OUP user menu

Evolutionary relationships between the former species Saccharomyces uvarum and the hybrids Saccharomyces bayanus and Saccharomyces pastorianus; reinstatement of Saccharomyces uvarum (Beijerinck) as a distinct species

Huu-Vang Nguyen, Claude Gaillardin
DOI: http://dx.doi.org/10.1016/j.femsyr.2004.12.004 471-483 First published online: 1 February 2005


Analysis of the nucleotide sequence of the GDH1 homologues from Saccharomyces bayanus strain CBS 380T and S. pastorianus strains showed that they share an almost identical sequence, SuGDH1*, which is a diverged form of the SuGDH1 from the type strain of the former species S. uvarum, considered as synonym of S. bayanus. SuGDH1* is close to but differs from SuGDH1 by the accumulation of a high number of neutral substitutions designated as Multiple Neutral Mutations Accumulation (MNMA). Further analysis carried out with three other markers, BAP2, HO and MET2 showed that they have also diverged from their S. uvarum counterparts by MNMA. S. bayanus CBS 380T is placed between S. uvarum and S. pastorianus sharing MET2, CDC91 sequences with the former and BAP2, GDH1, HO sequences with the latter. S. bayanus CBS 380T has been proposed to be a S. uvarum/S. cerevisiae hybrid and this proposal is confirmed by the presence in its genome a S. cerevisiae SUC4 gene. Strain S. bayanus CBS 380T, with a composite genome, is genetically isolated from strains of the former S. uvarum species, thus justifying the reinstatement of S. uvarum as a distinct species.

  • Saccharomyces bayanus
  • Saccharomyces uvarum
  • Saccharomyces pastorianus
  • Taxonomy
  • GDH1
  • MET2
  • BAP2

1 Introduction

Within the Saccharomyces sensu stricto species, the Saccharomyces pastorianus taxon includes strains isolated from beer and initially considered as distinct species, such as S. carlsbergensis, S. monacensis and all lager brewing strains. S. pastorianus strain CBS 1513 (S. carlsbergensis) was thought to be a hybrid between S. cerevisiae and either S. bayanus, based on DNA/DNA relatedness [1], or strain CBS 1503 (S. monacensis), based on sequences of MET2 or ACB1 genes [2,3]. Later work has shown that strain CBS 1503 is itself a hybrid [4,5] and brought again S. bayanus close to S. pastorianus because they share common chromosomes or genes [68].

The S. bayanus taxon is complex and includes several synonym species [1]; molecular analyses [9,10] have divided S. bayanus strains into two subgroups: (1) a prevalent strain group encompassing strains similar to the former species S. uvarum Beijerinck, typified by strain CBS 395T[11]; (2) a minor group with strain S. bayanus CBS 380T and a few ancestral strains. We have previously proposed that S. bayanus CBS 380T should be considered as a hybrid between S. uvarum and S. cerevisiae, because several chromosomes of CBS 380T are isomorphic with S. uvarum chromosomes and because this strain carries the S. cerevisiae Y' sequence [10]. Further molecular analysis by AFLP (Amplified Fragment Length Polymorphism) [12] has confirmed this proposal. RAPD (Random Amplified Polymorphism DNA) whole-genome analysis has placed S. bayanus CBS 380T outside of the S. uvarum strain group [13]. The hybrid nature of S. bayanus CBS 380T is also supported by its physiological profile, intermediate between that of S. uvarum and S. cerevisiae[14].

If S. pastorianus had resulted from hybridisation events between S. bayanus and S. cerevisiae, and if S. bayanus was itself a S. uvarum/S. cerevisiae hybrid, S. uvarum-specific sequences might be found in S. pastorianus. However, genome exploration carried out on partial gene sequences of S. uvarum and S. pastorianus has indicated only around 93% of sequence homology [5]. Therefore, there is a need to align the entire sequences of several markers to determine whether S. pastorianus sequences diverge from their S. uvarum homologues as a consequence of long adaptation of S. pastorianus to brewing conditions in the past. In this work we address this issue by comparing sequences of genes GDH1, MET2, BAP2 and HO in the principal strains of the complex S. uvarum, S. bayanus and S. pastorianus.

2 Materials and methods

2.1 Yeast strains and molecular techniques

Strains used in this study are listed in Table 1. S. uvarum reference strains are CBS 395T (CLIB 251) and CLIB 533 (623-6c), an ura 3-1 mutant [15] selected from a monosporic culture issued from strain MCYC 623 (also deposited as CBS 7001). MCYC 623 and 623-6c have been used in three sequencing projects from which sequences were labelled S. bayanus[1618]. Natural hybrids CID1 [19] and S6U [20] were from Piškur, and constructed hybrid (S. uvarum/S. cerevisiae) H1 was from Rainieri and co-workers [21]. Strains CBS 377, CBS 426, CBS 1462 and CBS 2156 are from H. Fukuhara, they were employed in this study after re-identification as S. uvarum or hybrids by PCR/RFLP of the MET2 or NTS2 marker and by karyotyping, followed by Southern blotting and hybridisation with S. cerevisiae ScGDH1 (Fig. 1). All other molecular techniques were as previously described [10].

View this table:
Table 1

List of yeast strains employed

CBS No.CLIB No.Species namesInitial namesSourcesYear of isolationRemarks
(a) S. uvarum subgroup
395251Type strainGrappes1898Considered as synonym of S. bayanus
377S. intermedius var. turicensisPear wine1934
426S. bayanusHoney?
111S. bayanus (exS. uvarum)Wine1986
398S. bayanus (ex S. uvarum)Cider1971
2946393S. bayanus??
431401S. tubiformisFermented pear juice?
7001(1) (623-6c)533S. abuliensisM. adopersus1978Derivative of CBS 7001
(b) S. bayanus
380181Type strainBeer1895
395B19-3C derived from CBS 380
424250S. globosusPear juice1924Synonym
425255S. heterogenicusFermented apple juice1924Synonym
1505254S. intermedius var. valdensisBeer1904Synonym
1546252S. inusitatusBeer1965Synonym
(c) Saccharomyces hybrids
8614620(S. bayanus xCiderCID1 natural
8615621S. cerevisiae)CiderS6U natural
477(S. uvarum x S. cerevisiae)WineH1 constructed
1462792(2)S. pastorianusBeer1924Natural
2156795(2)S. bayanus/pastorianusBeer1949Natural
(d) S. pastorianus
1538537Neotype strainBeer1895from CBS
(3)281Beer?NRRL Y-1551 from ARS
1513176S. carlsbergensisBeer1908
1503180S. monacensisBeer1908
(e) S. cerevisiae
1171227Type strain1883
112YNN295 (4)
  • (1) Known also as MCYC 623. (2) Hybrid determined in this study. (3) Not similar to CBS 1538 as generally admitted. (4) Standard strain for chromosome size determination. CBS: Centraalbureau voor Schimmelculture. CLIB: Collection de Levures d'Intérêt Biotechnologique. ARS: Agricultural Research Service.

Figure 1

Detection of GDH1 in Saccharomyces strains. Left panels: Electrophoretic karyotypes of yeast strains stained with ethidium bromide. Right panels: Southern-blot hybridisation with ScGDH1 probe. (a) Strains: Y: YNN 295; H1: hybrid S. cerevisiae×S. uvarum; Ci: CID1; S6: S6U; Su: S. uvarum CBS 395T; C21: CBS 2156; Sc: S. cerevisiae S288c; P1: S. pastorianus CBS 1538NT; P2: S. pastorianus NRRL Y-1551; Sb: S. bayanus CBS 380T. Arrows indicate chromosomes carrying GDH1 of S. cerevisae and S. uvarum. (b) Strains: Ca: S. carlsbergensis CBS 1513; gl: S. globosus CBS 424; it: S. intermedius CBS 1505; he: S. heterogenicus CBS 425; in: S. inusitatus CBS 1546; C4: S. uvarum CBS 426; C14: CBS 1462; Mo: S. monacensis CBS 1503; Mc: CLIB 533 (623-6c).

2.2 PCR amplification and sequence analyses

Deposited sequences and primers designed for PCR amplification in this work are listed in Table 2. Conditions for PCR were as follows: 4 min at 94 °C followed by 30 cycles with 30 s at 94 °C, 30 s at 45 °C, 2 min at 72 °C and a final elongation for 5 min at 72°. The mixture contained 5 pmoles of each primer, 10 nmol of dNTP, 1 unit ExTaq (TaKaRa Bio Inc., Otsu, Shiga, Japan) and 25 ng of DNA from yeast strains in a final volume of 25 μl. These conditions were used throughout, except with higher Tm in some markers (Table 2).

View this table:
Table 2

Primers developed for amplification and sequencing of markers

Marker/primersSequence (5′–3′)Tm (°C)Origins/sequencing primers
SuGDH1 ORF45AJ418037
SbgdhF1AACGGTAAGGAATCCTTCInternal primers for
SbgdhR1CGGAAATGTATTGGACTTTGS. bayanus/pastorianus
SuGDH1 IGS45AJ418037
SuCDC9148PORF 17213

Direct sequencing of PCR fragments was carried out (Genome Express, Meylan, France). Sequences obtained were analysed with the Staden package [22] and the GCG Wisconsin package (Genetics Computer Group, Madison, WI, USA); they are deposited in the EMBL sequence database (AJ sequences, Table 3). Existing sequence resources of Saccharomyces were retrieved from the databases: SGD (http://genome-www.stanford.edu), NCBI (http://www.ncbi.nlm.nih.gov/) and Génolevures (http://cbi.labri.u-bordeaux.fr/genolevures).

View this table:
Table 3

Sequences determined or used in this study

SpeciesStrain numbers
S. cerevisiaeNTCLIB 227 = CBS 1171AJ627632
S. uvarumTCLIB 251 = CBS 395AJ627640AJ627638AJ627876AJ630206
S. uvarumCLIB 533 = 623-6cAJ418037AY144803
S. uvarumCLIB 398AJ627874AJ627875
S. bayanusTCLIB 181 = CBS 380AJ627639AJ627635AB049009AB027451AJ630207AJ628136
S. bayanusCLIB 395 = B19-3CAJ628137
S. pastorianus NTCLIB 537 = CBS 1538AJ627636NA
S. pastorianusCLIB 281 = Y-1551AJ627637AJ630208
ex S. carlsbergensisCLIB 176 = CBS 1513AJ627641(*)NAAJ627633
ex S. monacensisCLIB 180 = CBS 1503AJ627642NAAJ627634
S. pastorianusM204L16688
S. pastorianusUnknownAB049008AB027450
  • AJ numbers are from EMBL database; other numbers are from GenBank.

  • (*) Identical to L16688.

  • NA, not amplifiable.

2.2.1 GDH1 gene of S. uvarum, S. bayanus CBS 380T and S. pastorianus strains

The Génolevures project [16] provided genomic data from strain S. uvarum CLIB 533 (623-6c). Clone AS0AA004C04DP1 contains partially the homologue (Accession No. AL397636) of the S. cerevisiae GDH1 (ScGDH1) gene, encoding an NADP-linked glutamate dehydrogenase. This clone was sequenced entirely on both strands and aligned with ScGDH1 to delimit the GDH1 homologue in S. uvarum (SuGDH1) and to design three primer pairs to amplify the coding sequence (CDS) of SuGDH1 (SugdhU/SugdhL) and its up/downstream intergenic sequences IGS (SugPrU/SugPrL and SugIgU/SugIgL, respectively). These primers were used to amplify and to sequence the GDH1 homologues, CDS and its up/down IGS from S. uvarum CBS 395T, CLIB 111, S. bayanus CBS 380T and S. pastorianus CBS 1503 and CBS 1513.

2.2.2 MET2, BAP2 and CDC91 of S. uvarum

These markers were amplified and sequenced with the primers designed from the MET2 sequence (Accession No. L16688) of S. pastorianus. The BAP2 gene of S. uvarum (SuBAP2) and its upstream IGS were amplified and sequenced with primers TatIgs/SubapL developed from the two contiguous genes TAT1 (PORF 852 MIT_Sbay_c110_852) and BAP2 (Accession No. AB049008).

2.2.3 GDH1, MET2, SUC2 and SUC4 of S. cerevisiae

The above S. cerevisiae markers were amplified and sequenced with primers based on the sequences available in the SGD database (Table 2). The GDH1 PCR fragment from strain S288c and the SUC4 PCR fragment from strain CBS 1171NT were used as probes in Southern-hybridisation experiments.

2.2.4 D1/D2 and NTS2 of ribosomal DNA

The D1/D2 and NTS2 (Non Transcribed Sequence) of rDNA of several strains were amplified and sequenced using the primer NL1/NL4 [23] and the up/lower primers described in [10], respectively.

2.3 Phylogenetic inference based on GDH1 and MET2 coding sequences

GDH1 and MET2 CDS of S. uvarum CBS 395T, S. bayanus CBS 380T and S. pastorianus sequenced in this work were aligned with their homologues from S. cerevisiae, S. paradoxus and S. pastorianus in databases using the Schizosaccharomyces pombe genes as outgroups. The phylogenetic trees were generated from the softwares at TreeTop-Phylogenetic Tree Prediction Service (http://www.genebee.msu.su/services/phtree_reduced.html).

3 Results and discussion

3.1 Detection of GDH1 markers in Saccharomyces strains

The ScGDH1 (YOR375c) probe was first hybridised to the chromosome blots of several strains. Fig. 1 shows that in S. cerevisiae strains YNN 295 and S288c this probe hybridised to chromosomes XV and IV, where GDH1 and GDH2, respectively, are located (Fig. 1(a), lanes Y and Sc). On chromosome blots of S. uvarum strains and S. bayanus CBS 380T (Fig. 1(a) and (b), lanes Su, Sb, C4 and Mc), the ScGDH1 probe hybridised with two chromosomes, the first one corresponding to S. uvarum chromosome 16, which has the same size as S. cerevisiae chromosome IV, and the second one to chromosome 10 [10]. In the hybrids, H1, CID1 and S6U the probe was detected at three positions: one at chromosome XV specific to S. cerevisiae, one at chromosome 10 specific to S. uvarum and one at the position corresponding to both S. cerevisiae chromosome IV and S. uvarum chromosome 16 (Fig. 1(a), lanes H1, Ci, S6). Strains CBS 2156 and CBS 1462 (Fig. 1(a), lane C21; 1B, lane C14) showed a GDH1 hybridisation profile similar to that of the hybrids, but two signals were detected at the position of chromosome XV, suggesting the presence of two copies of ScGDH1. In S. pastorianus strains, the ScGDH1 probe revealed two types of profile: the S. uvarum profile in strains CBS 1538NT, NRRL Y-1551 (Fig. 1(a), lanes P1, P2), and the hybrid profile in strains CBS 1503 (S. monacensis) and CBS 1513 (S. carlsbergensis) with, in this latter case, a fourth signal detected near the position of chromosome IV (Fig. 1(b), lanes Ca and Mo). The S. uvarum profile was also observed in the synonyms of S. bayanus: S. globosus, S. intermedius, S. heterogenicus and S. inusitatus (Fig. 1(b), lanes gl, it, he and in).

3.2 Sequence comparison of GDH1 from S. uvarum, S. bayanus and S. pastorianus. Evidence of multiple neutral mutations accumulation

By using specific primers, ScGDH1 and SuGDH1 could be amplified from S. cerevisiae or S. uvarum strains (Fig. 2 lanes Sc, Su and C111). In S. bayanus CBS 380T only SuGDH1 was amplified. On the other hand, both markers could be amplified in strain CBS 1503 (S. monacensis) and the hybrids CBS 1462 and CBS 2156 (Fig. 2, lanes Sb, Mo, C14 and C21). In strain CBS 1513 (S. carlsbergensis) only SuGDH1 could be amplified; ScGDH1, although evidenced by Southern hybridisation (Fig. 1), could not be amplified. For the S. pastorianus strains CBS 1538NT and NRRL Y-1551, only SuGDH1 was amplified (data not shown).

Figure 2

PCR amplification of ScGDH1 (upper panel) and SuGDH1 (lower panel). Strains: Sc: S. cerevisiae S288c; Su: S. uvarum CBS 395T; C111: S. uvarum CLIB 111; Sb: S. bayanus CBS 380T; Ca: S. carlsbergensis CBS 1513; Mo: S. monacensis CBS 1503; C14: CBS 1462; C21: CBS 2156.

Since hybridisation and PCR experiments suggested the presence of an S. uvarum GDH1 (SuGDH1) homologue in S. bayanus and S. pastorianus, sequencing of SuGDH1 PCR fragments was carried out. Alignment showed that there are two forms of this gene: the SuGDH1 form and its diverged form SuGDH1*. The SuGDH1 CDS, 1365 bp, and its up/downstream IGS (420 and 313 nucleotides, respectively) were identical in S. uvarum type strain CBS 395T (AJ627640), strains CLIB 111 and CLIB 533 (AJ418037). This sequence is identical to that in database (WashU_Sbay_Contig 673-22, MIT_Sbay_C773 PORF 24096) with strain sequences of 623-6c and MCYC 623 [16,17]. Four other S. uvarum strains of different origins: CBS 377, CBS 426, CBS 2946 and CLIB 398, exhibited also SuGDH 1 CDS.

From S. bayanus CBS 380T the GDH1 sequence obtained was designated as SuGDH1* because it presented 97% nucleotide identity with SuGDH1 over the CDS, while the up/downstream IGS presented only 86% nucleotide identity due to several indels. Alignment of SuGDH1/SuGDH1* CDS showed that they differ by 42 nucleotides, of which 39 were silent substitutions, 36 at the third position of the corresponding codons and three corresponding to different leucine codons: CTA(+481) to TTG and TTG(+1123) to CTG. Three other substitutions resulted in predicted functionally similar amino-acid changes: ACA(+499) to TCA (S to T) and GTC(+790) to ATT (V to I). The nucleotide sequences were thus 97% identical, while the deduced amino-acid sequences were 99.6% identical. Thus the SuGDH1* in S. bayanus CBS 380T is apparently the diverged form of SuGDH1, identical in several S. uvarum strains. In the case of S. pastorianus strains, SuGDH1* was found in strain NRRL Y-1551; or with one substitution: in strain CBS 1513 (S. carlsbergensis), T306 replaced C (silent), while in strains CBS 1538NT and CBS 1503 (S. monacensis), G89 was replaced by A, resulting in a R(AGA) to K(AAA) change. The pair SuGDH1/SuGDH1* indicates a 3% genetic distance separating S. bayanus/pastorianus from S. uvarum. This distance is less than that which separates two sibling species such as S. cerevisiae and S. paradoxus whose GDH1 homologues differ by 6%, presenting 84 substitutions of which 61 are silent (http://genome-www4.stanford.edu/). On the other hand, SuGDH1 and ScGDH1 CDS differ by 11%. The deduced 454 amino-acid sequence from SuGDH1 showed 95% identity with that of ScGDH1.

The ancient species classified as synonyms of S. bayanus: S. globosus CBS 424, S. heterogenicus CBS 425, S. intermedius var. valdensis CBS 1505 and S. tubiformis CBS 431, carry the SuGDH1 sequence (except for one silent substitution in S. globosus CBS 424). In contrast, the species S. inusitatus CBS 1546 carries the SuGDH1* allele. Thus the SuGDH1 CDS is conserved in S. uvarum strains while its diverged form SuGDH1* is present in S. bayanus CBS 380T, S. inusitatus CBS 1546 and S. pastorianus. Hence the SuGDH1* in these species has diverged from SuGDH1 by the accumulation of a number of neutral mutations, termed Multiple Neutral Mutations Accumulation (MNMA), far more than the natural polymorphism.

3.3 GDH1 alleles of S. uvarum and S. cerevisiae in the hybrids

In the S. uvarum/S. cerevisiae constructed hybrid H1 or in S6U (CBS 8615) and CID1 (CBS 8614), considered as natural S. bayanus/S. cerevisiae hybrids, the ScGDH1 allele was unchanged, while the SuGDH1 sequence was unchanged in H1 and S6U or with one silent substitution in CID1. CBS 8615 (S6U) and CBS 8614 (CID1) are thus redefined as hybrids between S. uvarum and S. cerevisiae. Strains CBS 1462 and CBS 2156 were recognised as hybrids because the former carries the combination ScGDH1/SuGDH1 while the latter carries the combination ScGDH1/SuGDH1*. All the comparisons of SuGDH1/ScGDH1 sequences in Saccharomyces strains are shown in Table 4.

View this table:
Table 4

Sequence variation in S. bayanus/pastorianus compared with S. uvarum and S. cerevisiae

Species/strain numberMarkers
%, SnSub%%, SnSub%, SnSub%, SnSub%, SnSub%, snSub%, SnSub%, Sn
CDS length (bp):1365136514611461176118301185
Reference sequences:AJ1627640YOR375cAJ1627638YNL277wE08858AJ1627876AJ630206AJ279065AJ243219
S. uvarum
CBS 395T/CLIB 251100A100ANd100100100100
CBS 377100A100ANd100(b)
CBS 4261/1A100ANd100(b)
CLIB 111100A100ANd100100100(b)
CLIB 3982/2A7/7ANd100100(b)
CBS 2946/CLIB 393100A100ANd10089% (c)
CLIB 533/623-6c100A100A1001001/1100100
S. bayanus
CBS 380T/CLIB 181SuGDH1*36/42A100A76/96 (a)Lg-BAP2 (d)AJ630207AF113892189% (c)
B19-3C/CLIB 395SuGDH1*36/42AANdNdNd89% (c)
S. globosus
CBS 424/CLIB 2501/1A5/5ANdLg-BAP2 (e)NdNd89% (c)
S. heterogenicus
CBS 425/CLIB 255100ASuMET2*74/91ANdLg-BAP2 (e)NdNdNd
S. intermedius
CBS 1505/CLIB 254100ASuMET2*74/91ANdLg-BAP2 (e)NdNdNd
S. inusitatus
CBS 1546/CLIB 252SuGDH1*36/42ASuMET2*74/93ANdNdNdNdNd
S. pastorianus
CBS 1538NT/CLIB 537SuGDH1*36/43ASuMET2*74/914/5NdNdNAAF113893189% (c)
syn S. carlsbergensis
CBS 1513/CLIB 176SuGDH1*37/43NASuMET2*74/914/576/96 (a)Lg-BAP2NAAF113893189%
syn S. monacensis
CBS 1503/CLIB 180SuGDH1*36/43100SuMET2*74/91A76/96 (a)Lg-BAP2NAAF113893189%
NRRLY-1551/CLIB 281SuGDH1*36/42A100ANdLgBAP2AJ630208AF113893189% (c)
CBS 8614/CLIB 620; CID11/1100PCR/P1/2NdNd100100Nd
CBS 8615/CLIB 621 S6U100PCR/P1/2NdNd100100Nd
CBS2156/CLIB795SuGDH1*36/42100SuMET2*74/913/3NdNdNdU44806 (Sc)Nd
H1100100100PCR/ENdNd100mixte Sc/SuNd
  • %: per cent of identity, Sn: sequence name.

  • Sub: number of silent substitutions versus total.

  • (a) Determined after alignment of the CDS from E08858 (S. uvarum) and from AB027451 (lager strain). (b) Identical AluI pattern with the NTS2 of S. uvarum CBS 395T. (c) 100% With the sequence NTS2 of strain CBS 1513 (AJ243214) and CBS 1503 (AJ243215). Nd, non-determined; A, absent; NA, not amplifiable. (d) Lg-BAP2 Accession No. AB049008. (e) Sequences determined in this study identical to Lg-BAP2.

  • PCR/P: identified by PCR amplification of the MET2 partial gene and presence of PstI (P) site for S. uvarum (Su) or EcoRI (E) site for S. cerevisiae (Sc).

3.4 MET2 gene of S. uvarum, S. pastorianus and S. bayanus CBS 380T

Saccharomyces uvarum CBS 395T and S. bayanus CBS 380T exhibited a conserved SuMET2 CDS over 1461 nucleotides (Accession No. AJ627635, AJ627638). In five other S. uvarum strains only CLIB 398 presented SuMET2 with seven silent substitutions (AJ627875). On the other hand MET2 of CBS1513 (S. carlsbergensis) differs from SuMET2 by 91 nucleotide substitutions (93.7% identity), of which 74 are silent; nevertheless, the two deduced amino-acid sequences have 97% identity. As for the case of SuGDH1, this figure allows us to consider that S. pastorianus strains carry the diverged form of SuMET2 designated as SuMET2*. Comparatively, the MET2 of S. cerevisiae and that of S. paradoxus are separated by 127 substitutions of which 106 are silent (http://genome-www4.stanford.edu/).

PCR/RFLP of the MET2 fragments from nucleotide +29 to +598 has been used to differentiate S. bayanus/S. uvarum from S. pastorianus[5,24], based on the presence or absence of a single restriction site: Pst restriction site identifies S. bayanus and S. uvarum strains while BamHI restriction site identifies S. monacensis and S. carlsbergensis. Sequence alignment indicated that the PstI site in SuMET2 at nucleotide +230 was absent from SuMET2* due to a silent substitution (G to A). Similarly, the BamHI site is present in SuMET2* but absent from SuMET2 due to an identical substitution at nucleotide +481. Thus identification based on a single-site indicator does not reflect the relatedness of two species (see Fig. 3).

Figure 3

Sequence changes in the SuMET2 alleles. Partial alignment of the SuMET2 ORF from S. uvarum (SuMET2) and S. pastorianus (SuMET2*), showing nucleotide substitutions, the loss of one PstI site and the gain of one BamHI site (boxes) in SuMET2* from S. pastorianus. Underlined are the sequences of the primers used to amplify the MET2 partial gene for identification of Saccharomyces species.

3.5 CDC91 conserved sequences in S. uvarum and S. bayanus CBS 380T

The CDC91 homologue of S. cerevisiae (encoding a GPI-anchored transamidase component) was amplified and sequenced from S. uvarum CBS 395T, S. bayanus CBS 380T and strain NRRL Y-1551. The CDC91 sequences found in S. bayanus CBS 380T (Accession No. AJ630207) and strain NRRL Y-1551 are identical to the PORF 17213 or Sbay_contig617.3 from strain MCYC 623 and 623-6c, respectively. The CDC91 sequence of S. uvarum CBS 395T presented one neutral substitution (Accession No. AJ630206), while it could not be amplified from S. pastorianus strains. As for the SuMET2 gene, the CDC91 sequence indicated that NRRL Y-1551 is different from S. pastorianus CBS 1538NT but similar to CBS 380T and closely related to S. uvarum. Strain NRRL Y-1551 exhibits also similar electrokaryotypes with CBS 380T and S. uvarum CBS 395T (Fig. 1(a), lanes P2, Sb, Su).

3.6 BAP2 and HO sequences from S. uvarum, S. bayanus CBS 380T and S. pastorianus

BAP2 encoding a branched-chain amino-acid permease and HO encoding the mating conversion endonuclease, were reported to be identical in S. bayanus CBS 380T and an S. pastorianus lager strain [7,8]. We amplified and sequenced an identical SuBAP2 from S. uvarum strains CBS 395T and CLIB 111. Alignment with the deposited Lg-BAP2 CDS revealed that, over 1830 nucleotides, there are 172 substitutions of which 104 are silent. The two CDS showed 90.2% identity both in nucleotide and in deduced amino-acid sequences.

The HO gene from strain S. uvarum EKB103 (Accession No. E08858) was also compared with the Lg-HO from the lager strain KBY001 (Accession No. AB027450): the two CDS (1761 nucleotides) differed by 96 substitutions of which 76 were silent. In comparison, between two strains EKB103 and MCYC 623 (MIT_Sbay_c627_3117), HO presented six silent substitutions. Thus, compared with MET2 genes, the HO genes of S. uvarum and S. pastorianus presented almost the same number of substitutions but over a longer sequence. Lg-BAP2 and Lg-HO are thus diverged from their homologues in S. uvarum, SuBAP2 and SuHO by MNMA.

3.7 Confirmation of the hybrid nature of S. bayanus CBS 380T by the presence of the S. cerevisiae SUC4 sequence

The presence of S. cerevisiae Y′, X and SUC2 sequence has previously been revealed by chromosomal blot in S. bayanus CBS 380T[5,10,25]. SUC genes are located at the telomeric regions in S. cerevisiae[26]. A SUC gene was indeed amplified and sequenced from S. bayanus CBS 380T, S. cerevisiae CBS 1171NT and from S. pastorianus strains, but it corresponds to the S. cerevisiae SUC4 gene. Hybridisation of the ScSUC4 amplified from strain CBS 1171NT with Southern-blotted electrokaryotypes revealed strong signals from three chromosomes of S. bayanus CBS 380T or its derivative B19-3C and from more chromosomes of S. pastorianus and S. cerevisiae strains S288C (Fig. 4). S. uvarum strains presented only a non-specific signal (Fig. 4, lanes Su and Mc).

Figure 4

Probing of S. cerevisiae SUC4 gene to the chromosomes of Saccharomyces strains. Left panel: Karyotypes stained with ethidium bromide. Right panel: S. cerevisiae SUC4 probing. Strains: ST: S. cerevisiae CBS 1171T; Sc: S. cerevisiae S288c; Sb: S. bayanus CBS 380T; Bc: B19-3C derivative of CBS 380T; Su: S. uvarum CBS 395T; Mc: S. uvarum CLIB 533 (623-6c); Ca: S. carlsbergensis CBS 1513; Mo: S. monacensis CBS 1503.

SUC4 nucleotide sequences from S. cerevisiae CBS 1171NT (Accession No. AJ826132) and S. bayanus CBS 380T (Accession No. AJ826136) are almost identical with the X07572 sequence (positions 210–2488). Thus, S. bayanus CBS 380T carries the SUC4 gene, CDS and up/downstream IGS, originating from S. cerevisiae. An identical ScSUC4 CDS (Accession No. AJ826132) was also amplified and sequenced from S. pastorianus CBS 1538NT, S. carlsbergensis, S. monacensis and strain NRRL Y-1551. In strain CLIB 395 (B19-3C), a derivative of S. bayanus CBS 380T[27], the ScSUC4 sequence (Accession No. AJ826137) has a (C) insertion at position +100.

3.8 Conclusion

As reported in bacteria, protein-encoding genes evolve more rapidly than rRNA genes [28]. In yeast, coding sequences (CDS) are more informative than intergenic sequences (IGS) to resolve closely related species [17]. In this study, GDH1, MET2, HO and BAP2 CDS served as basis for the elucidation of phylogenetic relations for S. uvarum and S. bayanus on the one hand and, interestingly, between S. uvarum and S. pastorianus on the other hand (Fig. 5). In the hybrids S. bayanus and S. pastorianus, sequence divergence is significant, but still less than that observed between closely related Saccharomyces sibling species that are thought to have diverged millions of years ago. Thus S. bayanus and S. pastorianus have diverged from S. uvarum recently. The GDH1 gene having few divergences indicates clearly the relatedness between S. bayanus and S. pastorianus with S. uvarum. The same relatedness can be seen with the MET2, HO, and to a lesser extent, with the BAP2 gene. All the markers used are dispersed among five S. cerevisae chromosomes and probably in the same number of S. uvarum chromosomes because of the high synteny (97%) between these two species [29]. This leads us to interpret that these divergences reflected a recent evolution from S. uvarum to S. bayanus CBS 380T and S. pastorianus rather than the results of lateral transfers of several markers. In the hybrid genome context, sequence comparison suggested that SuGDH1 and SuMET2 had evolved gradually but independently, because several strains carried both SuMET2*/SuGDH1*, but there were also combinations of either SuMET2/SuGDH1* as in strains S. bayanus CBS 380T, NRRL Y-1551 or SuMET2*/SuGDH1 as in strains CBS 425 and CBS 1505. This evolution is not observed with the hybrids CID1, S6U and CBS 1462, they carry sequences identical with S. uvarum as does the hybrid H1, recently constructed.

Figure 5

Phylogenetic trees established by multiple alignment of GDH1 (top) and MET2 (bottom) CDS determined in this study or from databases. The MET2 and GDH1 homologues of Sch. pombe are extracted from deposited sequences (AL023288 and AL033127).

Based on the sequences of GDH1 and MET2, a phylogenetic relationship between S. uvarum and S. pastorianus was established (Fig. 5). Taking all the data together, a plausible scheme is proposed to explain the evolution relationships between S. uvarum, S. bayanus and S. uvarum, S. pastorianus (Fig. 6).

Figure 6

Possible routes of evolution from S. uvarum to S. bayanus and S. pastorianus as suggested by sequence analysis.

3.8.1 S. uvarum strains form a homogeneous cluster; S. bayanus CBS 380T is genetically isolated from S. uvarum

As previously demonstrated, S. uvarum strains form a cluster: similar and characteristic karyotypes have been obtained with S. uvarum isolates [10,3032]. PCR/RFLP of NTS2 (rDNA) by AluI [9,10] and of MET2 by PstI [24] suggested that they share these same sequences. Sequence conservation in BAP2, CDC91, GDH1, HO and MET2 in many strains, identical D1/D2 sequences in different isolates (Table 4) [33] have demonstrated their genetic uniformity. Genetic inter-fertility between strain 623-6c, a derivative of MCYC 623, used as tester and all strains characterized as S. bayanus (ex S. uvarum) have demonstrated full genetic exchange between them [30]. All S. uvarum strains share common physiological characteristics: they are all Gal+, Mel+ and psychrotrophic (the growth is inhibited above 35 °C).

Until now S. bayanus CBS 380T and S. uvarum have been considered as conspecific because high DNA/DNA homology, but S. bayanus CBS 380T shares with S. uvarum common genes such as MET2 and CDC91 and diverged ones such as GDH1, BAP2 and HO together with S. cerevisiae Y′ and SUC4 sequences, a status corresponding to a hybrid S. uvarum/S. cerevisiae. Therefore CBS 380T is no longer conspecific with S. uvarum.

3.8.2 Reinstatement of S. uvarum as distinct species

Rossellö-Mora and Amann [34] have proposed the following phylo-phenetic species concept for prokaryotes: “The species could be described as a monophyletic and genomically coherent cluster of individual organisms that show a high degree of overall similarity in many independent characteristics, and is diagnosable by a discriminative phenotypic property”. This concept is entirely applicable to S. uvarum strains. They all share the same genetic and physiological characters defining a clade, allowing genetic exchange by inter-fertility and identifiable by a set of common characters. Therefore,we propose the reinstatement of S. uvarum Beijerinck as a distinct species, abolishing its current status as synonym of S. bayanus. In the reinstated S. uvarum taxon the Type strain is CBS 395. The former species S. abuliensis Santa Maria (CBS 7001T), S. intermedius E.C. Hansen var. turicensis Osterwalder (CBS 377T) and S. tubiformis Osterwalder (CBS 431T) are synonyms, while S. globosus Osterwalder (CBS 424T) and S. intermedius E.C. Hansen var. valdensis Osterwalder (CBS 1505T) are derivative strains (Table 5). S. inusitatus van der Walt (CBS 1546T) is reclassified as S. pastorianus. The name S. bayanus is retained for designation of partial hybrids S. uvarum/S. cerevisiae similar to strain CBS 380T, such as strain NRRL Y-1551 and S. heterogenicus Osterwalder (CBS 425T). The name S. bayanus var. uvarum (nomen invalidum) used by some authors to designate strains of S. uvarum species is no longer valid. The group Saccharomyces sensu stricto defined by Vaughan-Martini and Kurtzman [1] is thus redefined, it contains three sibling species S. cerevisiae, S. paradoxus, S. uvarum, and the hybrids between S. cerevisiae and S. uvarum, classified either as S. bayanus or as S. pastorianus.

View this table:
Figure 5

Reclassification of S. uvarum, S. bayanus and S. pastorianus strains

Old epithetsStrain numbersProposed namesStatus
S. uvarum var. uvarumCBS 395Saccharomyces uvarumType strain
S. abuliensisCBS 7001 (a)S. uvarumSynonym
S. abuliensis (derivative)CLIB 533 (b)S. uvarumMutant
S. intermedius var. turicensisCBS 377S. uvarumSynonym
S. tubiformisCBS 431S. uvarumSynonym
S. globosusCBS 424S. uvarumDerivative
S. bayanusCBS 2946S. uvarumDerivative
S. bayanusCBS 380Saccharomyces bayanusType strain
S. bayanusB19-3CS. bayanusMutant
S. pastorianusNRRL Y-1551S. bayanus
S. heterogenicusCBS 425S. bayanusSynonym (c)
S. intermedius var. valdensisCBS 1505S. bayanusDerivative (d)
S. inusitatusCBS 1546Saccharomyces pastorianusSynonym
  • (a) Also designated as MCYC 623 in the MIT sequencing project [18]. (b) Derivative of CBS 7001 designated as 623-6c. Used in the Genolevures and Washington University sequencing projects [16,18]. (c) Presence of S. cerevisiae Y′[10]. (d) Complex NTS2 AluI pattern [9,10].

3.8.3 S. pastorianus contains genetic material from more than one S. uvarum derivative

In S. pastorianus several genes, GDH1 or MET2, originating from S. cerevisiae, are well-conserved while the S. uvarum homologues have undergone divergence: SuGDH1*, SuMET2*, Lg-HO and Lg-BAP2. As S. uvarum has previously been classified as synonym of S. bayanus, the proposal of Vaughan-Martini and Kurtzman [1] is still validated. Recent studies using chromosomal or AFLP analysis [5,12] and gene sequences [7,8] have supported this proposal. As can be observed, SuGDH1*, SuMET2*, Lg-HO and Lg-BAP2 sequences diverged independently, suggesting that they may have been brought over by one or two different contributors, possibly by rare-mating which has been shown possible with several hybrids [12]. One of the contributors may be S. bayanus strains CBS 380T or NRRL Y-1551. However, proteome data correlate strain NRRL Y-1551 and CBS 380T with S. pastorianus: 69% of the proteins of the former and 35% of the proteins of the latter co-migrate with lager strain proteins in a 2D electrophoresis system [4]. Other S. uvarum derivatives such as strains CBS 424 and CBS 2946 (Table 4) may also be possible contributors. Further work will be necessary to designate precisely the partner(s) of S. cerevisiae in the composite genome of S. pastorianus.


We thank Dr. H. Fukuhara for providing us the CBS strains of his personal collection; Dr. J. Piskur for the gift of the CID1 and S6U natural hybrids; Dr. S. Rainieri for the H1 hybrid and its parent strains. N.H.V. is very indebted to C.R. Tinsley (INA-PG) for helpful discussion and language corrections of the manuscript; Dr. C. Neuvéglise for her help during the sequencing of the clone AS0AA004C04.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
View Abstract