Identification of European isolates of the lager yeast parent Saccharomyces eubayanus

Abstract

Lager brewing first occurred in Bavaria in the 15th century, associated with restrictions of brewing to colder months. The lager yeast, Saccharomyces pastorianus, is cold tolerant. It is a hybrid between Saccharomyces cerevisiae and Saccharomyces eubayanus, and has been found only in industrial settings. Natural isolates of S. eubayanus were first discovered in Patagonia 11 years ago. They have since been isolated from China, Tibet, New Zealand, and North America, but not from Europe. Here, we describe the first European strains UCD646 and UCD650, isolated from a wooded area on a university campus in Dublin, Ireland. We generated complete chromosome level assemblies of both genomes using long- and short-read sequencing. The UCD isolates belong to the Holarctic clade. Genome analysis shows that isolates similar to the Irish strains contributed to the S. eubayanus component of S. pastorianus, but isolates from Tibet made a larger contribution.

First isolation of European parents of lager yeasts from a wooded area on a university campus in Dublin, Ireland.

Introduction

Brewing is one of the oldest industries associated with humans, with evidence of fermented beverages from China from 7000 to 8000 years ago, and from Israel from up to 13 000 years ago (Liu et al. 2018, 2019). Until the Middle Ages in Europe, most beer brewing was associated with Saccharomyces cerevisiae yeast, a ‘top fermenter’. The majority of ales today are brewed using S. cerevisiae (Gallone et al. 2016). Lagers first appeared in the 13th century, but became particularly common from the 15th century in Bavaria, where fermentation and storage took place in cool cellars (Gallone et al. 2019). The Beer Purity Laws in Bavaria in 1516 followed by the edict from Albrecht V in 1553 defined the components of beer and restricted brewing to the colder winter months, which may have selected for the ‘bottom fermenting’ cold tolerant lager yeasts now known as Saccharomyces pastorianus (Dornbusch 1998). Lager brewing spread worldwide in the 19th century, and lagers now represent > 90% of beers sold.

It has been known for some time that S. pastorianus is a hybrid of two parents, one of which is S. cerevisiae (de Barros Lopes et al. 2002, Dunn and Sherlock 2008). However, the second parent, Saccharomyces eubayanus, was not isolated until 2011, from the Patagonian Andes in South America (Libkind et al. 2011). Saccharomyces eubayanus and S. pastorianus isolates are cold tolerant (Gibson et al. 2013). Isolates of S. eubayanus have now been found in North America (Peris et al. 2014, Langdon et al. 2020), the Tibetan Himalayas, Sichuan and West China (Bing et al. 2014), and New Zealand (Gayevskiy and Goddard 2016). By far the most isolates have come from South America, where the species is associated with Nothofagus trees (Eizaguirre et al. 2018, Langdon et al. 2020, Nespolo et al. 2020). Saccharomyces eubayanus probably originated in Patagonia, and subsequently colonized the world (Langdon et al. 2020, Nespolo et al. 2020).

Phylogenetic analysis shows that S. eubayanus isolates fall into two main clades, PA and PB (Peris et al. 2016). PA is further divided into two lineages, PA-1 and PA-2, whereas PB can be divided into at least four lineages, PB-1, PB-2, PB-3, and the Holarctic clade (Peris et al. 2016, Langdon et al. 2020, Nespolo et al. 2020). Some PA/PB admixed isolates have also been identified (Peris et al. 2016, Langdon et al. 2020). Isolates from South America are very diverse and belong to all the main lineages, apart from the Holarctic clade. Isolates from outside South America belong to the PB-1 or Holarctic lineages, or are admixtures between PA and PB. Genome sequences from only a handful of isolates from the Holarctic lineage are available; two from Tibet (Bing et al. 2014, Brouwers et al. 2019a), and two from North Carolina, USA (Peris et al. 2016). However, the S. eubayanus component of the hybrid lager yeasts (S. pastorianus) also belongs to the Holarctic lineage, suggesting that the S. eubayanus parent originated in the Holarctic (defined as the nontropical parts of Europe and Asia, Africa north of the Sahara, and North America south to the Mexican desert region; Peris et al. 2016).

It is generally assumed that the original hybridization between S. cerevisiae and S. eubayanus occurred in a brewery, because no nonindustrial isolates of S. pastorianus have been discovered, and the S. cerevisiae parent is more similar to ale yeasts than to natural isolates (Gibson and Liti 2015, Monerawela et al. 2015, Krogerus et al. 2017, Gallone et al. 2019). It is, however, not clear how many hybridizations occurred. There are two main groups of S. pastorianus strains—Group I (Saaz) and Group II (Frohberg)—with differing fermentation properties (Gibson et al. 2013). Saaz strains are generally triploid, with approximately two genomes from S. eubayanus and one from S. cerevisiae, and Frohberg strains are tetraploid (or higher) with approximately equal contributions from S. eubayanus and S. cerevisiae (Walther et al. 2014). Lager yeast genomes have some chimeric chromosomes that contain junctions between DNA of S. eubayanus and S. cerevisiae origin, and some of these junctions are identical between Saaz and Frohberg strains, which strongly supports a single origin from a shared hybrid parent (Hewitt et al. 2014, Walther et al. 2014, Okuno et al. 2016, Gallone et al. 2019, Salazar et al. 2019). However, there is also evidence that some of the genetic variation in S. eubayanus populations survived through the bottleneck of hybridization and is retained in modern lager yeast genomes (Peris et al. 2016). This retention of standing genetic variation could be attributable to heterozygosity in the original hybridizing S. eubayanus parent, or to extra hybridizations or backcrossings that occurred after the original hybridization event shared by Saaz and Frohberg (Peris et al. 2016). Gallone et al. (2019) proposed that the initial hybridization event(s) occurred in the 1500s, around the time of the Bavarian Beer Purity Law, and that the Saaz and Frohberg lineages separated later. Both lineages went through a bottleneck following the purification of isolates by E.C. Hansen at the Carlsberg Laboratory in 1883 (Hansen 1883), and subsequent sharing between brewers (Gorter de Vries et al. 2019).

Regardless of when the hybridization(s) between S. cerevisiae and S. eubayanus occurred, they are likely to have occurred in Europe, and possibly in Bavaria. It is, therefore, surprising that no European isolates of S. eubayanus have been described. Climate modelling suggests that Europe is a prime location (Langdon et al. 2020). An intriguing metagenomics study from Alsammar et al. (2019) identified a small number of rDNA reads from S. eubayanus in the Italian Alps, suggesting that there might be a reservoir of the species there. Another indirect indication is that many European isolates of Saccharomyces uvarum associated with human-driven fermentation show evidence of introgression from S. eubayanus (Almeida et al. 2014). Here, we describe the discovery of the first European isolates of S. eubayanus. They were isolated from a university campus in Dublin, Ireland. They belong to the Holarctic lineage. As predicted by Peris et al. (2016), they match the lager yeasts at some, but not all, loci. We estimate that the S. eubayanus parent strains of the Saaz and Frohberg lineages shared approximately 54% of S. eubayanus alleles with a Tibet-like population, 40% with an Irish-like population, and 6% with a North American-like population. This supports the hypothesis that standing genetic variation in S. eubayanus persists in lager yeasts (Peris et al. 2016), and suggests that a pool of European isolates remains to be discovered.

Methods

Isolation of S. eubayanus

Detailed instructions are provided in Text S1 (Supporting Information). In brief, approximately 2 g of soil from two nearby locations on the UCD campus were inoculated in 10 ml yeast extract–peptone–dextrose (YPD) medium with chloramphenicol [30 μg/ml] and ampicillin [100 μg/ml] and incubated at room temperature for 1 week. A volume of 10 µl of culture was then used to inoculate 10 ml of fresh media, and incubated at room temperature for 38 h. Diluted culture was plated on YPD agar with chloramphenicol [30 μg/ml] and ampicillin [100 μg/ml] 4 days. Single colonies were streaked on agar plates, and tentatively identified by sequencing the ITS region. Two potential S. eubayanus isolates (UCD646 and UCD650) were chosen for genome sequencing.

Genome sequencing

For short-read sequencing genomic DNA extraction, cells were harvested by centrifuging at 3000 rpm for 5 min and resuspending cell pellets in 200 µl Extraction buffer (2% Triton X 100, 100 mM NaCl, 10 mM Tris pH 7.4, 1 mM EDTA, and 1% SDS) in a screw-top tube. Nucleic acids were extracted by lysing cells and eluting DNA to an aqueous phase by adding ∼0.3 g acid washed beads and 200 µl phenol/chloroform/isoamyl alcohol 25:24:1, agitating the mixture with a 600 MiniG bead beater (Spex SamplePrep) for 30 s, and centrifuging at 14,000 rpm for 10 min. The top aqueous layer was extracted using 200 µl phenol/chloroform/isoamyl alcohol 25:24:1 and 200 µl TE buffer (10 mM TrisHCl, pH 8.0, 1 mM EDTA), and then extracted once more using 200 µl phenol/chloroform/isoamyl alcohol 25:24:1. A volume of 80 µl 7.5 M ammonium acetate and 1 ml 100% isopropanol were added to the total aqueous extraction to precipitate the DNA. DNA was pelleted by centrifugation at 14,000 rpm, washed using 70% ethanol and air-dried. Pellets were resuspended in 400 µl TE buffer with 1 µl RNase A (100 mg/ml) and incubated overnight at 37°C. DNA was reprecipitated and washed once more as above and resuspended in 150 µl water. Genomic DNA was sequenced by BGI Tech Solutions using DNBseq.

For long-read sequencing, the yeasts were cultured for 24–40 h in YPD medium. Genomic DNA was extracted using Qiagen’s Genomic Tip 100 G kit. High molecular weight was validated by gel electrophoresis, DNA concentration was quantified using a Qubit fluorometer, and quality was assessed using a Nanodrop. Multiplexed libraries using the SQK-RBK004 kit were made, and cleaned up with AMPure XP magnetic beads. Fresh MinION flow cells were primed with flow cell priming kit EXP-FLP002, and the libraries were sequenced using MinKNOW v4.1.22. Raw data was basecalled and demultiplexed after sequencing using Guppy v4.2.2.

Genome assembly and variant calling

MinION read quality was assessed using Nanoplot, after which Nanofilt was used to retain only reads with Phred quality score (Q) and length (L) values of Q > = 10 and L > = 5 kb for UCD650, and Q > = 7 and L > = 1 kb for UCD646. Reads were then assembled using Canu version 2.2 (Koren et al. 2017), after which contigs were error-corrected by five rounds of NextPolish (Chen et al. 2021) using the DNBseq short reads. We discarded five small contigs (≤ 61 kb) from the UCD646 assembly, and two small contigs (≤ 37 kb) from the UCD650 assembly. These contigs were derived from rDNA, telomeric sequences, or the alternative allele at the MAT locus. The total length of the UCD646 assembly is 12 050 494 bp with an N50 of 914 339 bp and a GC content of 39.74%. The total length of the UCD650 assembly is 12 006 963 bp with an N50 of 916 217 and a GC content of 39.75%. Additional genome statistics are presented in Table S2 (Supporting Information).

Following annotation with YGAP (Proux-Wera et al. 2012), 5682 protein coding genes and 283 tRNAs were predicted in UCD646, and 5664 protein coding genes and 283 tRNAs were predicted in UCD650. Using BUSCO v5.1.2, genome completeness was estimated at 99.7% for both strains (compared to the Saccharomycetes lineage data set).

In addition to sequencing reads from S. eubayanus UCD646 and UCD650, WGS data for 87 additional S. eubayanus genomes, and for two S. pastorianus genomes were downloaded from SRA (Table S3, Supporting Information). Data from several strains sequenced by Langdon et al. (2020) were not included because the quality of the sequence was insufficient for accurate variant calling. Illumina and DNBseq reads were trimmed with Skewer version 0.2.2 using parameters ‘-m pe -t 4 -l 35 -q 30 -Q 30’ (Jiang et al. 2014). The trimmed reads were aligned using BWA MEM (version 0.7.23) to a FASTA file combining both the Canu-assembled UCD646 genome and the S288c_R64-3-1 S. cerevisiae reference genome. Read pairs that aligned to the S. cerevisiae sequence, and reads that had an identity < 90% to the region they mapped to, were removed. This was done to exclude reads from the S. cerevisiae fraction of the S. pastorianus genomes. The filtered BAM files were sorted and duplicate reads were marked using GenomeAnalysisToolkit (GATK version 4.0.1.2) SortSam and MarkDuplicates tools respectively (McKenna et al. 2010). Variants were called with GATK HaplotypeCaller using the tag ‘–genotyping_mode DISCOVERY’, combined using GATK CombineGVCFs and joint-genotyped using GATK GenotypeGVCFs. Variant files were filtered for read depth (< 15) and genotype quality (< 40) using GATK VariantFiltration. Additionally, clusters of SNPs (5 SNPs in a 20 bp window) were excluded using GATK VariantFiltration. GATK SelectVariants was used to exclude multiallelic sites, and sites where a genotype could not be determined for > 30% of strains. Finally, indels were excluded for tree construction using the ‘–select-type-to-include SNP’ tag.

Phylogenetic analysis

A total of 1000 iterations of Random Repeated Haplotype Sampling (RRHS; Lischer et al. 2014) were used to generate SNP alignments where heterozygous sites were randomly resolved to either haplotype. SNP trees were then constructed from the 1000 alignment files using RAxML (v8.2.12) with the GTRGAMMA model of nucleotide substitution and the random number seed ‘-p 12345’ (Stamatakis 2014). The tree with the highest maximum likelihood was selected, and the remaining trees were used to generate branch support values.

Identification of shared variants in S. pastorianus strains

A custom script (https://github.com/CMOTsean/Eubayanus_shared_variants) was used to identify variants present in S. pastorianus strains CBS 1538 and W34/70 that could be assigned to subpopulations of the Holarctic clade of S. eubayanus. Briefly, if a variant was found in S. eubayanus UCD646 and in S. pastorianus CBS 1538 or W34/70 but not in S. eubayanus Holarctic strains from Tibet (CDFM21L.1 and ABFM5L) or from the USA (yHRVM107 and yHRVM108) the strains were assumed to have uniquely shared ancestry at that variant. The analysis was then repeated using S. eubayanus CDFM21L.1 and yHRVM107 to assign ancestry to the Tibet and North Carolina lineages, respectively. For each of strains CBS 1538 and W34/70, blocks of shared ancestry were defined by two or more consecutive variants that were uniquely shared between the S. pastorianus strain and one of the Holarctic populations. The blocks were expanded outwards to cover the entire genome such that borders between blocks were drawn at the midpoint between outermost sites. The total length of each type of region was calculated to provide an estimate of the portions of the S. eubayanus component of the genome that have shared ancestry with each Holarctic population. The underlying data for each SNP is provided in summary in Table S1A (Supporting Information), with all data in Table S1B (Supporting Information).

Growth on maltose

Three single colonies of S. eubayanus CBS 12357, UCD646, and UCD650 and S. cerevisiae S288C were inoculated in 5 ml of YPD and grown overnight with shaking at 200 rpm at 20°C. A total of two by 1 ml of each culture was centrifuged at 5000 rpm for 3 min, and cells were washed three times in 1 ml of PBS (phospho-buffered saline). One set of pellets was resuspended in 1 ml of YPD (2% glucose) and the other in 1 ml of YPM (2% maltose). The absorbance at 600 nm of each culture was adjusted to 0.1 and 200 µl was added to a round bottom 96-well plate. Cell growth was measured at room temperature every 10 min at 660 nm in a SYNERGY H1 microplate reader (BioTek Instruments, Inc.) The means of three biological replicates and the standard deviations are plotted. For spot assays, strains were grown overnight in 5 ml liquid YPD, washed three times in PBS and resuspended in 1 ml of PBS. The A₆₀₀ was adjusted to 0.4 and then serially diluted by ⅕ seven times before spotting. Strains were grown at room temperature for 72 h.

Results and discussion

Two S. eubayanus strains (UCD646 and UCD650) were isolated from soil samples from the Belfield campus of University College Dublin, as part of undergraduate research modules that identify and sequence the genomes of yeast species (Mullen et al. 2018, O'Boyle et al. 2018, Venkatesh et al. 2018, Almasoud et al. 2019, Faherty et al. 2019, Heneghan et al. 2019, O Cinneide et al. 2021, Ryan et al. 2021). A detailed description of the isolation method is provided in Text S1 (Supporting Information). The isolates came from soil on two sites on the university campus, about 17 m apart, collected in September 2021. One site (UCD650; GPS coordinates 53.306796, −6.233050) was soil under leaf litter from a sycamore tree (Acer pseudoplatanus), and the other site (UCD646; GPS coordinates 53.306563, −6.232683) was in soft soil close to oak (Quercus robur), beech (Fagus sylvatica), and cherry laurel (Prunus laurocerasus) trees (Fig. 1A). The isolates grow poorly at temperatures of 30°C or higher and grow well at cooler temperatures (e.g. 13°C), similar to S. uvarum and unlike S. cerevisiae (Fig. 1B). The type strain of S. eubayanus CBS 12357 grows better than either UCD646 or UCD650, particularly at 30°C (Fig. 1B).

Figure 1.

Identification and growth of S. eubayanus isolates. (A) Saccharomyceseubayanus UCD646 and UCD650 were identified in soil from a wooded area on the campus of University College Dublin. One soil sample was isolated from an area with oak, beech, and cherry laurel trees. (B) Isolates were streaked on YPD and incubated at the temperatures shown for 6 days. Sc = S. cerevisiae; Su = S. uvarum (UCD415 was also isolated from soil); and Se = S. eubayanus. CBS 12357 is the S. eubayanus type strain, obtained as VTT-C-72902. Room temperature is approximately 20°C.

The genomes of both isolates were sequenced using a combination of long-read (Oxford Nanopore MinION) and short-read (BGI DNBseq) sequencing. MinION reads were assembled using Canu (Koren et al. 2017), after which the assemblies were error-corrected by five rounds of polishing using NextPolish (Chen et al. 2021) with DNBseq reads. For both strains, each chromosome assembled as a single contig, except for chromosome XII where we manually made a join at the rDNA locus. Genomes were annotated using YGAP (Proux-Wera et al. 2012).

Both of the strains appear to be MATa/α diploids with very low heterozygosity. By mapping DNBseq reads to the final assemblies, we found eight heterozygous SNPs in UCD646, and 126 in UCD650. For UCD646, the MAT locus on chromosome III of the assembly has a MATa genotype, and the assembly also included a small contig with a MATα allele. For UCD650, chromosome III contained a MATα allele. There was no corresponding small contig with a MATa allele, but by BLASTN searches against a database of unassembled MinION reads we identified multiple reads that were derived from the MAT locus and were MATa. Chromosome III in both strains also contains HMLα and HMRa loci similar to the arrangement in S. cerevisiae.

Although the two strains were isolated from sites only a few metres apart, there are numerous small differences between the genomes of UCD646 and UCD650. They differ by 2517 homozygous SNPs. We identified 13 sites in the genome where one UCD strain contains a full-length Ty retrotransposon but the other does not (eight UCD646-specific Ty elements, and five UCD650-specific Ty elements). The strains also differ from each other by three inversions. On chromosome VII, UCD646 has an inversion of a 60-kb region spanning the centromere (all genes between COG7/YGL005C and TIM21/YGR033C) that is not present in UCD650, the Himalayan isolate CDFM21L.1 (Brouwers et al. 2019a), or the S. eubayanus type strain CBS 12357 from Patagonia (Baker et al. 2015, Brickwedde et al. 2018). The inversion is flanked by two Ty elements in opposite orientations in UCD646, only one of which is present in UCD650. On chromosome XI, UCD650 has an inversion of a 10-kb region (YKT6/YKL196C to CNB1/YKL190W) relative to UCD646, CDFM21L.1, and CBS 12357. On chromosome XIV, UCD650 differs from UCD646, CDFM21L.1, and CBS 12357 by a 25-kb inversion (SIW14/YNL032W to ARK1/YNL020C).

Both of the UCD strains contain the two reciprocal translocations that are shared by S. eubayanus and its sister species S. uvarum relative to S. cerevisiae: one translocation between chromosomes II and IV, and one between chromosomes VIII and XV (Scannell et al. 2011, Baker et al. 2015). Brouwers et al. (2019a) reported three major structural differences between the Himalayan isolate CDFM21L.1 and the type strain CBS 12357. We find that one of these (an 8-kb inversion in the chromosome VII-L subtelomere) is shared by the UCD strains and CDFM21L.1, and so differentiates the Holarctic from the Patagonian isolates. In the three Holarctic isolates, subtelomeres VI-R and VIII-L are highly similar to each other, whereas the type strain has a different sequence at VI-R, accounting for the second structural difference. The third structural difference noted by Brouwers et al. (2019a), involving an exchange between the chromosome V-R subtelomere and the rDNA region on chromosome XII, is specific to CDFM21L.1 and is not present in the UCD strains.

To determine the relationship of the UCD strains to other S. eubayanus isolates, we constructed a phylogenetic tree using variants from the genomes of 89 other isolates, mostly from the PB clade (Fig. 2). This tree recapitulated previous phylogenetic trees (Peris et al. 2016, Langdon et al. 2020, Nespolo et al. 2020), in which the PB clade is divided into at least four main lineages, PB-1, PB-2, PB-3, and the Holarctic lineage. UCD646 and UCD650 are closely related and belong to the Holarctic lineage, together with two North Carolina isolates (yHRVM107 and yHRVM108; Peris et al. 2016), two isolates from Tibet (CDFM21L.1 and ABFM5L; Bing et al. 2014, Brouwers et al. 2019a), and the S. eubayanus components of the S. pastorianus lager yeasts CBS 1538 (Saaz) and W34/70 (Frohberg) (Peris et al. 2016).

Figure 2.

The Irish S. eubayanus isolates belong to the Holarctic clade. Unrooted maximum-likelihood tree of 89 S. eubayanus and two S. pastorianus strains was constructed with RAxML using the GTRGAMMA model of nucleotide substitution, using a concatenated SNP alignment of 319 298 SNP sites. Coloured shapes highlight distinct populations identified previously (Nespolo et al. 2020), and the two Irish (UCD) isolates are highlighted in bold. Most strain labels have been omitted for clarity (see Figure S1, Supporting Information). Branch colour denotes support values on a scale from 0 (red) to 100 (green). All branches in the Holarctic clade have a support value of 100. Support values < 100 are labelled in Figure 1 (Supporting Information).

The phylogenetic tree suggests that the Tibet isolates are more closely related to the lager yeasts than either the Irish or the North Carolina isolates, which cluster together (Fig. 2). However, recombination within or between populations can mask ancestry (Peris et al. 2016). We, therefore, compared the alleles along each chromosome in S. pastorianus CBS 1538 and W34/70 to the alleles from the Tibetan, North Carolina, and Irish isolates (Fig. 3; Figure S2 and Table S1, Supporting Information), to identify SNP sites at which a lager strain shares an allele with only one of the three Holarctic populations (see ‘Methods’). All sites were mapped onto the genome of UCD646. Figure 3 shows an example of allele conservation across two chromosomes, and all chromosomes are shown in Figure S2 (Supporting Information). Blocks of shared ancestry were defined by dividing each chromosome into runs of consecutive variants assigned to the same population. The total length of these blocks was used to estimate the proportion of the S. eubayanus component of CBS 1538 and W34/70 that has shared ancestry with each of the Holarctic populations. For CBS 1538, 4 188 309 bp (37.69%) were assigned to the Irish population, 6 233 324 bp (56.09%) were assigned to the Tibet population, and 690 505 bp (6.21%) were assigned to the North Carolina population. For W34/70, 4 525 481 bp (42.28%) were assigned to the Irish population, 5 558 609 bp (51.93%) were assigned to the Tibet population, and 619 169 bp (5.78%) were assigned to the North Carolina population. The underlying data for each SNP is provided in summary in Table S1A (Supporting Information), with all data in Table S1B (Supporting Information).

Figure 3.

Shared ancestry of alleles between the S. eubayanus component of S. pastorianus genomes (CBS 1538 and W34/70) and Holarctic clade S. eubayanus isolates. Coloured offset vertical lines represent sites where an allele in a S. pastorianus strain is also present in only one other population (Tibet = Blue, Ireland = Magenta, or North Carolina = Green) in the Holarctic clade. Coloured rectangles represent blocks assigned to these populations, where blocks contain two or more neighboring alleles from the same population (borders are drawn at half the distance between the outermost sites). All alleles were called by alignment to the UCD646 assembly. Sites where a genotype could not be called or were filtered out are shown in black. All chromosome comparisons are shown in Figure S2 (Supporting Information), and supporting data is in Table S1 (Supporting Information). (A) Shared ancestry blocks of CBS 1538 and W34/70 on Chr XVI. CBS 1538 and W34/70 have similar patterns of ancestry except for a large region marked with a red box where CBS 1538 has shared ancestry with the Tibet population and W34/70 has shared ancestry with the Irish population. In addition, there are large regions in W34/70 where shared ancestry could not be reliably assigned. (B) Shared ancestry blocks of CBS 1538 and W34/70 on Chr XII. This example shows shared ancestry with all three Holarctic populations, although blocks assigned to North Carolina are shorter and fewer in number compared to Irish or Tibet blocks. Again, W34/70 has large blocks where no shared ancestry could be assigned.

As reported by Peris et al. (2016), there are some regions where the Saaz isolate of S. pastorianus (CBS 1538) appears to have a different origin than the Frohberg isolate (W34/70). For example, one region of Chromosome XVI in CBS 1538 is more closely related to Tibetan isolates, whereas the equivalent region in W34/70 is more closely related to Irish isolates (boxed in Fig. 3). Other examples using the whole genome comparisons are shown in Figure S2 (Supporting Information).

The ability of S. pastorianus isolates to consume the sugars glucose, maltose, and maltotriose is important for lager fermentation (Zastrow et al. 2001). Maltose and maltotriose are transported by related proteins. In S. pastorianus, maltotriose transporters are encoded by AGT1 and MTY1; MTY1 is probably derived from recombination between ancestral maltose transporters (Baker and Hittinger 2019, Brouwers et al. 2019b). Saccharomyces pastorianus isolates have two AGT1 alleles. One inherited from S. cerevisiae is truncated and nonfunctional, and the second full length allele is located on S. eubayanus chromosome XV (Vidgren et al. 2009). It was, therefore, assumed that S. pastorianus AGT1 originated from S. eubayanus, even though S. eubayanus isolates cannot utilize maltotriose, and most lack an AGT1 gene (Gibson et al. 2013, Baker and Hittinger 2019). Intact Agt1 maltotriose transporters were recently identified in Holarctic isolates of S. eubayanus from North America and from the Himalayas (Baker and Hittinger 2019, Brouwers et al. 2019a), and these are likely the source of the gene in S. pastorianus. There are no AGT1 orthologs in the genomes of S. eubayanus UCD646 or UCD650.

In several Saccharomyces species, maltose utilization genes are organized in MAL loci, consisting of a maltose transporter (called MALT in S. eubayanus), a maltase enzyme (MALS) and a transcriptional regulator (MALR; Charron et al. 1989). The MalR regulator binds to a bidirectional promoter between MALS and MALT (Needleman 1991). The numbers of MAL loci, which are located in subtelomeric regions, vary between strains (Naumov et al. 1994).

The S. eubayanus CBS 12357 type strain (isolated from Patagonia) encodes MAL transporters (MALT) at four subtelomeric regions (Brickwedde et al. 2018). Two (on Chrs V and XVI) are within intact MAL loci, with divergently transcribed MALT and MALS genes adjacent to MALR (Fig. 4A). Saccharomyces eubayanus CBS 12357 does not encode AGT1. The Tibetan (Himalayan) S. eubayanus isolates (e.g. CDFM21L.1, Fig. 4A) encode maltose utilization genes at three subtelomeric regions (Chrs II, V, and XIII) with shared synteny with CBS 12357, except that CDFM21L.1 has a rearrangement between Chr V and Chr XII, and has a fourth site on Chr XII (Brouwers et al. 2019a). There are AGT1 genes at three other subtelomeres. None of the MAL loci in CDFM21L.1 are intact. Saccharomyceseubayanus UCD646 and UCD650 encode maltose utilization genes at four subtelomeric locations, that share synteny with both CDFM21L.1 and CBS 12357 (Fig. 4A). There is a MAL locus (MALS, MALT, and MALR) on Chr XVI (Fig. 4A). All the genes are intact in UCD650, but in UCD646 the MALR open reading frame is truncated because of a frameshift (Fig. 4A). UCD646 does not appear to encode a functional MALR; other copies on chromosomes V and XIII are also truncated (Fig. 4A). Several maltose utilization genes in S. eubayanus CBS 12357 also have stop codons/frameshifts (Brouwers et al. 2019a). We find that unlike the S. eubayanus CBS 12357 type strain, S. eubayanus UCD646 and UCD650 grow very poorly on media containing maltose as the sole carbon source (Fig. 4B and C). Growth on maltose is almost as poor as that of S. cerevisiae S288C, which has a defective mal2 gene (Fig. 4B and C; Mortimer and Johnston 1986). It is, therefore, possible that the maltose transporters are not expressed in the Irish S. eubayanus isolates.

Figure 4.

Maltose utilization in S. eubayanus. (A).Organization of maltose utilization genes. Telomeres are indicated with red boxes, and chromosomes with black lines. Maltose utilization genes are shown in colour, and other genes in grey. Intact MAL loci are highlighted with black boxes. Truncated open reading frames are shown as outlines with black asterisks. Ty = Ty element; rDNA = ribosomal DNA locus. Information for S. eubayanus CBS 12357 is from the assembly from Brickwedde et al. (2018) and for CDFM21L.1 is from Brouwers et al. (2019a). (B) Cultures of S. eubayanus CBS 12357, UCD646, and UCD650 and S. cerevisiae S288C were grown at room temperature in YP containing either 2% glucose or 2% maltose as the sole carbon source. Growth was monitored every 10 min. The mean and standard deviation of three biological replicates is shown. (C) Cells from the indicated strains were diluted ⅕ and spotted on YP agar with 2% glucose or 2% maltose and grown at room temperature for 72 h.

Conclusion

Since the discovery of S. eubayanus isolates in Patagonia in 2011 (Libkind et al. 2011) it has been hypothesized that isolates would be found in Europe, and indeed modelling by Langdon et al. (2020) showed that Europe is a suitable location. Our discovery of isolates in Ireland is consistent with the ‘Out-of-Patagonia’ hypothesis, that S. eubayanus evolved in Patagonia where it adapted to cold and harsh conditions, and then spread to the rest of the world, probably in the postglacial period (Langdon et al. 2020, Nespolo et al. 2020). Our observation that some of the alleles in S. pastorianus isolates are closely related to alleles from the Irish S. eubayanus strains but that more of the genomes are close to alleles from Tibetan strains is consistent with the hypothesis that no one isolate of S. eubayanus is the direct ancestor of the parent of the lager yeasts, due to incomplete lineage sorting, backcrossing, or possibly multiple hybridization events (Peris et al. 2016). The maltotriose transporter gene AGT1, for example must have been acquired from isolates similar to the Tibetan strains. It is, therefore, likely that alleles in both Saaz and Frohberg strains of S. pastorianus are derived from standing variation in the S. eubayanus population (Peris et al. 2016).

However, it is also likely that isolates that share more similarities with S. pastorianus remain to be discovered in Europe. Langdon et al. (2020) speculate that competitive exclusion with S. uvarum or Saccharomyces paradoxus may have restricted the range of S. eubayanus in Europe, but it also possible that we are not looking in the right place, or the right ecological niche. This report, and the intriguing indication from metagenomics data from Italy (Alsammar et al. 2019), suggest that more European isolates from the Holarctic lineage will be discovered in the future.

Conflict of interest statement

None declared.

Data availability

The data for this study have been deposited in the European Nucleotide Archive (ENA) at EMBL-EBI under BioProject accession number PRJEB55404. The annotated chromosome sequences have been deposited at DDBJ/ENA/GenBank with accession numbers OX291632-OX291648 for UCD646 and OX291491-OX291507 for UCD650. The raw reads were deposited at SRA with accession number ERR10084971 for UCD646 and ERR10084972 for UCD650.