FEMS Yeast Research, 15, 2015, fov090 doi: 10.1093/femsyr/fov090 Advance Access Publication Date: 12 October 2015 Research Article
RESEARCH ARTICLE
2μ plasmid in Saccharomyces species and in Saccharomyces cerevisiae Pooja K. Strope1 , Stanislav G. Kozmin1,† , Daniel A. Skelly2 , Paul M. Magwene2 , Fred S. Dietrich1 and John H. McCusker1,∗ 1
Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC 27710, USA and 2 Department of Biology, Duke University, Durham, NC 27708, USA ∗ Corresponding author: Department of Molecular Genetics and Microbiology, DUMC, Jones Bldg., Rm. 239, 207 Research Drive, 3020, Durham, NC 27710, USA. Tel: +(919) 681-6744; Fax: +(919) 684-8735; E-mail: [email protected] † Present address: Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA. One sentence summary: We show extensive 2μ sequence variation in Saccharomyces cerevisiae and in other species of Saccharomyces and 2μ associations with S. cerevisiae populations, clinical origin and genotypes. Editor: Cletus Kurtzman
ABSTRACT We determined that extrachromosomal 2μ plasmid was present in 67 of the Saccharomyces cerevisiae 100-genome strains; in addition to variation in the size and copy number of 2μ, we identified three distinct classes of 2μ. We identified 2μ presence/absence and class associations with populations, clinical origin and nuclear genotypes. We also screened genome sequences of S. paradoxus, S. kudriavzevii, S. uvarum, S. eubayanus, S. mikatae, S. arboricolus and S. bayanus strains for both integrated and extrachromosomal 2μ. Similar to S. cerevisiae, we found no integrated 2μ sequences in any S. paradoxus strains. However, we identified part of 2μ integrated into the genomes of some S. uvarum, S. kudriavzevii, S. mikatae and S. bayanus strains, which were distinct from each other and from all extrachromosomal 2μ. We identified extrachromosomal 2μ in one S. paradoxus, one S. eubayanus, two S. bayanus and 13 S. uvarum strains. The extrachromosomal 2μ in S. paradoxus, S. eubayanus and S. cerevisiae were distinct from each other. In contrast, the extrachromosomal 2μ in S. bayanus and S. uvarum strains were identical with each other and with one of the three classes of S. cerevisiae 2μ, consistent with interspecific transfer. Keywords: 2-micron plasmid; 2μ plasmid; selfish genetic elements; population genomics; Saccharomyces cerevisiae; Saccharomyces genus
INTRODUCTION 2μ plasmid is a circular, multicopy, selfish genetic DNA element associated with the Saccharomyces cerevisiae genome. The canonical 6318 base pair 2μ plasmid contains four open reading frames (ORFs; FLP1, REP1, REP2, RAF1) and four cis-acting loci (an origin of replication (ORI); a stability (STB) locus; and 599 bp inverted repeats containing 48 bp Flp Recombination Targets (FRT)) (Futcher 1988; Jayaram et al. 2004; Chan et al. 2013). 2μ sexual/non-Mendelian inheritance, stability, site-specific re-
combination and copy number control in S. cerevisiae have been extensively studied; for reviews, see Futcher (1988), Jayaram et al. (2004) and Chan et al. (2013). 2μ is not present in all S. cerevisiae strains (Nakayashiki et al. 2005; Kelly et al. 2012) and exhibits sequence variation (Neuville, Bonneu and Aigle 1990; Wei, Pelcher and Rank 1991; Xiao, Pelcher and Rank 1991a,b; Rank, Xiao and Arndt 1994; Xiao and Rank 1996). Thus, 2μ presence/absence and sequence variation are both components of S. cerevisiae genetic variation and potentially so in other Saccharomyces species.
Received: 29 July 2015; Accepted: 1 October 2015 C FEMS 2015. All rights reserved. For permissions, please e-mail: [email protected]
1
2
FEMS Yeast Research, 2015, Vol. 15, No. 8
In this work, we describe for the 100-genomes S. cerevisiae strains (Strope et al. 2015) the presence/absence of 2μ and, where present, variation in 2μ copy number and sequences. We also examine the publicly available genome sequences of other Saccharomyces species to determine the presence/absence of both extrachromosomal and integrated 2μ; to establish 2μ phylogenies; and to assess the interspecific transfer of 2μ. Finally, we describe for the 100-genomes strains nuclear genotype, clinical/non-clinical origin and population associations with 2μ variation.
METHODS 2μ presence/absence, assembly, copy number determination and genotype class We sequenced 93 of the 100-genomes strains (Strope et al. 2015) and, by high read depth relative to single copy chromosomal genes, identified 67 as containing extrachromosomal 2μ; see Strope et al. (2015) for sequencing methodology. In brief, using the de novo assembler ABySS (v. 1.3.4), with parameters ‘k’: the k-mer length, ‘n’: the minimum number of pairs needed to join contigs, and ‘c’: the minimum mean k-mer coverage of a unitig, optimized for each strain, we assembled the 2μ from the 67 strains. We determined extrachromosomal 2μ copy number of these 67 2μ-containing strains by read depths relative to single copy chromosomal genes. The remaining seven of the 100genomes strains were sequenced in other studies: S288c (Goffeau et al. 1996), 1278b (Dowell et al. 2010), SK1 (Nishant et al. 2010), YJM789 (Wei et al. 2007), M22 (Doniger et al. 2008), RM11 (RM11 2004) and YPS163 (Doniger et al. 2008). The 100-genomes strains listed in Table S1 (Supporting Information) have been deposited in and should be requested from the Fungal Genetics Stock Center (http://www.fgsc.net). After the assembly of the 2μ sequences from the 2μcontaining 100-genomes S. cerevisiae strains, we manually converted the sequences to the same conformation as the reference S288c 2μ sequence in Saccharomyces Genome Database (SGD). We used BLAST (Altschul et al. 1990) and ssearch36 (Smith and Waterman 1981) to identify gene sequences in these assembled 2μ. Each of the four genes were extracted to build separate multiple alignments and phylogenies. Because the REP1–RAF1 and REP2–FLP1 pairs of gene phylogenies had similar topology (data not shown), we took those pairs of genes, and the intergenic sequence between those two genes, and built phylogenies. Specifically, for each of the 2μ-containing strains, sequences were gathered using blast hit coordinates for the REP1–RAF1 (S288c coordinates: 1887–3816 base pairs) and REP2–FLP1 (S288c coordinates: 5308–6318 and 1–1523 base pairs) gene pair regions. Sequence extraction and concatenation (for REP2–FLP1 sequences) was done using the EMBOSS tools seqret and union, respectively. Thus, we generated two ‘half circles’, consisting of the 3 end of RAF1 to the 3 end of REP1 (REP1–RAF1) and the 3 end of FLP1 to the 3 end of REP2 (REP2–FLP1). Multiple alignments were made using MAFFT (Katoh and Standley 2013). Phylogenies were estimated using neighbor joining in Clustal-w2 (Larkin et al. 2007). We also assessed the presence/absence of 2μ by PCR (2mF3: TGCATGATTAACTTCGAGAAGGG and 2m-R3: AAAGGACAAGGACCTGAGCG; RW319: GCCCTCAGTGATGGTGTTTT and RW320: CGTAAGCGTCGTTACCCAAT [Nakayashiki et al. 2005]) in the seven strains that we did not sequence. By both PCR and the absence of sequence reads, we found RM11 and YPS163 to lack 2μ. By both PCR and the presence of sequence reads, we found 1278b, SK1 and M22 to contain 2μ, which we assembled. For
the S288c background, which by PCR contains 2μ, we used the 2μ sequence in the SGD and in Hartley and Donelson (1980). For the YJM145 background, which is isogenic with YJM789 (Wei et al. 2007), PCR results showed that it contained 2μ. However, examination of YJM789 genome sequence data failed to identify 2μ sequences, which may be a consequence of the library preparation and/or sequencing technology. Because YJM145 is isogenic with YJM789 (Wei et al. 2007) and a segregant of YJM128 (McCusker et al. 1994), we used YJM128 sequence data (Magwene et al. 2011) to assemble 2μ. As 2μ interconverts between two forms by Flpmediated recombination at FRT sites (Futcher 1988; Jayaram et al. 2004; Chan et al. 2013), we assembled all 2μ into one form (Fig. 1). We screened short read archives and assembled sequences for extrachromosomal 2μ sequences from S. paradoxus (Liti et al. 2009; Bergstrom et al. 2014), S. kudriavzevii (Hittinger et al. 2010; Scannell et al. 2011), S. arboricolus (Liti et al. 2013), S. mikatae (Scannell et al. 2011) and S. uvarum (Kellis et al. 2003; Libkind et al. 2011; Scannell et al. 2011; Almeida et al. 2014), as well as S. eubayanus and the hybrid taxon S. bayanus (Cliften et al. 2003; Libkind et al. 2011; Almeida et al. 2014; Bing et al. 2014; Baker et al. 2015). In contrast to extrachromosomal 2μ, which is present in high copy number relative to single copy chromosomal sequences, 2μ sequences integrated into chromosomes should be present in approximately 1:1 ratio with known single copy chromosomal genes. On this basis, we screened for 2μ sequences integrated into nuclear chromosomes of S. cerevisiae and the other listed Saccharomyces species.
2μ genotype–phenotype and 2μ genotype-genotype associations in the 100-genomes strains We tested for associations between previously described phenotypes of the 100-genomes strains (Strope et al. 2015) and three aspects of 2μ genetic variation: 2μ presence/absence; 2μ genotype class; and, in the 67 2μ-containing 100-genomes strains with known copy number, 2μ copy number above/below median estimated copy number of 23. When testing 2μ presence/absence, we considered all of the 100-genomes strains. When testing 2μ genotype class, we excluded class C (n = 3) and included classes A (n = 46) and B (n = 23). When testing 2μ copy number, we considered only the 67 2μ-containing strains from our previous study (Strope et al. 2015). The program GEMMA version 0.94beta (Zhou and Stephens 2012) was used to conduct association tests. This program takes a linear mixed model approach to controlling population structure using a relatedness matrix (constructed from genotype data using GEMMA) constructed from 171 345 biallelic SNPs with minor allele frequency >5%, as previously described (Strope et al. 2015). We also tested for associations between the three aspects of 2μ genetic variation detailed above and genotypes at nuclear biallelic SNPs. The number of SNPs tested varied between 125 721 and 157 892 depending on the subset of strains assayed (the same subsets as explained above; only considering biallelic SNPs with minor allele frequency >5%). We examined associations with a P-value lower than an approximate Bonferroni threshold of 4 × 10−7 (following Strope et al. 2015).
Assessing differences in 2μ presence, copy number or genotype class between 100-genomes strains varying in clinical origin or population We tested whether 2μ presence, copy number or genotype class were associated with strains of particular population ancestry or clinical/non-clinical origin. Tests were conducted in R 2.15.1
Strope et al.
3
Class A (n = 46; S288c-type) RAF1 REP2
FRT
FLP1 REP1
Class B (n=23)
REP1-RAF1 91%
REP2-FLP1 99%
RAF1 REP2
FRT
FLP1 REP1
Class C (n=3)
REP2-FLP1 89%
REP1-RAF1 99%
RAF1 REP2
FRT
FLP1 REP1 Figure 1. The class A (red), class C (blue) and recombinant class B (red and blue) extrachromosomal 2μ sequences found in the 100-genomes S. cerevisiae strains.
4
FEMS Yeast Research, 2015, Vol. 15, No. 8
(R Core Team 2014) using Fisher’s exact test for tests between two categorical variables and Kruskal–Wallis one-way analysis of variance for tests between a categorical and a quantitative variable (i.e. 2μ copy number).
RESULTS Extrachromosomal and integrated 2μ in the 100-genomes S. cerevisiae strains and in other Saccharomycesspecies We determined the presence/absence of extrachromosomal 2μ in the 100-genomes S. cerevisiae strains and found 72 2μcontaining strains (Table S1, Supporting Information). For the 67 2μ-containing S. cerevisiae strains sequenced by us (Strope et al. 2015), we determined that 2μ copy number ranged from approximately 11 to 68 copies (median = 23) and that 2μ size ranged from 6094 to 6347 bp (Table S1, Supporting Information). We also determined extrachromosomal 2μ presence/absence in the genome sequences of 30 S. paradoxus strains (Liti et al. 2009; Bergstrom et al. 2014), 19 S. kudriavzevii strains (Hittinger et al. 2010; Scannell et al. 2011) and 58 S. uvarum strains (Kellis et al. 2003; Libkind et al. 2011; Scannell et al. 2011; Almeida et al. 2014), one S. mikatae strain (Scannell et al. 2011), one S. arboricolus strain (Liti et al. 2013) as well as two S. eubayanus and four S. bayanus strains (Cliften et al. 2003; Libkind et al. 2011; Almeida et al. 2014; Bing et al. 2014). Extrachromosomal 2μ is present in S. eubayanus and in the hybrid taxon S. bayanus, whereas it was absent in the S. mikatae and S. arboricolus strains. The 2μ sequence we assembled from S. eubayanus strain CBS12357 (SRR1613385), is 100% identical to the recently published 2μ of FM1318 (Baker et al. 2015), a derivative of CBS12357. In contrast to its frequent presence in S. cerevisiae and S. uvarum strains, extrachromosomal 2μ is almost or entirely absent from the S. paradoxus and S. kudriavzevii strains, respectively (Table 1). We also screened for 2μ sequences integrated into chromosomes of Saccharomyces species. We found no integrated 2μ sequences in any of the S. cerevisiae or S. paradoxus strains. While we found REP2—FLP1 2μ sequence integrated only in S. mikatae, which had been noted earlier in the same strain (Frank and Wolfe 2009), we found REP1–RAF1 2μ sequences integrated in some strains of S. uvarum, S. kudriavzevii and S. bayanus (Table S1, Supporting Information). The REP1–RAF1 2μ in-
tegrated sequence that had been discovered earlier in one S. bayanus strain (623-6C) and one S. uvarum strain (MCYC 623) (Frank and Wolfe 2009) is 95–100% identical to the REP1–RAF1 2μ integrated sequences from 46 S. uvarum and 2 S. bayanus strains in this study. Similarly, the S. kudriavzevii REP1–RAF1 integrated sequence in the two strains in this study is 99–100% identical to the sequence found in a different strain (IFO 1802) of S. kudriavzevii previously (Frank and Wolfe 2009).
2μ phylogenies in the 100-genomes S. cerevisiae strains and in other Saccharomyces species As described in Methods, we generated REP1–RAF1 and REP2– FLP1 ‘half circles’ that we used to create separate phylogenies. We found two classes of each half circle that occurred in three combinations to establish genotype classes A, B and C (Table S1, Supporting Information; Fig. 1). The structure of class B 2μ is consistent with one of two possible Flp-mediated recombination products between genotype classes A and C. We compared the extrachromosomal S. cerevisiae 2μ REP1– RAF1 and REP2–FLP1 sequences with extrachromosomal 2μ REP1–RAF1 and REP2–FLP1 sequences from other Saccharomyces species (Fig. 2). The extrachromosomal S. eubayanus 2μ is distinct from all other extrachromosomal 2μ. The extrachromosomal S. paradoxus 2μ is most closely related to, but is distinct from, S. cerevisiae class A extrachromosomal 2μ. In addition to its presence in S. cerevisiae, we found extrachromosomal class A 2μ in the 2μ-containing S. uvarum and S. bayanus strains (Table S1, Supporting Information; Fig. 2), consistent with interspecific transfer. Consistent with S. bayanus being an interspecific hybrid with one parent being S. uvarum (Libkind et al. 2011; Almeida et al. 2014), the chromosomally integrated REP1–RAF1 2μ sequences of S. uvarum and S. bayanus are equivalent. However, the chromosomally integrated REP1–RAF1 2μ sequences of S. uvarum/S. bayanus and S. kudriavzevii are distinct from each other, consistent with separate integration events of distinct 2μ, and are also distinct from all of the extrachromosomal 2μ, consistent with their being other as yet unidentified extrachromosomal 2μ (Fig. 2A). Similarly, the nuclearly integrated REP2–FLP1 sequence in S. mikatae is distinct from all other integrated and extrachromosomal 2μ sequences (Fig. 2B).
2μ associations in the 100-genomes strains Table 1. List of Saccharomyces species that were examined for the presence/absence of 2μ sequence. The total number of strains for each species and the number of strains with 2μ sequences present are listed. Genome sequences examined: S. cerevisiae (Strope et al. 2015); S. paradoxus (Liti et al. 2009; Bergstrom et al. 2014); S. kudriavzevii (Hittinger et al. 2010; Scannell et al. 2011); S. uvarum (Kellis et al. 2003; Libkind et al. 2011; Scannell et al. 2011; Almeida et al. 2014); S. mikatae (Scannell et al. 2011); S. arboricolus (Liti et al. 2013); S. eubayanus and S. bayanus (Cliften et al. 2003; Libkind et al. 2011; Almeida et al. 2014; Bing et al. 2014). Species S. cerevisiae S. paradoxus S. kudriavzevii S. uvarum S. eubayanus S. bayanus S. mikatae S. arboricolus
No. of genome sequences
No. of 2μ+ plasmid
100 30 19 58 2 4 1 1
72 1 0 13 1 2 0 0
We found no associations between extrachromosomal 2μ genotype class and 2μ copy number; we also found no associations between phenotypes previously measured in the 100-genomes strains (Strope et al. 2015) and extrachromosomal 2μ genetic variation (presence, genotype class). However, we found associations between extrachromosomal 2μ presence and genotype class with nuclear genome SNP genotypes in the 100-genomes strains (Table S2, Supporting Information). We found 2μ presence to be associated with S. cerevisiae populations and that 2μ genotype classes were non-randomly distributed between populations (Bonferroni-corrected P = 3.18 × 10−5 and 4.86 × 10−8 , respectively). Specifically, nearly all S. cerevisiae strains possessing extrachromosomal 2μ plasmid were wine/European population or mosaic group strains, including 90% (37/41) of wine/European population and 72% (33/46) of mosaic group strains. In contrast, only two of five West African population strains and no strains from any other population possessed extrachromosomal 2μ. Among strains with 2μ, 94% (35/37) of wine/European strains were genotype class A, while
Strope et al.
5
(A) 2μ plasmid + nuclearly integrated REP1-RAF1 S. eubayanus (n=1) 100 Nuclear: S. uvarum (n=47) S. bayanus (n=3)
100
S. paradoxus (n=1) 100 100 100
100
100
S. cerevisiae (n=46) S. uvarum (n=13) S. bayanus (n=2) S. cerevisiae (n=26) Nuclear: S. kudriavzevii (n=3)
0.03
(B) 2μ plasmid REP2-FLP1 S. eubayanus (n=1) 100 100
S. cerevisiae (n=69) S. uvarum (n=13) S. bayanus (n=2)
83 S. paradoxus (n=1)
100
S. cerevisiae (n=3)
100 Nuclear: S. mikatae (n=1) 0.03 Figure 2. Phylogeny of 2μ sequences in S. cerevisiae and other Saccharomyces species. (A) Extrachromosomal and integrated REP1–RAF1 2μ sequences. Red taxa belong to class A and blue taxa belong to class B and C. (B) Extrachromosomal and integrated REP2–FLP1 2μ sequences. Red taxa belong to classes A and B, and blue taxa belong to class C. The previously identified nuclearly integrated 2μ sequences REP1–RAF1 from S. uvarum, S. bayanus and S. kudriavzevii, and REP2–FLP1 from S. mikatae (Frank and Wolfe 2009) are included in the respective clades.
64% (21/33) of mosaic strains were genotype class B. The three class C-containing strains, YJM1400, 1401, 1479, are closely related mosaic group strains isolated from fruit in the Philippines. Finally, we tested for 2μ associations with clinical/nonclinical origin and found significant differences between clinical and non-clinical strains (Bonferroni-corrected P = 0.0019 and 0.0035, respectively). Specifically, while clinical strains lacking 2μ are uncommon in our dataset (9.5%, 4/42), a significantly larger fraction of non-clinical strains lack 2μ (42%, 24/57). Most non-clinical strains with 2μ (n = 33) possess genotype class A
(79%, 26/33), while clinical strains with 2μ (n = 39) are approximately evenly divided between genotype classes A (n = 20) and B (n = 19).
DISCUSSION We placed the 72 extrachromosomal 2μ plasmids found in the 100-genomes S. cerevisiae strains into genotype classes A, B and C (Fig. 1). Class C 2μ, which we found in three closely related mosaic group strains, may be present in an under-represented
6
FEMS Yeast Research, 2015, Vol. 15, No. 8
or unsequenced population of S. cerevisiae or may have introgressed from an under-represented or unsequenced species of Saccharomyces. Class A 2μ (n = 46) is found primarily in the wine/European population while class B 2μ (n = 23) is found mostly in the ‘outbred’ mosaic group. The wine/European population is the largest contributor to the outbred mosaic group genomes (Strope et al. 2015). We hypothesize that class B 2μ arose from an Flp-mediated recombination event between a class A 2μ, possibly from a wine/European strain, and a class C 2μ, which is consistent with the frequent presence of class B 2μ in the outbred mosaic group. In the 100-genomes strains, the 2μ presence and genotype class associations with SNP genotypes (Table S2, Supporting Information), which have no known biological relevance to 2μ, may be due to population structure. We also found 2μ presence and genotype class associations with populations and with clinical/non-clinical origin. In the case of clinical origin, given that all clinical strains belong to the mosaic (63%, 27/43) group or wine/European (37%, 16/43) population (Strope et al. 2015), the 2μ-clinical origin associations may reflect the population differences in 2μ presence and genotype classes. We hypothesize that 2μ presence and genotype class associations with populations may reflect population history and outcrossing frequencies in different S. cerevisiae populations. In addition to S. cerevisiae, we screened genome sequences for extrachromosomal 2μ in S. uvarum (58 strains), S. paradoxus (30 strains), S. kudriavzevii (19 strains) and S. eubayanus (2 strains), S. mikatae (1 strain), S. arboricolus (1 strain), as well as the hybrid taxon S. bayanus (4 strains), finding four extrachromosomal 2μ lineages (Fig. 2A). While we cannot determine directionality, the presence of class A 2μ in S. cerevisiae and in S. uvarum is clear evidence of introgression. Among multiple sequenced S. uvarum strains, only three strains had nuclear chromosomal introgressions from S. cerevisiae (Almeida et al. 2014). When we searched introgressions in our S. cerevisiae genome sequences (Strope et al. 2015) for S. uvarum introgressions, we found none. Such results suggest that, at least for the S. cerevisiae-S. uvarum species pair, 2μ may introgress more readily than nuclear chromosomes, possibly similar to what has been described in S. cerevisiae heterokaryons (Sigurdson, Gaarder and Livingston 1981). In contrast, as described in our initial description of the 100genomes strains (Strope et al. 2015), we found many S. paradoxus introgressions in S. cerevisiae. In this work, we find a 2μ in only one S. paradoxus strain, which is distinct from all 2μ found in S. cerevisiae. Such results suggest that, at least for the S. cerevisiae– S. paradoxus species pair, 2μ may introgress less readily than nuclear chromosomes. What is perhaps surprising is why 2μ, with its sexual/nonMendelian inheritance and low loss rate of 10−4 –10−5 in S. cerevisiae (Futcher 1988; Chan et al. 2013), is not present in all members of a Saccharomyces species and/or in all species of Saccharomyces. In S. cerevisiae, experimental integration of 2μ into specific chromosomes in the nuclear genome of strains containing extrachromosomal 2μ results in chromosome-specific genome instability (Falco et al. 1982; Falco and Botstein 1983; Wakem and Sherman 1990). Thus, one hypothesis is that integrated 2μ could be responsible for the absence of extrachromosomal 2μ in some but not all strains of a species or for the absence of extrachromosomal 2μ in a species. However, we found no integrated 2μ sequences in the S. cerevisiae or S. paradoxus strains. Similar to some other yeast species (Frank and Wolfe 2009), we identified REP1–RAF1 2μ sequences, with sequence identity >70% with S288c 2μ sequence, integrated in 3 of 19 S. kudriavzevii and in 47 of 58 S. uvarum strains (Table S1, Supporting Infor-
mation; Fig. 2A), and REP2–FLP1 2μ sequence integrated in S. mikate (Table S1, Supporting Information; Fig. 2B). However, of the 13 extrachromosomal 2μ-containing S. uvarum strains, all contain integrated REP1–RAF1 sequences (Table S1, Supporting Information). We suggest that integrated 2μ sequences are unlikely to be responsible for the absence of extrachromosomal 2μ in some strains of a species, such as S. cerevisiae, S. paradoxus, or S. uvarum, or for the absence of extrachromosomal 2μ in all sequenced strains of a species, such as S. kudriavzevii. Instead, we suggest that the absence of extrachromosomal 2μ may be due to strain-specific and/or species-specific fitness costs of extrachromosomal 2μ, which may be Mendelian (Holm 1982a,b) or quantitative in nature. Despite the absence of integrated 2μ in S. cerevisiae, 2μ has a fitness cost, a 1% reduced growth rate (Futcher and Cox 1983; Mead, Gardner and Oliver 1986; Futcher 1988; Futcher, Reid and Hickey 1988; Harrison et al. 2012), which likely contributes to the absence of 2μ in some S. cerevisiae strains. The 1% reduced growth rate of 2μ-containing strains may be due to the utilization by 2μ of host nuclear genome-encoded functions for its stability and copy number control (Holm 1982a,b; Sleep et al. 2001; Zhao, Wu and Blobel 2004; Dobson et al. 2005; Hajra, Ghosh and Jayaram 2006; Cui, Ghosh and Jayaram 2009; Sau et al. 2014). Consistent with its utilization of host nuclear genome-encoded functions to control its copy number, 2μ is lethal in some mutant backgrounds (Holm 1982a,b; Zhao, Wu and Blobel 2004; Dobson et al. 2005). One factor that could clearly lead to differences in 2μ presence/absence between populations or species of Saccharomyces is differences in the fitness cost of 2μ. In addition, we hypothesize that 2μ presence/absence may reflect species-specific or population-specific differences in outcrossing frequencies or population structure. Both outcrossing frequencies and the extent of population structure vary among Saccharomyces yeasts (e.g. [Johnson et al. 2004; Ruderfer et al. 2006; Liti et al. 2009]). From a population genetic perspective, both outcrossing and non-random mating (population structure) can influence effective population size (Crow and Kimura 1970). Since effective population size is an important parameter governing the efficacy of selection (Crow and Kimura 1970), differences in outcrossing and/or population structure could allow a moderately deleterious polymorphism (such as 2μ presence) to behave neutrally in a population with reduced efficacy of selection.
SUPPLEMENTARY DATA Supplementary data are available at FEMSYR online.
FUNDING This work was supported by the National Institutes of Health (grant R01 GM098287). PKS was supported as a fellow of the Tri-Institutional Molecular Mycology and Pathogenesis Training Program (MMPTP) supported by the National Institutes of Health/National Institute of Allergy and Infectious Diseases (T32 award AI052080-11). Conflict of interest. None declared.
REFERENCES Almeida P, Goncalves C, Teixeira S, et al. A Gondwanan imprint on global diversity and domestication of wine and cider yeast Saccharomyces uvarum. Nat Commun 2014;5:4044.
Strope et al.
Altschul SF, Gish W, Miller W, et al. Basic local alignment search tool. J Mol Biol 1990;215:403–10. Baker E, Wang B, Bellora N, et al. The Genome Sequence of Saccharomyces eubayanus and the Domestication of LagerBrewing Yeasts. Mol Biol Evol 2015;32:2818–31. Bergstrom A, Simpson JT, Salinas F, et al. A high-definition view of functional genetic variation from natural yeast genomes. Mol Biol Evol 2014;31:872–88. Bing J, Han PJ, Liu WQ, et al. Evidence for a Far East Asian origin of lager beer yeast. Curr Biol 2014;24:R380–1. Chan KM, Liu YT, Ma CH, et al. The 2 micron plasmid of Saccharomyces cerevisiae: a miniaturized selfish genome with optimized functional competence. Plasmid 2013;70:2–17. Cliften P, Sudarsanam P, Desikan A, et al. Finding functional features in Saccharomyces genomes by phylogenetic footprinting. Science 2003;301:71–6. Crow JF, Kimura M. An Introduction to Population Genetics Theory. New York: Harper, Row, 1970. Cui H, Ghosh SK, Jayaram M. The selfish yeast plasmid uses the nuclear motor Kip1p but not Cin8p for its localization and equal segregation. J Cell Biol 2009;185:251–64. Dobson MJ, Pickett AJ, Velmurugan S, et al. The 2 microm plasmid causes cell death in Saccharomyces cerevisiae with a mutation in Ulp1 protease. Mol Cell Biol 2005;25:4299–310. Doniger SW, Kim HS, Swain D, et al. A catalog of neutral and deleterious polymorphism in yeast. PLoS Genet 2008;4:e1000183. Dowell RD, Ryan O, Jansen A, et al. Genotype to phenotype: a complex problem. Science 2010;328:469. Falco SC, Botstein D. A rapid chromosome-mapping method for cloned fragments of yeast DNA. Genetics 1983;105:857–72. Falco SC, Li Y, Broach JR, et al. Genetic properties of chromosomally integrated 2 mu plasmid DNA in yeast. Cell 1982;29:573–84. Frank AC, Wolfe KH. Evolutionary capture of viral and plasmid DNA by yeast nuclear chromosomes. Eukaryot Cell 2009;8:1521–31. Futcher AB. The 2 micron circle plasmid of Saccharomyces cerevisiae. Yeast 1988;4:27–40. Futcher AB, Cox BS. Maintenance of the 2 microns circle plasmid in populations of Saccharomyces cerevisiae. J Bacteriol 1983;154:612–22. Futcher B, Reid E, Hickey DA. Maintenance of the 2 micron circle plasmid of Saccharomyces cerevisiae by sexual transmission: an example of a selfish DNA. Genetics 1988;118:411–5. Goffeau A, Barrell BG, Bussey H, et al. Life with 6000 genes. Science 1996;274:546,563–7. Hajra S, Ghosh SK, Jayaram M. The centromere-specific histone variant Cse4p (CENP-A) is essential for functional chromatin architecture at the yeast 2-microm circle partitioning locus and promotes equal plasmid segregation. J Cell Biol 2006;174:779–90. Harrison E, Koufopanou V, Burt A, et al. The cost of copy number in a selfish genetic element: the 2-mum plasmid of Saccharomyces cerevisiae. J Evolution Biol 2012;25:2348–56. Hartley JL, Donelson JE. Nucleotide sequence of the yeast plasmid. Nature 1980;286:860–5. Hittinger CT, Goncalves P, Sampaio JP, et al. Remarkably ancient balanced polymorphisms in a multi-locus gene network. Nature 2010;464:54–8. Holm C. Clonal lethality caused by the yeast plasmid 2 mu DNA. Cell 1982a;29:585–94. Holm C. Sensitivity to the Yeast Plasmid 2mu DNA is conferred by the nuclear allele nibl. Mol Cell Biol 1982b;2:985–92. Jayaram M, Mehta S, Uzri D, et al. Site-specific recombination and partitioning systems in the stable high copy propagation
7
of the 2-micron yeast plasmid. Prog Nucleic Acid Re 2004;77: 127–72. Johnson LJ, Koufopanou V, Goddard MR, et al. Population genetics of the wild yeast Saccharomyces paradoxus. Genetics 2004;166:43–52. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol 2013;30:772–80. Kellis M, Patterson N, Endrizzi M, et al. Sequencing and comparison of yeast species to identify genes and regulatory elements. Nature 2003;423:241–54. Kelly AC, Shewmaker FP, Kryndushkin D, et al. Sex, prions, and plasmids in yeast. P Natl Acad Sci USA 2012;109: E2683–90. Larkin MA, Blackshields G, Brown NP, et al. Clustal W and Clustal X version 2.0. Bioinformatics 2007;23:2947–8. Libkind D, Hittinger CT, Valerio E, et al. Microbe domestication and the identification of the wild genetic stock of lager-brewing yeast. P Natl Acad Sci USA 2011;108: 14539–44. Liti G, Carter DM, Moses AM, et al. Population genomics of domestic and wild yeasts. Nature 2009;458:337–41. Liti G, Nguyen Ba AN, Blythe M, et al. High quality de novo sequencing and assembly of the Saccharomyces arboricolus genome. BMC Genomics 2013;14:69. McCusker JH, Clemons KV, Stevens DA, et al. Genetic characterization of pathogenic Saccharomyces cerevisiae isolates. Genetics 1994;136:1261–9. Magwene P, Kayikci O, Granek JA, et al. Outcrossing, mitotic recombination, and life-history trade-offs shape genome evolution in Saccharomyces cerevisiae. P Natl Acad Sci USA 2011;108:1987–92. Mead DJ, Gardner DC, Oliver SG. The yeast 2 micron plasmid: strategies for the survival of a selfish DNA. Mol Gen Genet 1986;205:417–21. Nakayashiki T, Kurtzman CP, Edskes HK, et al. Yeast prions [URE3] and [PSI+] are diseases. P Natl Acad Sci USA 2005;102:10575–80. Neuville P, Bonneu M, Aigle M. Heterogeneity among the 2 microns plasmids in Saccharomyces cerevisiae: a new sequence for the REP1 gene. Gene 1990;89:139–44. Nishant KT, Wei W, Mancera E, et al. The Baker’s Yeast diploid genome is remarkably stable in vegetative growth and meiosis. PLoS Genet 2010;6:e1001109. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2014. http://www.Rproject.org/ (19 October 2015, date last accessed). Rank GH, Xiao W, Arndt GM. Evidence for Darwinian selection of the 2-micron plasmid STB locus in Saccharomyces cerevisiae. Genome 1994;37:12–8. RM11. Saccharomyces Cerevisiae RM11-1a Sequencing Project. Cambridge, MA, USA: Broad Institute of Harvard and MIT. 2004. http://www.broadinstitute.org/annotation/genome/ saccharomyces cerevisiae/Home.html (15 October 2015, date last accessed). Ruderfer DM, Pratt SC, Seidel HS, et al. Population genomic analysis of outcrossing and recombination in yeast. Nat Genet 2006;38:1077–81. Sau S, Conrad MN, Lee CY, et al. A selfish DNA element engages a meiosis-specific motor and telomeres for germ-line propagation. J Cell Biol 2014;205:643–61. Scannell DR, Zill OA, Rokas A, et al. The awesome power of yeast evolutionary genetics: new genome sequences and strain
8
FEMS Yeast Research, 2015, Vol. 15, No. 8
resources for the saccharomyces sensu stricto genus. G3 2011;1:11–25. Sigurdson DC, Gaarder ME, Livingston DM. Characterization of the transmission during cytoductant formation of the 2 micrometers DNA plasmid from Saccharomyces. Mol Gen Genet 1981;183:59–65. Sleep D, Finnis C, Turner A, et al. Yeast 2 microm plasmid copy number is elevated by a mutation in the nuclear gene UBC4. Yeast 2001;18:403–21. Smith TF, Waterman MS. Identification of common molecular subsequences. J Mol Biol 1981;147:195–7. Strope PK, Skelly DA, Kozmin SG, et al. The 100-genomes strains, an S. cerevisiae resource that illuminates its natural phenotypic and genotypic variation and emergence as an opportunistic pathogen. Genome Res 2015;25: 762–74. Wakem LP, Sherman F. Chromosomal assignment of mutations by specific chromosome loss in the yeast Saccharomyces cerevisiae. Genetics 1990;125:333–40.
Wei W, McCusker JH, Hyman RW, et al. Genome sequencing and comparative analysis of Saccharomyces cerevisiae strain YJM789. P Natl Acad Sci USA 2007;104:12825–30. Wei XA, Pelcher LE, Rank GH. DNA sequence divergence and functional conservation at the STB locus of yeast 2 microns circle variants. J Bacteriol 1991;173:1181–6. Xiao W, Pelcher LE, Rank GH. Evidence for cis- and trans-acting element coevolution of the 2-microns circle genome in Saccharomyces cerevisiae. J Mol Evol 1991a;32:145–52. Xiao W, Pelcher LE, Rank GH. Sequence diversity of yeast 2 microns RAF gene and its co-evolution with STB and REP1. Gene 1991b;101:75–80. Xiao W, Rank GH. The 2 micron plasmid of laboratory yeast strains is a type-1/type-2 hybrid. Yeast 1996;12:809–13. Zhao X, Wu CY, Blobel G. Mlp-dependent anchorage and stabilization of a desumoylating enzyme is required to prevent clonal lethality. J Cell Biol 2004;167:605–11. Zhou X, Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nat Genet 2012;44:821–4.
RESEARCH ARTICLE
2μ plasmid in Saccharomyces species and in Saccharomyces cerevisiae Pooja K. Strope1 , Stanislav G. Kozmin1,† , Daniel A. Skelly2 , Paul M. Magwene2 , Fred S. Dietrich1 and John H. McCusker1,∗ 1
Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC 27710, USA and 2 Department of Biology, Duke University, Durham, NC 27708, USA ∗ Corresponding author: Department of Molecular Genetics and Microbiology, DUMC, Jones Bldg., Rm. 239, 207 Research Drive, 3020, Durham, NC 27710, USA. Tel: +(919) 681-6744; Fax: +(919) 684-8735; E-mail: [email protected] † Present address: Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA. One sentence summary: We show extensive 2μ sequence variation in Saccharomyces cerevisiae and in other species of Saccharomyces and 2μ associations with S. cerevisiae populations, clinical origin and genotypes. Editor: Cletus Kurtzman
ABSTRACT We determined that extrachromosomal 2μ plasmid was present in 67 of the Saccharomyces cerevisiae 100-genome strains; in addition to variation in the size and copy number of 2μ, we identified three distinct classes of 2μ. We identified 2μ presence/absence and class associations with populations, clinical origin and nuclear genotypes. We also screened genome sequences of S. paradoxus, S. kudriavzevii, S. uvarum, S. eubayanus, S. mikatae, S. arboricolus and S. bayanus strains for both integrated and extrachromosomal 2μ. Similar to S. cerevisiae, we found no integrated 2μ sequences in any S. paradoxus strains. However, we identified part of 2μ integrated into the genomes of some S. uvarum, S. kudriavzevii, S. mikatae and S. bayanus strains, which were distinct from each other and from all extrachromosomal 2μ. We identified extrachromosomal 2μ in one S. paradoxus, one S. eubayanus, two S. bayanus and 13 S. uvarum strains. The extrachromosomal 2μ in S. paradoxus, S. eubayanus and S. cerevisiae were distinct from each other. In contrast, the extrachromosomal 2μ in S. bayanus and S. uvarum strains were identical with each other and with one of the three classes of S. cerevisiae 2μ, consistent with interspecific transfer. Keywords: 2-micron plasmid; 2μ plasmid; selfish genetic elements; population genomics; Saccharomyces cerevisiae; Saccharomyces genus
INTRODUCTION 2μ plasmid is a circular, multicopy, selfish genetic DNA element associated with the Saccharomyces cerevisiae genome. The canonical 6318 base pair 2μ plasmid contains four open reading frames (ORFs; FLP1, REP1, REP2, RAF1) and four cis-acting loci (an origin of replication (ORI); a stability (STB) locus; and 599 bp inverted repeats containing 48 bp Flp Recombination Targets (FRT)) (Futcher 1988; Jayaram et al. 2004; Chan et al. 2013). 2μ sexual/non-Mendelian inheritance, stability, site-specific re-
combination and copy number control in S. cerevisiae have been extensively studied; for reviews, see Futcher (1988), Jayaram et al. (2004) and Chan et al. (2013). 2μ is not present in all S. cerevisiae strains (Nakayashiki et al. 2005; Kelly et al. 2012) and exhibits sequence variation (Neuville, Bonneu and Aigle 1990; Wei, Pelcher and Rank 1991; Xiao, Pelcher and Rank 1991a,b; Rank, Xiao and Arndt 1994; Xiao and Rank 1996). Thus, 2μ presence/absence and sequence variation are both components of S. cerevisiae genetic variation and potentially so in other Saccharomyces species.
Received: 29 July 2015; Accepted: 1 October 2015 C FEMS 2015. All rights reserved. For permissions, please e-mail: [email protected]
1
2
FEMS Yeast Research, 2015, Vol. 15, No. 8
In this work, we describe for the 100-genomes S. cerevisiae strains (Strope et al. 2015) the presence/absence of 2μ and, where present, variation in 2μ copy number and sequences. We also examine the publicly available genome sequences of other Saccharomyces species to determine the presence/absence of both extrachromosomal and integrated 2μ; to establish 2μ phylogenies; and to assess the interspecific transfer of 2μ. Finally, we describe for the 100-genomes strains nuclear genotype, clinical/non-clinical origin and population associations with 2μ variation.
METHODS 2μ presence/absence, assembly, copy number determination and genotype class We sequenced 93 of the 100-genomes strains (Strope et al. 2015) and, by high read depth relative to single copy chromosomal genes, identified 67 as containing extrachromosomal 2μ; see Strope et al. (2015) for sequencing methodology. In brief, using the de novo assembler ABySS (v. 1.3.4), with parameters ‘k’: the k-mer length, ‘n’: the minimum number of pairs needed to join contigs, and ‘c’: the minimum mean k-mer coverage of a unitig, optimized for each strain, we assembled the 2μ from the 67 strains. We determined extrachromosomal 2μ copy number of these 67 2μ-containing strains by read depths relative to single copy chromosomal genes. The remaining seven of the 100genomes strains were sequenced in other studies: S288c (Goffeau et al. 1996), 1278b (Dowell et al. 2010), SK1 (Nishant et al. 2010), YJM789 (Wei et al. 2007), M22 (Doniger et al. 2008), RM11 (RM11 2004) and YPS163 (Doniger et al. 2008). The 100-genomes strains listed in Table S1 (Supporting Information) have been deposited in and should be requested from the Fungal Genetics Stock Center (http://www.fgsc.net). After the assembly of the 2μ sequences from the 2μcontaining 100-genomes S. cerevisiae strains, we manually converted the sequences to the same conformation as the reference S288c 2μ sequence in Saccharomyces Genome Database (SGD). We used BLAST (Altschul et al. 1990) and ssearch36 (Smith and Waterman 1981) to identify gene sequences in these assembled 2μ. Each of the four genes were extracted to build separate multiple alignments and phylogenies. Because the REP1–RAF1 and REP2–FLP1 pairs of gene phylogenies had similar topology (data not shown), we took those pairs of genes, and the intergenic sequence between those two genes, and built phylogenies. Specifically, for each of the 2μ-containing strains, sequences were gathered using blast hit coordinates for the REP1–RAF1 (S288c coordinates: 1887–3816 base pairs) and REP2–FLP1 (S288c coordinates: 5308–6318 and 1–1523 base pairs) gene pair regions. Sequence extraction and concatenation (for REP2–FLP1 sequences) was done using the EMBOSS tools seqret and union, respectively. Thus, we generated two ‘half circles’, consisting of the 3 end of RAF1 to the 3 end of REP1 (REP1–RAF1) and the 3 end of FLP1 to the 3 end of REP2 (REP2–FLP1). Multiple alignments were made using MAFFT (Katoh and Standley 2013). Phylogenies were estimated using neighbor joining in Clustal-w2 (Larkin et al. 2007). We also assessed the presence/absence of 2μ by PCR (2mF3: TGCATGATTAACTTCGAGAAGGG and 2m-R3: AAAGGACAAGGACCTGAGCG; RW319: GCCCTCAGTGATGGTGTTTT and RW320: CGTAAGCGTCGTTACCCAAT [Nakayashiki et al. 2005]) in the seven strains that we did not sequence. By both PCR and the absence of sequence reads, we found RM11 and YPS163 to lack 2μ. By both PCR and the presence of sequence reads, we found 1278b, SK1 and M22 to contain 2μ, which we assembled. For
the S288c background, which by PCR contains 2μ, we used the 2μ sequence in the SGD and in Hartley and Donelson (1980). For the YJM145 background, which is isogenic with YJM789 (Wei et al. 2007), PCR results showed that it contained 2μ. However, examination of YJM789 genome sequence data failed to identify 2μ sequences, which may be a consequence of the library preparation and/or sequencing technology. Because YJM145 is isogenic with YJM789 (Wei et al. 2007) and a segregant of YJM128 (McCusker et al. 1994), we used YJM128 sequence data (Magwene et al. 2011) to assemble 2μ. As 2μ interconverts between two forms by Flpmediated recombination at FRT sites (Futcher 1988; Jayaram et al. 2004; Chan et al. 2013), we assembled all 2μ into one form (Fig. 1). We screened short read archives and assembled sequences for extrachromosomal 2μ sequences from S. paradoxus (Liti et al. 2009; Bergstrom et al. 2014), S. kudriavzevii (Hittinger et al. 2010; Scannell et al. 2011), S. arboricolus (Liti et al. 2013), S. mikatae (Scannell et al. 2011) and S. uvarum (Kellis et al. 2003; Libkind et al. 2011; Scannell et al. 2011; Almeida et al. 2014), as well as S. eubayanus and the hybrid taxon S. bayanus (Cliften et al. 2003; Libkind et al. 2011; Almeida et al. 2014; Bing et al. 2014; Baker et al. 2015). In contrast to extrachromosomal 2μ, which is present in high copy number relative to single copy chromosomal sequences, 2μ sequences integrated into chromosomes should be present in approximately 1:1 ratio with known single copy chromosomal genes. On this basis, we screened for 2μ sequences integrated into nuclear chromosomes of S. cerevisiae and the other listed Saccharomyces species.
2μ genotype–phenotype and 2μ genotype-genotype associations in the 100-genomes strains We tested for associations between previously described phenotypes of the 100-genomes strains (Strope et al. 2015) and three aspects of 2μ genetic variation: 2μ presence/absence; 2μ genotype class; and, in the 67 2μ-containing 100-genomes strains with known copy number, 2μ copy number above/below median estimated copy number of 23. When testing 2μ presence/absence, we considered all of the 100-genomes strains. When testing 2μ genotype class, we excluded class C (n = 3) and included classes A (n = 46) and B (n = 23). When testing 2μ copy number, we considered only the 67 2μ-containing strains from our previous study (Strope et al. 2015). The program GEMMA version 0.94beta (Zhou and Stephens 2012) was used to conduct association tests. This program takes a linear mixed model approach to controlling population structure using a relatedness matrix (constructed from genotype data using GEMMA) constructed from 171 345 biallelic SNPs with minor allele frequency >5%, as previously described (Strope et al. 2015). We also tested for associations between the three aspects of 2μ genetic variation detailed above and genotypes at nuclear biallelic SNPs. The number of SNPs tested varied between 125 721 and 157 892 depending on the subset of strains assayed (the same subsets as explained above; only considering biallelic SNPs with minor allele frequency >5%). We examined associations with a P-value lower than an approximate Bonferroni threshold of 4 × 10−7 (following Strope et al. 2015).
Assessing differences in 2μ presence, copy number or genotype class between 100-genomes strains varying in clinical origin or population We tested whether 2μ presence, copy number or genotype class were associated with strains of particular population ancestry or clinical/non-clinical origin. Tests were conducted in R 2.15.1
Strope et al.
3
Class A (n = 46; S288c-type) RAF1 REP2
FRT
FLP1 REP1
Class B (n=23)
REP1-RAF1 91%
REP2-FLP1 99%
RAF1 REP2
FRT
FLP1 REP1
Class C (n=3)
REP2-FLP1 89%
REP1-RAF1 99%
RAF1 REP2
FRT
FLP1 REP1 Figure 1. The class A (red), class C (blue) and recombinant class B (red and blue) extrachromosomal 2μ sequences found in the 100-genomes S. cerevisiae strains.
4
FEMS Yeast Research, 2015, Vol. 15, No. 8
(R Core Team 2014) using Fisher’s exact test for tests between two categorical variables and Kruskal–Wallis one-way analysis of variance for tests between a categorical and a quantitative variable (i.e. 2μ copy number).
RESULTS Extrachromosomal and integrated 2μ in the 100-genomes S. cerevisiae strains and in other Saccharomycesspecies We determined the presence/absence of extrachromosomal 2μ in the 100-genomes S. cerevisiae strains and found 72 2μcontaining strains (Table S1, Supporting Information). For the 67 2μ-containing S. cerevisiae strains sequenced by us (Strope et al. 2015), we determined that 2μ copy number ranged from approximately 11 to 68 copies (median = 23) and that 2μ size ranged from 6094 to 6347 bp (Table S1, Supporting Information). We also determined extrachromosomal 2μ presence/absence in the genome sequences of 30 S. paradoxus strains (Liti et al. 2009; Bergstrom et al. 2014), 19 S. kudriavzevii strains (Hittinger et al. 2010; Scannell et al. 2011) and 58 S. uvarum strains (Kellis et al. 2003; Libkind et al. 2011; Scannell et al. 2011; Almeida et al. 2014), one S. mikatae strain (Scannell et al. 2011), one S. arboricolus strain (Liti et al. 2013) as well as two S. eubayanus and four S. bayanus strains (Cliften et al. 2003; Libkind et al. 2011; Almeida et al. 2014; Bing et al. 2014). Extrachromosomal 2μ is present in S. eubayanus and in the hybrid taxon S. bayanus, whereas it was absent in the S. mikatae and S. arboricolus strains. The 2μ sequence we assembled from S. eubayanus strain CBS12357 (SRR1613385), is 100% identical to the recently published 2μ of FM1318 (Baker et al. 2015), a derivative of CBS12357. In contrast to its frequent presence in S. cerevisiae and S. uvarum strains, extrachromosomal 2μ is almost or entirely absent from the S. paradoxus and S. kudriavzevii strains, respectively (Table 1). We also screened for 2μ sequences integrated into chromosomes of Saccharomyces species. We found no integrated 2μ sequences in any of the S. cerevisiae or S. paradoxus strains. While we found REP2—FLP1 2μ sequence integrated only in S. mikatae, which had been noted earlier in the same strain (Frank and Wolfe 2009), we found REP1–RAF1 2μ sequences integrated in some strains of S. uvarum, S. kudriavzevii and S. bayanus (Table S1, Supporting Information). The REP1–RAF1 2μ in-
tegrated sequence that had been discovered earlier in one S. bayanus strain (623-6C) and one S. uvarum strain (MCYC 623) (Frank and Wolfe 2009) is 95–100% identical to the REP1–RAF1 2μ integrated sequences from 46 S. uvarum and 2 S. bayanus strains in this study. Similarly, the S. kudriavzevii REP1–RAF1 integrated sequence in the two strains in this study is 99–100% identical to the sequence found in a different strain (IFO 1802) of S. kudriavzevii previously (Frank and Wolfe 2009).
2μ phylogenies in the 100-genomes S. cerevisiae strains and in other Saccharomyces species As described in Methods, we generated REP1–RAF1 and REP2– FLP1 ‘half circles’ that we used to create separate phylogenies. We found two classes of each half circle that occurred in three combinations to establish genotype classes A, B and C (Table S1, Supporting Information; Fig. 1). The structure of class B 2μ is consistent with one of two possible Flp-mediated recombination products between genotype classes A and C. We compared the extrachromosomal S. cerevisiae 2μ REP1– RAF1 and REP2–FLP1 sequences with extrachromosomal 2μ REP1–RAF1 and REP2–FLP1 sequences from other Saccharomyces species (Fig. 2). The extrachromosomal S. eubayanus 2μ is distinct from all other extrachromosomal 2μ. The extrachromosomal S. paradoxus 2μ is most closely related to, but is distinct from, S. cerevisiae class A extrachromosomal 2μ. In addition to its presence in S. cerevisiae, we found extrachromosomal class A 2μ in the 2μ-containing S. uvarum and S. bayanus strains (Table S1, Supporting Information; Fig. 2), consistent with interspecific transfer. Consistent with S. bayanus being an interspecific hybrid with one parent being S. uvarum (Libkind et al. 2011; Almeida et al. 2014), the chromosomally integrated REP1–RAF1 2μ sequences of S. uvarum and S. bayanus are equivalent. However, the chromosomally integrated REP1–RAF1 2μ sequences of S. uvarum/S. bayanus and S. kudriavzevii are distinct from each other, consistent with separate integration events of distinct 2μ, and are also distinct from all of the extrachromosomal 2μ, consistent with their being other as yet unidentified extrachromosomal 2μ (Fig. 2A). Similarly, the nuclearly integrated REP2–FLP1 sequence in S. mikatae is distinct from all other integrated and extrachromosomal 2μ sequences (Fig. 2B).
2μ associations in the 100-genomes strains Table 1. List of Saccharomyces species that were examined for the presence/absence of 2μ sequence. The total number of strains for each species and the number of strains with 2μ sequences present are listed. Genome sequences examined: S. cerevisiae (Strope et al. 2015); S. paradoxus (Liti et al. 2009; Bergstrom et al. 2014); S. kudriavzevii (Hittinger et al. 2010; Scannell et al. 2011); S. uvarum (Kellis et al. 2003; Libkind et al. 2011; Scannell et al. 2011; Almeida et al. 2014); S. mikatae (Scannell et al. 2011); S. arboricolus (Liti et al. 2013); S. eubayanus and S. bayanus (Cliften et al. 2003; Libkind et al. 2011; Almeida et al. 2014; Bing et al. 2014). Species S. cerevisiae S. paradoxus S. kudriavzevii S. uvarum S. eubayanus S. bayanus S. mikatae S. arboricolus
No. of genome sequences
No. of 2μ+ plasmid
100 30 19 58 2 4 1 1
72 1 0 13 1 2 0 0
We found no associations between extrachromosomal 2μ genotype class and 2μ copy number; we also found no associations between phenotypes previously measured in the 100-genomes strains (Strope et al. 2015) and extrachromosomal 2μ genetic variation (presence, genotype class). However, we found associations between extrachromosomal 2μ presence and genotype class with nuclear genome SNP genotypes in the 100-genomes strains (Table S2, Supporting Information). We found 2μ presence to be associated with S. cerevisiae populations and that 2μ genotype classes were non-randomly distributed between populations (Bonferroni-corrected P = 3.18 × 10−5 and 4.86 × 10−8 , respectively). Specifically, nearly all S. cerevisiae strains possessing extrachromosomal 2μ plasmid were wine/European population or mosaic group strains, including 90% (37/41) of wine/European population and 72% (33/46) of mosaic group strains. In contrast, only two of five West African population strains and no strains from any other population possessed extrachromosomal 2μ. Among strains with 2μ, 94% (35/37) of wine/European strains were genotype class A, while
Strope et al.
5
(A) 2μ plasmid + nuclearly integrated REP1-RAF1 S. eubayanus (n=1) 100 Nuclear: S. uvarum (n=47) S. bayanus (n=3)
100
S. paradoxus (n=1) 100 100 100
100
100
S. cerevisiae (n=46) S. uvarum (n=13) S. bayanus (n=2) S. cerevisiae (n=26) Nuclear: S. kudriavzevii (n=3)
0.03
(B) 2μ plasmid REP2-FLP1 S. eubayanus (n=1) 100 100
S. cerevisiae (n=69) S. uvarum (n=13) S. bayanus (n=2)
83 S. paradoxus (n=1)
100
S. cerevisiae (n=3)
100 Nuclear: S. mikatae (n=1) 0.03 Figure 2. Phylogeny of 2μ sequences in S. cerevisiae and other Saccharomyces species. (A) Extrachromosomal and integrated REP1–RAF1 2μ sequences. Red taxa belong to class A and blue taxa belong to class B and C. (B) Extrachromosomal and integrated REP2–FLP1 2μ sequences. Red taxa belong to classes A and B, and blue taxa belong to class C. The previously identified nuclearly integrated 2μ sequences REP1–RAF1 from S. uvarum, S. bayanus and S. kudriavzevii, and REP2–FLP1 from S. mikatae (Frank and Wolfe 2009) are included in the respective clades.
64% (21/33) of mosaic strains were genotype class B. The three class C-containing strains, YJM1400, 1401, 1479, are closely related mosaic group strains isolated from fruit in the Philippines. Finally, we tested for 2μ associations with clinical/nonclinical origin and found significant differences between clinical and non-clinical strains (Bonferroni-corrected P = 0.0019 and 0.0035, respectively). Specifically, while clinical strains lacking 2μ are uncommon in our dataset (9.5%, 4/42), a significantly larger fraction of non-clinical strains lack 2μ (42%, 24/57). Most non-clinical strains with 2μ (n = 33) possess genotype class A
(79%, 26/33), while clinical strains with 2μ (n = 39) are approximately evenly divided between genotype classes A (n = 20) and B (n = 19).
DISCUSSION We placed the 72 extrachromosomal 2μ plasmids found in the 100-genomes S. cerevisiae strains into genotype classes A, B and C (Fig. 1). Class C 2μ, which we found in three closely related mosaic group strains, may be present in an under-represented
6
FEMS Yeast Research, 2015, Vol. 15, No. 8
or unsequenced population of S. cerevisiae or may have introgressed from an under-represented or unsequenced species of Saccharomyces. Class A 2μ (n = 46) is found primarily in the wine/European population while class B 2μ (n = 23) is found mostly in the ‘outbred’ mosaic group. The wine/European population is the largest contributor to the outbred mosaic group genomes (Strope et al. 2015). We hypothesize that class B 2μ arose from an Flp-mediated recombination event between a class A 2μ, possibly from a wine/European strain, and a class C 2μ, which is consistent with the frequent presence of class B 2μ in the outbred mosaic group. In the 100-genomes strains, the 2μ presence and genotype class associations with SNP genotypes (Table S2, Supporting Information), which have no known biological relevance to 2μ, may be due to population structure. We also found 2μ presence and genotype class associations with populations and with clinical/non-clinical origin. In the case of clinical origin, given that all clinical strains belong to the mosaic (63%, 27/43) group or wine/European (37%, 16/43) population (Strope et al. 2015), the 2μ-clinical origin associations may reflect the population differences in 2μ presence and genotype classes. We hypothesize that 2μ presence and genotype class associations with populations may reflect population history and outcrossing frequencies in different S. cerevisiae populations. In addition to S. cerevisiae, we screened genome sequences for extrachromosomal 2μ in S. uvarum (58 strains), S. paradoxus (30 strains), S. kudriavzevii (19 strains) and S. eubayanus (2 strains), S. mikatae (1 strain), S. arboricolus (1 strain), as well as the hybrid taxon S. bayanus (4 strains), finding four extrachromosomal 2μ lineages (Fig. 2A). While we cannot determine directionality, the presence of class A 2μ in S. cerevisiae and in S. uvarum is clear evidence of introgression. Among multiple sequenced S. uvarum strains, only three strains had nuclear chromosomal introgressions from S. cerevisiae (Almeida et al. 2014). When we searched introgressions in our S. cerevisiae genome sequences (Strope et al. 2015) for S. uvarum introgressions, we found none. Such results suggest that, at least for the S. cerevisiae-S. uvarum species pair, 2μ may introgress more readily than nuclear chromosomes, possibly similar to what has been described in S. cerevisiae heterokaryons (Sigurdson, Gaarder and Livingston 1981). In contrast, as described in our initial description of the 100genomes strains (Strope et al. 2015), we found many S. paradoxus introgressions in S. cerevisiae. In this work, we find a 2μ in only one S. paradoxus strain, which is distinct from all 2μ found in S. cerevisiae. Such results suggest that, at least for the S. cerevisiae– S. paradoxus species pair, 2μ may introgress less readily than nuclear chromosomes. What is perhaps surprising is why 2μ, with its sexual/nonMendelian inheritance and low loss rate of 10−4 –10−5 in S. cerevisiae (Futcher 1988; Chan et al. 2013), is not present in all members of a Saccharomyces species and/or in all species of Saccharomyces. In S. cerevisiae, experimental integration of 2μ into specific chromosomes in the nuclear genome of strains containing extrachromosomal 2μ results in chromosome-specific genome instability (Falco et al. 1982; Falco and Botstein 1983; Wakem and Sherman 1990). Thus, one hypothesis is that integrated 2μ could be responsible for the absence of extrachromosomal 2μ in some but not all strains of a species or for the absence of extrachromosomal 2μ in a species. However, we found no integrated 2μ sequences in the S. cerevisiae or S. paradoxus strains. Similar to some other yeast species (Frank and Wolfe 2009), we identified REP1–RAF1 2μ sequences, with sequence identity >70% with S288c 2μ sequence, integrated in 3 of 19 S. kudriavzevii and in 47 of 58 S. uvarum strains (Table S1, Supporting Infor-
mation; Fig. 2A), and REP2–FLP1 2μ sequence integrated in S. mikate (Table S1, Supporting Information; Fig. 2B). However, of the 13 extrachromosomal 2μ-containing S. uvarum strains, all contain integrated REP1–RAF1 sequences (Table S1, Supporting Information). We suggest that integrated 2μ sequences are unlikely to be responsible for the absence of extrachromosomal 2μ in some strains of a species, such as S. cerevisiae, S. paradoxus, or S. uvarum, or for the absence of extrachromosomal 2μ in all sequenced strains of a species, such as S. kudriavzevii. Instead, we suggest that the absence of extrachromosomal 2μ may be due to strain-specific and/or species-specific fitness costs of extrachromosomal 2μ, which may be Mendelian (Holm 1982a,b) or quantitative in nature. Despite the absence of integrated 2μ in S. cerevisiae, 2μ has a fitness cost, a 1% reduced growth rate (Futcher and Cox 1983; Mead, Gardner and Oliver 1986; Futcher 1988; Futcher, Reid and Hickey 1988; Harrison et al. 2012), which likely contributes to the absence of 2μ in some S. cerevisiae strains. The 1% reduced growth rate of 2μ-containing strains may be due to the utilization by 2μ of host nuclear genome-encoded functions for its stability and copy number control (Holm 1982a,b; Sleep et al. 2001; Zhao, Wu and Blobel 2004; Dobson et al. 2005; Hajra, Ghosh and Jayaram 2006; Cui, Ghosh and Jayaram 2009; Sau et al. 2014). Consistent with its utilization of host nuclear genome-encoded functions to control its copy number, 2μ is lethal in some mutant backgrounds (Holm 1982a,b; Zhao, Wu and Blobel 2004; Dobson et al. 2005). One factor that could clearly lead to differences in 2μ presence/absence between populations or species of Saccharomyces is differences in the fitness cost of 2μ. In addition, we hypothesize that 2μ presence/absence may reflect species-specific or population-specific differences in outcrossing frequencies or population structure. Both outcrossing frequencies and the extent of population structure vary among Saccharomyces yeasts (e.g. [Johnson et al. 2004; Ruderfer et al. 2006; Liti et al. 2009]). From a population genetic perspective, both outcrossing and non-random mating (population structure) can influence effective population size (Crow and Kimura 1970). Since effective population size is an important parameter governing the efficacy of selection (Crow and Kimura 1970), differences in outcrossing and/or population structure could allow a moderately deleterious polymorphism (such as 2μ presence) to behave neutrally in a population with reduced efficacy of selection.
SUPPLEMENTARY DATA Supplementary data are available at FEMSYR online.
FUNDING This work was supported by the National Institutes of Health (grant R01 GM098287). PKS was supported as a fellow of the Tri-Institutional Molecular Mycology and Pathogenesis Training Program (MMPTP) supported by the National Institutes of Health/National Institute of Allergy and Infectious Diseases (T32 award AI052080-11). Conflict of interest. None declared.
REFERENCES Almeida P, Goncalves C, Teixeira S, et al. A Gondwanan imprint on global diversity and domestication of wine and cider yeast Saccharomyces uvarum. Nat Commun 2014;5:4044.
Strope et al.
Altschul SF, Gish W, Miller W, et al. Basic local alignment search tool. J Mol Biol 1990;215:403–10. Baker E, Wang B, Bellora N, et al. The Genome Sequence of Saccharomyces eubayanus and the Domestication of LagerBrewing Yeasts. Mol Biol Evol 2015;32:2818–31. Bergstrom A, Simpson JT, Salinas F, et al. A high-definition view of functional genetic variation from natural yeast genomes. Mol Biol Evol 2014;31:872–88. Bing J, Han PJ, Liu WQ, et al. Evidence for a Far East Asian origin of lager beer yeast. Curr Biol 2014;24:R380–1. Chan KM, Liu YT, Ma CH, et al. The 2 micron plasmid of Saccharomyces cerevisiae: a miniaturized selfish genome with optimized functional competence. Plasmid 2013;70:2–17. Cliften P, Sudarsanam P, Desikan A, et al. Finding functional features in Saccharomyces genomes by phylogenetic footprinting. Science 2003;301:71–6. Crow JF, Kimura M. An Introduction to Population Genetics Theory. New York: Harper, Row, 1970. Cui H, Ghosh SK, Jayaram M. The selfish yeast plasmid uses the nuclear motor Kip1p but not Cin8p for its localization and equal segregation. J Cell Biol 2009;185:251–64. Dobson MJ, Pickett AJ, Velmurugan S, et al. The 2 microm plasmid causes cell death in Saccharomyces cerevisiae with a mutation in Ulp1 protease. Mol Cell Biol 2005;25:4299–310. Doniger SW, Kim HS, Swain D, et al. A catalog of neutral and deleterious polymorphism in yeast. PLoS Genet 2008;4:e1000183. Dowell RD, Ryan O, Jansen A, et al. Genotype to phenotype: a complex problem. Science 2010;328:469. Falco SC, Botstein D. A rapid chromosome-mapping method for cloned fragments of yeast DNA. Genetics 1983;105:857–72. Falco SC, Li Y, Broach JR, et al. Genetic properties of chromosomally integrated 2 mu plasmid DNA in yeast. Cell 1982;29:573–84. Frank AC, Wolfe KH. Evolutionary capture of viral and plasmid DNA by yeast nuclear chromosomes. Eukaryot Cell 2009;8:1521–31. Futcher AB. The 2 micron circle plasmid of Saccharomyces cerevisiae. Yeast 1988;4:27–40. Futcher AB, Cox BS. Maintenance of the 2 microns circle plasmid in populations of Saccharomyces cerevisiae. J Bacteriol 1983;154:612–22. Futcher B, Reid E, Hickey DA. Maintenance of the 2 micron circle plasmid of Saccharomyces cerevisiae by sexual transmission: an example of a selfish DNA. Genetics 1988;118:411–5. Goffeau A, Barrell BG, Bussey H, et al. Life with 6000 genes. Science 1996;274:546,563–7. Hajra S, Ghosh SK, Jayaram M. The centromere-specific histone variant Cse4p (CENP-A) is essential for functional chromatin architecture at the yeast 2-microm circle partitioning locus and promotes equal plasmid segregation. J Cell Biol 2006;174:779–90. Harrison E, Koufopanou V, Burt A, et al. The cost of copy number in a selfish genetic element: the 2-mum plasmid of Saccharomyces cerevisiae. J Evolution Biol 2012;25:2348–56. Hartley JL, Donelson JE. Nucleotide sequence of the yeast plasmid. Nature 1980;286:860–5. Hittinger CT, Goncalves P, Sampaio JP, et al. Remarkably ancient balanced polymorphisms in a multi-locus gene network. Nature 2010;464:54–8. Holm C. Clonal lethality caused by the yeast plasmid 2 mu DNA. Cell 1982a;29:585–94. Holm C. Sensitivity to the Yeast Plasmid 2mu DNA is conferred by the nuclear allele nibl. Mol Cell Biol 1982b;2:985–92. Jayaram M, Mehta S, Uzri D, et al. Site-specific recombination and partitioning systems in the stable high copy propagation
7
of the 2-micron yeast plasmid. Prog Nucleic Acid Re 2004;77: 127–72. Johnson LJ, Koufopanou V, Goddard MR, et al. Population genetics of the wild yeast Saccharomyces paradoxus. Genetics 2004;166:43–52. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol 2013;30:772–80. Kellis M, Patterson N, Endrizzi M, et al. Sequencing and comparison of yeast species to identify genes and regulatory elements. Nature 2003;423:241–54. Kelly AC, Shewmaker FP, Kryndushkin D, et al. Sex, prions, and plasmids in yeast. P Natl Acad Sci USA 2012;109: E2683–90. Larkin MA, Blackshields G, Brown NP, et al. Clustal W and Clustal X version 2.0. Bioinformatics 2007;23:2947–8. Libkind D, Hittinger CT, Valerio E, et al. Microbe domestication and the identification of the wild genetic stock of lager-brewing yeast. P Natl Acad Sci USA 2011;108: 14539–44. Liti G, Carter DM, Moses AM, et al. Population genomics of domestic and wild yeasts. Nature 2009;458:337–41. Liti G, Nguyen Ba AN, Blythe M, et al. High quality de novo sequencing and assembly of the Saccharomyces arboricolus genome. BMC Genomics 2013;14:69. McCusker JH, Clemons KV, Stevens DA, et al. Genetic characterization of pathogenic Saccharomyces cerevisiae isolates. Genetics 1994;136:1261–9. Magwene P, Kayikci O, Granek JA, et al. Outcrossing, mitotic recombination, and life-history trade-offs shape genome evolution in Saccharomyces cerevisiae. P Natl Acad Sci USA 2011;108:1987–92. Mead DJ, Gardner DC, Oliver SG. The yeast 2 micron plasmid: strategies for the survival of a selfish DNA. Mol Gen Genet 1986;205:417–21. Nakayashiki T, Kurtzman CP, Edskes HK, et al. Yeast prions [URE3] and [PSI+] are diseases. P Natl Acad Sci USA 2005;102:10575–80. Neuville P, Bonneu M, Aigle M. Heterogeneity among the 2 microns plasmids in Saccharomyces cerevisiae: a new sequence for the REP1 gene. Gene 1990;89:139–44. Nishant KT, Wei W, Mancera E, et al. The Baker’s Yeast diploid genome is remarkably stable in vegetative growth and meiosis. PLoS Genet 2010;6:e1001109. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2014. http://www.Rproject.org/ (19 October 2015, date last accessed). Rank GH, Xiao W, Arndt GM. Evidence for Darwinian selection of the 2-micron plasmid STB locus in Saccharomyces cerevisiae. Genome 1994;37:12–8. RM11. Saccharomyces Cerevisiae RM11-1a Sequencing Project. Cambridge, MA, USA: Broad Institute of Harvard and MIT. 2004. http://www.broadinstitute.org/annotation/genome/ saccharomyces cerevisiae/Home.html (15 October 2015, date last accessed). Ruderfer DM, Pratt SC, Seidel HS, et al. Population genomic analysis of outcrossing and recombination in yeast. Nat Genet 2006;38:1077–81. Sau S, Conrad MN, Lee CY, et al. A selfish DNA element engages a meiosis-specific motor and telomeres for germ-line propagation. J Cell Biol 2014;205:643–61. Scannell DR, Zill OA, Rokas A, et al. The awesome power of yeast evolutionary genetics: new genome sequences and strain
8
FEMS Yeast Research, 2015, Vol. 15, No. 8
resources for the saccharomyces sensu stricto genus. G3 2011;1:11–25. Sigurdson DC, Gaarder ME, Livingston DM. Characterization of the transmission during cytoductant formation of the 2 micrometers DNA plasmid from Saccharomyces. Mol Gen Genet 1981;183:59–65. Sleep D, Finnis C, Turner A, et al. Yeast 2 microm plasmid copy number is elevated by a mutation in the nuclear gene UBC4. Yeast 2001;18:403–21. Smith TF, Waterman MS. Identification of common molecular subsequences. J Mol Biol 1981;147:195–7. Strope PK, Skelly DA, Kozmin SG, et al. The 100-genomes strains, an S. cerevisiae resource that illuminates its natural phenotypic and genotypic variation and emergence as an opportunistic pathogen. Genome Res 2015;25: 762–74. Wakem LP, Sherman F. Chromosomal assignment of mutations by specific chromosome loss in the yeast Saccharomyces cerevisiae. Genetics 1990;125:333–40.
Wei W, McCusker JH, Hyman RW, et al. Genome sequencing and comparative analysis of Saccharomyces cerevisiae strain YJM789. P Natl Acad Sci USA 2007;104:12825–30. Wei XA, Pelcher LE, Rank GH. DNA sequence divergence and functional conservation at the STB locus of yeast 2 microns circle variants. J Bacteriol 1991;173:1181–6. Xiao W, Pelcher LE, Rank GH. Evidence for cis- and trans-acting element coevolution of the 2-microns circle genome in Saccharomyces cerevisiae. J Mol Evol 1991a;32:145–52. Xiao W, Pelcher LE, Rank GH. Sequence diversity of yeast 2 microns RAF gene and its co-evolution with STB and REP1. Gene 1991b;101:75–80. Xiao W, Rank GH. The 2 micron plasmid of laboratory yeast strains is a type-1/type-2 hybrid. Yeast 1996;12:809–13. Zhao X, Wu CY, Blobel G. Mlp-dependent anchorage and stabilization of a desumoylating enzyme is required to prevent clonal lethality. J Cell Biol 2004;167:605–11. Zhou X, Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nat Genet 2012;44:821–4.
