Selection of yeasts from bee products for alcoholic beverage production

Abstract

The use of appropriate yeast strains allows to better control the fermentation during beverage production. Bee products, especially of stingless bees, are poorly explored as sources of fermenting microorganisms. In this work, yeasts were isolated from honey and pollen from Tetragonisca angustula (Jataí), Nannotrigona testaceicornis (Iraí), Frieseomelitta varia (Marmelada), and honey of Apis mellifera bees and screened according to morphology, growth, and alcohol production. Bee products showed to be potential sources of fermenting microorganisms. From 55 isolates, one was identified as Papiliotrema flavescens, two Rhodotorula mucilaginosa, five Saccharomyces cerevisiae, and nine Starmerella meliponinorum. The S. cerevisiae strains were able to produce ethanol and glycerol at pH 4.0–8.0 and temperature of 10–30 °C, with low or none production of undesirable compounds, such as acetic acid and methanol. These strains are suitable for the production of bioethanol and alcoholic beverages due to their high ethanol production, similar or superior to the commercial strain, and in a broad range of conditions like as 50% (m/v) glucose, 10% (v/v) ethanol, or 500 mg L⁻¹ of sodium metabisulfite.

Introduction

Non-traditional alcoholic beverages, including cider, mead, palm wines, and sorghum beer, are produced with yeasts available for the production of traditional beverages [1–4]. The use of appropriate strains enables better control of the production process, allowing standardization and improving the quality of these beverages [2]. Therefore, the search for new microorganisms to improve the production of non-traditional alcoholic beverages is necessary.

Bee products are abundant sources of microorganisms [5–10], however, they are poorly explored as sources of fermenting microorganisms for industrial processes. Moreover, these products are traditionally consumed by humans and considered to be safe and microbiologically stable [11]. This work brought novelty in isolating yeasts from food (pollen and honey) of brazilian native stingless bees such as Tetragonisca angustula (Jataí), Nannotrigona testaceicornis (Iraí), and Frieseomelitta varia (Moça-Branca) (Hymenoptera, Apidae, Meliponini). We have found that the diversity of microorganisms associated with these bees is greater than observed in Africanized Apis mellifera (Honeybee). The microbial biodiversity present in stingless bees is important and yet vulnerable since these bees are in danger of extinction due to loss of neotropical habitat by deforestation and excessive use of pesticides [12–16].

Yeasts used for alcoholic beverage production must be innocuous, osmotolerant, able to produce and tolerate ethanol [17], resistant to the sulfur compounds used for microbial control during fermentation [18], and genetically stable [19]. In this work, we isolated and identified yeasts from honey and pollen of stingless bees (social, honey-producing tropical bees of the family Apidae, which have a nonfunctional stinger) and Honeybee. They were selected according to their potential to develop a new ferment for the production of non-traditional alcoholic beverages.

Material and methods

Isolation and maintenance of the yeasts

Yeasts from stingless bees (Apidae, Meliponini) were isolated from pollen and honey samples from beehives of species Tetragonisca angustula (Jataí), Frieseomelitta varia (Moça-Branca), and Nannotrigona testaceicornis (Iraí). Samples of freshly harvested honey were also collected from Africanized Apis mellifera (Honeybee). The samples were collected in the main apiary at the Department of Entomology, Universidade Federal de Viçosa—MG. Commercial lyophilized S. cerevisiae r.f. bayanus E491 (Perdomini/Blastosel Delta®) was used as control.

First, 5 g of each sample was diluted in 45 mL of saline solution. The dilutions were plated in acidified PDA (Potato Dextrose Agar—BIOMEDH®) medium, pH 3.5, and incubated at 25 ± 3 °C for 48 h. Colonies presenting distinct morphologies were selected [20]. The isolates were cultivated in YEPG (0.5% yeast extract +1% peptone +2% glucose) medium at 25 ± 3 °C for 48 h. The yeasts that were able to grow under these conditions were concentrated by centrifugation, suspended in YEPG broth with 15% (v/v) of glycerol, and frozen at − 20 °C for subsequent application. Before each analysis, the isolates were activated in YEPG broth at 25 ± 3 °C for 12 h, followed by centrifugation at 2236×g for 10 min at 4 °C. Posteriorly, they were inoculated at approximately 10⁶ cells mL⁻¹.

Yeast identification

DNA extraction was performed as described by Lachance (1990), with modifications. DNA precipitation was performed with 100 μL of 95% ice-cold ethanol and 20 μL of sodium acetate solution 3 mol L⁻¹ [21]. The genomic DNAs were used as template for the sequencing of the D1/D2 regions of the ITS genes using the primers NL1 (5′-GCATATCAATAAGCGGAGGAAAAG-3′) and NL4 (5′-GGTCCGTGTTTCAAGACGG-3′) [22] by Macrogen® Inc. (Korea). The sequences were analyzed using the CLC Genomics® software system (QIAGEN Bioinformatics) and aligned to sequences downloaded from the GenBank database using the Blastn Suite software system (NCBI National Center for Biotechnology Information/US Department of Health & Human Services—https://www.ncbi.nlm.nih.gov/blast/).

Preliminary selection of the isolates

The isolates were submitted to stress conditions that commonly occur during fermentative processes for alcoholic beverage production: (I) 4% (v/v) ethanol, the lowest alcohol content in beverages from fermented fruits, cider, and mead; (II) 20% (m/v) glucose; and (III) 4% (v/v) ethanol +20% (m/v) glucose. The base culture medium was YEP (0.5% yeast extract +1% peptone). Control contained 2% glucose. The microbial growth was monitored (OD₆₀₀) during 72 h at 25 ± 3 °C. The treatments were performed in triplicates, and the results were expressed as a percentage of growth relative to the commercial yeast. Yeasts that showed a final OD₆₀₀ of at least 50% in comparison to the standard were selected.

Tolerance to stress conditions

The selected strains were screened for their ability to grow in media containing glucose, ethanol, or sodium metabisulfite at high concentrations. They were inoculated in the respective medium in 96-well plates and incubated at 25 ± 3 °C for 24 h. The microbial growth (OD₆₀₀) was monitored at intervals of 30 min. To assess the osmotolerance of the strains, media composed by 10, 20, 30, 40, and 50% (m/v) of glucose were prepared, whereas ethanol tolerance was evaluated with media containing 5, 10, 15, and 20% (v/v) of ethanol and 2% of glucose. Media for the sulfite tolerance test contained 100, 250, and 500 mg L⁻¹ of sodium metabisulfite (Trademark®). The yeasts specific growth rate (h⁻¹) was calculated (Eq. 1) for each condition, and three strains were selected for technological characterization

$$\mathrm{lnN}={\mathrm{lnN}}_0+\mathrm{\mu }.\mathrm{t}$$ 1

Where N = concentration of viable cells (cells mL⁻¹); N₀ = concentration of viable cells at initial time (cells mL⁻¹); t = elapsed time (h); μ = specific growth rate (h⁻¹)

Technological characterization of the isolates

The three selected strains were tested for their ability to grow and ferment under different pH and temperature conditions in 10-mL test tubes. The pH values of YEPG broths were adjusted to 2.5, 4.0, and 5.5 by adding citrate buffer 0.1 mol L⁻¹ (SIGMA®), or to 8.0 and 9.5 by adding hydroxymethyl amino methane buffer (INLAB®). The solutions were inoculated and incubated at 25 ± 3 °C for 12 h. For the temperature test, each yeast was inoculated in YEPG broth pH 7.0 and incubated at 10, 15, 20, 25, and 30 °C. The yeast growth was monitored at 600 nm, at intervals of 5 h during 25 h, and the specific growth rates (Eq. 1) and the ethanol yields (Eq. 2) were calculated.

$$Y_{\left(\mathit{et}/\mathit{glu}\right)}=\frac{P_f-P_i}{S_i-S_f}\ast 100$$ 2

Where P_f = product concentration at the end of fermentation (g L⁻¹); P_i = product concentration at the beginning of fermentation (g L⁻¹); S_i = substrate concentration at the beginning of fermentation (g L⁻¹); S_f = substrate concentration at the end of fermentation (g L⁻¹).

The effect of pH and temperature on the production of ethanol, methanol, glycerol, and acetic acid was assessed by high-performance liquid chromatography (HPLC) (Shimadzu 20-AT), equipped with refractive index (UV 210 nm RI). The samples were injected in an ion exclusion column Aminex® HPX-87H (300 × 7.8 mm), at 50 °C, with 5 mM H₂SO₄ (VETEC®) as mobile phase and flow rate of 0.6 mL min⁻¹. Glucose (SYNTH®), ethanol (VETEC®), methanol (VETEC®), glycerol (SYNTH®), and acetic acid (VETEC®) were used as patterns to produce a calibration curve to correlate the retention time and the concentration [23].

Statistical analysis of the results

The data were analyzed by ANOVA and Tukey’s test at 5% probability, using MINITAB 2010 [24, 25]. The graphs were made using the OriginPro® software system [26].

Results and discussion

The stingless bee Iraí and Jataí were the most abundant sources of cultivable yeasts in the YEPG broth. The pollen collected from these bees presented 5.34 and 5.20 log CFU mL⁻¹, respectively, while their honey presented 4.34 and 3.22 log CFU mL⁻¹, respectively. Meanwhile, the pollen from Moça-Branca presented 2.80 log CFU mL⁻¹ and no yeasts could be isolated from its honey. Finally, the honey from honeybee contained only 1.59 log CFU mL⁻¹ of cultivable yeasts. Stingless bees incorporate fermenting microorganisms to the harvested pollen to metabolize it, allowing its conservation and developing particular sensorial characteristics in this product [27] which explains the high number of yeasts isolated from this pollen. Similarly, these bees add acid-producing microorganisms to their honey to increase its acid content since the moisture of the honey from these species can reach up to 31% [28, 29]. On the other hand, A. mellifera honey has only 17.23% moisture and acidity of 14.53 m e q kg⁻¹, conditions hostile to microorganisms [30], which justifies the low number of colonies found in this product.

Initially, 55 wild yeasts were isolated: 15 from Jataí honey, 14 from Jataí pollen, eight from Iraí honey, 12 from Iraí pollen, three from Moça-Branca pollen, and three from Honeybee honey. However, most of these yeasts were not able to grow in YEPG broth. Nineteen isolates presenting distinct morphologies were selected: nine from Jataí honey (JM4, JM5, JM7, JM9, JM10, JM11, JM12, JM13, JM15), four from Jataí pollen (JP3, JP6, JP12, JP14), one from Iraí honey (IM8), four from Iraí pollen (IP5, IP6, IP9, IP12) and one from Honeybee honey (M1). From these, nine were identified as Starmerella meliponinorum, two as Rhodotorula mucilaginosa, five as Saccharomyces cerevisiae, one as Papiliotrema flavescens, and two as Candida apicola (Table 1). Some of these species have been isolated from similar products in Brazil. Strains of S. meliponinorum have been found in material discarded by bees in the beehive, in pollen, honey, and propolis from Jataí bees, in honey from Melipona quadrifasciata, and in adults of Melipona rufiventris and Trigona fulviventris [31]. Strains of Cryptococcus (Papilotrema) flavescens, originally called C. laurentii, can be found in soil, leaves, and swamps [32, 33] and can be able to produce ethanol from glucose [34] and high amounts of lipids from dairy byproducts [35]. The species R. mucilaginosa comprises yeasts able to produce carotenoids [36] or fatty acids, products of the biodiesel industry [37]. Some strains control gray filamentous fungi in fruit post-harvest [38]. This species has been already described in A. mellifera honey in Portugal [39]. Finally, the yeast C. apicola, isolated from A. mellifera honey, is a non-Saccharomyces species associated with the development of aroma in wine production [40].

Table 1.

Identification of the isolated yeasts by sequencing portion of the 16S rRNA

Strain	Species	E-value	% Similarity	GenBank accession no.
JM4	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JM5	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JM7	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JM9	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JM10	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JM11	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JM12	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JM13	Saccharomyces cerevisiae strain YI1	0.0	98%	KX428527.1
JM15	Starmerella meliponinorum strain SB147	0.0	100%	AB568334.1
JP3	Papiliotrema flavescens, strain DMKU-CE139	0.0	100%	LC178830.1
JP6	Starmerella meliponinorum strain CBS 9117	0.0	99%	KJ630492.1
JP12	Rhodotorula mucilaginosa strain SK0809R 2	0.0	100%	KP990660.1
JP14	Saccharomyces cerevisiae strain S2-6	0.0	100%	KU862641.1
IM8	Saccharomyces cerevisiae isolate yazhong W3 N4	1,00E-35	97%	KX119945.1
M1	Candida apícola	0.0	100%	U45703.1
IP5	Candida apicola strain BEE-2b	0.0	99%	KT718105.1
IP6	Rhodotorula mucilaginosa isolate SM6-1	0.0	100%	KU316790.1
IP9	Saccharomyces cerevisiae strain4	0.0	96%	HM107797.1
IP12	Saccharomyces cerevisiae strain V3	9.00E-177	86%	KX428526.1

IM8, Iraí honey strain 8; JP9, Iraí pollen strain 9; JP14, Jatai pollen strain 14

Preliminary selection of strains

The source from which the yeasts were isolated was not a determinant for yeast resistance to stress. Yeasts are considered poorly resistant when they are not able to grow at concentrations of 20% (m/v) hexoses, while the most resistant grow in media containing up to 70% (m/v) hexoses [41]. Strains of S. cerevisiae are able to grow up to 15% (v/v) ethanol, whereas some non-Saccharomyces species can multiply in media containing up to 10% (v/v) ethanol and 20% (m/v) glucose [41]. Ten strains isolated in this work (C. apicola (IP5, M1), S. meliponinorum (JM4, JM10, JM11), S. cerevisiae (JM13, IP12, JP14), and R. mucilaginosa (IP6, JP12)) were able to grow in 4% (m/v) ethanol similarly to the commercial yeast (Fig. 1a). Among these, seven (C. apicola M1, S. meliponinorum (JM4, JM10), S. cerevisiae (JM13, JP14, IP12) and R. mucilaginosa IP6) also multiply even when submitted to high osmolarity (Fig. 1b).

Fig. 1.

Relative resistance of the isolated yeasts strains to a ethanol 4% (v/v), b glucose 20% (m/v), and c 4% (v/v) ethanol +20% (m/v) glucose (cross-stress). The values refer to the final population of each strain relative to the final population of the commercial strain (S) after incubation in culture medium during 72 h at 25 ± 3 °C. The error bars refer to the standard deviation of the measurements. Means followed by the same letters in same graphic does not differ significantly by de Tukey´s test (p < 0.05). S, commercial Saccharomyces cerevisiae; M, Apis mellifera honey; IM, Nannotrigona testaceicornis honey; IP, Nannotrigona testaceicornis pollen; JP, Tetragonisca angustula pollen; Sm, Starmerella meliponinorum; Sc, Saccharomyces cerevisiae; Pf, Papiliotrema flavescens; Rm, Rhodotorula mucilaginosa; Ca, Candida apícola

Throughout the beverage production process, the yeast is simultaneously subjected to thermal, osmotic, and oxidative stress [17]. The worts used for the production of alcoholic beverages have 10 to 25% (m/v) sugars, mainly glucose, fructose, and sucrose. Under these conditions, ethanol is produced up to 4 to 16% (v/v). Ethanol increases the porosity of cell membrane through lipid solubilization, while glucose causes water molecules to escape from inside the cell [17], thus affecting the membranes fluidity, protein stability, and energy production, inhibiting cell multiplication [42]. Therefore, in crossed stress, the process of dehydration is higher. In general, when the conditions were simultaneously imposed, all strains isolated from the bee products presented lesser growth when compared to the commercial strain (Fig. 1c). Six strains of three species (C. apicola (M1 and IP5), S. cerevisiae (JM13, JP14 and IP9), and R. mucilaginosa IP6) presented at least 50% of the growth of the commercial yeast and were selected for further testing. Saccharomyces cerevisiae IM8 was also selected since it produced pleasant aroma during cultivation (not shown).

Tolerance tests

The addition of 10% (m/v) glucose to the culture medium did not cause significant deviations in the yeasts’ growth kinetic (Fig. 2a), although they differed from the commercial strain during the adaptation period. At higher glucose concentrations, S. cerevisiae JM13 presented reduced growth, while JP14 and IP9, which belong to the same species, presented higher growth rates even when compared to the commercial strain (Fig. 2a–f).

Fig. 2.

Growth curve of the seven pre-selected isolates and the commercial yeast in YEP medium at different glucose concentrations. The dots represent values’ means and error bars refer to the standard deviation of the measurements. S, commercial Saccharomyces cerevisiae; M1, Apis mellifera honey strain 1; IM8, Nannotrigona testaceicornis honey strain 8; IP5, IP6, and IP9, Nannotrigona testaceicornis pollen strain 5, 6, and 9, respectively; JP14, Tetragonisca angustula pollen strain 14

The isolates were able to withstand up to 10% (v/v) ethanol, mainly S. cerevisiae JP14 and IP9, while the commercial strain resisted up to 15% (v/v) but with longer lag phase. The concentration of 20% (v/v) was limiting for all yeasts (Fig. 3a–d). In general, yeasts of the genus Saccharomyces are tolerant of media containing ethanol, especially the species S. cerevisiae [43]. During fermentation there is an increase in the concentration of fermentation products and some yeasts can respond to ethanol stress by increasing their tolerance [44]. Cells are able to increase the levels of unsaturated fatty acids and ergosterol in their membranes, or to express factors to repair denatured proteins [44]. The medium composition also affects yeast resistance to ethanol [44]. Under the conditions used in this work, the strains S. cerevisiae IM8, C. apicola IP5, S. cerevisiae IP9, and S. cerevisiae JP14 have adapt and multiply in medium with up to 10% (v/v) ethanol (Fig. 3b) and were considered satisfactory for the production of fermented beverages, in which ethanol concentration varies between 4 and 18% (v/v) [45]. On the other hand, yeasts IP6, JM13, and M1 presented reduced growth rate and were considered as less suitable for alcoholic beverage production.

Fig. 3.

Growth curve of the seven pre-selected isolates and the commercial yeast in YEP medium at different ethanol and sodium metabisulfite concentrations. The dots represent values’ means and error bars refer to the standard deviation of the measurements. S, commercial Saccharomyces cerevisiae; M1, Apis mellifera honey strain 1; IM8, Nannotrigona testaceicornis honey strain 8; IP5, IP6, and IP9: Nannotrigona testaceicornis pollen strain 5, 6, and 9, respectively; JP14, Tetragonisca angustula pollen strain 14

Sodium or potassium metabisulfite is an important additive used to avoid bacterial growth during the production of fermented beverages [46]. Thus, the resistance to sulfur compounds is essential when selecting yeasts for this purpose. Sulfite decreases ATP production by the cell and, at high concentrations, accumulates in the cytoplasm and dissociates to sulfurous acid, which leads to cell death [47]. As a defense mechanism, some yeasts synthesize an enzyme, sulfite reductase, which transforms sulfite [47]. The strains R. mucilaginosa IP6, S. cerevisiae JM13, and C. apicola M1 presented tolerance to metabisulfite only at concentrations higher than 500 mg L⁻¹, while the S. cerevisiae strains JP14, IM8, and IP9 grew at 100 and 250 mg L⁻¹ (Fig. 3e–g). These concentrations are ideal for the production of wines and mead, whose worts are commonly added of metabisulfite from 80 to 235 mg°L⁻¹ [17, 48]. Thus, the S. cerevisiae strains IM8, JP14, and IP9 were selected for technological characterization.

Technological characterization

The potential of the S. cerevisiae strains IM8, JP14, and IP9 for fermentative processes was corroborated by the assessment of their growth and production of compounds under different pH and temperature conditions. In general, the yeasts exhibited the highest specific growth rate at 20 °C and their growth kinetic was less affected by lower temperatures in comparison to the commercial yeast, which presented maximum growth at 25 ± 3 °C. Moreover, no significant acetic acid or methanol production was verified in the fermented medium (< 1 g L⁻¹) for all treatments.

The specific growth rate of the strain IM8 was not affected by the medium pH, except at pH 2.5 (Table 2; Table 3). The commercial strain and the strain JP14 presented maximum growth at pH 7.0, although they still maintained replicative potential in acidic media. Under this condition, the commercial yeast presented higher ethanol yield, but its potential for ethanol production was lower than that of the yeasts isolated from stingless bees (Table 3).

Table 2.

The specific growth rate (μ_max) and ethanol yield (Y_(et/glue)) of the isolated strains in medium with different pH and temperature

Strain		pH
		4.0	5.5	7.0	8.0	9.0
S	μmax (h−1)	0.13 ± 0.01 b	0.13 ± 0.02 b	0.25 ± 0.03 a	0.11 ± 0.06 b	0.04 ± 0.03 c
	Y(et/glu) (%)	12.53 ± 0.08 ab	10.90 ± 2.58 b	15.80 ± 0.03 a	12.89 ± 1.80 ab	12.10 ± 2.29 ab
IM8	μmax (h−1)	0.10 ± 0.03 a	0.17 ± 0.02 a	0.12 ± 0.00 a	0.12 ± 0.05 a	0.16 ± 0.10 b
	Y(et/glu) (%)	39.44 ± 6.17 a	23.00 ± 0.43 b	23.72 ± 3.91 b	24.32 ± 2.84 b	17.39 ± 1.73 b
IP9	μmax (h−1)	0.17 ± 0.01 ab	0.19 ± 0.07 a	0.26 ± 0.07 a	0.04 ± 0.01 c	0.05 ± 0.00 bc
	Y(et/glu) (%)	34.45 ± 3.58 ab	40.44 ± 8.62 a	21.65 ± 1.61 abc	19.21 ± 5.61 bc	14.20 ± 3.33 c
JP14	μmax (h−1)	0.03 ± 0.01 b	0.07 ± 0.01 b	0.20 ± 0.10 a	0.08 ± 0.01 b	0.08 ± 0.01 ab
	Y(et/glu) (%)	36.18 ± 6.15 a	22.03 ± 0.23 b	23.92 ± 1.10 b	24.09 ± 0.13 b	18.01 ± 2.04 b
Strain		Temperature (°C)
		10	15	20	25	30
S	μmax (h−1)	0.20 ± 0.01 b	0.20 ± 0.00 a	0.18 ± 0.00 b	0.25 ± 0.03 a	0.18 ± 0.01 b
	Y(et/glu) (%)	21.03 ± 3.24 ab	19.44 ± 3.66 ab	25.38 ± 1.42 a	18.30 ± 4.28 ab	14.99 ± 3.99 b
IM8	μmax (h−1)	0.18 ± 0.03 a	0.18 ± 0.03 a	0.20 ± 0.01 a	0.12 ± 0.01 b	0.11 ± 0.00 b
	Y(et/glu) (%)	14.33 ± 4.70 b	19.69 ± 0.75 ab	18.72 ± 2.37 ab	23.79 ± 3.93 a	17.61 ± 2.81 ab
IP9	μmax (h−1)	0.23 ± 0.14 a	0.29 ± 0.01 a	0.37 ± 0.12 a	0.23 ± 0.14 a	0.16 ± 0.10 a
	Y(et/glu) (%)	15.47 ± 2.07 a	7.37 ± 1.58 b	17.60 ± 3.11 a	21.71 ± 1.62 a	8.36 ± 2.92 b
JP14	μmax (h−1)	0.20 ± 0.16 a	0.14 ± 0.02 a	0.14 ± 0.02 a	0.29 ± 0.02 a	0.12 ± 0.01 a
	Y(et/glu) (%)	24.13 ± 3.31 a	21.51 ± 1.18 a	24.42 ± 2.51 a	23.99 ± 1.10 a	22.91 ± 3.88 a

Data represent the mean ± SD. Values followed by the same letters in the same column (same test) do not differ significantly by the means test Tukey (p < 0.05). S, commercial Saccharomyces cerevisiae; IM8, Iraí honey strain 8; JP9, Iraí pollen strain 9; JP14, Jatai pollen strain 14

Table 3.

Glucose consumption and ethanol, glycerol, and methanol production by commercial and isolated yeasts strains in culture medium at different pH and temperature conditions after 24 h of fermentation

Temperature
Strain	Temp. (°C)	Glucose (g L−1)				Ethanol (g L−1)				Glycerol (g L−1)				Methanol (g L−1)
S	10	17.69	±	0.02	a	2.34	±	0.75	bcdefg	0.31	±	0.02	e	0.26	±	0.02	abcd
IM8		17.63	±	0.91	a	1.28	±	0.70	fgh	0.37	±	0.05	de	0.00	±	0.00	e
IP9		16.71	±	0.51	ab	0.28	±	0.30	h	0.15	±	0.02	e	0.00	±	0.00	e
JP14		11.87	±	0.02	d	3.18	±	0.85	abcdef	0.43	±	0.02	de	0.30	±	0.00	abcd
S	15	17.69	±	2.48	a	3.31	±	0.12	abcdef	0.97	±	0.08	ab	0.22	±	0.00	abcd
IM8		17.60	±	1.67	a	1.62	±	1.20	efgh	0.95	±	0.05	abc	0.21	±	0.01	abcd
IP9		16.45	±	0.35	abc	0.21	±	0.68	h	0.45	±	0.05	cde	0.19	±	0.03	cd
JP14		11.14	±	0.06	de	3.18	±	0.25	abcdef	0.95	±	0.09	abc	0.20	±	0.03	abcd
S	20	7.88	±	1.14	e	4.46	±	0.36	a	0.47	±	0.03	bcde	0.19	±	0.04	cd
IM8		17.69	±	2.14	a	1.92	±	1.27	defgh	0.40	±	0.01	de	0.43	±	0.06	ab
IP9		17.58	±	0.89	a	2.10	±	0.69	cdefgh	0.31	±	0.02	e	0.14	±	0.01	cd
JP14		13.72	±	0.02	bcd	3.86	±	0.13	abcd	0.46	±	0.04	cde	0.19	±	0.01	bcd
S	25	8.24	±	0.00	e	3.22	±	0.05	abcdef	0.36	±	0.07	de	0.23	±	0.02	abcd
IM8		2.78	±	2.51	f	4.18	±	0.33	ab	0.48	±	0.05	bcde	0.44	±	0.23	a
IP9		17.68	±	1.31	a	3.10	±	1.14	abcdef	0.45	±	0.01	cde	0.33	±	0.25	abcd
JP14		17.06	±	0.07	ab	3.53	±	0.60	abcde	0.42	±	0.03	de	0.21	±	0.01	abcd
S	30	13.30	±	0.02	cd	2.63	±	0.86	abcdefg	1.08	±	0.07	a	0.25	±	0.02	abcd
IM8		7.93	±	0.03	e	2.87	±	0.43	abcdef	0.86	±	0.69	abcd	0.45	±	0.03	a
IP9		2.24	±	0.01	f	0.65	±	0.36	gh	0.50	±	0.03	bcde	0.08	±	0.05	d
JP14		17.65	±	0.16	a	4.02	±	0.02	abc	1.00	±	0.06	a	0.18	±	0.01	cd
pH
Strain	pH	Glucose (g L−1)				Ethanol (g L−1)				Glycerol (g L−1)				Methanol (g L−1)
S	4.0	9.19	±	2.48	ef	1.57	±	1.04	d	0.90	±	0.16	ab	0.05	±	0.02	a
IM8		10.83	±	0.56	def	4.29	±	0.84	a	1.42	±	0.21	ab	0.09	±	0.05	a
IP9		7.13	±	2.65	fg	3.51	±	1.30	abcd	1.76	±	0.48	ab	0.08	±	0.02	a
JP14		11.69	±	0.75	cde	4.21	±	0.38	ab	0.77	±	0.05	ab	0.08	±	0.02	a
S	5.5	16.39	±	0.13	ab	1.79	±	0.43	cd	0.83	±	0.61	ab	0.42	±	1.44	a
IM8		12.60	±	0.41	bcde	2.90	±	0.06	abcd	1.02	±	0.75	ab	0.02	±	0.00	a
IP9		3.60	±	2.50	g	1.34	±	0.89	d	2.85	±	0.95	a	0.76	±	0.68	a
JP14		13.58	±	2.33	abcd	2.99	±	0.48	abcd	2.11	±	0.34	ab	0.72	±	0.01	a
S	7.0	17.61	±	0.00	a	3.22	±	0.75	abcd	0.36	±	0.07	b	0.23	±	0.02	a
IM8		17.61	±	0.00	a	4.18	±	0.69	ab	0.48	±	0.05	b	0.44	±	0.23	a
IP9		17.60	±	0.01	a	3.81	±	0.29	abc	0.45	±	0.01	b	0.33	±	0.25	a
JP14		17.61	±	0.00	a	4.21	±	0.19	ab	0.42	±	0.03	b	0.21	±	0.01	a
S	8.0	17.61	±	0.00	a	2.27	±	0.32	abcd	0.42	±	0.03	b	0.21	±	0.01	a
IM8		17.41	±	0.35	a	4.23	±	0.44	ab	0.74	±	0.18	ab	0.11	±	0.04	a
IP9		17.61	±	0.00	a	3.38	±	0.99	abcd	1.14	±	1.43	ab	0.56	±	0.04	a
JP14		17.61	±	0.00	a	4.24	±	0.02	a	0.87	±	0.65	ab	0.18	±	0.17	a
S	9.0	16.84	±	0.28	a	2.03	±	0.35	bcd	0.80	±	0.06	ab	0.25	±	0.42	a
IM8		17.12	±	0.84	a	2.99	±	0.41	abcd	1.05	±	0.65	ab	0.40	±	0.17	a
IP9		16.42	±	2.06	ab	2.30	±	0.48	abcd	1.11	±	0.59	ab	0.64	±	0.26	a
JP14		15.64	±	1.77	abcd	2.84	±	0.65	abcd	2.00	±	2.00	ab	0.39	±	0.11	a

Data represent the mean ± SD. Values followed by the same letters in the same column (same test) do not differ significantly by the means test Tukey (p < 0.05). S, commercial Saccharomyces cerevisiae; IM8, Iraí honey strain 8; JP9, Iraí pollen strain 9; JP14, Jatai pollen strain 14

Yeasts IM8 and JP14 presented higher ethanol yields at pH 4.0, while IP9 produced ethanol in higher quantity at pH 5.0 and at broad temperature range (Table 3). These pH conditions are common in fruit-, vegetable-, and honey-based worts [45]. The yeast IM8 presented high ethanol production at pH 4.0, 7.0, and 8.0. However, at pH 9.0, ethanol production was inhibited in all strains (Table 2, Table 3). At pH 8.0 and 9.0, the isolates consumed all the glucose present in the medium. However, this high consumption was not proportionally converted to ethanol.

IM8 also produced constant ethanol yields between 15–30 °C, while fermentation by IP9 was stable, but slow between 10–25 °C (Table 2, Table 3). Fermentation by JP14 was not affected by temperature, which increases the options of products and processes in which it could be employed. The higher glucose consumption occurred at 25 °C, with increased ethanol production by the yeasts (Table 3). On the other hand, the highest ethanol production was observed only from 20 °C for the commercial strain. Ethanol production at room temperature is aimed during bioethanol production, either due to the costs of refrigeration or the possibility of simultaneous cellulose hydrolysis and fermentation [49]. For beverage production, the use of these temperatures favors the development of mesophilic contaminants and induces the production of higher alcohols by the yeast [18]. However, at low temperatures, yeast metabolism is reduced, which can lead to low ethanol yields [50]. In this context, JP14 has competitive advantages due to the high yields of ethanol obtained at 10 and 15 °C, when compared with the other strains.

The presence of acetic acid in alcoholic beverages indicates wort contamination, mainly by yeasts of the species Brettanomyces bruxellensis/Dekkera bruxellensis and bacteria of the genus Acetobacter spp. [51]. Thus, the failure to detect this compound in the wort indicates the absence of contamination during the fermentation process. Methanol was produced at concentrations between 0.1 and 0.5 g L⁻¹ in media at pH 8.0 and 9.0 (Table 3). Methanol is mainly formed from pectin, which is present in fruits and vegetables. Pectin is transformed by the enzyme pectin-methyl esterase [52]. The medium used in this study contained no pectin, so the production of this compound was low and occurred exclusively in alkaline media.

Under conditions similar to those of worts traditional for alcoholic beverages production (pH 3.5–4.5), sugars were efficiently converted by all strains, with low glycerol production and absence of methanol (Table 3). The highest glycerol concentration was produced at pH 5.5 (Table 3). This product confers softness and viscosity to beverages and is desirable at the concentration of 1–15 g L⁻¹ [53]. Therefore, the amount of glycerol produced by all strains at pH 4.0 and 5.0 is ideal for alcoholic beverage production.

In general, the strain S. cerevisiae IP9 presented sensitivity to sodium metabisulfite and produced almost no ethanol at 10, 15, and 30 °C. On the other hand, the S. cerevisiae IM8 presented high specific growth rate at lower temperatures and produced ethanol in concentrations similar to those of the commercial strain at 25 and 30 °C and over a wide pH range. Moreover, IM8 was also able to grow in medium containing up to 40% (m/v) glucose or 500 mg L⁻¹ sodium metabisulfite, and its growth was not affected by ethanol at the concentrations tested. Finally, glucose consumption and ethanol production by S. cerevisiae JP14 at different temperatures were similar to those of the commercial yeast. This similarity was also observed throughout the tests of resistance and technological characterization. The dynamicity and high ethanol yields of this strain at different temperatures make it the ideal candidate to research on fermentation processes aiming at biofuel or alcoholic beverage production.

Conclusions

Products from stingless bees (Jataí and Iraí) have been identified as satisfactory sources of fermentative yeast species, including strains of the species S. cerevisiae, which presented the best potential for ethanol production between the 55 isolates. These strains are promising candidates for researches regarding the production of non-traditional alcoholic beverages since they can grow in synthetic media containing up to 50% (m/v) glucose, 10% (v/v) ethanol, or 500 mg L⁻¹ of sodium metabisulfite. They can also produce ethanol and glycerol under conditions commonly found in worts used for the production of wines, mead, and fermented fruits, especially at pH 4.0–5.0. The yeast behavior depends on the wort composition and the production process. So, specific evaluations are required for each aimed product.

Funding information

This work was supported by Fundação Arthur Bernardes (FUNARBE), Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).

Compliance with ethical standards

Disclaimer

The funders had no role in the study design, data collection, analysis, publishing decisions, or preparation of this manuscript.