrate on which conidia were produced affected the endogenous reserves stored in conidia during conidiogenesis (Rangel et al. 2004). Corroborating what was reported above, Bulla et al. (2013) evaluated the toxicity of the antihypertensive agent perindopril on MT strain by conidia germination speed parameter; no toxicity was detected and concentrations of 200 and 20 μg/ml increased the germination speed. On the contrary, Almeida et al. (2014) reported that EPs 7630® (an ethanolic root extract from the plant Pelargonium sidoides) delayed MT conidia germination when compared to controls, although the conidia viability was preserved.