2.1. Characterization of Areas and Collection of Samples

This study was conducted in two municipalities in northern Minas Gerais State, Brazil (Figure 1). The two collection areas originally corresponded to the ecosystems of Cerrado sensu stricto, with one area (Area I: 16°29 ^′40,93 ^″ S, 42°57 ^′19,16 ^″ W; altitude 825 m) located in Grão Mogol, a municipality whose climate is classified as humid subtropical (Cwa), while the other area (Area II: 16°23 ^′20,28 ^″ S, 44°9 ^′30,00 ^″ W; altitude 812 m) is located in Montes Claros, a municipality located in a region with a tropical savanna climate (Aw), according to the Köppen-Geiger classification [24].

In addition to climatic differences, there are noticeable differences in ecological and socioenvironmental aspects between the areas, including altitude, vegetation, and degree of anthropization. Area I is surrounded by a rural matrix with abandoned pasture areas, monocultures, and some fragments of preserved Cerrado. In contrast, Area II is located in a region close to an urban matrix, and the fragments of preserved vegetation are very small and have a higher degree of anthropization.

A total of 96 flowers of Q. grandiflora were collected in different stages of dehiscence, with 48 flowers in each area. The flowers were collected aseptically and deposited, immediately upon being collected, in Falcon tubes containing 25 mL of sterile saline–peptone solution (0.9% of NaCl and 1% peptone), supplemented with chloramphenicol (0.05%, w/v), for transport.

2.2. Isolation

Yeast isolates were obtained by shaking the Falcon tubes vigorously in a vortex shaker and then spreading 100 μL of the liquid on plates containing YM agar medium (0.3% yeast extract, 0.3% malt extract, 0.5% peptone, 1.0% glucose, 2% agar, and 5 μg/μL of chloramphenicol), which were then incubated in a microbial growth oven at 28°C (±3°C) for 3–8 days. The growth of yeast colonies on the plates was observed during incubation, after which purification was performed by morphological typing using the streak-by-depletion technique on YEPD agar medium (1.0% yeast extract, 2.0% peptone, and 2.0% dextrose). The isolates were then kept refrigerated at 4°C until identification. The strains examined in this study are deposited in the Collection of Microorganisms and Cells of the Federal University of Minas Gerais (UFMG), Belo Horizonte, Minas Gerais, Brazil.

2.3. DNA Sequencing and Phylogenetic Analysis

For the extraction of DNA and amplification by polymerase chain reaction (PCR), standard methods were used, according to [25, 26], respectively. D1/D2 divergent domains were amplified by PCR using primers NL-1 (5 ‘GCATATCAATAAGCGGAGGAAAAG3’) and NL-4 (5 ‘GGTCCGTGTTTCAAGAGGG3’) and sequenced using BigDye Version 3.1 (Applied Biosystems, United States) in combination with the ABI 3730 automated sequencing system. Yeast isolates were grown in liquid YM culture medium for a period of 1–3 days. After growth, cells were precipitated by centrifugation of 1.5 mL of the culture. DNA was extracted by the chloroform/isoamyl alcohol method, using glass beads for mechanical lysis of the cells and extraction buffer (2% Triton X, 1% SDS, 100 mM NaCl, 10 mM Tris pH 8, 1 mM EDTA pH 8). The DNA fragment corresponding to the D1/D2 domain of the 26S ribosomal RNA gene was amplified by PCR using primers in a reaction with a final volume of 25 μL containing 20–30 ng of template DNA, 20 pmol of each primer, 1.5 mM MgCl2, and 0.2 mM dNTPs (dATP, dCTP, dGTP, and dTTP). The PCR amplification criteria used were denaturation at 92°C for 40 s, annealing at 57°C for 1 min and 30 s, extension at 72°C for 2 min, and final extension at 72°C for 5 min. The amplified fragments were treated to remove impurities and unwanted materials. After treatment, the fragments were subjected to sequencing.

The resultant sequences were analyzed using the PHPH tool (http://asparagin.cenargen.embrapa.br/phph/) [27] for quality verification and then compared with sequences deposited in GenBank using the Basic Local Alignment Search Tool (BLAST) of the National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov) [28]. Alignment and phylogenetic relationships among sequences of the isolates and the sequences of GenBank were determined with the MEGA X program [29] using the maximum likelihood method with the Kimura-2-parameter evolutionary model and a discrete gamma distribution (+ G) with five categories, assuming that a certain fraction of sites is evolutionary invariable (+ I) [30]. Bootstrap analysis [31] to estimate confidence limits was based on 1000 replications. The nucleotide sequences obtained there were deposited in GenBank under the accession numbers given in Table 1.

2.4. Extracellular Enzyme Activities

The isolated yeasts were tested for their ability to degrade pectin, starch, carboxymethylcellulose (CMC), skim milk proteins (casein), xylan, and tannic acid using the agar diffusion method. For this procedure, the isolates were previously cultured at 25°C for 24–48 h on medium (to check for purity and metabolic activity). After growth, a portion of the colony was resuspended in a test tube containing 2 mL of sterile distilled water. Inocula standardization was performed using an optical density at 600 nm (OD600) of 0.05 of each yeast species suspension. After resuspending the cultures, 5 mL of each sample (in triplicate) was inoculated into plates containing YNB agar culture medium (Difco), enriched with specific carbon sources as mentioned above for the detection of each corresponding enzyme activity [32–35]. After inoculation, the plates were incubated from 5 to 8 days at 25°C, as some of the isolated yeasts (specimens) were not growing well at the initial temperature of 28°C, requiring adaptation to minimize sample loss. Verification of positive results was made by observing halos formed around the colonies, and negative results by their absence. The enzymatic activity for each microorganism was determined semiquantitatively using the enzymatic index (EI) represented by the equation: EI = dh/dc, where dh is the average diameter of the substrate hydrolysis halos and dc is the average diameter of the colonies, both measured and expressed in millimeters [34, 36].

3.1. Molecular Isolation and Identification

From the 96 samples from flowers of Q. grandiflora, a total of 58 yeast isolates were recovered: 56 yeasts and two yeast-like fungi. Flowers, in general, are seen as optimal niches for the proliferation and survival of microorganisms, especially yeasts, as they are rich in organic compounds such as monosaccharides, polysaccharides, and other carbon and nitrogen compounds, which favor microbiological growth [14]. In addition to physiology, flowers have a morphological apparatus that is evolutionarily attractive to insects and other pollinators and floral visitors. This characteristic, in addition to promoting the dispersion of yeasts, facilitates genetic exchange between established yeast communities.

The yeasts and yeast-like fungi recovered from the flowers of Q. grandiflora were identified by sequencing the D1/D2 region of the large subunit (26S) rRNA gene and comparing the sequences with those deposited in GenBank, where the values of identity and coverage are considered (Table 1).

The microbial community isolated from the flowers of Q. grandiflora was found to be composed of 18 different yeast species and two species of yeast-like fungi. Yeasts were represented by 14 genera, with Candida being the most prevalent with five species. The other genera represented by more than one species were Meyerozyma and Wickerhamomyces, both with two species. In addition to being represented by a greater number of species, the genus Candida also had a higher number of occurrences among the isolates, followed by the genera Starmerella, Wickerhamomyces, and Coniochaeta.

The Candida genus is represented by species that easily establish themselves in the most diverse functional niches, habitats, and hosts, as long as these locations offer the minimum conditions necessary for the survival and adaptation of this group of yeasts. Thus, the identification of more than one species of this genus among the identified Ascomycetic yeasts was not surprising, since the genus Candida incorporates a large group of yeasts from the same phylum, in many cases not even correlated based solely on taxonomic characters, sometimes somewhat ambiguous. In accordance with the results of the present work, other authors have also reported the occurrence of species of the genus Candida in association with flowers [37], as well as all other species and/or genera of yeast identified from the sampling of Q. grandiflora flowers carried out in this work [13, 23].

In addition to the genus Candida, we observed the occurrence of lineages of the genus Starmerella in the flowers of Q. grandiflora. Corroborating this finding, we verify in the literature that ascomycetic yeast species systematically isolated from ephemeral flowers and their associated insects in neotropical, nearctic, and Australian biogeographic regions belong mainly to the Starmerella, Metschnikowia, Kodamaea, and Wickerhamiella clades [15, 38]. Additionally, in agreement with this information, we also identified species of the genera Kodamaea and Wickerhamiella among the isolates. Many of the genera identified in this study have already been isolated from flowers of other plant species around the world. Some examples are the Hanseniaspora and Rhodotorula genera, which were also identified by [39] associated with orchid flowers, and the species of the genera Pichia, Meyerozyma, Kodamaea, Candida, and Wickerhamomyces, also isolated by [40] in flowers of wild and domestic plants. In addition to species of these genera, we isolated the dematiaceous yeast Exophiala dermatitidis; it is a species generally found in association with soil and dead plants worldwide [40, 41]; however, we could observe its presence in association with the flowers of Q. grandiflora, which increases and corroborates the distribution and occurrence of this species in different habitats. This species is melaninogenicus and commonly associated with phaeohyphomycosis, rarer respiratory and systemic infections in immunocompromised patients, and also with cases of osteomyelitis and septic arthritis [42, 43] and has already been isolated in cherry blossoms.

The two fungal species associated with Q. grandiflora and that did not fit phylogenetically as yeasts belong to the genera Anthracocystis and Myrmaecium. Although these taxa have been identified among yeast isolates and have yeast morphologies, according to [44], the sequences obtained from these yeasts were distributed in different subclades of Anthracocystis (Ustilaginales) and none of them can be directly associated with a teleomorphic species. Only one of these yeast anamorphs was assigned to a species, namely, Pseudozyma flocculosa. Following the recent practice of fusing anamorphic yeast species with a sexual stage (teleomorphic) under the oldest generic name, this new yeast taxon originated from recombination with Anthracocystis, resulting in Anthracocystis flocculosa. The phytopathogenic species Ustilago maydis and Ustilago esculenta, of the same family as Anthracocystis heteropogonicola (Ustilaginaceae), are known examples of dimorphic or yeast-like fungi since their cells can develop as yeasts or take on mycelial form according to the environmental signals received [45, 46]. No records of A. heteropogonicola growing as a dimorphic fungus were found, which may be a relevant contribution to the description of this species. A similar fact was observed for Myrmaecium rubrum (Valsariaceae), as no direct references to its growth as a dimorphic fungus were found. However, it is known that this characteristic exists for other species of the family, as observed by [47] for Valsaria insitiva, a phytopathogenic Dothideomycetes that causes perennial cancer in apples and almonds [48].

The isolates JGM1 (Priceomyces), PPA10 and PPA21 (Candida), PP18 (Kazachstania), and PPB13 (Wickerhamomyces) could not be identified to the species level. The molecular phylogenetic analyses based on sequences of the D1/D2 region of large subunit (26S) rDNA, as well as the internal transcribed spacer (ITS) region of these strains, showed that three of them fit as possible new species (PPA10, PPA18, and PPA21). According to available data from previously deposited sequences, PPA18 is a possible new species closely related to Kazachstania servazzii, Kazachstania africana, and Kazachstania aerobia (all with 96% similarity). In addition, the isolates PPA10 and PPA21 correspond to a possible new species of the Clavispora clade, called Candida ruelliae (85% identity) isolated originally from flowers of Ruellia sp. [49]. They also have genetic similarity with two other records of the genus Candida (both with 97% identity, unpublished data).

Corroborating the findings of the present work and in agreement with reports available in the literature, the genus Qualea is a very promising niche for the isolation of new mycological species. This can be seen in the bioprospecting studies carried out by [50], who described two new fungal species of the genus Uncinula associated with the leaves of Q. grandiflora and Qualea multiflora, respectively. In another study, Araújo et al. [51] describe three other new mycological species belonging to the genera Alternaria, Passalora, and Periconiella associated with the leaves of Q. grandiflora and Qualea parviflora.

Identifying different species of yeasts in Q. grandiflora flowers, in addition to reinforcing the status of this plant as a super host for biodiversity [50], is an incentive for conservation at the natural and social scales of the Cerrado biome. As an example, we identified species of the genera Meyerozyma and Wickerhamomyces. Species of these genera have physiological characteristics that allow them to perform with great performance in the spontaneous fermentation of distilled drinks [52]. This fact is of particular importance, given that the northern region of the state of Minas Gerais has worldwide prominence in the production of artisanal cachaças and sugar cane spirits. In this context, it is extremely desirable to characterize new yeast species in order to select the most efficient and productive strains. In line with the aforementioned data, studies of this nature come as an incentive to further efforts aimed at protecting the Cerrado biome, as well as its associated macro- and microbiodiversity.

Based on isolates per sampled area, the number of species was significantly higher for Area II (Figure 2), even considering that the sampling effort was equivalent in both locations. This significant difference in the number of species between the two areas may be correlated with climatic, ecological, socioenvironmental, or land use differences. Despite the more preserved aspect of the surrounding vegetation, Area I had a lower number of yeast species than Area II. Thus, it can be speculated that floral visitors (who act as dispersers of yeasts) of Area II have a limited variety of resources and have adopted generalist habits of foraging. Furthermore, to meet their nutritional and physiological needs, they need to visit a greater variety of plant species, resulting in greater dispersion and favoring an increase in the number of vectorized yeast species. Despite the significant difference in the number of species, some occurred in both areas, suggesting vertical transmission, floral visitors of the same species, or similarities in the floristic composition between the studied regions.

We emphasize that only four of all the species identified were shared between the two areas (Figure 2). Although premature, we deduce that these species may be more evolutionarily associated with Q. grandiflora in terms of occurrence in its flowers, response to visiting insects or specialized pollinators, and more related to foraging in that niche. More broadly, the isolates obtained there were separated into ascomycetes and basidiomycetes, with the phylum Ascomycota representing 91.3% of all the identified isolates. Considering that ascomycetic yeasts use the sugars offered by the different floral organs, and that they obtain energy preferably through the fermentative route, this percentage may be related to the supply of nectar and the consequent availability of sugars in the flowers of Q. grandiflora [53]. This observation corroborates other studies of yeast diversity in flowers, where the percentage of ascomycete yeasts was substantially higher than that of basidiomycete yeasts [14].

In order to demonstrate how the identified species are phylogenetically correlated, a maximum likelihood tree was generated (Figure 3). The highest bootstrap values of the phylogenetic analysis determined clades of isolates significantly corresponding to their respective genera. This demonstrates the phylogenetic proximity among the organisms identified in both sampled areas.

3.2. Screening for Extracellular Enzymatic Activity

Table 2 indicates the extracellular enzyme production profiles of the tested yeast isolates. Of these, 44.8% were able to produce at least one of the studied enzymes.

The use of solid media supplemented with specific substrates is a standard methodology for detecting enzyme production by microorganisms. With this method, the radial growth of colonies and hydrolysis halos are measured and, through this direct correlation, the EI is established. Bocchese et al. and Bandeira et al. [55, 56] recommend an EI ≥ 2.0 to show the ability of a microorganism to synthesize enzymes in a solid medium, whereas Ferreira et al. [57] consider strains with EI > 1.50 to be potential producers. Here, like [58], we consider EI ≥ 2.0 to infer the character of potential enzyme producer.

In this context, the best enzyme production among the microorganisms tested was observed with the cellulase test, since among the 20 isolates that presented hydrolysis halos in this test, eight presented EI ≥ 2.0, with values ranging from 2 to 6.44. For amylase production, halo formation was observed in 12 isolates with EIs ranging from 1.33 to 3.6. In the protease assays, EIs ranging from 1.21 to 3.0 were observed among the 12 isolates with some hydrolysis halo formation, while xylanase production was observed in only five isolates with general EIs ranging from 1.5 to 2.12. In the tannase test, enzyme production was also observed in only five isolates; however, the general EIs ranged from 1.5 to 2.0. None of the isolates tested produced pectinolytic enzymes.

When analyzing EIs by substrate tested, the highest values for enzymatic production in the cellulase test were expressed by the fungus M. rubrum and the yeast Papiliotrema laurentii (6.44 and 5.42, respectively). However, in terms of isolates, the species Coniochaeta rhopalochaeta had the highest number with cellulolytic activity. These findings corroborate the data presented by [59] who demonstrated the ability of Coniochaeta pulveracea to produce cellulases that degrade cellulose, suggesting that this species and its closest relatives (e.g., Coniochaeta boothii, C. rhopalochaeta, and Coniochaeta subcorticalis) are some of the main yeasts responsible for the hydrolysis of wood given their systems of extracellular hydrolytic enzymes. The EIs expressed by the isolates of P. laurentii are interesting since they were all > 2.0. This observation supports the fact that basidiomycetes are the most potent degraders of the cellulose polymer [60].

The highest EIs observed in the amylase test were for an isolate of the species C. rhopalochaeta and for an isolate of M. rubrum. In the xylanase test, production was low in both the total number of isolates and by individual isolate, with no EI ≥ 2.0. In the tannic acid test, however, it is interesting to note that of the five isolates that showed some EI value, four were of the species C. rhopalochaeta. Additionally, among all the isolates that showed some enzymatic activity, C. rhopalochaeta proved to be the species with a more heterogeneous application potential, as it was the only species among the tested strains that showed some enzymatic production in all the tests with positive results.

The analysis of enzymatic production by microorganism isolation area revealed that most of the isolates with positive results come from Area II, which can be correlated with the greater number of yeast species from that area (Figure 1). In addition to this variation between isolates by area, there was variation in enzymatic production among strains of the same species. For example, the EI values for the isolates of C. rhopalochaeta and Priceomyces sp., whose tests showed positive results, varied significantly. For this reason, regarding the data used from the Coniochaeta isolates, each one was isolated from different floral samples; for this reason, their enzyme profile was studied individually.

Some studies have demonstrated the ability of microbial strains to produce enzymes with potential for industrial use, and hydrolytic enzymes such as cellulases and amylases have stood out [61]. Although cellulases are produced especially by bacteria and filamentous fungi [62], we can see that epiphytic yeast isolated from flowers of Q. grandiflora also has the ability to synthesize enzymes of this complex. In addition, some studies with yeasts isolated from different substrates, such as flowers [63], fruits [61, 64], and decomposing leaves [65], have shown the potential that this group of microorganisms has for the production of cellulase.

Amylases and proteases form one of the main groups of industrial enzymes, given their broad spectrum of biotechnological applications [66]. Amylases are responsible for the hydrolysis of starch into maltose, glycogen, dextrins, and progressively smaller polymers composed of glucose units [67] and are closely related to fruit ripening and sweet taste. Proteases, on the other hand, are responsible for the proteolytic cleavage of peptides. Proteolytic enzymes produced by microorganisms can act in the hydrolysis of proteins of the cell membrane and wall of the host plants, facilitating penetration and infection [68]. In the present study, we observed that, among the strains with some enzymatic expression, the isolates of the species C. rhopalochaeta were the best producers of both amylase and protease, with emphasis on the amylolytic synthesis by the PPB4 isolate, which had an EI well above the reference value for potential producers.

Xylanases produced by microorganisms, as well as proteases, are commonly related to phytopathogens, since they have the ability to decompose hemicellulose, which is one of the main components of the plant cell wall [69]. Tannin acyl hydrolase or tannase, in turn, catalyzes the hydrolysis of tannins by producing gallic acid and glucose [70]. Tannins are part of the plant defense system against microorganisms, so the production of tannase can be considered a microbial counterattack [71]. Our results revealed that only 8.62% of the tested yeast isolates expressed some enzymatic production in the xylanase and tannase tests, respectively. This low enzymatic productivity may be related to the isolation loci of the microorganisms [72], as they were isolated from the surface of flowers and not from internal parts, as expected for enzyme-secreting microorganisms involved in the process of phytopathogenesis.

In addition to demonstrating one of the biochemical interfaces of the studied group of microorganisms, the analysis of the data obtained for the production of the aforementioned enzymes subsidizes knowledge of microbial bioprospecting for enzyme synthesis, which is an area under increasing expansion. Furthermore, despite the fact that enzymes from traditional sources, such as animals and plants, have greater acceptance in the world market, recent years have seen a trend to replace these enzyme sources with microbial sources, given not only their susceptibility to genetic manipulation but also their broad biochemical diversity [19]. Enzymes obtained by natural means, in addition to being less financially costly, are more easily approved by regulatory agencies than those obtained by synthetic or genetically modified routes [73]. Despite the fact that filamentous fungi are more exploited in terms of enzyme synthesis, yeasts have advantages over them because they are easy to handle, fast-growing, do not have great nutritional needs, and the cost of materials for cultivating them is less [74]. In addition, they are likely to present greater production through genetic manipulation, and their enzymes have the ability to catalyze various reactions, giving them not only biochemical versatility but also more specific application conditions [49, 75].

From this perspective, the search for new yeast strains in tropical biomes that produce enzymes is essential for the area of bioprocesses. In addition, the discovery of wild yeast lineages with potential for enzymatic production means finding new microbial functions for the described taxa, so that it is fully possible and desirable to expand the use of microorganisms in new bioassays aiming at directional selection, that is, as a form of wild species screening [76, 77]. While culture-independent methods have the potential to offer a clearer and more accurate understanding of yeast diversity in tropical forest biomes—and to determine if these ecosystems host greater diversity than temperate forests—research in this area remains limited. Despite the rich variety of habitats and intricate ecological interactions characteristic of tropical forests, relatively few studies have explored their yeast biodiversity [78]. Yeasts and yeast-derived products exhibit broad applicability across a wide range of industrial and biotechnological sectors. These include the production of fermented foods and beverages, pharmaceuticals, industrial enzymes, vitamins, carotenoids, lipids, steroids, polysaccharides, β-glucans, nucleotides, flavor compounds, biofuels, and various platform chemicals—also referred to as chemical building blocks. Moreover, yeasts play a role in environmental applications such as biocontrol and bioremediation [79].