Already more than twenty years ago some authors [5,6] anticipated the potential use of NSYs during the first step of fermentation applied to improve the final flavor of wines. Today, knowledge has exponentially increased and the aromatic enhancement of wines due to the use of NSYs is only one of the many other features identified for their potential application. In view of the possible use in agri-food industry, it is a common prerogative that NSYs must possess specific traits to be selected. Indeed, after the general re-valuation of the role of NSYs, several studies have focused the attention on their multiple advantages, particularly in winemaking. Among these, the antagonistic activity against undesired microorganisms is of paramount importance.

It has been widely demonstrated that some NSY strains in winemaking can control spoilage yeasts or filamentous fungi both in the vineyard and during the early stages of fermentation [7,8,9]. In this regard, spoilage species producing off-flavors or films in bulk wine such as Zygosaccharomyces rouxii, Kloeckera apiculata, Pichia spp. and Candida spp. may be prevented by the use NSYs through released active antimicrobial extracellular molecules [10]. Effectively, the growing interest in reducing and rationalizing the use of SO₂ in winemaking has urged scientific investigations towards the antimicrobial activity of NSYs as an alternative to conventional chemical additives [11,12]. About this, natural control strategies may involve the use of killer toxins (mycocins), antimicrobial peptides such as Lactoferricin B or volatile compounds produced by NSYs as a biological strategy to counteract contamination [13].

The term “biocontrol” related to the use of microorganisms as natural biological agents was defined as the reduction in pathogen or disease activities through organisms or their molecules. In agri-food, this concept is related to an alternative strategy to the use of chemical products and the use of microorganisms with antagonist action against other microorganisms lowering the use of pesticides and boosting food quality and safety [14,15,16].

In recent years, to avoid the huge losses of fruit and vegetables due to pathogens, biological strategies have been studied as alternatives to products of a chemical nature. The addition of microorganisms as bio-protective agents or their antimicrobial products has already been identified as “bio-protection”. This practice is firstly used in agriculture and then in food industries for the control of fruit decay from post-harvest spoilage microorganisms and the relative extension of their shelf life. This strategy consists in the inoculation of viable antagonistic microorganisms (bacteria, yeasts or a mixture of them) or the addition of their antimicrobial products in complete or partial purified form during, at the end or after the chain production [17,18,19]. Biological control based on the use of beneficial microorganisms is receiving increasing recognition, although the number of registered and approved marketed bioproducts containing bioactive yeasts or bacteria remains scarce. Czajkowski and collaborators [20] argue that this situation arises from objective difficulties encountered during registration and marketing, but also from problems of understanding the specific roles of each member of a consortium and their biological activity. Regarding the application of NSYs in bio-protection strategies against grapevine trunk diseases in the vineyard (the fungal pathogen Botrytis cinerea causing bunch sour rot), research has increased significantly in recent years [21,22]. In addition, there is a growing interest in the application of this approach on the protection against undesirable microorganisms during the entire wine production chain as well as during the conservation and maturation phases [23,24]. The renewed interest in biological control is due to the growing attention to the use of sulfites in food and particularly in wine. In this regard, NSYs have been proposed as a possible effective natural alternative. Several recent studies have been carried out by using selected strains of M. pulcherrima and T. delbrueckii at the pre-fermentative stage in the red winemaking process [23,25,26]. Other studies focused attention on the use of other selected strains of M. pulcherrima during the cold clarification stage of the Italian Verdicchio white variety underlining the double role of this yeast as a biocontrol agent and wine aroma enhancer [27].

The antimicrobial activity of NSYs has also been investigated in other sectors because the re-evaluation of the use of these yeasts with biocontrol purposes has been extended across the food industry. Some applicative studies also showed their effectiveness in breadmaking or cheesemaking. Valsaraj and colleagues [28] highlighted the biological load of some NSYs and how their toxic killer effects could be used in the food and beverage industry, maintaining food quality and improving the safety in beer, cheese or bread. They suggested that the biocontrol strategy of NSYs isolated from foods and beverages that are naturally fermented may be effective in suppressing wild yeast strains during another fermentation.

No less important is the involvement of the antimicrobial role of yeast strains in the medical field. For over thirty years, Polonelli and Morace [29] reported the killer phenomenon against bacteria, seeing the possibility of using these toxins to counteract the growing phenomenon of antibiotic resistance. In a more recent study, the yeasts D. hansenii, Pichia fermentans, Candida tropicalis and Wickerhamomyces anomalus have been shown to induce bacterial lysis [30]. Chen et al. [31] isolated mycocin-producing strains of Kluyveromyces marxianus and demonstrated that crude extracts were effective in preventing Escherichia coli disease in mice. In addition, the yeasts Kluyveromyces lactis and K. marxianus, isolated from cheese, were able to inhibit the growth of pathogenic microorganisms, such as Listeria monocytogenes and Candida albicans [32]. Finally, C. albicans strains isolated from children had inhibitory effects on strains of E. coli ATCC 25 922 and Staphylococcus aureus ATCC 25 923 [33].

The regulation and the control of the growth of undesired microorganisms could be exploited through different mechanisms such as space and nutrient competition, cell-to-cell contact or antagonistic action mediated by antimicrobial compounds production such as mycocins, small peptides, VOCs or extracellular vesicles.

The main mode of action of yeasts in biocontrol is the competition for space and nutrients [34]. Most organisms can starve pathogens or parasites through iron withdrawal [35]. Aureobasidium pullulans possesses a siderophore identified as fusarinin C which has been shown to exhibit antibacterial activity [36,37]. Pulcherriminic acid is a cyclic dipeptide that complexes iron in M. pulcherrima. Studies have shown that pigment-deficient mutants of M. pulcherrima exhibited reduced or null antifungal activity and iron deprivation of the fungal pathogen. This suggested that the production of this pigment is one of several mechanisms by which this yeast antagonizes plant pathogenic fungi [38]. However, mutants lacking the ability to synthesize pulcherriminic acid still strongly inhibited filamentous fungi, confirming that the antifungal activity was only due to iron deprivation. Therefore, the exact contribution of iron chelators to the yeast biocontrol activity remains to be clarified in detail. Another competition for nutrients was found in Saccharomycopsis schoenii in that it is unable to assimilate sulfur, a specific feature of the genus Saccharomycopsis. In addition, phytopathogenic fungi and Trichoderma species exhibited a similar phenomenon, which may indicate that methionine is an important target for such organisms and has been hotly contested [39]. Pioneering experiments were aimed at evaluating the suitability of an easily transformable Pichia (formerly Ogataea) angusta haploid strain to identify biocontrol-minus mutant clones: while the wild-type strain proved effective in reducing brown rot lesion caused by Monilinia fructicola on apple fruit, its derivate leucine-auxotrophic mutant L1 had no significant effect in controlling the pathogen. The addition of exogenous leucine fully restored the biocontrol capability of mutant L1, whereas a leucine stand-alone treatment showed no significant biocontrol effect [40].

Another strategy to compete for space is the formation of biofilm. Biofilms are microbial communities consisting of a single or more species and are considered virulence factors for pathogenic microbes [41,42]. The development of a yeast biofilm begins with the adhesion of single cells to a surface and usually involves cell wall modifications, secretion of an extracellular matrix, and often the formation of hyphae or pseudo-hyphae [43]. In yeasts to be used in biocontrol activities, biofilm formation in wounds is now considered an important mode of action even though the molecular basis of the process and the composition of different biofilms have only been studied in detail for P. fermentans. Biofilm formation of P. fermentans in apple wounds protects against post-harvest disease, while on peaches it changes from a yeast-like to hyphal growth form and causes rapid decay of inoculated fruit in the absence of a plant pathogen [44]. Biofilm formation has been demonstrated besides that of S. cerevisiae also in P. kudriavzevii, W. anomalus and M. pulcherrima [45,46,47,48].

The saprophytic yeast-like fungus A. pullulans has been well documented for over 60 years in microbiological literature for its ubiquitarian presence in soil, water, rock surfaces and in both cold and warm areas. A lot of A. pullulans strains are known to produce a wide range of natural antimicrobials that are useful for biocontrol applications against plant pathogens mediated by nutrient and space competition and/or VOC production (see Section 2.2.6). A. pullulans can be used at both vineyard, for the management of grey mold disease caused by B. cinerea, and winery levels, through the synthesis of antifungal compounds, providing a versatile tool for the viticulturist/farmer as well as for the oenologist to combat problems in the field and create a high-quality wine.

Mycocins (killer toxins) are the most investigated yeast antimicrobial compounds. Since Bevan and Makower [49] discovered the killer phenomenon in S. cerevisiae, several other yeast species have been found to produce a toxic proteinaceous factor that kills sensitive strains [50,51]. Several potential applications for the killer phenomenon have been suggested in the food industry to control spoilage yeasts in the preservation of food and beverages [52,53].

The mycocin Kpkt produced by Tetrapisispora phaffii was first described as an anti-spoilage yeast [52]. Kpkt acts through a specific β-glucanase activity causing irreversible modifications on the cell wall structure and it is codified by the TpBGL2 chromosomal gene [17,54,55]. The recombinant toxin (rKpkt) was recently obtained by transferring the Kpkt coding gene in Komagataella phaffii (formerly Pichia pastoris) [56]. The recombinant rKpkt, when expressed in K. phaffii, displayed a wider spectrum of action than its native yeast [57], reinforcing the idea of the possible application of mycocins in the food and beverages industries. T. phaffii was used in mixed fermentation at the pre-fermentative stage to control wild yeasts such as Hanseniaspora, Zygosaccharomyces and Saccharomycodes in the place of sulfur dioxide [11].

Several works have been focused on the study of NSYs able to counteract the development of Brettanomyces spp., a relevant dangerous yeast in winemaking [58]. In this context, Pikt and Kwkt mycocins, produced by W. anomalus (formerly Pichia anomala) and Kluyveromyces wickerhamii, respectively [54], are able to counteract Dekkera/Brettanomyces. Pikt is an ubiquitin-like protein of about 8 kDa able to interact with the β-1,6-glucan of the cell wall of sensitive yeasts [59], while Kwkt is a protein of about 72 kDa of molecular mass, without any glycosyl residue [11] and β-1,6-glucosidase activity that seems to be involved in the act of blocking the cell cycle function of sensitive yeasts [12]. Cytofluorimetric evaluation showed that both Pikt and Kwkt caused irreversible death of this yeast in a different way from sulfur dioxide that induced a viable but non-cultivable (VBNC) state of Brettanomyces with a consequent recovery of yeasts when fresh medium was replaced [12]. Another P. membranifaciens strain showed a killer action producing two mycocins denominated PMKT and PMKT2. PMKT binds linear (1→6)-β-d-glucans in the cell wall and Cwp2p plasma membrane receptor of sensitive yeasts, leading to alterations in ionic exchange via plasma membrane [60]. PMKT2, a protein with an apparent molecular mass of 30 kDa, binds mannoproteins and induces cell cycle blockage in the early S-phase of sensitive yeasts and stimulates markers of cellular apoptosis such as the cytochrome c release, DNA strand breaks, metacaspase activation and production of reactive oxygen species at a low dose [61]. The killer activity of P. membranifaciens was exploited in winemaking to control the B. bruxellensis economic relevant spoilage yeast using mixed fermentation with S. cerevisiae and B. bruxellensis (inoculum ratio of 1:1). P. membranifaciens inhibited B. bruxellensis growth without any effects on the fermentation activity of S. cerevisiae.

Other mycocins (CpKT1 and CpKT2) active against B. bruxellensis are produced by Candida pyralidae. Both mycocins were active and stable at pH 3.5–4.5 and with the general conditions of the winemaking environment [62]. Both mycocin CpKT1 and C. pyralidae viable yeasts were used in mixed fermentation in red grape juice containing B. bruxellensis, determining a decrease in spoilage yeast concentration [63].

Additionally, a strain of W. anomalus was proposed as a biocontrol agent against Brettanomyces/Dekkera spp. [18]. The killer activity of W. anomalus is expressed through the release of KTCf20, a mycocin. This mycocin was able to counteract the growth of different spoilage yeasts such as Brettanomyces/Dekkera, Pichia guilliermondii and P. membranifaciens. Moreover, they showed that W. anomalus in mixed fermentation did not negatively affect S. cerevisiae strains. Another mycocin named WA18 and active against B. bruxellensis is produced by an autochthonous W. anomalus strain isolated from soil pit, and it exhibited 99% identity with UDP-glycosyltransferase protein [64]. In accordance with de Ullivarri and coworkers [18], the compatibility of this W. anomalus strain in mixed fermentation with S. cerevisiae yeast was confirmed. Another killer strain belonging to W. anomalus has been investigated for its wide potential of antibacterial activity against numerous human pathogenic agents [65]. Interest in these antibacterial mycocins was revealed by Muccilli and Restuccia, [66] who highlighted the potential use against pathogens resistant to conventional antibiotics, such as Staphylococcus aureus.

All reported studies contributed to demonstrate the global exigence to reduce conventional chemicals by using selected NCYs to ensure high-quality agri-food products with increased aromatic features and/or longer shelf life.

Some peptides produced by yeasts have shown antimicrobial effects against several grape-must/wine-contaminating yeasts. In general, these peptides show lengths of up to 100 amino acids, sorted into variable sequences, and the mode of action involves the disruption of the cell wall in sensitive strains. For example, small peptides with molecular mass below 5 kDa produced by Candida intermedia have shown greater antimicrobial specific effects against B. bruxellensis [70]. In addition, the antibacterial activity of the same strain against Escherichia coli, L. monocytogenes and S. typhimurium was demonstrated [71]. Similar observations have been reported by Younis et al. [72], where three isolates of C. intermedia from raw milk and fruit yoghurt showed antimicrobial activity against E. coli, S. aureus and Pseudomonas aeruginosa.

Another mechanism of interaction in mixed fermentation using NSY is the possible involvement of extracellular vesicles (EVs). In this regard, a recent work on the exo-proteome of EV-enriched fractions in pure and mixed fermentation with six different species of NSYs and S. cerevisiae showed a wide diversity of proteins secreted, indicating the presence of interactions and the possible involvement of EVs [73]. The EV-enriched fractions from different species such as S. cerevisiae, T. delbrueckii and Lachancea thermotolerans showed enrichment in glycolytic enzymes and cell-wall-related proteins, particularly the enzyme exo-1,3-β-glucanase. However, this protein was not involved in the here-observed negative impact of the T. delbrueckii extracellular fractions on the growth of other yeast species. These findings suggest that EVs may play a role in fungal interactions.

Some enzymes secreted may be involved in biocontrol action. Indeed, the secretion of enzymes degrading cellular components such as chitinases, glucanases or proteases is a common feature in all kinds of host–pathogen interactions and has been intensively studied. Chitinolytic enzymes allow the degrading of fungal cell walls [74]. Yeasts belonging to genera Aureobasidium, Candida, Debaryomyces, Metschnikowia, Meyerozyma, Pichia, Saccharomyces, Tilletiopsis, Wickerhamomyces and Saccharomycopsis exhibited this enzymatic activity [16,39]. Chitinases from sources, i.e., fungi and filamentous bacteria, have demonstrated biocontrol activity against plant pathogenic fungi and chitinases are extensively studied as potential biopesticides, targets for resistance breeding or as transgenes in genetically modified plants. Chitinases, probably in an indirect manner, influence biocontrol activity because Chito-oligosaccharides (CHOSs) resulting from chitin degradation are potent inducers of plant immune responses [51].

Glucans are major cell wall components in fungi and exoglucanases are involved in cell wall modification, cell adhesion and resistance to mycocins [75]. A 1,3-β-glucanase from Candida oleophila was the first gene cloned in this organism and overexpression or deletion of this gene did not significantly affect Penicillium digitatum spore germination, but subsequent studies have documented a reduced inhibitory activity of the β-exoglucanase deletion mutant compared to the wild type and overexpressing strain (in vitro and in fruit), thus demonstrating the involvement of glucanases in the biocontrol activity of yeast [76]. In W. anomalus, the deletion of two exo-β-glucanases (PaEXG1 and PaEXG2) significantly reduced the fruit biocontrol activity against B. cinerea [77], while the single deletion of PaEXG2 did not reduce biocontrol performance. Exoglucanase activity was also detected in several biocontrol yeasts and was linked to antagonist activity, but without demonstrating a causal involvement.

It has long been proven [78] that in Rhodotorula glutinis and Cryptococcus laurentii, β-1,3-glucanase activity did not correlate with their respective inhibitory activity against B. cinerea. The pathogenicity of yeast species belonging to Candida, Cryptococcus or Malassezia is related to lipase activity. Several studies have also correlated the role of lipases with the biocontrol action of fungi and bacteria against plate diseases. For this reason, the lipolytic activity of yeasts may represent an aspect to be investigated in relation to biological control [79].

The alkaline serine protease Alp5 of A. pullulans reduced spore germination and germ tube length of Penicillium expansum, B. cinerea, M. fructicola and Alternaria alternata in vitro and showed a concentration-dependent inhibitory effect on these pathogens on the apple tree [80]. Protease activity has been reported but not confirmed or investigated in the genera Metschnikowia, Pichia and Wickerhamomyces, but has not been further studied or confirmed.

Mycoparasitism is little studied in yeasts, but some studies have shown that P. guilliermondii adheres to the hyphae of B. cinerea and causes the collapse of the hyphae, presumably due to the secretion of hydrolytic enzymes such as glucanases [81].

Species belonging to the genus Saccharomycopsis have been studied against the biocontrol of several clinically relevant Penicillium species and yeasts [39].

Volatile organic compounds (VOCs) are the most relevant metabolites that show a biocontrol action. These metabolites are molecules < 300 Da with a low solubility in water and a high vapor pressure and include molecules such as hydrocarbons, alcohols, thioalcohols, aldehydes, ketones, thioesters, cyclohexenes, heterocyclic compounds, phenols and benzene derivatives. Volatiloma is specific for each yeast species as well as the spectrum of action against pathogenic microorganisms [82,83]. VOCs are species-specific and are produced by fungi, bacteria and yeast during their primary and secondary metabolism, limiting the growth of other microorganisms. The chemical composition of VOCs strongly depends on the environment and the pathogen being antagonized. The chemical composition includes alcohols, aldehydes, cyclohexenes, benzene derivatives, heterocyclic compounds, hydrocarbons, ketones, phenols, thioalcohols and thioesters. VOCs are produced by several yeast species to reduce the growth of pathogen molds. In this context, Di Francesco and coworkers [84] have demonstrated both in vitro and in vivo that the VOCs produced by A. pullulans reduced the growth and the infection of B. cinerea, Colletotrichum acutatum, P. expansum, P. digitatum and P. italicum.

Additionally, VOCs formed by NSYs W. anomalus, M. pulcherrima and A. pullulans as well as S. cerevisiae showed a biocontrol action against B. cinerea on table grape berries [85]. Selected strains of Cyberlindnera jadinii, Candida friedrichii, C. intermedia and L. thermotolerans inhibited the formation of both mycelial growth and ochratoxin A in Aspergillus carbonarius and Aspergillus ochraceus identifying β-phenyl ethanol as the active compound [86,87]. The VOCs of W. anomalus prevented spore germination, mycelial growth and toxin production of Aspergillus flavus [88]. Similarly, VOCs released by Candida sake reduced the incidence of apple rot caused by P. expansum and B. cinerea [89]. The inhibitory activity of Sporidiobolus pararoseus on spore germination and mycelial growth of B. cinerea was mainly attributed to 2-ethyl-1-hexanol, whereas C. intermedia produced 1,3,5,7-cyclooctatetraene, 3-methyl1-butanol, 2-nonanone and phenylethyl alcohol as the major components of its volatilome during the interaction with this pathogen. VOCs released by W. anomalus, Pichia kluyveri and H. uvarum inhibited A. ochraceus growth and ochratoxin A production during the fermentation process of coffee [90].