Effect of Bacillus spp. on Aspergillus westerdijkiae growth, sporulation and ochratoxin A production in green-coffee medium

Coffee is the most consumed beverage worldwide, with the estimative of 600 billion cups served every day. Brazil is the largest coffee producer and exporter, being responsible for producing more than 43 million sacs in 2015 and having approximately 2.25 million hectares destined to this culture (Associação Brasileira de Indústria de Café, 2016). The quality of the coffee beans affect directly on its price on the market and may have profound economic impacts when those grains are infested with spoilage agents such as filamentous fungi.


INTRODUCTION
Coffee is the most consumed beverage worldwide, with the estimative of 600 billion cups served everyday.Brazil is the largest coffee producer and exporter, being responsible for producing more than 43 million sacs in 2015 and having approximately 2.25 million hectares destined to this culture (ABIC, 2015).The quality of the coffee beans affect directly on its price on the market and may have profound economic impacts when those grains are infested with spoilage agents such as filamentous fungi.
Several fungi can contaminate the coffee beans and apart of causing direct spoilage, different toxins may be produced.Ochratoxin A (OTA) is one of the major mycotoxins found in coffee worldwide (Ostry et al. 2013;Taniwaki et al. 2014;Geremew et al. 2016).This mycotoxin has nefrotoxic, teratogenic and mutagenic in animals and is considered by the International Agency of Research on Cancer (IARC) as Group 2B: Possibly carcinogenic to humans (IARC, 2002(IARC, , 2012)).On warm climates such as those found in Brazil, OTA contamination in coffee is mainly associated with Aspergillus section Circumdati members A. steynii and A. westerdijkiae, with the latter being reported as the main responsible of OTA production in this product by several studies (Gil-Serna et al. 2011;Ostry et al. 2013;Taniwaki et al. 2014;Geremew et al. 2016).Aspergillus westerdijkiae is an opportunistic fungus that may grow and produce OTA under a different set of water activities and temperatures and can invade coffee beans both at pre-harvest and during storage, characterizing this species as a flexible contaminant (Gil-Serna et al. 2014).
Traditionally, fungal control during pre-harvest is mainly conducted by synthetic fungicide applications, but due to the adverse effects caused by those products on the environment and the consumers, alternative ecological-friendly methods are being employed to prevent fungal contamination.Biological control is nowadays broadly used as a prophylactic method for phytopatogens control on several agricultural cultures (Ongena and Jacques, 2008;Pérez-Garcia et al. 2011).
Among different biocontrol agents used worldwide, soil bacteria from Bacillus genus present promising characteristics such as low nutritional requirements, resistance to adverse environmental conditions, low or none toxicity to environment and the production of a large number of antifungal compounds (Ongena and Jacques, 2008;Pérez-Garcia et al., 2011).Although species from Bacillus genus are already being used as biocontrol agents to a diverse number of phytopathogens and contaminants, its efficacy towards Aspergillus westerdijkiae growth and OTA production remains commercially unexplored.Therefore, the aims of this study were to evaluate the effect of the interaction with three Bacillus spp. on A. westerdijkiae mycelial growth rate and conidial and OTA production in a green-coffee based medium.

Bacterial and fungal isolates
Three Bacillus sp.isolates previously selected on our laboratory were used in this trials: Bacillus safensis RF69; Bacillus amyloliquefaciens RP103 and Bacillus subtilis RP242.One isolate of OTA producer A. westerdijkiae previously isolated from coffee beans and made available by ITAL -Campinas was also used.

Inoculum preparation and culture media
The assays were performed in green-coffee medium as proposed by Gil-Serna et al. (2014), which contains 3% (w/v) of green-coffee beans and 20 g L -1 bacteriological agar.The medium preparation was conducted by boiling 30g of grounded green-coffee beans on 1 L of distilled water for 30 mins with subsequent filtration of the mixture and re-adjustment of the volume to 1 L of water.For inoculum preparation, bacterial isolates were grown for 24h on 523 broth (Kado and Heskett, 1970) and cell concentrations were adjusted for 10 9 CFU mL -1 by serial dilutions.A. westerdijkiae isolate was grown on Malt Extract Agar (MEA) at 25 °C for 7 days and the colonies were used for removal of agar plugs (5 mm) from the border of growing colonies with a cork borer (Pitt and Hocking, 2009).

Effect on fungal mycelial growth rate
For the evaluation of the effect of Bacillus sp.isolates on A. westerdijkiae mycelium growth rate, we used the method proposed by Bluma and Etcheverry (2006).Briefly, aliquots (1 mL) of bacterial suspensions were inoculated on Petri plates (final concentration of 10 9 cells mL -1 ) and after that, 19 mL of melted Green-Coffee Based Medium (GCM) was inoculated and homogenized.After solidification, agar plugs of A. westerdijkiae mycelium were inoculated in the center of the plate.
Treatment and control plates (A.westerdijkiae growing alone) were incubated on polyethilene bags for 10 days at 25 ºC in triplicates.This assay was repeated three times.Fungal growth in diameter, was evaluated daily by measuring two radii at right angles one to another with a ruler (Cavaglieri et al., 2005, Bluma andEtcheverry, 2006).The radial growth rate (mm day -1 ) was calculated by linear regression.

Effect on conidial production
The ten-day cultures described on item 2.3 were also used to evaluate the effect of interaction with Bacillus spp. on A. westerdijkiae conidial production.Cultures were washed twice with 5 mL of peptone water (0.01%) and harvested three times using a glass handle.
Conidial counts of the 10 mL suspensions were performed using a Neubauer chamber and the number of conidia mL -1 was calculated.Subsequently, total conidial number was calculated and determined the concentration of conidia per cm² of mycelium.This evaluations were repeated three times for validation of the obtained results.

OTA extraction and quantification
Ochratoxin A extraction from the cultures conducted by crumpling the whole cultures on 10 mL of chloroform twice and subsequently the extracts were filtered and evaporated to dryness.The extracts were resuspended in 250 µL of the initial mobile phase (Acetonitrile:H2O:formic acid, 10:90:0,2%), transferred to 10 µL vials and injected on Ultra-Performance Liquid Chromatography and Sequential Mass Spectometry (UPLC-MS/MS) system as described by Castro et al. (2014).

Statistical analyses
One-way ANOVA was applied to data analyzes, using Statistica 10.0 (StatSoft, Inc. ( 2011) software for Windows.Tuckey test was conducted for treatments mean comparison using p<0.05 significance and linear regression analyzes were conducted to calculate fungal mycelial growth rate.

RESULTS AND DISCUSSION
Results of the effect of interaction with Bacillus isolates on A. westerdijkiae mycelial growth rate and conidial and OTA production are showed on Table 1.The fungal diameters per day are showed on Figure 1.All three Bacillus sp.isolates were able to significantly (p<0.05)reduce A. westerdijkiae mycelial growth in a range between 76 and 96%.Growth per day was significantly reduced by all Bacillus isolates, which B. amyloliquefaciens RP103 and B. subtilis RP242 showed to be more effective in restraining A. westerdijkiae propagation than B. safensis RF69.The ability to produce conidia was also negatively affected by all Bacillus isolates, which significantly reduced the conidia per cm².While direct antagonism has already been reported by Petchkongkaew et al. (2008), this is the first study that show reduction of A. westerdijkiae mycelial growth rate and the reduction of the conidial production by Bacillus spp.Inhibition of growth rate often lengthens fungal lag phase, and therefore, decrease the its capacity to compete and survive, characterizing a possible control method for this fungus (Cavaglieri et al., 2005).The ability to reduce conidial production is also an interesting characteristic to potential biocontrol agents, since the propagation of spores carried by the air is the main dissemination method used by filamentous fungi and its inhibition affects the fungal capacity to invade other substrates.
Ochratoxin A final concentrations were also affected by interaction with the biocontrol agents.All the three Bacillus sp.isolates showed to reduce final OTA concentrations in a range between 62 and 96% (Table 1).Accordingly to Choudhary (1992), reduction of toxins levels can be attributed to physical competition for space and nutrition, competition with the fungus for a substrate required for toxin production, induction for a change in the biochemical environment influencing the metabolic pathway available to the toxigenic fungi or degradation of OTA following its formation.
Isolates B. amyloliquefaciens RP103 and B. subtilis RP242 were the most effective on reducing OTA production.Similar results were observed by Petchkongkaew et al. (2008), which reported in vitro reduction of A. westerdijkiae's production of OTA by several Bacillus sp.isolates and found a B. licheniformis isolate that reduced OTA production by 92%.Our findings and those reported by Petchkongkaew et al. (2008) confirm the potential biotechnological applications of these bacteria on reducing OTA production by A. westerdijkiae on coffee beans.

CONCLUSIONS
In summary, the results obtained in this study present the promising characteristics of Bacillus spp. to be used as biocontrol agents to ochratoxigenic Aspergillus westerdijkiae.Our findings show that isolates B. amyloliquefaciens RP103 and B. subtilis RP242 are extremely efficient on inhibiting vegetative growth, the dissemination capacity and OTA production of A. westerdijkiae in a coffee based medium as substrate and may be used as an alternative control method aiming the improvement of coffee microbiological and toxicological quality.Further studies must be conducted to evaluate the effect of the bacterial treatment on A. westerdijkiae growth and OTA production directly in coffee beans to confirm the commercial usage of the selected biocontrol agents.

Figure 1 -
Figure 1 -Aspergillus westerdijkiae mycelial diameter per day when interacting with biological control agents Bacillus safensis RF69, B. amyloliquefaciens RP103 and B. subtilis RP242 in a Green-Coffee Based medium.

Table 1 -
Effect of interaction of Bacillus isolates and A. westerdijkiae on fungal mycelial growth rate and conidial and ochratoxin A production.