Документ взят из кэша поисковой машины. Адрес оригинального документа : http://mccmb.belozersky.msu.ru/2015/proceedings/abstracts/110.pdf Дата изменения: Mon Jun 15 15:36:58 2015 Дата индексирования: Sat Apr 9 23:30:46 2016 Кодировка:

Sequencing genomes of Saccharomyces cerevisiae strains belonging to the Peterhof Genetic Collection helps elucidate the origin of several widely used laboratory strains Oleg Tarasov, Polina Drozdova, E. Radchenko, D. Polev, P. Dobrynin, Sergey IngeVechtomov
Saint Petersburg State University, Russia, Saint-Petersburg, Universitetskaya emb. 7-9 ovtarasov@gmail.com

Introduction Saccharomyces cerevisiae is a widely used model organism. The haplo id S. cerevisiae strain S288c is the progenitor to many o f co mmo nly used yeast laboratory strains and gave the first sequenced eukaryot ic geno me. S288c and its relatives originate fro m Carbondale Breeding stocks of O. Winge and C. Lindegren, which resulted from crosses between not only S. cerevisiae itself but also other Saccharomyces species [1, 2]. The Peterhof genet ic co llect ion of yeast is unrelated to S288c and originates from an industrial dist illery strain [3]. Some strains of this co llect ion are widely used in the field o f yeast prion research [4, 5, and other works]. A number of genet ic variations between Peterhof and S288c-related strains was ident ified but the who le geno me data for Peterhof strains are scarce. To date, genomes of more than 150 yeast strains of different origin have been sequenced. Comparison o f such a variet y of genomes helps clarify the natural history of yeast populat ions and allow to characterize geno mic elements that are selected under specific condit ions. Thus, we aimed to characterize the geno mes o f so me yeast strains fro m Peterhof genet ic co llect ion. Results In this work, we analyzed genomes of four S. cerevisiae strains. 15V-P4 is one o f the haplo id progenitors of the Peterhof genet ic co llect ion; it originates from the init ial industrial strain XII through 7 generat ions of intertetrad self-fert ilization and 3 subsequent inbred crosses. 2525-2V-P3982 is a laboratory strain o f pure Peterhof origin, and 1B-D1606 and 74-D694 are hybrid descendants of both Peterhof and S288c-derived strains. All strains sequenced are haplo id.


Geno mes of 15V-P4, 25-25-2V-P3982, and 1B-D1606 were sequenced with Ion Torrent PGM using unpaired reads. Raw reads for the 74-D694 geno me produced with Illumina GAII were retrieved fro m http://bio inf.nuim.ie/ wp-content/uploads/2011/10/74D_sequence.txt.zip. Trimming o f reads was performed wit h fastx_toolkit v0.0.13.1. Trimming length was chosen according to the basic statistics calculated with FastQC. Genome coverage was from about 12 to 40X for different strains. We de novo assembled the reads with SPAdes v3.1.0 and Mira v4.0 and found that qualit y o f the SPAdes assemblies was generally higher (Table 1 and data not shown). Thus, we used SPAdes assemblies for further analys is. Table 1. Quast statist ics of SPAdes assemblies. 15V-P4 Number of contigs Largest bases N50 Number of S288c genes found 1,188 contig, 92,741 11,686,072 18,500 2,753 + 743 partial 25-25-2V-P3982 927 102,076 11,613,898 25,573 92.68 4,832 + 863 partial 1B-D1606 480 252,839 11,567,380 73,204 94.33 3,363 + 206 partial 74-D694 1,684 71,636 11,310,898 11,948 92.36 2,848 + 630 partial

Total length, bases

Geno me fraction, % 92.10

In order to assess the difference between our strains and the reference strain S288c, alignment of short reads to the reference geno me was performed wit h bowtie v2.1.0. Then, SNP calling was performed with samtools v1.0 mpileup co mmand with subsequent filtering wit h vcftools v1.0. All indels were filtered out with vcftools, and variat ions in the repeat regions (ident ified with RepeatMasker v4.0.2) were also filtered out. We est imated genetic difference between our strains and S288c based on number o f pairwise SNPs (fig. 1). 15V-P4 and S288c differs by 45,781 SNPs which is co mparable to the level o f divergence between distant S. cerevisiae populations reported previously [6]. As we could predict, "pure Peterhof" 25-25-2V-P3982 strain is the most similar to 15V-P4. However, these two strains have much more pairwise SNPs than we expected. We suppose that this


difference may reflect laboratory evo lut ion of the strain. 1B-D1606 and 74-D694 are roughly half as distant from S288c as 15V-P4 which is consistent with their hybrid origin. Figure 1. NJ clustering of our strains and S288c based on number of pairwise SNPs. Shown in right are numbers of SNPs in comparison to S288c.

We tested whether any SNPs can be attributed to known phenotypic differences between Peterhof and S288c-derived strains. Amn1 and Flo8 are transcript ional regulators of cell aggregation in yeast. Amn1
Asp368Val

and Flo8T

rp142Stop

alleles are known to contribute much

into change from clumping to non-clumping phenotype [7]. We observe the same tendency in our strains (data not shown). De novo geno me assemblies annotated with exonerate v2.2.0 were used to search for specific genes known to be lacking in S288c but present in other strains primarily o f industrial or environmental origin [8]. All four strains studied possess the KHR and RTM1 genes. The KHR gene encodes killer toxin of unknown nature. In 25-25-2V-P3982 and 1B-D1606, this gene is annotated on the same cont igs as known genes o f chro mosome IX. In 15V-P4 and 74-D694, it is annotated on its own contig wit hout any neighbouring ORFs. The RTM1 gene is a member of a three-gene cluster associated with the subtelo meric sucrose utilizat ion (SUC) locus that is present in several clinical, industrial, and environmental iso lates. It encodes a lipid-translocating exporter and is known to be advantageous for strains growing on mo lasses. In 15V-P4, we also found the cluster of five genes init ially identified in wine strains [9]. Based on sequence, we suppose that 5-oxo-L-prolinase gene is a pseudogene while other four genes may be act ive. Interest ingly, 15V-P4 appear to be the first yeast strain reported to obtain simultaneously the RTM1 gene and the wine-specific cluster. It can be associated with


dist illery origin o f Peterhof genet ic co llect ion. Wine cluster is supposed to move in yeast geno mes easily, therefore it could be quickly lost during laboratory evo lut ion. Other widely distributed genes we looked for (biot in biosynt hesis genes BIO1 and BIO6, fructose transporter FSY1, cell wall component AWA1, epoxide hydrolase EHL, and Nacetyltransferase MPR1) were found in none of four strains analyzed. Conclusions The results presented above show that geno mes of the strains o f the Peterho f genet ic collect ion and of the S288c-based laboratory strains differ significant ly and provide insight into some physio logical differences o f these strains such as clumping. The geno mic data obtained are consistent with known origin o f Peterho f strains. These data could form the basis for planning future work in these strains. The work was supported by RFBR grants 14-04-31265, 14-04-32213, and 15-29-02526. Geno me sequencing was performed in the Research Resource Center for mo lecular and cell techno logies, Research park of Saint Petersburg State Universit y. References 1. R.K.Mortimer, J.R.Johnston (1986), Genetics, 113(1):35­43. 2. F. Sherman (2002), Methods Enzymol, 350:3­41. 3. S.G.Inge-Vechtomov (1963), Vestn. LGU, 21:117­125 (in Russian). 4. Y.O.Cherno ff et al. (1993), Curr Genet, 24:268­270. 5. L.Westergard, H.L.True (2014), Mol Microbiol, 92:183­193. 6. G.Lit i et al. (2009), Nature, 458:337­341. 7. J.Li et al. (2013), DNA Res, 20(1):55­66. 8. A.R.Borneman, I.S.Pretorius (2015), Genetics, 199(2):281­291. 9. A.R.Borneman et al. (2011), PLoS Genet, 7(2):e1001287.