2003 Copyright by the Genetics Society of America
Genetic Diversity in Yeast Assessed With Whole-Genome Oligonucleotide Arrays
Elizabeth A. Winzeler,*,1 Cristian I. Castillo-Davis, Guy Oshiro,* David Liang,* Daniel R. Richards, Yingyao Zhou* and Daniel L. Hartl
*Genomics Institute of the Novartis Research Foundation, San Diego, California 92121, Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, Massachusetts 02138 and Department of Genetics, Stanford University School of Medicine, Stanford, California 95307 Manuscript received August 4, 2002 Accepted for publication October 21, 2002 ABSTRACT The availability of a complete genome sequence allows the detailed study of intraspecies variability. Here we use high-density oligonucleotide arrays to discover 11,115 single-feature polymorphisms (SFPs) existing in one or more of 14 different yeast strains. We use these SFPs to dene regions of genetic identity between common laboratory strains of yeast. We assess the genome-wide distribution of genetic variation on the basis of this yeast population. We nd that genome variability is biased toward the ends of chromosomes and is more likely to be found in genes with roles in fermentation or in transport. This subtelomeric bias may arise through recombination between nonhomologous sequences because full-gene deletions are more common in these regions than in more central regions of the chromosome.
ITH few exceptions, only one strain or an individual of a particular species is sequenced while hundreds of other variants, which may be important to public health, scientic research, or commercial applications, remain undeciphered. In the bakers yeast, Saccharomyces cerevisiae, a derivative of strain S288c was sequenced. Despite the availability of the sequence information for this strain, many full-genome studies, including gene expression studies (Chu et al. 1998), genome-wide chromatin-binding studies (Wyrick et al. 2001), and studies of the replication dynamics of the yeast genome (Raghuraman et al. 2001), have been conducted using alternative but commonly used yeast strains. In some cases, the strain may be directly related to S288c (A364, W303, and 1278b derivatives) and in some cases the strain may be completely unrelated (SK1). As many of these studies rely on oligonucleotide or cDNA arrays that were derived from S288c sequence information, the quality of the data may differ depending on the region of the genome under investigation and on whether or not the region is identical by descent to that of S288c. These strain differences could contribute to some of the inconsistencies in genomewide data sets (Grunenfelder and Winzeler 2002). Mortimer and Johnston (1986) have traced the pedigrees of some laboratory yeast strains as far as is known, but direct ascertainment of the relationships between laboratory strains, in the absence of full-genome sequencing, has been impossible.
W
1 Corresponding author: Department of Cell Biology, The Scripps Research Institute, 10550 Torrey Pines Rd., La Jolla, CA 92037.
Single-base changes between two sequences 25 bp in length, especially those found in the central region of a 25mer, disrupt their hybridization (Chee et al. 1996; Gingeras et al. 1998; Troesch et al. 1999; Lockhart and Winzeler 2000). Thus, with oligonucleotide arrays carrying large numbers of probes of this length (termed features), the approximate location of allelic variation between two strains can be discovered (Winzeler et al. 1998). Since the locations of all the features in the genome are generally known, the approximate position of a predicted polymorphism between the two strains can also be ascertained when the genomic DNA hybridization patterns are compared (Figure 1). Such hybridization differences have been termed single-feature polymorphisms (SFPs; Borevitz et al. 2003). The Affymetrix S98 oligonucleotide array contains 285,156 different 25mers from the yeast genomic sequence. Although this array was designed primarily to be a tool for gene expression analysis, it was also created to maximize the amount of yeast genome sequence covered. In addition to probes to all annotated genes in the yeast genome, probes to small nonannotated genes (Oshiro et al. 2002), probes to nontranslated RNA, and probes to noncoding regions were included in the design. Although a percentage of these probes overlap one another, altogether 16% of the yeast genome is probed by this array design. We reasoned that because of the high degree of coverage, these arrays could be used to identify a signicant proportion of the genetic variation existing between strains and that this information could then be used to determine strain relationships. In addition, through the inclusion of several wild isolates, we have characterized the distribution of allelic
Genetics 163: 7989 ( January 2003)
80
E. A. Winzeler et al.
Figure 1.Example of hybridization pattern for probes that detect nonvariant (A) or variant (B) alleles. The ...
Winzeler.Genetics.2003