In the Saccharomyces cerevisiae strains used for genome sequencing and functional analysis, the mitochondrial DNA replicase Mip1p contains a single nucleotide polymorphism changing the strictly conserved threonine 661 to alanine. This substitution is responsible for the increased rate of mitochondrial DNA point mutations and deletions in these strains.
THE yeast Saccharomyces cerevisiae is one of the few eukaryotic organisms that can survive in the absence of mitochondrial DNA (mtDNA). When grown on glucose-containing medium, S. cerevisiae produces respiratory-deficient mutants that form small colonies and thus are referred to as “cytoplasmic petites” (Slonimski and Ephrussi 1949). They either contain deletions in their mtDNA (rho−) or are devoid of mtDNA (rho0) (Dujon 1981). petite frequency varies within a large range among laboratory strains (Marmiroli et al. 1980), a not surprising feature with regards to the numerous genes that control petite accumulation (Contamine and Picard 2000). One of these is MIP1, which encodes the mitochondrial polymerase (DNA polymerase gamma) (Foury 1989). Previous studies using the MIP1 gene isolated from a genomic library constructed with strain Σ1278b (Grenson at al. 1966), referred to as MIP1[Σ], have shown that Mip1p is an accurate replicase (Foury and Vanderstraeten 1992; Hu et al. 1995; Vanderstraeten et al. 1998). However, recent works using the MIP1 allele from S288c-related strains, referred to as MIP1[S], have pointed to higher mutation rates of the mitochondrial genome (Baruffini et al. 2006; Stuart et al. 2006). Here we show that this increase in mtDNA instability results from a single nucleotide substitution in the MIP1 gene, changing a strictly conserved threonine at position 661 to alanine.
MIP1[S] allele increases mtDNA instability:
petite mutant frequency was determined in the DWM-5A-Δmip1 strain (Baruffini et al. 2006) carrying pFL39 plasmid-borne versions of either MIP1[S] or MIP1[Σ]. A fourfold increase in petite accumulation was observed at 28° in the presence of MIP1[S] compared to MIP1[Σ] (Figure 1A). Moreover, in the MIP1[S] context, and in contrast to the MIP1[Σ] context, petite accumulation was highly dependent on temperature (Figure 1A). On the basis of the capacity of independent petite clones to restore mit− mutations to wild type (Slonimski and Tzagoloff 1976), it was concluded that, in both MIP1 contexts, most petites had retained mtDNA fragments and were rho−, even though, in the MIP1[S] context, the fraction of rho0 clones at 36° was significantly higher (Figure 1B). The frequency of EryR mutants, which are caused by mtDNA mutations in the 21S rRNA gene (Sor and Fukuhara 1984), gives an estimate of the frequency of mtDNA point mutations. A sixfold increase in EryR mutants was observed in the MIP1[S] context compared to the MIP1[Σ] context (Figure 1C). These data show that in the same nuclear background the mitochondrial genome is less stable in the presence of the MIP1[S] allele, suggesting that this trait could be ascribed to differences in the amino acid sequence of the two Mip1p isoforms.
T661A substitution in the MIP1[S] allele is the cause of temperature-sensitive petite accumulation:
DNA sequence analysis revealed 26 single nucleotide polymorphisms (SNPs) between MIP1[S] and MIP1[Σ] alleles (Table 1). The MIP1 sequence was identical in W303-1B and S288c, confirming the close origin of these strains (Schacherer et al. 2007). Sixteen SNPs were silent and 10 produced nonsynonymous substitutions. Multiple amino acid alignments with several fungi and animal pol g sequences showed that these modifications affected mostly poorly conserved residues (data not shown). However, Thr661, which is found in Mip1p[Σ] and is strictly conserved in all species from yeasts to humans, is changed to alanine in Mip1p[S] due to an A-to-G transition at position 1981 of the MIP1 sequence. Moreover, Thr661 has been found in two recently sequenced S. cerevisiae strains, YJM789 (Gu et al. 2005) and RM11-1a (Lee et al. 2006), which have a MIP1[S]-like allele.
To determine whether the T661A substitution was the cause of petite accumulation in the MIP1[S] context, a MIP1[S]A661T variant was constructed. In the presence of this new allele, petite frequency was similar to that observed with the MIP1[Σ] allele and, moreover, the temperature-sensitive trait had disappeared (Figure 1A). The frequency of EryR mutants was also reduced (Figure 1C). In a heteroallelic MIP1[S]A661T/MIP1[S] strain, petite and EryR mutant frequencies were similar to those observed in a strain containing MIP1[S]A661T only, indicating that the MIP1[S] allele is recessive (data not shown).
We measured petite frequency in several laboratory strains. Σ1278b, D273-10B/A1 (Sherman 1964), and FL100 (Lacroute 1968), which possess the MIP1[Σ] allele (E. Baruffini, unpublished data), had low levels of petites (Table 2). W303-1B (Thomas and Rothstein 1989) and BY4742 (Brachmann et al. 1998), which possess the MIP1[S] allele (and thus Ala661), accumulated petites at higher levels and in a temperature-dependent manner (Table 2). To further demonstrate that increased instability of the mitochondrial genome could be ascribed to Ala661 rather than to the genetic background of these strains, the pFL38 plasmid-borne MIP1[S]A661T and MIP1[S] alleles were introduced in a mip1Δ derivative of D273-10B/A1. petite frequency was low in the presence of MIP1[S]A661T and increased in the presence of MIP1[S] (Table 3). Altogether, these data demonstrate that the T661A substitution in Mip1p is the cause of the higher accumulation of petites in S288c, BY4742, and W303-1B strains.
Natural isolates contain a threonine residue at position 661:
To determine whether T661A was a widespread mutation, we analyzed 20 strains isolated from grapes of different Italian regions or used as industrial starters for wine or bread production. Adenine at position 1981 of the MIP1 gene is associated with a BsaAI site while with guanine the BsaAI site is lost. To distinguish these polymorphisms, a DNA fragment encompassing nucleotides 1046–2329 was amplified and digested by BsaAI. The presence of adenine at position 1981 generates two DNA fragments of 937 and 347 bp, whereas guanine generates a single DNA fragment of 1284 bp (Figure 2A). In all strains, the MIP1 gene possessed the BsaAI site and thus Thr661.
On the other hand, the partial genome sequence of haploid segregants from 36 strains of different origins has recently been published (Saccharomyces Genome Resequencing Project at the Sanger Institute, http://www.sanger.ac.uk/Teams/Team71/durbin/sgrp/index.shtml). Most Western isolates have the MIP1[S]A661T allele; the MIP1 gene of Asian isolates is a complex mosaic of the MIP1[S]A661T and MIP1[Σ] alleles, but none of the strains has the MIP1[Σ] allele, raising the question of the origin of this allele. However, all isolates have a threonine at position 661.
In contrast, analysis of the MIP1 gene of the diploid EM93 strain, which is the progenitor of S288c and derivative strains and has been estimated to share 88% of its genome with S288c (Mortimer and Johnston 1986), revealed that, in addition to the two DNA fragments generated by the BsaAI digest, an undigested fragment of 1284 bp was also present (Figure 2A), a sign of heterozygosity at the BsaA1 site. MIP1 gene sequencing and tetrad analysis (data not shown) confirmed that both A and G were present at position 1981 (Figure 2B). Therefore, EM93 is heterozygous for the MIP1 allele. These data led to the conclusion that the T661A substitution is unique and specific to EM93, the founder of several commonly used laboratory strains.
This work provides an explanation for the high frequency of mtDNA deletions and point mutations occurring in commonly used laboratory strains: a missense mutation in the mitochondrial DNA replicase brought about by EM93, the founder strain (Mortimer and Johnston 1986). The fact that S288c and its derivatives BY4741, BY4742, and BY4743, the strains used in the genome sequencing and large-scale functional analysis projects, contain the MIP1[S] allele should be taken into consideration in genomic studies focused on alterations of the mitochondrial metabolism. Moreover, the MIP1[S] allele has previously been used to establish the impact of human pathological POLG mutations on the stability of mtDNA (Baruffini et al. 2006; Stuart et al. 2006). It must be stressed that even though this allele, which exacerbates certain defects, can be useful in detecting subtle effects of mip1 mutations, it may also cause unreliable phenotypes.
We thank I. Ferrero (University of Parma) for stimulating discussion, D. Cavalieri, M. Polsinelli (University of Florence, Italy), and M. Budroni (University of Sassari, Italy) for providing natural wine strains, and G. Liti and E. J. Louis (University of Nottingham, UK) for useful information about the S. cerevisiae strains sequenced at the Sanger Institute. We thank Roberto Silva for his skillful technical assistance. This work was funded by the Belgian National Fund for Scientific Research (to F.F.), the Interuniversity Attraction Programme, Belgian Science Policy (to F.F.), the Association Française contre les Myopathies (to F.F.), and the Fondazione Telethon-Italy no. GGP030039 and National Benefit Research Project 2006 2006069034_003 (to E.B. and T.L.).
Communicating editor: A. Nicolas
- Received July 25, 2007.
- Accepted August 5, 2007.
- Copyright © 2007 by the Genetics Society of America