Expression patterns of Mal genes and association with differential maltose and maltotriose transport rate of two Saccharomyces pastorianus yeasts

ABSTRACT

Beer brewing is a well-known process that still faces great challenges, such as the total consumption of sugars present in the fermentation media. Lager-style beer, a major worldwide beer type, is elaborated by Saccharomyces pastorianus (Sp) yeast, which must ferment high maltotriose content worts, but its consumption represents a notable problem, especially among Sp strains belonging to group I. Factors, such as fermentation conditions, presence of maltotriose transporters, transporter copy number variation, and genetic regulation variations contribute to this issue. We assess the factors affecting fermentation in two Sp yeast strains: SpIB1, with limited maltotriose uptake, and SpIB2, known for efficient maltotriose transport. Here, SpIB2 transported significantly more maltose (28%) and maltotriose (32%) compared with SpIB1. Furthermore, SpIB2 expressed all MAL transporters (ScMALx1, SeMALx1, ScAGT1, SeAGT1, MTT1, and MPHx) on the first day of fermentation, whereas SpIB1 only exhibited ScMalx1, ScAGT1, and MPH2/3 genes. Some SpIB2 transporters had polymorphic transmembrane domains (TMD) resembling MTT1, accompanied by higher expression of these transporters and its positive regulator genes, such as MAL63. These findings suggest that, in addition to the factors mentioned above, positive regulators of Mal transporters contribute significantly to phenotypic diversity in maltose and maltotriose consumption among the studied lager yeast strains.

IMPORTANCE

Beer, the third most popular beverage globally with a 90% market share in the alcoholic beverage industry, relies on Saccharomyces pastorianus (Sp) strains for lager beer production. These strains exhibit phenotypic diversity in maltotriose consumption, a crucial process for the acceptable organoleptic profile in lager beer. This diversity ranges from Sp group II strains with a notable maltotriose-consuming ability to Sp group I strains with limited capacity. Our study highlights that differential gene expression of maltose and maltotriose transporters and its upstream trans-elements, such as MAL gene-positive regulators, adds complexity to this variation. This insight can contribute to a more comprehensive analysis needed to the development of controlled and efficient biotechnological processes in the beer brewing industry.

INTRODUCTION

Beer represents a highly popular fermented beverage globally (1), with production reaching 189 million kiloliters in 2022 (https://www.statista.com). The important role of microorganisms in fermented food biotechnology depends on the enzymatic conversion of raw substrates to human edible and nutritive products, especially in beer brewing (2, 3). Yeasts are crucial in beer brewing, with Saccharomyces cerevisiae (Sc) used for ale-style beer, with fermentation temperature ranging from 15°C to 26°C, and Saccharomyces pastorianus for lager-style beer, with fermentation temperatures between 5°C and 16°C (4, 5). S. pastorianus (Sp), an interspecies hybrid of S. cerevisiae (Sc), and a cryotolerant yeast, Saccharomyces eubayanus (Se) (6–8), is divided into Saaz type (group I) and Frohberg type (group II) based on chromosome number content and sugar consumption (9). There are two hypotheses that explain the lager yeast’s origin: (i) the existence of two independent hybridization events involving different yeasts, Sc, and Se (6, 10, 11); and (ii) both groups of lager yeast share a single hybridization event, and the evolution resulting from that hybridization shaped genomic divergences between both groups (12). The prevalence of Sp in the brewing industry implies that this hybrid possesses a selective advantage over its parent strains. Key characteristics contributing to this advantage include enhanced biological fitness under stress conditions and efficient maltotriose consumption (10, 13, 14). However, physiological differences between lager yeast highlight the defective maltotriose uptake seen in some members of group I (9). The transport of maltose and maltotriose stands as a bottleneck for achieving effective fermentation (15), given their substantial presence, accounting for 50%–60% and 15%–20% of the fermentable sugars in a standard brewing wort, respectively (16, 17). Despite their abundance, the transport of maltose and maltotriose is regulated by a carbon catabolite repression mechanism, where glucose is prioritized over maltose and maltotriose (18–20). Consequently, maltose is more easily consumed by most brewing yeasts, leaving maltotriose more prevalent in later stages of fermentation, where some yeast strains exhibit a deficiency in its uptake (11, 15, 21, 22).

Different factors are involved in deficient maltotriose consumption phenotype. These include (i) the stressful conditions on the last days of fermentation (23), (ii) the presence or absence of specific transporters for maltotriose, such as AGT1 and MTT1/MTY1 (24–26), and (iii) regulatory differences in maltose and maltotriose transporter genes (27, 28). Furthermore, the emergence of chimeric genes, resulting from genomic recombination, are also involved in the mechanism of generating variation in closely related yeast, leading to the development of highly efficient maltotriose transporters (29, 30).

As set before, alpha glycoside consumption varies among lager yeast, and its study lack global picture as in RNA-seq transcriptomic level analysis. Here, we investigate the sugar transport system in two lager yeasts, specifically SpIB1, characterized by a limited maltotriose uptake, and SpIB2, distinguished by its effective maltotriose transport. We investigate if fermentations conditions and/or differences in the genetic architecture of MAL genes are associated with this phenotypic diversity of maltose and maltotriose consumption. By integrative analysis of transcriptome, qualitative detections of MAL genes and transcripts, and biochemical determination of maltose and maltotriose transport rate, we discuss molecular events, such as polymorphic regions in maltose and maltotriose transporters, and differential MAL gene expression patterns are associated with the underlying phenomena involved in the phenotypic diversity of sugar utilization in two lager yeasts. To our knowledge, this work shows the first RNA-seq data approach to get insight into sugar consumption in two lager yeasts, enabling us to detect the regulatory MAL genes as a key feature in this difference.

RESULTS

Abundance of permeases in the strains studied

To investigate the possible gene copy number disparities among permeases, potentially influencing distinct sugar utilization patterns, we leveraged genome sequencing data to identify gene loci associated with MAL31, AGT1, MPH2/3, and MTT1 in both SpIB2 and SpIB1 strains. Although congruence in copy number was observed for AGT1, MPH2/3, and MTT1, MAL31 exhibited variation in copy number between SpIB2 (four loci) and SpIB1 (six loci) (Table 1; Fig. S1). Despite both strains having the MTT1 transporter in their genome, the nucleotide sequence was incomplete compared with the reference length (25, 26). The length of nucleotides and amino acids in the permeases significantly varied (Table 1), with only AGT1- and MAL31-specific isoforms being complete in each genome. Additionally, the structural distribution of MAL loci differed from the canonical pattern, which was observed only at scaffold15 in SpIB2 (11, 31) (Fig. S1). Transcriptomic data confirmed the completeness of permeases, showing that both strains had the MAL31p transporter complete in each transcriptome, whereas the presence of the MTT1p transporter seemed to be absent. To resolve these conflicting results, multiple alignment analyses were conducted to determine the phylogenetic distribution of these transporters using a maximum likelihood algorithm.

TABLE 1.

The location and size of permeases for maltose and maltotriose in the studied strains

Gene name	Systematic name	Locus (scaffold number)	Coordinates	Strand	Length in genome nt/aa	Length in transcriptome (aa)
SpIB2 permeases
AGT1	YGR289C	Scaffold3	4536–6368	+	1,851/610	558
AGT1	YGR289C	Scaffold50	71200–73051	−	1,185/394
MAL31	YBR298C	Scaffold102	3–581	−	581/193	615
MAL31	YBR298C	Scaffold6	612004–613788	−	1,784/594
MAL31	YBR298C	Scaffold11	492091–492738	−	2,165/721
MAL31	YBR298C	Scaffold15	20813–21508	+	1,793/597
MPH3	YJR160C	Scaffold29	946600–948408	−	1,808/602	374
MTY1	MTY1	Scaffold13	658031–659338	−	1,307/435	Not detected
SpIB1 permeases
AGT1	YGR289C	Scaffold11	3868–5700	+	1,833/611	610
AGT1	YGR289C	Scaffold23	4299–6150	+	1,185/394
MAL31	YBR298C	Scaffold8	14887–15582	+	1,825/608	614
MAL31	YBR298C	Scaffold10	282142–283926	−	1,784/594
MAL31	YBR298C	Scaffold26	9677–11624	+	1,947/614
MAL31	YBR298C	Scaffold30	645561–645869	−	309/103
MAL31	YBR298C	Scaffold82	1–792	+	792/264
MAL31	YBR298C	Scaffold85	308–757	+	448/149
MPH3	YJR160C	Scafold24	556421–558229	−	1,809/602	462
MTY1	MTY1	Scaffold64	204–1553	−	1,349/449	Not detected

Structural analysis, expression levels, and phylogenetic distribution of the permeases

To detect specific variations in the structure, expression level, and phylogenetic relationships of maltose and maltotriose transporters that could approximate the differential sugar consumption efficiency, we conduct multiple sequence alignment to detect polymorphic regions related to maltose and maltotriose translocation mechanism. Also, we conduct transcriptomic analysis of the second and fifth days of fermentation for each yeast strain to detect differences that could explain the maltose and maltotriose transport efficiency. Finally, we conduct a maximum likelihood phylogeny to estimate structural/function relationships between maltose and maltotriose transporters of both strains.

We detect that the AGT1 and MAL31 transporters of both strains showed 12 transmembrane domains (TMD1-12), which is characteristic of MSF transporters (32–34). Despite its partial sequence (558 aa), the SpIB2 AGT1 transporter displayed the complete set of canonical transmembrane domains typical of MSF family members. However, some transcripts associated with MPH2/3 and MAL31 genes in both strains lacked transmembrane domain predictions, whereas others exhibited deletions affecting TMD structure (Fig. S2 and S3). For the remaining permeases, the 3D structure prediction matches with alpha-glycoside transporters, showing a high confidence value between TMD structure and predicted aligned error distributions similar between them (Fig. S2 and S3). Taking into account these results, those proteins that showed deletions or not transmembrane domains predictions were excluded from multiple sequence analysis and phylogenetic tree estimations (35).

Regarding the conservation pattern, TMD1 (110–129 aa), TMD2 (157–178 aa), TMD7 (371-399), TMD11 (501–524 aa), and TMD12 (532–550 aa) showed >50% of conservation. However, we also observed some variations in MAL31- and AGT1-specific regions evidencing a polymorphic region that depends on Mal transporter (Fig. 1; Fig. S4). Permease amino acid conservation patterns revealed a consistent presence (≥50%) of E120, D123, E167, and R505 in their respective TMD (Fig. 1). As drawn at Fig. S4, asparagine (N) is conserved ≥50% in all TMD11 sequences, except for specific MAL11/AGT1 isoforms present in SpIB2 and SpIB1, where there is an isoleucine (I). Additionally, an alanine (A) is conserved ≥50% in TMD12 sequences, but not for SpIB2 and SpIB1 MAL11/AGT1, where threonine (T557) was observed in SpIB2 isoform, whereas SpIB1 isoforms had A379 and Y384 (Fig. S5).

Fig 1.

Sections of multiple amino acid sequence alignments of maltose and maltotriose transporters showing the charged residues involved in the maltose translocation mechanism. Only the predicted transmembrane domains TMD-1, TMD-2, and TMD-11 are represented.

In the expression level analysis, SpIB2 exhibits higher transcriptional activity for most transporters, except for MAL61, which shows more expression in the SpIB1 strain. Specifically, the AGT1 isoform from SpIB2 demonstrates a higher median of fragment per kilobase of transcript per million of mapped reads values (FPKM ), 2.6 and 6.08, on the second and fifth days of fermentation, compared with the isoform from SpIB1 with FPKM values of 0.22 and 1.39 on the same days (Fig. 2A). For MAL31 genes, the SpIB2 isoform displays the highest expression values on the second and fifth days (329.26 and 180.08 FPKM), contrasting with the highest expressed SpIB1 isoform showing 34.52 and 50.12 FPKM on the respective days (Fig. 2B). These results are consistent with the regulatory region analysis where, in general, the yeast SpIB2 has more functional binding sites for MAL63 regulatory proteins (Tables S2 and S3).

Fig 2.

Expression of maltose and maltotriose transporters depicted in FPKM values. Date is represented as boxplot MAL gene expression FPKM values extracted from triplicate transcriptomic data of each strain. Note that in (A) AGT1, (B) MAL31, (D) MPH2, and (E) MPH3 genes the strain SpIB2 have more expression than the SpIB1 strain. The strain SpIB1 has major expression of (C) MAL61 transporter only on the fifth day of fermentation.

Concerning MAL61 transporters, the SpIB1 isoform exhibits the highest expression (67.27 FPKM) at the fifth day of fermentation (Fig. 2C). Transcriptional activity of MPH2 and MPH3 is higher in SpIB2 than SpIB1 (Fig. 2D and E). However, both strains display transcript isoforms with deletions affecting numerous transmembrane domains (Fig. S2).

To obtain the relationship between MAL transporters of SpIB2, SpIB1, and those reported for S. cerevisiae, S. eubayanus, and S. pastorianus, a maximum likelihood phylogenetic tree was estimated. Figure 3 reveals the presence of three principal clades observed previously (30, 36), where very divergent AGT1 proteins formed their own group. Here, we observed that MAL11 proteins from SpIB2 and SpIB1 grouped in this cluster, suggesting a similar function. The MPH3 and MPH2 groups are constituted of two proteins from S. cerevisae and one protein from S. pastorianus; however, we do not include the proteins from SpIB2 and SpIB1 as we observed deletions in the important transmembrane domains for maltose translocation (Fig. S2). The largest clade constituted of MALT transporters form S. eubayanus and MAL31_6 from SpIB1, MALx1 from both Sc and Sp and lager hybrid-specific MTT1/MTY1 transporters. It is important to note that SpIB2 MAL31 proteins grouped, supported with 98.6 bootstrap value, in the MTT1 cluster, suggesting a very related maltotriose transport function. Finally, we include a very distant MAL11 transporter from Torulaspora delbrueckii to obtain a more robust interpretation of the tree (35).

Fig 3.

Maximum likelihood phylogeny with 1,000 times bootstraping for maltose and maltotriose transporter proteins reported for S. cereviciae, S. eubayanus, S. pastorianus, SpIB2, and SpIB1. The scale bar indicates the number of amino acid substitutions per site. Note how the SpIB2 MAL31 proteins, marked in red, grouped with the MTT1/MTY1 permeases reported in Sp.

Global expression analysis

To determine the gene expression affected only by the type of yeast strain and not by the fermentation condition and also to obtain statistical insights from the expression level analysis (FPKM) section, we decided to perform differential gene expression analysis by reading alignment of transcriptomic reads from SpIB1 and SpIB2 against the reference genome of the yeast S. pastorianus CBS 1483. The MAL loci analyzed in this work showed 99%–100% identity and a coverage of 87%–100%, with chromosomal regions of the Sp CBS1483 (Table S1). In general terms, we observed a considerable effect depending on the strain type. In the second day of fermentation, we determined 1613 down-regulated genes and 1215 up-regulated genes in SpIB1. On the fifth day of fermentation, we observed 2494 down-regulated genes and 2183 up-regulated genes in SpIB1. Here we can conversely interpret these differential expressed genes (37–39), where the down-regulated genes in SpIB1 are up-regulated in SpIB2, and up-regulated genes in SpIB1 are down-regulated genes in SpIB2.

From these differentially expressed genes, we focused on MAL loci genes, resulting in the second-day fermentation comparison (F2_820 vs F2_790), with two differentially expressed genes in SpIB1; MAL11_2 and MAL11_1 (Fig. 4A). In contrast, strain SpIB2 exhibits differential expression in genes MAL33_1, MAL33_2, MAL32_1, MAL31_2, MAL33_3, MAL63, and MAL13 (Fig. 4A; Table S4). These genes perform positive regulatory functions, and its global expression is confirmed by FPKM measurements, indicating major expression in SpIB2 (Fig. 2; Fig. S5). However, is important to note that the MAL63, MAL13, and MAL33_3 genes are statistically significant but show a low expression in SpIB2 (log2 of fold change <1, gray dots in Fig. 4A), which indicate that their abundance is less than double in the strain SpIB2.

Fig 4.

Volcano plot representing differentially expressed genes in the (A) second and (B) fifth days of fermentation. Up- and down-regulated genes are represented in blue and red, respectively (absolute log2 fold change >1 and false-discovery rate [FDR] < 0.05). Genes with significant but low expression change (absolute log2 fold change <1 and FDR < 0.05) are depicted in gray, and non-significant genes are in black (FDR > 0.05).

The same pattern of expression is seen on the fifth day of fermentation (F5), where strain SpIB1 shows expression of maltose and maltotriose transporters, MAL11 (AGT1) and MAL31, as well as some maltases, such as MAL32 and MAL62 (Fig. 4B; Table S5). Similarly, strain SpIB2 has more positive regulatory genes, such as MAL63 and MAL33_1–3, and some transporters, such as MAL11_3 and MAL31_4. The first four genes, which are positive regulators, showed more than double the abundance in SpIB2 compared with SpIB1, highlighting the significant utilization of these genes by SpIB2.

Biochemical analysis of the transport rate and molecular detection of maltose and maltotriose transporters

To demonstrate if fermentation conditions have a significative effect in the maltose and maltotriose transport, we determined the p-nitrophenyl-α-D-glucopyraonside (pNP-glucose) and p-nitrophenyl-α-d-glucopyranosil-(1,4)-d-glucopyranoside (pNP-maltose) transport rate curves for the strains SpIB2 and SpIB1 in the same reaction conditions. The fundamentals of this assay are that chemically labeled pNP substrates are transported into the cell by alpha-glycoside transporters (Malx1, AGT1, MPHx, and MTT1), where alpha-amylases hydrolyze the alpha bond of maltose and maltotriose, releasing the pNP. After permeabilization and pNP extraction, spectrophotometric measurements were performed at 500 nm for quantitative determination (40). Because the stoichiometric ratio of pNP to the alpha-glucoside of interest is 1:1 (41), it can extrapolate the concentration of transported pNP to the concentration of labeled alpha glycoside.

Although both strains take maltose at a higher rate than maltotriose, SpIB2 demonstrates higher transport rates of both alpha-glycosides than SpIB1 (Fig. 5A and B). Specifically, for maltose, SpIB2 has a transport rate of 282 nmol min⁻¹ mg⁻¹ dry weight, whereas SpIB1 has 202 nmol min⁻¹ mg⁻¹ dry weight (Fig. S6).

Fig 5.

Cellular transport rate determined by uptake of pNP-glucose and pNP-maltose for 7 min in the strain (A) SpIB1 and (B) SpIB2. (C) Residual maltose and maltotriose on the second and fifth days of fermentation for strain SpIB1. Data for strain SpIB2 were undetectable.

Similarly, in maltotriose transport, SpIB2 exhibits a higher rate of 183 nmol min⁻¹ mg⁻¹ dry weight compared with SpIB1 124 nmol min⁻¹ mg⁻¹ dry weight. These biochemical results also reflect the phenotypic behavior in the same fermentation condition where we observed 0.49 ± 0.04% p/p and 1.44 ± 0.27%P/P of residual maltose and maltotriose at the fifth day, respectively, in SpIB1, whereas the sugars were not detectable in SpIB2 (Fig. 5C).

Genetic amplification data suggest that both strains carry ScMALx1, SeMALx1, ScAGT1, SeAGT1, MTT1, and MPH2/3 genes, where the presence of a MTT1-like in SpIB1 was confirmed (Fig. 6A). Transcript gene detection indicates that both strains express all analyzed permeases. Notably, SpIB2 exhibits all permeases on the first day, whereas SpIB1 only shows ScMalx1, ScAGT1, and MPHx. On the second day of fermentation, we observed SeMalx1, ScAGT1, MPHx, and MTT1 in SpIB2 and all analyzed transcripts in SpIB1 (Fig. 6B). From these outputs, we highlight that SpIB2 expresses all MAL transporters analyzed in these conditions, and SpIB1 expresses all MAL transporters, especially the MTT1 on the second day of fermentation, suggesting a probable transcription regulation delay caused by differences in the expression level of the MAL regulatory proteins.

Fig 6.

Molecular detection of maltose and maltotriose permeases in SpIB2 and SpIB1. (A) Genomic detection with PCR of permeases ScMALx1, SeMALx1, ScAGT1, SeAGT1, MPHx, and MTT1/MTY1 indicates that each strain has all the MAL transporters included in this analysis. (B) Transcript detection by RT-PCR on the first and second days of the fermentation of the studied strains. Notice that on the first day, the strain SpIB2 showed all Mal transcripts, whereas SpIB1 only showed ScMalx1, ScAGT1, and MPHx.

Overall, these findings support our bioinformatic inferences, standing out genetic regulation as the predominant factor in the phenotypic variation of maltose and maltotriose transport in the studied strains.

DISCUSSION

Structural variability accounts for the phenotypic variation in alpha-glycoside transport

Spontaneous mutation is a significant driving force behind protein diversity under natural selection, potentially allowing for ~50%–80% divergence without effecting protein structure (42, 43). However, changes in specific residues, particularly those related to substrate-binding sites, could impact protein function (42). Mal permease displays a highly conserved amino acid pattern (≥50%) in TMD1, TMD2, and TMD11, which were reported as relevant for maltotriose transportation (41, 44). Additionally, an alanine (A) is conserved ≥50% in TMD12 sequences but not for SpIB2 and SpIB1 MAL11/AGT1, where threonine (T557) was observed (45). However, it is important to mention that these amino acids are modeled in the context of AGT1 transporters and may not represent the big picture of maltotriose specificity of MTT1/MTY1-like transporters.

Recent studies have reported some crucial polymorphic sites in TMD7 and TMD11 in MAL61, MTT1, and MTT1-like transporters, where the presence of T379 and N384 residues in TMD7 are involved in maltotriose specificity (46). Here, the SpIB2 MAL31 showed these amino acid substitutions but not present in any SpIB1 MAL31 proteins. This result strongly suggests that SpIB2 has an efficient maltotriose transport system present in its transcriptome, whereas SpIB1 has an AGT1-like landscape in its alpha glycoside transport architecture.

Overall, amino acid and synteny changes found in the Mal loci (Fig. S1) indicate that industrial brewing fermentation conditions act as a selective force, and the concomitant hybrid nature of Sp may cause the maltose and maltotriose transport plasticity observed in this yeast (11, 12, 31, 36, 46).

Expression patterns and identity of the transporters efficiently explain the phenotypic variation of maltose and maltotriose transport in the analyzed strains

Currently, diverse S. pastorianus strains use MTT1 and/or AGT1 proteins for maltotriose transport (21–26). Sugar depletion on the second (F2) and fifth (F5) days for strain SpIB2 and SpIB1 reflects an illustration of differences in genetic regulation of MAL-type genes. This notion is supported by the transcriptomic data of putative MTT1 and AGT1 genes in SpIB2 but just AGT1 in SpIB1. Transcriptomic results confirmed that MAL31 and MTT1 have more expression values (FPKM) in SpIB2 than in SpIB1. However, the expression values (FPKM) for AGT1 were low in both strains, even though its importance has been previously reported (Fig. 2) (28, 47, 48). This outcome makes sense because these lager yeasts have AGT1 in its genetic background, suggesting an evolutionary path related to lower expression of AGT1 and the concomitant presence of MTT1-like transporters (30).

The maltose transporters MAL61 were more expressed in SpIB1 on the fifth day of fermentation (Fig. 2C), and this is consistent with its maltose consumption (Fig. 5C), which resemble the interval of maltose and maltotriose residuals observed previously (24). Finally, the expression level of MPHx genes was more in SpIB2 than in SpIB1; however, the MPHx genes were only maltose-specific transporters and its function in maltotriose transport is ambiguous (24, 36).

Regarding the phylogenetic relationships of maltose and maltotriose transporters, our tree is consistent with previous reports (30, 36). In this sense, the SpIB2 strain showed three MAL31 proteins clustered with MTT1/MTY1 transporters, evidencing that this strain probably has an efficient maltotriose transport system, whereas no MTT1/MTY1 clustering was observed in SpIB1 strain. It is important to note that SpIB1 MAL31_6 grouped with S. eubayanus maltose transporters, resembling a function similar to maltose permease of this yeast.

As observed here and in other studies, the maltose and maltotriose transporters AGT1 are a very distant group from the largest MAL31 clade, which has a MTY1/MTT1 maltotriose-specific transporter subclade, suggesting that the evolution of the maltotriose consumption feature has redundancy and frequent occurrence in MAL transporter landscape. However, contrary to the rapid evolution of MAL loci, the maltotriose-specific transporter proteins are a low-frequency event in the evolution pathway of this crucial industrial trait (29, 30, 36), which explains the rich phenotypic diversity observed in the alpha glycoside transport in SpIB2, SpIB1, and other important industrial yeasts (9, 24).

Under the same conditions, the yeasts show differences in their maltose and maltotriose transport rates

To support the bioinformatic predictions related with the differential expression of MAL transporters and regulators and prove the hypothetic scenario where the fermentation conditions exert an effect in the maltose and maltotriose consumption of SpIB2 and SpIB1, the cellular transport rate of pNP-glucose (structural analog of maltose) and pNP-maltose (structural analog of maltotriose) was calculated in the same reaction conditions (22, 40, 44, 49).

Despite the null hypothesis being not different in the cellular transport rate in the same reaction conditions, carbohydrate transport results showed important differences between the analyzed strains. SpIB2 strain was 28% and 32% more efficient to ingress maltose and maltotriose, respectively (Fig. 5; Fig. S6). Contrary to previous results, fermentation conditions make little or no contribution to the alpha-glucoside transport phenotype (23).

In the context of genetic regulation imparted by catabolite repression and the previous transporter affinity, we can reevaluate the importance of maltose rather than maltotriose transport. Even so, AGT1 is responsible for transport of both sugars, and it shows two affinity systems, a higher system for maltose (Km = 14 mM) and a lower system for maltotriose (Km = 27 mM) (47, 49–51). These biochemical information and bioinformatic predictions could help us to explain that the reduced maltotriose transport observed in SpIB1 is caused by the differences in the expression patterns of MAL transporters and regulators (Fig. 5C).

Physical evidence of permeases suggests differential regulation among yeasts studied

Genomic detection indicates that both strains contain the same molecular machinery for maltose and maltotriose transport (Fig. 6A). Transcript detection through reverse PCR assays demonstrated that all tested genes are induced in SpIB2 on the first day, even with glucose presence, including MTT1. However, in SpIB1, the last gene was absent. Reduced transport rate seen in SpIB1 may be associated with a lower expression of MTT1-like and other putative auxiliary proteins as the ScAGT1 allele (47, 50) and MPH2/3 (24). However, it has been reported that the ScAGT1 allele is truncated in S. pastorianus yeast (48), and the MPHx genes have been associated with maltose but not with maltotriose transport (24, 36, 52). This molecular behavior matches the maltose concentrations at the second day of fermentation (Fig. 5C), where only the ScMalx1 and MPHx transcripts, expressed on the first day, are functionally competent for this consumption (Fig. 6B).

On the second day, SpIB1 expressed all analyzed genes (Fig. 6B) that reflect the nearly complete depletion of maltose but not maltotriose (Fig. 5C). Under these circumstances, it is reasonable to speculate that both strains have different regulations patterns related to the hierarchy of sugar consumption.

Conclusion: Difference in genetic regulation plays a significant role in the variability of maltose and maltotriose transport in both strains of S. pastorianus

Many efforts have been addressed to mitigate the limitations of the new sequencing technologies in resolving repetitive and conflicting regions in the genomes of some yeasts (12, 24). Nonetheless, complementing bioinformatics approaches with the transcriptome of interest and experimental evidence allows for even higher resolving power of the omics information in question (53, 54). The present study investigates factors influencing maltose and maltotriose consumption in two industrial lager yeasts. We can identify not only genetic variations in two important regions involved in maltotriose specificity but also differences in expression levels of MAL transporter and positive MAL regulatory genes. These results indicate that the phenotypic diversity related to maltotriose consumption is more complex than a single-factor phenomenon, highlighting the significant utilization of MAL regulatory genes. The major expression of MAL-positive regulatory genes stands as novel contributor to the complexity of maltose and maltotriose consumption phenotypes in the studied yeast. This additional factor must be included in the holistic analysis to improve the beer brewing process.

MATERIALS AND METHODS

Strains

Strains SpIB1 and SpIB2, representing groups I and II, were obtained from Cervecería Cuauhtémoc Moctezuma (Monterrey, NL, México). Inoculated into YP-M medium (1% yeast extract, 2% peptone, and 2% maltose), the yeast strains were stored in YP-M medium with 40% glycerol at −20°C after 48 h or upon reaching stationary phase for future use.

Cultivation conditions and media

Both strains were inoculated in YP-M medium and brewing wort with the following composition: 138 ppm free amino nitrogen, 0.11% (w/w) fructose, 4.85% (w/w) glucose, 11% (w/w) maltose, and 3% (w/w) maltotriose. Fermentation proceeded for 5 days at 16°C, and the experiment was performed in triplicates. Sampling occurred on the first, second, and fifth days, chosen for their significant phenotypic expression in carbon dioxide production and flocculation phenotype.

Inspection and pair-based comparisons of permeases present in the genomic and transcriptomic data of the studied yeasts

We utilized previously obtained genomics (55) to gather insights into maltose and maltotriose transporters within SpIB1 and SpIB2 strains. By examining genome annotation files (31), we identified transporter genes in each strain. Employing the getfasta function of bedtools version 2.30.0 (56), grep UNIX utility, and ggplot2 R package, we generated gene plots and loci information.

Permease structure analysis

We conducted structural analysis by performing multiple sequence alignment for obtaining amino acids in the maltose translocation mechanism (44). Sequence conservation analysis was performed using the msa package version 1.32.0 (57), with CLUSTAL-W and BLOSUM80 substitution matrix, to identify these amino acids in the MAL transporter proteins of both yeasts. The inclusion criteria for permease amino acid sequence involved complete sequences and/or with complete transmembrane domain predictions.

The phylogenetic distribution of each permease was obtained through the maximum likelihood method in ape version 5.7-1 (58) and phangorn version 2.11.1 packages with a bootstrap sample space of 1,000 replicates (59). The topology of transmembrane alpha-helix domains (TMD) in the MAL permeases of the studied strains was determined using CCTOP software (60). Finally, 3D models of the MAL permeases were constructed using the AlphaFold DB methodology (61, 62).

Mal gene regulatory region analysis

We used YEASTRACT to identify putative binding sites of the transcription factors MAL63p and Mig1p for each Mal upstream regulatory region (63). Based on a previous scientific model (27, 28), in the absence of binding sites to the activator of Mal genes, MAL63 shows reduced transcription levels for reporter genes or low growth in culture media with maltotriose as the sole carbon source and vice versa.

FPKM values of the mal genes found in each strain

From the previously generated transcriptomic data (64), we derived FPKM values for the transporters MALx1, AGT1, MTT1, and MPH2/3 to obtain MAL gene expression. Transcriptomic data were obtained from the second and fifth days of fermentation. Due to the hybrid nature of the yeast, potential specific recombination events, and the absence of a reliable reference genome, we performed de novo assembly.

Transcriptomic paired-end reads for both yeast strains were quality-trimmed using Trimmomatic version 0.38 (parameters LEADING:5 TRAILING:5 MINLEN:50) (65) and Sortmerna for contaminant RNA removal (66). Subsequently, forward and reverse reads for each strain were concatenated independently and assembled using Trinity tool version 2.14.0 with default parameters (67, 68). The assembled transcripts were annotated using Trinotate tool version 3.1.0 (69). For mapping trimmed reads to the de novo assembled transcript of each strain, Bowtie2 tool version 2.4.4 with default parameters (70) was employed. RSEM software version 1.3.3 with default values determined expression values (FPKM) (71). Notably, in the context of Trinity assembly output, the term “isoform” is used, leaving the biological significance regarding alternative splicing. Additionally, the unambiguous use of gene nomenclature AGT1/MAL11 and MTT1/MTY1 is emphasized in this study.

Global expression analysis

We conducted a differential gene expression analysis for both yeast strains using samples collected on the second and fifth days of fermentation. To assess statistical differences in Mal gene expression, we employed read alignment with the reference genome of S. pastorianus cbs 1483 (genome and annotations available at NCBI under bioproject prjna522669). Star version 2.7.3a with default parameters was used for alignment (72). Htseq-count version 0.11.1 (73) converted bam alignment files to read count matrices, and edger version 3.34.0 (39) performed differential expression analysis. Genes with low expression levels (cpm > 5) were filtered, contrasting the groups on the second day of fermentation (f2) for both strains (contrast group f2_820-f2_790), and similarly on the fifth day of fermentation (f5) (contrast group f5_820-f5_790). A gene was considered significantly expressed if it showed at least a twofold abundance in each strain [log2(2) =1] with a false discovery rate (FDR) of <0.05. We focus on Mal genes from the list of differentially expressed genes.

Cellular transport rate test

To determine the transport rate of pNP-α-glucose (structurally related to maltose) and pNP-α-maltose (structurally related to maltotriose) in our studied strains, we performed a cellular transport experiment in standard conditions as previously described (40). Briefly, SpIB2 and SpIB1 were suspended at 16 gL⁻¹ in 50 mM succinate-Tris pH 5.0 at 30°C for 5 min; pNP-glucose or pNP-maltose (40 mM in water) was independently added to each strain at a final concentration of 5 mM. After 1-min interval, 100-µL aliquots were sampled and immediately placed in boiling water bath for 3 min. After cooling at room temperature, one volume of 2 M of NaHCO₃ was added to each sample, and the cells were removed by centrifugation, and the p-nitrophenol present in supernatant was measured at 400 nm. The experiments were performed in triplicate, and controls of previously boiled cells were used. Student’s t-test was performed in R version 4.2.3.

DNA and RNA extraction

DNA purification was carried out as previously mentioned (74). Briefly, the SpIB1 and SpIB2 yeasts were inoculated in YP-M medium, and after 48 h, 2-mL aliquots of cells were spun down at 3,000 rpm; 400 µL of sterile extraction buffer was added and incubated at 65°C for 10 min; 130 µL of 3 M sodium acetate pH5.2 was added in each sample followed vortexing for 30 s and incubated at −20°C. Finally, the lysate was centrifugated at 13,000 rpm at 4°C for 15 min; the supernatant was precipitated with isopropanol. For RNA extraction, the hot phenol protocol was performed (75), with the only difference being the use of AES solution (50 mM sodium acetate, 10 mM EDTA, and 0.5% SDS) instead of the proposed TES solution. Briefly, 10-mL samples from first and second fermentation days were centrifugated for 5 min at 3,000 rpm at 4°C; the cell pellets were washed with 5 mL of sterile cold water and centrifugated at 3,000 rpm at 4°C for 5 min; the supernatant was removed, and the cell pellet was frozen in liquid nitrogen and stored at −70°C until use. The frozen cell pellet was thawed on ice and resuspended in 400 µL of AE solution; 400 µL of acid phenol was added followed by 30 s vortexing; the extraction samples were incubated for 60 min at 65°C and then centrifugated for 5 min at 13,000 rpm at 4°C. The supernatant was placed in new sterile microcentrifuge tubes, and a second hot acid phenol extraction procedure was performed. The resultant supernatant was treated with one volume of chloroform and centrifugated for 5 min at 13,000 rpm at 4°C; the resulting supernatant was precipitated with two volumes of ice-cold isopropanol and 40 µL of 3 M sodium acetate at pH 5.3, then ice-cold incubated for 60 min. Then, the samples were centrifugated for 10 min at 13,000 rpm at 4°C, and the total RNA pellet was washed with 70% ice-cold ethanol in diethyl pyrocarbonate (DPCE)-treated sterile water. Finally, the resultant washed pellet was resuspended in 50 µL DPCE-treated water, seen in agarose gel electrophoresis, and concentrations were determined in nanodrop equipment.

Molecular detection of permeases in each strain

For the genomic detection of gene encoding Malx1, AGT1, MTT1, and MPH2/3 permeases in both genomes, primers, and PCR experiments were performed as described previously (24). Briefly, the following PCR program was used: pre-incubation (95°C for 5 min), amplification cycle repeated 35 times (95°C for 10 s, Tm °C for each specific primer for 10 s, 72°C for 10 s).

A reverse transcription PCR (RT-PCR) experiment was used to detect the Malx1, AGT1, MTT1, and MPH2/3 transcripts. Briefly, a commercial reverse transcriptase (M-MLV Reverse Transcriptase, Promega) was used, following the manufacturer instructions, and using the same PCR conditions as mentioned above (24). The primers for ACT1, TAF10, and TFC1 were used as reference control housekeeping genes (76).

Wort and alpha glycoside consumption analytics

Wort samples were analyzed by HPLC Series 1200 Agilent Technologies using infrared detector (IR). The mobile phase was H₂SO₄ (5 mM). The flow rate was established to 1 mL/min using carbohydrate column Aminex HPX-87C Hi-Plex Na, 300 × 7.7 mm, and the column was equilibrated with 5 mM H₂SO₄ in water at 55°C (24). The experiments were done in triplicates and the Student’s t-test was done for statistical significance in the difference of maltose and maltotriose uptake means.

DATA AVAILABILITY

The data are available on the GEO NCBI platform, and the accession number for the data is GSE269873.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.00397-24.

Supplemental material. aem.00397-24-s0001.docx.

Figures S1 to S6; Tables S1 to S5.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.