In Mediterranean countries, table olives are among the most commonly consumed fermented food. They are considered a functional food due to their nutritional value, content in bioactive compounds and dietary fiber, fatty acid composition and presence of several antioxidants (Campus et al., 2018).

In table olive fermentation, mold growth can cause huge economic losses and reduce the product safety (El Adlouni et al., 2006; Ghitakou et al., 2006). Molds of the genera Aspergillus and Penicillium have been described in different olive fermentation processes (Heperkan et al., 2006) and both are considered undesirable microorganisms. They can produce mycotoxins and cause the softening of fruits because of their cellulase and xylanase activities (Arroyo-López et al., 2016). Penicillium crustosum is one of the most prevalent molds in fermented black table olives and is able to produce toxic metabolites such as dehydrocyclopeptin, andrastin A, cyclopenol, penitrem A, roquefortine C, viridicatol (Bavaro et al., 2017) and thomitrem A and E (Rundberget and Wilkins, 2002). Aspergillus section Nidulantes includes several species able to produce mycotoxins, like aflatoxins, sterigmatocystin, emestrin, fumitremorgins, asteltoxins, and paxillin; four species (A. astellatus, A. miraensis, A. olivicola, A. venezuelensis) produce aflatoxin B1 (Chen et al., 2016).

Different chemical, physical and biological methods have been proposed to prevent mold growth or to eliminate or reduce aflatoxins during table olive fermentation. For example, the spoilage of table olive by molds can be reduced using chemical preservatives, such as potassium sorbate and sodium benzoate (Turantaş et al., 1999) or natamycin (Hondrodimou et al., 2011). Değirmencioğlu et al. (2014) studied the effect of washing solutions -acetic acid, lactic acid, chlorine dioxide- on dry-salted olives during 7 months of storage at 4 °C. The chlorine dioxide solution (10 ppm) combined with vacuum packaging was the best approach for controlling mold and yeast growth. In addition, the effectiveness of high hydrostatic pressures and citrinin against mold growth in table olives has been demonstrated (Tokuşoğlu et al., 2010; Argyri et al., 2014).

In view of growing consumer concern for food safety issues, including additive content (Bearth et al., 2014), the food industry is seeking biological alternatives in order to inhibit undesirable molds. Among the novel approaches, biopreservation and, more specifically, the selection of protective cultures has been identified as one of the more promising alternatives.

It is well known that lactic acid bacteria (LAB) can produce several antifungal metabolites (Schnürer and Magnusson, 2005). LAB are able to inhibit molds related to bread spoilage (Corsetti et al., 1998); consequently, the use of antifungal sourdough in the bakery industry is now a common practice to ensure the microbiological safety of bread (Gerez et al., 2009). Antifungal lactobacilli can also be found in raw milk (Delavenne et al., 2012); therefore, specifically selected LAB strains can also be used as biopreservatives in fresh cheeses (Fernandez et al., 2017). The antifungal activity of LAB, specifically Lactobacillus plantarum, and their ability to reduce aflatoxin B1 during olive storage has also been demonstrated (Kachouri et al., 2014).

Considering the state of the art, the aim of the present study was to find LAB able to inhibit molds commonly associated with table olive fermentation and spoilage.

In the present study, 59 LAB exhibited antifungal activity. Both in agar and in co-culture, the LAB exhibited greater inhibitory activity against Penicillium crustosum compared to Aspergillus section Nidulantes. The antifungal LAB strains so identified belong to two species (Lactobacillus pentosus and Lactobacillus sanfranciscensis), whose ability to inhibit or reduce mold growth is well known (Corsetti et al., 1998; Schnürer and Magnusson, 2005).

The antifungal activity of LAB is related to their ability to produce antifungal metabolites, e.g. organic acids, proteinaceous compounds, reuterin, and 3-hydroxylated fatty acids (Schnürer and Magnusson, 2005). For example, the inhibitory properties of phenyllactic acid (PLA) produced by LAB against several fungal species isolated from food have been demonstrated (Valerio et al., 2004). Regarding proteinaceous compounds, bacteriocin-producing LAB were already isolated from fermented olives and it was observed that their ability to produce bacteriocins was affected by NaCl, pH and temperature (Hurtado et al., 2011). Moreover, the bacteriocin production by LAB seemed to be affected by the presence of other bacteria during olive fermentation (Ruiz-Barba et al., 2010).

The antifungal activity of LAB could also be due to a synergic effect between the sodium acetate present in the MRS medium and lactic acid and other compounds produced by LAB (Cabo et al., 2002). Schillinger and Villareal (2010) demonstrated how sodium acetate in a culture medium can influence the inhibitory activity of LAB; in their experiment, LAB which exhibited antifungal activity in MRS agar with sodium acetate did not have the capacity to inhibit molds in MRS agar without sodium acetate. Also, Cheong et al., (2014) reported that LAB which exhibited antifungal activity in MRS agar with sodium acetate did not exhibit the same behavior in MRS agar without sodium acetate; however, the same strains were able to inhibit Penicillium commune in cottage cheese.

Lind et al., (2005) demonstrated how the absence of sodium acetate in MRS agar did not influence the antifungal activity of Propionibacterium against Penicillium roqueforti and Aspergillus fumigatus. Magnusson et al., (2003) performed an HPLC analysis of the supernatants of the LAB exhibiting antifungal activity, previously tested in MRS broth, and the concentration of lactic acid was equal to or even higher than concentrations in strains devoid of inhibitory activity; in addition, the concentration of acetic acid was similar to that found in MRS broth. This demonstrated that the activity was probably due to the production of other antifungal compounds.

In the present study we decided to carry out the tests using MRS with sodium acetate. Penicillium crustosum and Aspergillus section Nidulantes grew well in all the control samples, so the inhibition was probably due to the production of antifungal substances by LAB.

In our opinion, this study makes a useful contribution to solving the problem of fungal growth and potential mycotoxin accumulation during table olive fermentation, thereby improving its safety.

Summarizing: