The mitochondrial genome of wild-type yeast cells. VI. Genome organization

A maturase-like coding sequence downstream of the OXI2 gene of yeast mitochondrial DNA is interrupted by two GC clusters and a putative end-of-messenger signal

By completing and correcting the sequence of a 1.8 kb DNA segment downstream of the oxi2 gene of Saccharomyces cerevisiae, a long, potentially coding sequence ("RF2") has been identified. The sequence is rather closely related to the RF1 open reading frame, downstream of the oxil gene, and, further, to the major family of intronic open reading frames. The RF2 open reading frame is not continuous, however, for it is interrupted by two GC clusters, both of which ultimately result in a -1 frameshift. Comparison with RF1 reveals a third insertion. This is centered on an oligo nucleotide, AATAATATTCTTA, which is found (sometimes in a slightly modified form) downstream of ten proven or suspected protein coding genes, including RF1 and RF2, and is known to terminate the apocytochrome b messenger RNA. It is suggested from the known distribution of this putative "end-of-messenger" signal, that it could play an essential part in controlling the expression of several minor proteins, both intronic and non-intronic. The possibility of the RF2 sequence being functional in spite of its interruptions is also discussed.

Lessons from a small, dispensable genome: the mitochondrial genome of yeast

This article reviews the investigations on the mitochondrial genomes of yeast carried out in the author's laboratory during a quarter of a century (to be precise between 1966 and 1992). Our studies dealt with the structural basis for the cytoplasmic petite mutation, the replication, the transcription and the recombination of the mitochondrial genome, a genome which is dispensable and which comprises abundant non-coding sequences. This work led to some general conclusions on the nuclear genome of eukaryotes. Some recent results in apparent contradiction with our conclusions on ori sequences will also be briefly discussed.

Transmission of mitochondrial DNA in Ustilago violacea

Mitochondrial DNA (mtDNA) restriction fragment length polymorphisms (RFLPs) were used as genetic markers for following mitochondrial transmission in the basidiomycete Ustilago violacea. Yeast-like cells of opposite mating types (a1 and a2) were mated on 2% water agar and were treated with alpha-tocopherol to induce formation of dikaryotic hyphae. Upon depletion of the alpha-tocopherol, the hyphae budded off haploid cells with parental nuclear genotypes. These cells were examined for mitochondrial RFLP phenotype. In progeny expressing the a1 mating type, mitochondria from either parent were observed equally frequently. In progeny with the a2 mating type, mitochondria were almost exclusively (94%) from the a2 parent.

Indication for deletion of two introns in the oxi-3 gene of a respiratory-competent Saccharomyces cerevisiae strain

Physical mapping of the mitochondrial DNA of the wild-type Saccharomyces cerevisiae strain RXII revealed that most of the restriction sites as well as the location of the apocytochrome b gene were identical in comparison with the known maps of the mitochondrial genome in other Saccharomyces cerevisiae strains. In the middle of the SalI linearized map of the RXII mitochondrial DNA, a deletion was detected which resulted in the loss of two EcoRI and one BamHI restriction sites. The corresponding region, however, exists in most other laboratory strains of Saccharomyces mapped so far. This region overlaps the introns aI2 and aI3 surrounding exon A3 sequences of the subunit 1 of the cytochrome oxidase gene. The nucleotide sequence of the subunit 1 gene showed that the BamHI site was located close to the aI3-A4 intron-exon junction and the distal EcoRI site close to the aI2-A2 boundary. I therefore conclude that these two introns are deleted in the mitochondrial genome of strain RXII. The exon A3 must have been conserved since this strain was respiratory competent. This result, while being a good example of the morphological diversity of a genome with the same function, may contribute to an understanding of the role of introns in the mitochondrial split genes in yeast.

In vivo double-strand breaks occur at recombinogenic G + C-rich sequences in the yeast mitochondrial genome

An optional 46-base-pair G + C-rich element (GC cluster) in the coding region of the yeast mitochondrial var1 gene inserts preferentially in crosses into recipient alleles that lack the sequence. Unlike a similar gene conversion event involving the insertion of an optional 1143-base-pair intron, the mitochondrial 21S rRNA gene, which requires the action of a protein encoded by a gene within that intron, conversion of the var1 GC cluster does not require any protein product of the mitochondrial genome. We have detected double-strand breaks in the var1 gene in mitochondrial DNA isolated from unmated haploid rho+ and rho- strains at or near the boundaries of the optional GC cluster, as well as at a conserved copy of that sequence 160 base pairs upstream. No double-strand breaks were detected in the recipient var1 DNA molecules in the vicinity of the optional GC cluster target sequence. This contrasts with 21S rRNA-encoding DNA (rDNA) intron conversion where the recipient, but not the donor DNA, is cleaved at the element insertion site. These results suggest that although the 21S rDNA intron and the var1 GC cluster are preferentially inserted into their respective short alleles, these conversions probably occur by different mechanisms.

The AT spacers and the var1 genes from the mitochondrial genomes of Saccharomyces cerevisiae and Torulopsis glabrata: evolutionary origin and mechanism of formation

Intergenic sequences represent 63% of the mitochondrial 'long' (85 kb) genome of Saccharomyces cerevisiae. They comprise 170-200 AT spacers that correspond to 47% of the genome and are separated from each other by GC clusters, ORFs, ori sequences, as well as by protein-coding genes. Intergenic AT spacers have an average size of 190 bp, and a GC level of 5%; they are formed by short (20-30 nt on the average) A/T stretches separated by C/G mono- to trinucleotides. An analysis of the primary structures of all intergenic AT spacers already sequenced (32 kb; 80% of the total) has shown that they are characterized by an extremely high level of short sequence repetitiveness and by a characteristic sequence pattern; the frequencies of A/T isostichs conspicuously deviate from statistical expectations, and exponentially decrease when their (AT + TA)/(AA + TT) ratio, R, decreases. A situation basically identical was found in the AT spacers of the mitochondrial genome (19 kb) of Torulopsis glabrata. The sequence features of the AT spacers indicate that they were built in evolution by an expansion process mainly involving rounds of duplication, inversion and translocation events which affected an initial oligodeoxynucleotide (endowed with a particular R ratio) and the sequences derived from it. In turn, the initial oligodeoxynucleotide appears to have arisen from an ancestral promoter-replicator sequence which was at the origin of the nonanucleotide promoters present in the mitochondrial genomes of several yeasts. Common sequence patterns indicate that the AT spacers so formed gave rise to the var1 gene (by linking and phasing of short ORFs), to the DNA stretches corresponding to the untranslated mRNA sequences and to the central stretches of ori sequences from S. cerevisiae.

The unusual varl gene of yeast mitochondrial DNA

The var1 gene specifies the only mitochondrial ribosomal protein known to be encoded by yeast mitochondrial DNA. The gene is unusual in that its base composition is nearly 90 percent adenine plus thymine. It and its expression product show a strain-dependent variation in size of up to 7 percent; this variation does not detectably interfere with function. Furthermore, var1 is an expandable gene that participates in a novel recombinational event resembling gene conversion whereby shorter alleles are preferentially converted to longer ones. The remarkable features of var1 indicate that it may have evolved by a mechanism analogous to exon shuffling, although no introns are actually present.

A mitochondrial reading frame which may code for a maturase-like protein in Saccharomyces cerevisiae

In S. cerevisiae, the large oxi3/oli2 mitochondrial transcript contains the products of the oxi3, aap1 and oli2 genes and an unassigned reading frame, RF3. In the work presented here, we have completed the nucleotide sequence of RF3. We have shown that RF3 is composed of four fairly large ORFs which overlap within GC rich sequences. Furthermore, a shift of +1 base was found between each pair of consecutive reading frames. We discuss how these frameshifts could be removed to produce a 500 aminoacid long protein containing the two well conserved P1 and P2 oligopeptide sequences featuring several mitochondrial intron reading frames, suggesting, thereby, a RNA-maturase-like activity for the putative RF3 protein. In addition, we suggest that the insertion of GC clusters in a gene could provide a novel way of regulating its expression.

A new putative gene in the mitochondrial genome of Saccharomyces cerevisiae

The 2200-bp ori2-ori7 region of the mitochondrial (mt) genome of Saccharomyces cerevisiae has been sequenced on the genome of a petite, b7, excised at those ori sequences from wild-type strain B. The region contains an open reading frame, ORF5, which is transcribed into a 900-nucleotide (nt) RNA in both the parental wild-type strain and its derived petite, b7. This RNA uses as a template the strand used by most mt transcripts. Its start point is located 337 nt upstream of ORF5; and a messenger termination site has been found 900 nt downstream of the initiation site. These data suggest that ORF5 is a new mitochondrial gene. The G + C content of ORF5 is only 15.7%; 90% of the G + C base pairs of ORF5 are comprised in a palindromic G + C cluster similar to that present in the varl gene. The coding capacity of ORF5 is 46 amino acids (aa), mainly represented by methionine, phenylalanine, arginine, valine, asparagine, isoleucine and tyrosine. The aa composition and the codon usage of ORF5 are reminiscent of those of varl and of other intergenic ORFs.

Polymorphic variations in the ori sequences from the mitochondrial genomes of different wild-type yeast strains

We determined the restriction maps and primary structures of two as yet poorly characterized regions of the mitochondrial genomes of different wild-type strains of Saccharomyces cerevisiae. These regions respectively comprised the ori1 sequence and the newly identified ori8 sequence. Ori1 and ori8, together with their flanking sequences, exhibit a large polymorphism, resulting from specific variations due to insertions or deletions of optional GC clusters at different locations. The mechanisms underlying such sequence rearrangements are discussed.

The ori sequences of the mitochondrial genome of a wild-type yeast strain: number, location, orientation and structure

We have investigated the number, the location, the orientation and the structure of the seven ori sequences present in the mitochondrial genome of a wild-type strain, A, of Saccharomyces cerevisiae. These homologous sequences are formed by three G + C-rich clusters, A, B and C, and by four A + T-rich stretches. Two of the latter, p and s, are located between clusters A and B; one, l, between clusters B and C; and one r, either immediately follows cluster C (in ori 3-7), or is separated from it by an additional A + T-rich stretch, r', (in ori 1 and ori 2). The most remarkable differences among ori sequences concern the presence of two additional G + C-rich clusters, beta and gamma, which are inserted in sequence l of ori 4 and 6 and in the middle of sequence r of ori 4, 6 and 7, respectively. Neglecting clusters beta and gamma and stretch r', the length of ori sequences is 280 +/- 1 bp, and that of the l stretch 200 +/- 1 bp. Hairpin structures can be formed by the whole A-B region, by clusters beta and gamma, and (in ori 2-6) by a short AT sequence, lp, immediately preceding cluster beta. An overall tertiary folding of ori sequences can be obtained. Some structural features of ori sequences are shared by the origins of replication of the heavy strands of the mitochondrial genomes of mammalian cells.

Two modules from the hypersuppressive rho- mitochondrial DNA are required for plasmid replication in yeast

In the yeast hypersuppressive (HS) rho- mutants most of the mitochondrial genome is deleted, but the remainder containing one of the three rep sequences is amplified. One of these sequences, rep2, and its flanking regions have been previously cloned and reported to promote autonomous plasmid replication in yeast. The present study suggests that the Ars activity associated with this HS rho- mitochondrial DNA (mtDNA) fragment is due to the presence in cis of at least two modules: (i) the 11-bp consensus sequence 5'-ATAAACTATAAAAT-3', common to several ars sequences, and (ii) a palindromic sequence of the mitochondrial replicator. Proper spacing between the two modules, which varies from about 100 to 200 bp, is required for the Ars+ activity.

Organization and processing of the mitochondrial oxi3/oli2 multigenic transcript in yeast

In the present article, we confirm our previous proposal (Faye and Simon 1983a, b) that the oxi3 and oli2 genes belong to the same transcription unit. Furthermore, we have shown that a primary polycistronic transcript covers oxi3, aap1, oli2 and extends beyond URF2. Transcriptional analysis of this region revealed several cleavage points. The examination of the DNA sequence at and surrounding these cleavage points disclosed that some of them take place at or near specific sequences found also in other known multigenic transcripts. Two of the major cleavages involve the stem-loop structure of GC rich clusters. We discuss the possibility that some of these cleavage sites serve as post-transcriptional processing signals and may be necessary for the maturation of the precursor RNA.

Expandable var1 gene of yeast mitochondrial DNA: in-frame insertions can explain the strain-specific protein size polymorphisms

The var1 locus of yeast mitochondrial DNA encodes a protein of the small mitochondrial ribosome subunit, denoted var1 protein. The size of var1 protein was previously shown to exhibit a strain-dependent polymorphism, determined by various combinations of at least three genetic elements. We report here that the var1 gene is itself polymorphic and that the six forms of this gene examined here differ by various combinations of three in-frame insertions into the coding region of the smallest allele. These insertions, which appear to be the molecular basis for the genetic elements, could increase the size of var1 protein by 8, 10, 16, 24, or 26 amino acid residues, accounting for the observed protein polymorphisms. Furthermore, we have characterized three additional sources of sequence variation located outside of the coding region but within major transcripts of the var1 gene.

Transcriptional analysis of the Saccharomyces cerevisiae mitochondrial var1 gene: anomalous hybridization of RNA from AT-rich regions

A family of mitochondrial RNAs hybridizes specifically to the var1 region on Saccharomyces cerevisiae mitochondrial DNA (Farrelly et al., J. Biol. Chem. 257:6581-6587, 1982). We constructed a fine-structure transcription map of this region by hybridizing DNA probes containing different portions of the var1 region and some flanking sequences to mitochondrial RNAs isolated from var1-containing petites. We also report the nucleotide sequence of more than 1.2 kilobases of DNA flanking the var1 gene. Our primary findings are: (i) The family of RNAs we detect with homology to var1 DNA is colinear with the var1 gene. Their direction of transcription is olil to cap, as it is for most other mitochondrial genes. (ii) Extensive hybridization anomalies are present, most likely due to the high A-T (A-U) content of the hybridizing species and to the asymmetric distribution of their G-C residues. An important conclusion is that failure to detect transcripts from A-T-rich regions of the yeast mitochondrial genome by standard blot transfer hybridizations cannot be interpreted to mean that such sequences, which are commonly supposed to be spacer DNA, are noncoding or lack direct function in the expression of mitochondrial genes.

The mitochondrial genome of Saccharomyces cerevisiae contains numerous, densely spaced autonomously replicating sequences

Restriction fragments produced by a complete Sau3A cleavage of Saccharomyces cerevisiae grande mitochondrial DNA were ligated into the yeast-Escherichia coli shuttle vector YIp5 to establish a clone library representing the mitochondrial genome. 30 hybrid plasmids with an average insert size of 1200 bp were chosen at random and tested for the presence of an autonomously replicating sequence (ars). Over two-thirds of these plasmids transformed yeast at high frequency, indicating the mitochondrial genome contains a large number of ars elements. Our calculations suggest there may be over 40 ars elements contained within the mitochondrial DNA with an average spacing of less than 1700 bp. Mapping experiments indicate that ars elements can be found at many locations on the mitochondrial genome, and in the initial example we have tested, the locations of ars elements derived from grande and petite mtDNAs appear to coincide. If we assume that these ars elements represent mitochondrial DNA replication origins used in vivo, these observations would explain in part the fact that petite mtDNAs can be derived from any location on the grande mitochondrial genome.

Transcriptional analysis of the Saccharomyces cerevisiae mitochondrial var1 gene: anomalous hybridization of RNA from AT-rich regions

A family of mitochondrial RNAs hybridizes specifically to the var1 region on Saccharomyces cerevisiae mitochondrial DNA (Farrelly et al., J. Biol. Chem. 257:6581-6587, 1982). We constructed a fine-structure transcription map of this region by hybridizing DNA probes containing different portions of the var1 region and some flanking sequences to mitochondrial RNAs isolated from var1-containing petites. We also report the nucleotide sequence of more than 1.2 kilobases of DNA flanking the var1 gene. Our primary findings are: (i) The family of RNAs we detect with homology to var1 DNA is colinear with the var1 gene. Their direction of transcription is olil to cap, as it is for most other mitochondrial genes. (ii) Extensive hybridization anomalies are present, most likely due to the high A-T (A-U) content of the hybridizing species and to the asymmetric distribution of their G-C residues. An important conclusion is that failure to detect transcripts from A-T-rich regions of the yeast mitochondrial genome by standard blot transfer hybridizations cannot be interpreted to mean that such sequences, which are commonly supposed to be spacer DNA, are noncoding or lack direct function in the expression of mitochondrial genes.

Transcription of mitochondrial DNA

While mitochondrial DNA (mtDNA) is the simplest DNA in nature, coding for rRNAs and tRNAs, results of DNA sequence, and transcript analysis have demonstrated that both the synthesis and processing of mitochondrial RNAs involve remarkably intricate events. At one extreme, genes in animal mtDNAs are tightly packed, both DNA strands are completely transcribed (symmetric transcription), and the appearance of specific mRNAs is entirely dependent on processing at sites signalled by the sequences of the tRNAs, which abut virtually every gene. At the other extreme, gene organization in yeast (Saccharomyces) is anything but compact, with long stretches of AT-rich DNA interspaced between coding sequences and no obvious logic to the order of genes. Transcription is asymmetric and several RNAs are initiated de novo. Nevertheless, extensive RNA processing occurs due largely to the presence of split genes. RNA splicing is complex, is controlled by both mitochondrial and nuclear genes, and in some cases is accompanied by the formation of RNAs that behave as covalently closed circles. The present article reviews current knowledge of mitochondrial transcription and RNA processing in relation to possible mechanisms for the regulation of mitochondrial gene expression.

Nature of an inserted sequence in the mitochondrial gene coding for the 15S ribosomal RNA of yeast

The small ribosomal RNA, or 15S RNA, or yeast mitochondria is coded by a mitochondrial gene. In the central part of the gene, there is a guanine-cytosine (GC) rich sequence of 40 base-pairs, flanked by adenine-thymine sequences. The GC-rich sequence is (5') TAGTTCCGGGGCCCGGCCACGGAGCCGAACCCGAAAGGAG (3'). We have found that this sequence is absent in the 15S rRNA gene of some strains of yeast. When present, it is transcribed into the mature 15S rRNA to produce a longer variant of the RNA. Sequences identical or closely related to this GC-rich sequence are present in many regions of the mitochondrial genome of Saccharomyces cerevisiae. The 5' and 3' terminal structures of all these sequences are highly constant.

Properties of a Saccharomyces cerevisiae mtDNA segment conferring high-frequency yeast transformation

The bakers' yeast Saccharomyces cerevisiae is a facultative anaerobe, tolerant to mutations in its mitochondrial genome. Individual cytoplasmic petite mutants retain genetic information derived from any portion of the parenteral mtDNA, prompting questions concerning distribution of the DNA replication origin(s) on the yeast mitochondrial genome. The experiments described in this paper were designated to test the possibility of using high-frequency yeast transformation as a selection for yeast mtDNA sequences conferring autonomously replicating function. A complete petite mitochondrial genome was inserted into the yeast vector YIp5, and the hybrid plasmid (YRMp1) was used to transform yeast. YRMp1 promoted high-frequency transformation of both wild-type yeast cells and petite mutant hosts lacking mtDNA and was maintained in each of these strains as a high-copy-number extrachromosomal element. The stability and copy-number properties of YRMp1 are similar to those of YRp12, a recombinant plasmid containing a yeast chromosomal autonomously replicating sequence.

Coding capacity of complementary DNA strands

A Fortran computer algorithm has been used to analyze the nucleotide sequence of several structural genes. The analysis performed on both coding and complementary DNA strands shows that whereas open reading frames shorter than 100 codons are randomly distributed on both DNA strands, open reading frames longer than 100 codons ("virtual genes") are significantly more frequent on the complementary DNA strand than on the coding one. These "virtual genes" were further investigated by looking at intron sequences, splicing points, signal sequences and by analyzing gene mutations. On the basis of this analysis coding and complementary DNA strands of several eukaryotic structural genes cannot be distinguished. In particular we suggest that the complementary DNA strand of the human epsilon-globin gene might indeed code for a protein.

Gene conversion at the var1 locus on yeast mitochondrial DNA

Alleles of the var1 locus on yeast mtDNA determine the apparent size of the mitochondrial translation product, var1 polypeptide. We have analyzed most of the different var1 alleles in our collection, which number at least 15, and have developed procedures and a genetic rationale for determining their origin and predicting their behavior in crosses. The var1 alleles are characterized by two genetically defined segments, designated a and b, which can move from one var1 allele to another by asymmetric gene conversion. We show that the a segment behaves as an entity in recombination; it is either present in or absent from different var1 alleles. The b segment usually, but not always, recombines as an entity; in some cases, only portions of the b segment recombine by gene conversion. Thus, the total number of electrophoretically resolvable var1 species we observe is explained by the assortment of a, b, and partial b segments. Each segment recombines at a characteristic frequency; however, one example is presented which shows that the recipient can modulate the frequency of gene conversion. Finally, we show that, like the 21S rDNA region (omega), there is polarity of gene conversion within var1.

Extreme chemical mutagen sensitivity of respiratory adaptation in yeast

The respiratory adaptation process (i.e. essentially mitochondrial biogenesis) in the cells of both wild-type Saccharomyces cerevisiae and strains sensitive to ultraviolet radiation (UV) undergoing transition from the anaerobic to the aerobic state (1-2 h aeration) could be arrested by a prior incubation for 15--30 min with several chemical mutagens and other DNA-acting chemicals at very low concentrations (10-7 to 10-6 M added to cells suspended at the density of 10(7) cells/ml). At the same concentrations, these chemicals also inhibited DNA and RNA biosynthesis in maturing mitochondria during respiratory adaptation. This provides suggestive evidence for the view that the inhibitory effect of the chemical mutagens on respiratory adaptation could be due to lesions introduced into the DNA of promitochondria in the anaerobic cells. The system of respiratory adaptation in S. cerevisiae cells could serve as a rapid test for ascertaining the potentiality of a chemical to affect cellular DNA and probably, in turn, its potentiality to be mutagenic.

Transfer RNA genes in the cap-oxil region of yeast mitochondrial DNA

A cytoplasmic "petite" (rho-) clone of Saccharomyces cerevisiae has been isolated and found through DNA sequencing to contain the genes for cysteine, histidine, leucine, glutamine, lysine, arginine, and glycine tRNAs. This clone, designated DS502, has a tandemly repeated 3.5 kb segment of the wild type genome from 0.7 to 5.6 units. All the tRNA genes are transcribed from the same strand of DNA in the direction cap to oxil. The mitochondrial DNA segment of DS502 fills a sequence gap that existed between the histidine and lysine tRNAs. The new sequence data has made it possible to assign accurate map positions to all the tRNA genes in the cap-oxil span of the yeast mitochondrial genome. A detailed restriction map of the region from 0 to 17 map units along with the locations of 16 tRNA genes have been determined. The secondary structures of the leucine and glutamine tRNAs have been deduced from their gene sequences. The leucine tRNA exhibits 64% sequence homology to an E. coli leucine tRNA.