Identification of mitochondrial gene products by DNA-directed protein synthesis in vitro

Mitochondrial protein synthesis: figuring the fundamentals, complexities and complications, of mammalian mitochondrial translation

Mitochondrial protein synthesis is essential for all mammals, being responsible for providing key components of the oxidative phosphorylation complexes. Although only thirteen different polypeptides are made, the molecular details of this deceptively simple process remain incomplete. Central to this process is a non-canonical ribosome, the mitoribosome, which has evolved to address its unique mandate. In this review, we integrate the current understanding of the molecular aspects of mitochondrial translation with recent advances in structural biology. We identify numerous key questions that we will need to answer if we are to increase our knowledge of the molecular mechanisms underlying mitochondrial protein synthesis.

RNA processing in yeast mitochondria: characterization of mit(-) mutants disturbed in the synthesis of subunit I of cytochrome c oxidase

Mit(-) mutants disturbed in the synthesis of cytochrome c oxidase subunit I lack the mRNA for this protein and accumulate longer RNAs still containing intron sequences. We have analyzed the patterns of transcripts occurring in several such mutants in an attempt to define a pathway of processing events and to demarcate intron-sequences involved in RNA splicing. We find that processing does not follow a strictly ordered pathway and, in contrast to the situation for the cytochrome b gene, that a block in the processing of an intron does not necessarily lead to a block in the processing of introns downstream. Although in some cases, this may result from overlapping specificities of intronic-URF encoded RNA maturases, an internal start of translation on precursor RNAs seems more likely.M5-16, a mutant deleted for a large part of the central portion of the subunit I gene exhibits delayed processing and a highly simplified pattern of intermediates. The lengths of these indicate that maturation of the mRNA for subunit I involves processing, as well as splicing.

In-vitro translation of mitochondrial mRNAs by yeast mitochondrial ribosomes is hampered by the lack of start-codon recognition

In an attempt to reconstitute an homologous in-vitro translation system for yeast mitochondrial mRNAs, we have isolated ribosomes, supernatant factors, and tRNAs from mitochondria of Saccharomyces carlsbergensis. While poly(U) is translated faithfully in this system, no translation of in-vitro synthesised cytochrome c oxidase subunit II (COX2) mRNA could be detected. Formation of formylmethionyl-puromycin on mitochondrial ribosomes is stimulated by ApUpG, but not by COX2 mRNA, although mitochondrial small ribosomal subunits bind to this mRNA in vitro, even without added tRNA and initiation factors. We conclude, therefore, that the inability to faithfully translate mitochondrial mRNAs in vitro may be the result of an inability of mitochondrial ribosomes to recognize the initiation codon.

Translation of polyuridylic acid in lysed mitochondria

After osmotic shock with 50 mM Tricine buffer (pH 7.9), isolated mitochondria from D. Melanogaster embryos are treated with a low concentration of Triton X-100 (25 micrograms/mg of protein). The lysed mitochondria are still capable of RNA and protein synthesis. While incorporation of labeled precursor is often higher in lysed than in intact mitochondria, neosynthesized proteins exhibit similar electrophoretic patterns. Studies of labeled precursor incorporation in the presence of various effectors indicate a better accessibility to the translation machinery in lysed mitochondria than in intact mitochondria. Such a system has proven capable of translating an exogenous synthetic mRNA, i.e., poly (U).

Kinetoplast DNA segments function as promoters in Escherichia coli cells

By means of coupled transcription-translation in Escherichia coli cell-free system, an open reading frame was found in the Crithidia oncopelti maxi-circle kDNA segment cloned in the hybrid plasmid pCo1. Subfragments from this region were tested for their ability to function as promoters in E. coli cells. For this purpose the vector pVE8 was constructed using the aminoglycoside phosphotransferase II (APTII) gene from the Tn5 transposon, and the pHC79 cosmid. After cloning of Sau3A fragments of pCo1 by insertion into pVE8 three types of plasmids containing promoter sequences were obtained. Two of these plasmids displayed promoter strength in E. coli cells greater than that of the normal promoter of the APTII gene. However the promoters found are not necessary for the coupled transcription-translation of kDNA in the E. coli cell-free system.

The local expression of adult chicken heart myosins during development. I. The three days embryonic chicken heart

Immunofluorescence studies were performed on serial sections of three days embryonic chicken hearts using antibodies specific for adult atrial and ventricular myosin heavy chains respectively. The anti-ventricular myosin serum reacted with the entire myocardium showing a decreasing intensity going from the truncus arteriosus to the atrial part; however, the antiatrial myosin serum reacted weakly with the myocardium of the atrial part. Two other interesting observations were made, i) the anti-atrial myosin serum reacted with non-myocardial cells in the cardiac jelly, ii) both antisera reacted with a thin myocardial layer, extending from the ventral wall of the atrial part via the medio-dorsal wall of the atrio-ventricular canal to the dorsal wall of the ventricular part.

Variation, transcription and circular RNAs of the mitochondrial gene for subunit I of cytochrome c oxidase

The gene for subunit I of cytochrome c oxidase, contained within the OX13 region of yeast mitochondrial DNA, is split and shows a remarkable variation in structure, which is strain-dependent. The most complex form so far characterized is that of the Saccharomyces cerevisiae strain KL14-4A, in which nine or possibly ten exons are separated by eight to nine introns. At least four of these are facultative, two being absent from S. cerevisiae strain D273-10B (sequenced by Bonitz et al., 1980) and a further two lacking from the gene in Saccharomyces carlsbergensis. The complexity of the gene in KL14-4A is also reflected in its transcript pattern. RNA blot hybridization with isolated and cloned DNA fragments of the OX13 region permits visualization of more than 60 RNAs, which show overlapping and discontinuous hybridization behaviour. In the less complex strains D273-10B and S. carlsbergensis, this number is 20 and 11, respectively. These RNAs are most likely intermediates in processing events leading to the appearance of the mature messenger RNA for cytochrome c oxidase subunit I, which we identify as a 2100-nucleotide transcript (18SE). Most of the processing events are dependent on mitochondrial protein synthesis and do not constitute a single obligatory processing pathway. Like other yeast mitochondrial mRNAs, the 18 S RNA contains a long, untranslated 5' flanking sequence (approximately 400 nucleotides). One unusual aspect of splicing events involving OX13 transcripts is the accumulation of three of the excised introns as single-stranded RNA circles. These abundant and stable transcripts appear to be covalently closed. The simplest assumption is that they arise as (by)-products of splicing, but secondary ligation events have not been excluded. Their function is as yet unknown.

Initiation of transcription of genes for mitochondrial ribosomal RNA in yeast: comparison of the nucleotide sequence around the 5'-ends of both genes reveals a homologous stretch of 17 nucleotides

The DNA sequence around the beginning of the genes coding for the large and small ribosomal RNAs in yeast mitochondria has been established. In order to determine the 5'-end points of the ribosomal RNAs, DNA fragments were labelled in vitro at a restriction site within each gene and hybridized with ribosomal RNA. The hybrids were then treated with S1 nuclease and the products analysed for size by gel electrophoresis. This enabled us to identify where in the determined DNA sequence the 21S ribosomal RNA and the precursor for 15S ribosomal RNA (15.5S rRNA) start, since both transcripts are initiated de novo (Levens et al. (1981) J.Biol.Chem., 256, 5226-5232). Comparison of the DNA sequences around the start points of transcription reveals the existence of a homologous stretch of 17 nucleotides. This conserved sequence may be an essential element of a promoter in mtDNA.

Use of a synthetic DNA oligonucleotide to probe the precision of RNA splicing in a yeast mitochondrial petite mutant

In some strains of Saccharomyces cerevisiae the mitochondrial gene coding for 21S rRNA is interrupted by an intron of 1143 bp. This intron contains a reading frame for 235 amino acids: Unassigned Reading Frame (URF). In order to check whether expression of this URF is required for proper splicing of precursors to 21S rRNA, the precision of RNA splicing was analysed in a petite mutant, where no mitochondrial protein synthesis is possible anymore. We have devised a new assay to monitor the precision of the splicing event. The method is of general application, provided that the sequence of the splice boundaries is known. In the case of the 21S rRNA it involves the synthesis of the DNA oligonucleotide d(CGATCCCTATTGTC( complementary to the 5' d(CGATCCCTAT) and 3' d(TGTC) borders flanking the intron in the 21S rRNA gene. The oligonucleotide is labelled with 32p at the 5'-end, hybridised to RNA and subsequently subjected to digestion with S1 nuclease. Resistance to digestion will only be observed if the correct splice-junction is made. The petite mutant we have studied contains a 21S rRNA with the same migration behaviour as wildtype 21S rRNA. In RNA blotting experiments, using an intron specific hybridisation probe, the same intermediates in splicing are found both in wild type and petite mutant. Finally the synthetic oligonucleotide hybridises to petite 21S rRNA and its thermal dissociation behaviour is indistinguishable from a hybrid formed with wildtype 21S rRNA. We conclude that expression of the URF, present in the intron of the 21S rRNA gene, is not required for processing and correct splicing of 21S ribosomal precursor RNA.

In vitro reconstruction of the mitochondrial translation system of yeast

We have isolated the translation system from yeast mitochondria and have reconstructed it in vitro. This submitochondrial system, composed of mitochondrial ribosomes, tRNA, pH 5 fraction and mRNA, is maximally active at 10 mM Mg2+ and 100 mM KCl or NH4Cl. NH4+ is more stimulatory than K+. Added Escherichia coli tRNA gives less than half the activity obtained with added mitochondrial tRNA. Activity is enhanced with protease inhibitors but not with Ca2+, spermine, or spermidine. In contrast to heterologous translation systems, the present system produces products with molecular weights similar to those of products synthesized by yeast mitochondria in vivo and by intact yeast mitochondria in vitro. The results support the idea that the unique coding features of the mitochondrial genome require a unique translation system for accurate translation of mitochondrial mRNAs.

A putative precursor for the small ribosomal RNA from mitochondria of Saccharomyces cerevisiae

We have characterized a putative precursor RNA (15.5S) for the 15S ribosomal RNA in mitochondria of Saccharomyces cerevisiae. Hybrids were formed with mitochondrial RNA and mtDNA fragments terminally labelled at restriction sites located within the gene coding for 15S ribosomal RNA and treated with S1 nuclease (Berk, A.J. and Sharp, J.A. (1977) 12, 721-732). Sites of resistant hybrids were measured by agarose gel electrophoresis and end points of RNAs determined. The 15.5S RNA is approximately 80 nucleotides longer than the 15S ribosomal RNA, with the extra sequences being located at the 5'-end. Both 15S ribosomal RNA and 15.5S RNA are fully localised within a 2000 base pair HapII fragment. This putative precursor and the mature 15S ribosomal RNA are also found in petite mutants which retain the 15S ribosomal RNA gene. The petite mutant with the smallest genetic complexity has its end point of deletion (junction) just outside the HapII site located in the 5' flank of the 15S ribosomal RNA genes as determined by S1 nuclease analysis. This leaves a DNA stretch approximately 300 base pairs long where an initiation signal for mitochondrial transcription may be present.

In vitro suppression of UGA codons in a mitochondrial mRNA

Although both prokaryotic and eukaryotic messenger RNAs can be easily translated in heterologous protein-synthesizing systems, attempts to achieve correct synthesis of mitochondrial proteins by translation of mitochondrial mRNAs in such systems have failed. In general, the products of synthesis are of low molecular weight and presumably represent fragments of mitochondrial proteins. These fragments display a strong tendency to aggregate. Explanations have included the use by mitochondria of codons requiring a specialized tRNA population and the fortuitous occurrence within genes of purine-rich sequences resembling bacterial ribosome binding sites. In addition, the long 5'-leader sequences present in many mitochondrial (mt) RNAs may also contribute to difficulties in mRNA recognition by heterologous ribosomes. Recent sequence analysis of human mtDNA suggests that the genetic code used by mammalian mitochondria deviates in a number of respects from the 'universal' code, the most striking of these being the use of the UGA termination codon to specify tryptophan. That this may also apply in yeast mitochondria has been shown by Fox and Macino et al., thus providing an obvious and easily testable explanation for the inability of heterologous systems to synthesize full-length mitochondrial proteins. We confirm this explanation and describe here the in vitro synthesis of a full-length subunit II of yeast cytochrome c oxidase in a wheat-germ extract supplemented with a partially purified mitochondrial mRNA for this protein and a UGA-suppressor tRNA from Schizosaccharomyces pombe.

Splice point sequence and transcripts of the intervening sequence in the mitochondrial 21S ribosomal RNA gene of yeast

By S1 nuclease mapping we have located the intervening sequence in the large ribosomal RNA gene of Saccharomyces cerevisiae omega+ strains 570 bp from the 3' end of the rRNA gene. No intervening sequence was detected at this position in S. carlsbergensis, but the sequences of the mature 21S rRNAs of these two strains appear to be identical in this region. By comparing the DNA sequence of the region of the intervening sequence in an omega+ strain with the corresponding sequence in S. carlsbergensis, we have determined the splice points of the 21S rRNA gene. These sequences show no homology with splice points in nuclear and viral genes or with the splice points in the chloroplast 23S rRNA gene of Chlamydomonas. The external borders of the splice points have a complementary sequence in the intervening sequence. The largest transcript hybridizing with the probe of the intervening sequence has a size corresponding to that expected for an rRNA precursor still containing the intervening sequence; the smallest transcript corresponds in size to the intervening sequence itself.

Mutations affecting RNA splicing and the interaction of gene expression of the yeast mitochondrial loci cob and oxi-3

In Saccharomyces cerevisiae strains KL14-4A and 777-3A, four intervening sequences of 1900 (l alpha beta), 1400 (l beta gamma), 1300 (l gamma delta) and 650 bp (l delta epsilon) separate the five coding sequences (alpha-epsilon) of the structural gene (cob) for cytochrome b. Its major transcript is an 18S RNA (2200 nucleotides) which is likely to be the functional mRNA. The lengths of a series of larger transcripts and their hybridization with probes specific for different intervening sequences are consistent with their being intermediates in a splicing process which generates 18S RNA from a giant primary transcript (greater than or equal to 7.5 kb) covering the whole cob region. There is no absolute order of splicing. The intervening sequence l alpha beta is excised in two stages. The first generates a stable 10S RNA, coded for by sequences immediately downstream of the 18S RNA coding segment alpha. The function of this RNA is unknown. Its excision is an early step in processing, whereas excision of the remainder of l alpha beta is a late event. We have studied four cytochrome b-deficient mutants. These map in intervening sequences and are splicing-defective. They accumulate 22S-28S RNAs which contain one or more intervening sequences. The l alpha beta mutants synthesize long, novel polypeptides, antigenically related to cytochrome b, possibly as a result of read-through into the intervening sequences. Several cob mutants also display alterations in their transcripts of oxi-3, the locus which codes for cytochrome c oxidase subunit I. This indicates that interactions between cob and oxi-3 exist at the level of RNA processing.

Transcription maps of mtDNAs of two strains of saccharomyces: transcription of strain-specific insertions; Complex RNA maturation and splicing

We have developed a two-dimensional method for simultaneously mapping on the yeast mtDNA genome all the transcripts representing more than 0.01% of mtRNA. In two yeast strains, Saccharomyces carlsbergensis NCYC-74 and Saccharomyces cerevisiae KL14-4A, about 25 discrete transcripts were found apart from tRNAs. The mtDNAs of these strains differ by the absence (NCYC-74) or presence (KL 14-4A) of various large insertions located within genetically active regions. The transcripts can all be related to known loci on the genetic map. In nearly all cases the RNAs are much longer than required to specify the known protein product of the locus concerned. The organization of the transcripts is similar in the two strains except at the positions of the large insertions (500-3300 bp) in the oxi-3 and cob loci. The sequences of these insertions are present in RNA species larger than 25S, but are absent from smaller transcripts of the same regions. This is probably due to splicing, since the coding sequences for most of these smaller transcripts are noncontiguous. The smaller transcripts of other loci also seem to arise from processing of larger RNA species. The oxi-3 locus, containing the structural gene for cytochrome c oxidase subunit l, is transcribed in a very complex fashion that suggests differential splicing into partially overlapping transcripts. This may indicate that oxi-3 has additional genetic functions, including possible control of the biosynthesis of cytochrome c oxidase holoenzyme or its assembly into the mitochondrial inner membrane. As in the case of the eucaryote nucleus, the regulation of mitochondrial gene expression seems to occur more at the level of RNA processing than has been recognized thus far.