The sorting of mitochondrial DNA and mitochondrial proteins in zygotes: preferential transmission of mitochondrial DNA to the medial bud.

Green fluorescent protein (GFP) was used to tag proteins of the mitochondrial matrix, inner, and outer membranes to examine their sorting patterns relative to mtDNA in zygotes of synchronously mated yeast cells in rho+ x rho0 crosses. When transiently expressed in one of the haploid parents, each of the marker proteins distributes throughout the fused mitochondrial reticulum of the zygote before equilibration of mtDNA, although the membrane markers equilibrate slower than the matrix marker. A GFP-tagged form of Abf2p, a mtDNA binding protein required for faithful transmission of rho+ mtDNA in vegetatively growing cells, colocalizes with mtDNA in situ. In zygotes of a rho+ x rho+ cross, in which there is little mixing of parental mtDNAs, Abf2p-GFP prelabeled in one parent rapidly equilibrates to most or all of the mtDNA, showing that the mtDNA compartment is accessible to exchange of proteins. In rho+ x rho0 crosses, mtDNA is preferentially transmitted to the medial diploid bud, whereas mitochondrial GFP marker proteins distribute throughout the zygote and the bud. In zygotes lacking Abf2p, mtDNA sorting is delayed and preferential sorting is reduced. These findings argue for the existence of a segregation apparatus that directs mtDNA to the emerging bud.

Cytoplasmic type 80S ribosomes associated with yeast mitochondria. IV. Attachment of ribosomes to the outer membrane of isolated mitochondria.

Growing yeast spheroplasts were shown to have, on the average, four times the number of cytoplasmic ribosomes in contact with the outer mitochondrial membrane compared to starved spheroplasts. Ribosomes in contact with mitochondria in the growing spheroplast preparation, like free cytoplasmic ribosomes, exist primarily as polysome structures. In the starved spheroplast preparation, both mitochondria-bound and free cytoplasmic ribosomes exist primarily as monosomes. Mitochondria isolated from growing spheroplasts in a medium containing lmM Mg++ have cytoplasmic ribosomes bound directly to the outer membrane. These ribosomes can be quantitatively removed by washing the mitochondria with 2 mM EDTA. Mitochondria from starved spheroplasts are capable of accepting either free cytoplasmic polysomes or cytoplasmic polysomes extracted from mitochondria. However, the extent of polysome binding to mitochondria was shown to be a direct function of the Mg++ concentration; a smaller percentage of the input polysomes bind as the Mg++ concentration is lowered. At 1 mM Mg++, neither free cytoplasmic nor mitochondria-bound polysomes bind to mitochondria. Nevertheless, when growing spheroplasts are broken and mitochondria isolated in medium containing 1 mM Mg++, the mitochondria are seen to have cytoplasmic ribosomes firmly attached to the outer membrane. This result, in addition to our earlier data (Kellems, R. E., and R. A. Butow. 1974. J. Biol. Chem. 249:3304-3310), support the view that cytoplasmic ribosomes attached to the outer membrane of purified mitochondria were attached in vivo. In preparations of mitochondria isolated from growing spheroplasts, ribosomes appear to be found to specific regions of the outer membrane, namely those regions which are in close association or in contact with the inner mitochondrial membrane. This is particularly evident with mitochondria in a condensed configuration. This finding suggests a mechanism whereby cytoplasmically synthesized mitochondrial protein could be transferred by a process of vectorial translation across both membranes of the organelle.

Euglena gracilis: formation of giant mitochondria.

The addition of antimycin A (0.5 microgram per milliliter) to cultures of a bleached strain of Euglena gracilis in the logarithmic phase of growth on succinate as a carbon source results in (i) an interruption of growth for 24 hours and (ii) an increase in whole-cell respiration and the emergence of a novel succinoxidase activity within 2 to 4 hours. After 3 to 5 hours, the mitochondria enlarge, fuse, and form a sheathlike structure situated close to the periphery of the cell.

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.

The mitochondrial genotype can influence nuclear gene expression in yeast.

Isochromosomal, respiratory-deficient yeast strains, such as a mit-, a hypersuppressive petite, and a petite lacking mitochondrial DNA, are phenotypically identical in spite of differences in their mitochondrial genomes. Subtractive hybridizations of complementary DNA's to polyadenylated RNA isolated from derepressed cultures of these strains reveal the presence of nuclear-encoded transcripts whose abundance varies not only between them and their respiratory-competent parent, but among the respiratory-deficient strains themselves. Transcripts of some nuclear-encoded mitochondrial proteins, like cytochrome c and the alpha and beta subunits of the mitochondrial adenosine triphosphatase, whose abundance is affected by glucose or heme, do not vary. In the absence of major metabolic variables, yeast cells seem to respond to the quality and quantity of mitochondrial DNA and modulate the levels of nuclear-encoded RNA's, perhaps as a means of intergenomic regulation.

Mitochondrial transformation in yeast by bombardment with microprojectiles.

The genetic transformation of mitochondria and chloroplasts has been an intractable problem. The newly developed "biolistic" (biological ballistic) process was used to deliver DNA into yeast cells to stably transform their mitochondria. A nonreverting strain, which is respiratory deficient because of a deletion in the mitochondrial oxi3 gene, was bombarded with tungsten microprojectiles coated with DNA bearing sequences that could correct the oxi3 deletion. Respiratory-competent transformants were obtained in which the introduced oxi3 DNA is integrated at the homologous site in the mitochondrial genome. Organelle genomes can now be manipulated by molecular genetic techniques in the same way as nuclear genomes.

RNA turnover and the control of mitochondrial gene expression.

Recent evidence suggests that RNA turnover in yeast mitochondria is important, not only to regulate RNA abundance, but also to facilitate group I intron splicing and suppress the potentially toxic effect of high levels of excised group I intron RNAs. Protein-assisted splicing of group I introns requires that splicing factors are 'actively' recycled, because of their tight binding to the intron RNA. The putative NTP-dependent RNA helicase Suv3p might promote this recycling and, at the same time, suppress intron overaccumulation because of the functional association of this protein with mtEXO, a novel 3'-5' exoribonuclease that can degrade excised group I intron RNAs.

Organelle transformation: shoot first, ask questions later.

Transformation of mitochondria and chloroplasts is now possible by the application of microprojectile bombardment for the direct delivery of DNA to these organelles in living cells. Recent experiments have shown the feasibility of reverse genetic studies on mitochondrial and chloroplast genomes. Moreover, foreign or novel genes can now be stably introduced into mitochondria and chloroplasts, allowing the study of a much wider range of molecular genetic problems involving organelles than was previously possible.

Adaptive thermogenesis: orchestrating mitochondrial biogenesis.

The biogenesis of mitochondria requires products of the nuclear and mitochondrial genomes. Recent studies of adaptive thermogenesis have shown how mitochondrial proliferation and respiratory activity in brown fat and skeletal muscle are directed by the transcriptional coactivator PGC-1.

A polymerase switch in the synthesis of rRNA in Saccharomyces cerevisiae.

Transcription of ribosomal DNA by RNA polymerase I is believed to be the sole source of the 25S, 18S, and 5.8S rRNAs in wild-type cells of Saccharomyces cerevisiae. Here we present evidence for a switch from RNA polymerase I to RNA polymerase II in the synthesis of a substantial fraction of those rRNAs in respiratory-deficient (petite) cells. The templates for the RNA polymerase II transcripts are largely, if not exclusively, episomal copies of ribosomal DNA arising from homologous recombination events within the ribosomal DNA repeat on chromosome XII. Ribosomal DNA contains a cryptic RNA polymerase II promoter that is activated in petites; it overlaps the RNA polymerase I promoter and produces a transcript equivalent to the 35S precursor rRNA made by RNA polymerase I. Yeast cells that lack RNA polymerase I activity, because of a disruption of the RPA135 gene that encodes subunit II of the enzyme, can survive by using the RNA polymerase II promoter in ribosomal DNA to direct the synthesis of the 35S rRNA precursor. This polymerase switch could provide cells with a mechanism to synthesize rRNA independent of the controls of RNA polymerase I transcription.

A transcriptional switch in the expression of yeast tricarboxylic acid cycle genes in response to a reduction or loss of respiratory function.

The Hap2,3,4,5p transcription complex is required for expression of many mitochondrial proteins that function in electron transport and the tricarboxylic acid (TCA) cycle. We show that as the cells' respiratory function is reduced or eliminated, the expression of four TCA cycle genes, CIT1, ACO1, IDH1, and IDH2, switches from HAP control to control by three genes, RTG1, RTG2, and RTG3. The expression of four additional TCA cycle genes downstream of IDH1 and IDH2 is independent of the RTG genes. We have previously shown that the RTG genes control the retrograde pathway, defined as a change in the expression of a subset of nuclear genes, e.g., the glyoxylate cycle CIT2 gene, in response to changes in the functional state of mitochondria. We show that the cis-acting sequence controlling RTG-dependent expression of CIT1 includes an R box element, GTCAC, located 70 bp upstream of the Hap2,3,4,5p binding site in the CIT1 upstream activation sequence. The R box is a binding site for Rtg1p-Rtg3p, a heterodimeric, basic helix-loop-helix/leucine zipper transcription factor complex. We propose that in cells with compromised mitochondrial function, the RTG genes take control of the expression of genes leading to the synthesis of alpha-ketoglutarate to ensure that sufficient glutamate is available for biosynthetic processes and that increased flux of the glyoxylate cycle, via elevated CIT2 expression, provides a supply of metabolites entering the TCA cycle sufficient to support anabolic pathways. Glutamate is a potent repressor of RTG-dependent expression of genes encoding both mitochondrial and nonmitochondrial proteins, suggesting that it is a specific feedback regulator of the RTG system.

A basic helix-loop-helix-leucine zipper transcription complex in yeast functions in a signaling pathway from mitochondria to the nucleus.

The expression of some nuclear genes in Saccharomyces cerevisiae, such as the CIT2 gene, which encodes a glyoxylate cycle isoform of citrate synthase, is responsive to the functional state of mitochondria. Previous studies identified a basic helix-loop-helix-leucine zipper (bHLH/Zip) transcription factor encoded by the RTG1 gene that is required for both basal expression of the CIT2 gene and its increased expression in respiratory-deficient cells. Here, we describe the cloning and characterization of RTG3, a gene encoding a 54-kDa bHLH/Zip protein that is also required for CIT2 expression. Rtg3p binds together with Rtg1p to two identical sites oriented as inverted repeats 28 bp apart in a regulatory upstream activation sequence element (UASr) in the CIT2 promoter. The core binding site for the Rtg1p-Rtg3p heterodimer is 5'-GGTCAC-3', which differs from the canonical E-box site, CANNTG, to which most other bHLH proteins bind. We demonstrate that both of the Rtg1p-Rtg3p binding sites in the UAS(r) element are required in vivo and act synergistically for CIT2 expression. The basic region of Rtg3p conforms well to the basic region of most bHLH proteins, whereas the basic region of Rtg1p does not. These findings suggest that the Rtg1p-Rtg3p complex interacts in a novel way with its DNA target sites.

RNA processing and expression of an intron-encoded protein in yeast mitochondria: role of a conserved dodecamer sequence.

The 3' ends of most Saccharomyces cerevisiae mitochondrial mRNAs terminate at a conserved dodecamer sequence, 5'-AAUAAUAUUCUU-3', of unknown function. We have studied the consequences of mutations within a dodecamer found in an 1,143-base-pair optional intron of the mitochondrial large (21S) rRNA gene on RNA processing. The dodecamer is situated at the 3' end of an expressed open reading frame (ORF) within that intron, and the mutations are two adjacent transversions that extend the intron ORF by 51 nucleotides. The strain harboring these mutations, L5-10-1, is defective in biased intron transmission in crosses to strains that lack the intron, as are other mutants which contain nucleotide changes within the ORF (I. G. Macreadie, R. M. Scott, A. R. Zinn, and R. A. Butow, Cell 41:395-402, 1985). However, unlike these other mutants, wild-type strains, or petites which retain the intron allele, L5-10-1 is defective in processing at the intron dodecamer. In addition, L5-10-1 lacks a prominent 2.7-kilobase RNA containing both intron and exon sequences and at least two of four RNAs that correspond to various forms of the excised intron. We propose that these RNAs, missing in L5-10-1 but present in all other strains examined, arise in part by processing at the intron dodecamer. In addition, in all strains examined, we have detected a novel processing activity in which precursor 21S rRNA transcripts are cleaved in the upstream exon, about 1,500 nucleotides from the 5' end of the RNA. This activity, together with 3' intron dodecamer cleavage, probably accounts for the 2.7-kilobase RNA species, a candidate for the mRNA for the intron-encoded protein.

Intramitochondrial functions regulate nonmitochondrial citrate synthase (CIT2) expression in Saccharomyces cerevisiae.

We have examined the effects of perturbation of mitochondrial function on expression of two nuclear genes encoding the mitochondrial and peroxisomal forms of citrate synthase in Saccharomyces cerevisiae, CIT1 and CIT2. CIT2 expression was as much as 30-fold higher in [rho0] petites, than in isochromosomal [rho+] cells, whereas CIT1 expression was slightly down regulated in [rho0] cells. CIT2 expression was also increased in [rho+] cells by inhibition of respiration with antimycin A or in [rho+] cells containing a disruption of the CIT1 gene. These effects were additive, and together they approached the level of CIT2 expression seen in [rho0] cells. Experiments using heterologous gene fusions showed that all of the effects leading to increased expression of CIT2 were transcriptionally controlled through 5'-flanking CIT2 DNA sequences. Analysis of [rho+] and [rho0] cells containing disruptions of CIT1 and CIT2, singly and in combination, showed that the peroxisomal citrate synthase could partially spare the mitochondrial isoform for growth yield in [rho+] but not in [rho0] cells. These studies suggest a physiological role for increased expression of CIT2 in cells with altered mitochondrial function. They also provide additional evidence for a retrograde path of communication from mitochondria to the nucleus in yeast cells.

Interaction between the yeast mitochondrial and nuclear genomes influences the abundance of novel transcripts derived from the spacer region of the nuclear ribosomal DNA repeat.

We have identified stable transcripts from the so-called nontranscribed spacer region (NTS) of the nuclear ribosomal DNA repeat in certain respiration-deficient strains of Saccharomyces cerevisiae. These RNAs, which are transcribed from the same strand as is the 37S rRNA precursor, are 500 to 800 nucleotides long and extend from the 5' end of the 5S rRNA gene to three major termination sites about 1,780, 1,830, and 1,870 nucleotides from the 3' end of the 26S rRNA gene. A survey of various wild-type and respiration-deficient strains showed that NTS transcript abundance depended on the mitochondrial genotype and a single codominant nuclear locus. In strains with that nuclear determinant, NTS transcripts were barely detected in [rho+] cells, were slightly more abundant in various mit- derivatives, and were most abundant in petites. However, in one petite that was hypersuppressive and contained a putative origin of replication (ori5) within its 757-base-pair mitochondrial genome, NTS transcripts were no more abundant than in [rho+] cells. The property of low NTS transcript abundance in the hypersuppressive petite was unstable, and spontaneous segregants that contained NTS transcripts as abundant as in the other petites examined could be obtained. Thus, respiration deficiency per se is not the major factor contributing to the accumulation of these unusual RNAs. Unlike RNA polymerase I transcripts, the abundant NTS RNAs were glucose repressible, fractionated as poly(A)+ RNAs, and were sensitive to inhibition by 10 micrograms of alpha-amanitin per ml, a concentration that had no effect on rRNA synthesis. Abundant NTS RNAs are therefore most likely derived by polymerase II transcription.

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.

Functional expression of a yeast mitochondrial intron-encoded protein requires RNA processing at a conserved dodecamer sequence at the 3' end of the gene.

All mRNAs of yeast mitochondria are processed at their 3' ends within a conserved dodecamer sequence, 5'-AAUAAUAUUCUU-3'. A dominant nuclear suppressor, SUV3-I, was previously isolated because it suppresses a dodecamer deletion at the 3' end of the var1 gene. We have tested the effects of SUV3-1 on a mutant containing two adjacent transversions within a dodecamer at the 3' end of fit1, a gene located within the 1,143-base-pair intron of the 21S rRNA gene, whose product is a site-specific endonuclease required in crosses for the quantitative transmission of that intron to 21S alleles that lack it. The fit1 dodecamer mutations blocked both intron transmission and dodecamer cleavage, neither of which was suppressed by SUV3-1 when present in heterozygous or homozygous configurations. Unexpectedly, we found that SUV3-1 completely blocked cleavage of the wild-type fit1 dodecamer and, in SUV3-1 homozygous crosses, intron conversion. In addition, SUV3-1 resulted in at least a 40-fold increase in the amount of excised intron accumulated. Genetic analysis showed that these phenotypes resulted from the same mutation. We conclude that cleavage of a wild-type dodecamer sequence at the 3' end of the fit1 gene is essential for fit1 expression.

Complex recognition site for the group I intron-encoded endonuclease I-SceII.

We have characterized features of the site recognized by a double-stranded DNA endonuclease, I-SceII, encoded by intron 4 alpha of the yeast mitochondrial COX1 gene. We determined the effects of 36 point mutations on the cleavage efficiency of natural and synthetic substrates containing the Saccharomyces capensis I-SceII site. Most mutations of the 18-bp I-SceII recognition site are tolerated by the enzyme, and those mutant sites are cleaved between 42 and 100% as well as the wild-type substrate is. Nine mutants blocked cleavage to less than or equal to 33% of the wild-type, whereas only three point mutations, G-4----C, G-12----T, and G-15----C, block cleavage completely. Competition experiments indicate that these three substrates are not cleaved, at least in part because of a marked reduction in the affinity of the enzyme for those mutant DNAs. About 90% of the DNAs derived from randomization of the nucleotide sequence of the 4-bp staggered I-SceII cleavage site are not cleaved by the enzyme. I-SceII cleaves cloned DNA derived from human chromosome 3 about once every 110 kbp. The I-SceII recognition sites in four randomly chosen human DNA clones have 56 to 78% identity with the 18-bp site in yeast mitochondrial DNA; they are cleaved at least 50% as well as the wild-type mitochondrial substrate despite the presence of some substitutions that individually compromise cleavage of the mitochondrial substrate. Analysis of these data suggests that the effect of a given base substitution in I-SceII cleavage may depend on the sequence at other positions.

Mobile group II introns of yeast mitochondrial DNA are novel site-specific retroelements.

Group II introns aI1 and aI2 of the yeast mitochondrial COXI gene are mobile elements that encode an intron-specific reverse transcriptase (RT) activity. We show here that the introns of Saccharomyces cerevisiae ID41-6/161 insert site specifically into intronless alleles. The mobility is accompanied by efficient, but highly asymmetric, coconversion of nearby flanking exon sequences. Analysis of mutants shows that the aI2 protein is required for the mobility of both aI1 and aI2. Efficient mobility is dependent on both the RT activity of the aI2-encoded protein and a separate function, a putative DNA endonuclease, that is associated with the Zn2+ finger-like region of the intron reading frame. Surprisingly, there appear to be two mobility modes: the major one involves cDNAs reverse transcribed from unspliced precursor RNA; the minor one, observed in two mutants lacking detectable RT activity, appears to involve DNA level recombination. A cis-dominant splicing-defective mutant of aI2 continues to synthesize cDNAs containing the introns but is completely defective in both mobility modes, indicating that the splicing or the structure of the intron is required. Our results demonstrate that the yeast group II intron aI2 is a retroelement that uses novel mobility mechanisms.

Patterns of mitochondrial sorting in yeast zygotes.

Inheritance of mitochondrial DNA (mtDNA) in Saccharomyces cerevisiae is usually biparental. Pedigree studies of zygotic first buds indicate limited mixing of wild-type (p+) parental mtDNAs: end buds are frequently homoplasmic for one parental mtDNA, while heteroplasmic and recombinant progeny usually arise from medial buds. In crosses involving certain petites, however, mitochondrial inheritance can be uniparental. In this study we show that mitochondrial sorting can be influenced by the parental mtDNAs and have identified intermediates in the process. In crosses where mtDNA mixing is limited and one parent is prelabeled with the matrix enzyme citrate synthase 1 (CS1), the protein freely equilibrates throughout the zygote before the first bud has matured. Furthermore, if one parent is p0 (lacking mtDNA), mtDNA from the p+ parent can also equilibrate; intracellular movement of mtDNA is unhindered in this case. Surprisingly, in zygotes from a p0 CS1+ x p+ CS1- cross, CS1 is quantitatively translocated to the p+ end of the zygote before mtDNA movement; subsequently, both components equilibrate throughout the cell. This initial vectorial transfer does not require respiratory function in the p+ parent, although it does not occur if that parent is p-. Mouse dihydrofolate reductase (DHFR) present in the mitochondrial matrix can also be vectorially translocated, indicating that the process is general. Our data suggest that in zygotes mtDNA movement may be separately controlled from the movement of bulk matrix constituents.