Assembly of the mitochondrial membrane system. Analysis of structural mutants of the yeast coenzyme QH2-cytochrome c reductase complex

Coenzyme QH2-cytochrome c reductase is a multisubunit complex of the mitochondrial respiratory chain. Mutants of Saccharomyces cerevisiae with lesions in cytochromes b, c1, the non-heme iron protein, and the noncatalytic subunits have been used to study several aspects of the assembly of the complex. Strains with mutations in single subunits exhibit a variety of different phenotypes. Mutants in the 17-kDa (core 3) subunit grow normally on a nonfermentable substrate indicating that this component is not essential for either enzymatic activity or assembly of the enzyme. Mutations in all the other subunits express a respiratory-deficient phenotype and the absence of detectable enzyme activity. Among the respiratory-defective strains, some have mature cytochrome b (non-heme iron protein and cytochrome c1 mutants), while other mutants lack spectrally detectable cytochrome b and have reduced levels of the apoprotein (mutants in the 44-, 40-, 14-, and 11-kDa core subunits). Mutations in single subunits exert different effects on the concentrations of their partner proteins. These may be summarized as follows: 1) No substantial loss in the 44- or 40-kDa core subunits is seen in single mutants; 2) the concentration of cytochrome c1 is also relatively unaffected by mutations in the other subunits except for the cytochrome b mutant which has 60% of the wild type level of cytochrome c1; 3) all the single mutants have only 15-20% of the normal amount of non-heme iron protein; 4) mutations in the non-heme iron protein have no appreciable effect on the concentrations of the other subunits; 5) mutations in single subunits cause parallel decreases in the concentrations of cytochrome b, the 14-, and the 11-kDa subunits. These results indicate that the synthesis or stability of a subset of subunits depends on the presence of other subunit polypeptides of the complex. At present we favor the idea that the observed changes in the concentrations of some subunits are due to higher turnover rates of the proteins in a partially assembled complex. Based on the mutant phenotypes, a tentative model for the assembly of coenzyme QH2-cytochrome c reductase is proposed. According to this model it is envisioned that the subunits interact with one another in the lipid bilayer. Maturation of apocytochrome b occurs after it is assembled with the nonstructural subunits to form a core structure. This intermediate complex interacts with the non-heme iron protein to form the active holoenzyme.     

Nucleotide sequence analysis of the nuclear gene coding for manganese superoxide dismutase of yeast mitochondria, a gene previously assumed to code for the Rieske iron-sulphur protein

We have previously reported the isolation of the gene coding for a 25-kDa polypeptide present in a purified yeast QH2:cytochrome c oxidoreductase preparation, which was thus identified as the gene for the Rieske iron-sulphur protein [Van Loon et al. (1983) Gene 26, 261-272]. Subsequent DNA sequence analysis reported here reveals, however, that the encoded protein is in fact manganese superoxide dismutase, a mitochondrial matrix protein. Comparison with the known amino acid sequence of the mature protein indicates that it is synthesized with an N-terminal extension of 27 amino acids. In common with the N-terminal extensions of other imported mitochondrial proteins, the presequence has several basic residues but lacks negatively charged residues. The function of these positive charges and other possible topogenic sequences are discussed. Sequences 5' of the gene contain two elements that may be homologous to the suggested regulatory sites, UAS 1 and UAS 2 in the yeast CYC1 gene [Guarente et al. (1984) Cell 36, 503-511]. The predicted secondary structures in manganese superoxide dismutase appear to be very similar to those reported for iron superoxide dismutase, suggesting similar three-dimensional structures. Making use of the known three-dimensional structure of the Fe enzyme, the Mn ligands are predicted.     

Molecular characterization of two novel crystal protein genes from Bacillus thuringiensis subsp. thompsoni.

Two genes encoding the predominant polypeptides of Bacillus thuringiensis subsp. thompsoni cuboidal crystals were cloned in Escherichia coli and sequenced. The polypeptides have electrophoretic mobilities of 40 and 34 kDa, with the deduced amino acid sequences predicting molecular masses of 35,384 and 37,505 Da, respectively. No statistically significant similarities were detected between the 40- or 34-kDa crystal protein and any other characterized B. thuringiensis crystal protein, nor were they detected between the 40- and 34-kDa crystal proteins. A 100-MDa plasmid carries both crystal protein genes, which appear to be part of an operon, with the 40-kDa gene 64 nucleotides upstream of the 34-kDa gene. Both crystal proteins are synthesized in approximately the same amounts. Even though small compared with other crystal proteins, the 34-kDa crystal protein has insecticidal activity against lepidopteran larvae (Manduca sexta). The 40-kDa polypeptide appears to have no insecticidal activity, but it could have a role in crystal structure.     

Yeast ubiquinol: cytochrome c oxidoreductase is still active after inactivation of the gene encoding the 17-kDa subunit VI

The single nuclear gene encoding the 17-kDa subunit VI of yeast ubiquinol: cytochrome c oxidoreductase has been inactivated by one-step gene disruption. Disruption was verified by Southern blot analysis of nuclear DNA and immunoblotting. Cells lacking the 17-kDa protein are still capable of growth on glycerol and they contain all other subunits of complex III at wild-type levels, implying that the 17-kDa subunit is not essential for either assembly of complex III, or its function. In vitro, electron transport activity of complex III of mutant cells is about 40% of the wild-type complex, but for the total respiratory chain no significant differences in activity was measured between mutant and wild type. The energy-transducing capacity of the complex is not reduced in the absence of the 17-kDa protein. In a relatively high proportion of the transformants, disruption of the 17-kDa gene was accompanied by the appearance of a second mutation causing a petite phenotype. In these cells which lack cytochrome b, the presence of the 17-kDa protein (after complementation) results in stabilization of cytochrome c1.     

Membrane topography of the subunits of ubiquinol-cytochrome-c oxidoreductase of Saccharomyces cerevisiae. The 14-kDa and the 11-kDa subunits face opposite sides of the mitochondrial inner membrane

The topology of the subunits of ubiquinol-cytochrome-c oxidoreductase of the yeast Saccharomyces cerevisiae has been determined using a digitonin/proteinase K assay. With this assay we were able selectively to disrupt the mitochondrial membranes and to identify the subunits which became proteinase-K sensitive after disruption of either the outer or both outer and inner membranes. This approach confirmed previous indications for the localization of the core I protein, cytochrome c1, cytochrome b, the FeS protein and the 17-kDa subunit, while it also provided direct evidence for the site of accessibility to proteinase K of the 14-kDa and 11-kDa subunits. The 14-kDa subunit faces the mitochondrial matrix and the 11-kDa subunit faces the intermembrane space.     

Rpa4, a homolog of the 34-kilodalton subunit of the replication protein A complex.

Replication protein A (RPA) is a complex of three polypeptides of 70, 34, and 13 kDa isolated from diverse eukaryotes. The complex is a single-stranded DNA-binding protein essential for simian virus 40-based DNA replication in vitro and for viability in the yeast Saccharomyces cerevisiae. We have identified a new 30-kDa human protein which interacts with the 70- and 13-kDa subunits of RPA, with a yeast two-hybrid/interaction trap method. This protein, Rpa4, has 47% identity with Rpa2, the 34-kDa subunit of RPA. Rpa4 associates with the 70- and 13-kDa subunits to form a trimeric complex capable of binding to single-stranded DNA. Rpa4 is preferentially expressed in placental and colon mucosa tissues. In the placenta, Rpa4 is more abundant than the 70-kDa Rpa1 subunit and is not associated with either Rpa1 or with any other single-stranded DNA-binding protein. In proliferating cells in culture, Rpa4 is considerably less abundant than Rpa1 and Rpa2. Northern (RNA) blot analysis suggest that there are alternatively processed forms of the RPA4 mRNA, and Southern blot analysis indicates that beside RPA4 there may be other members of the RPA2 gene family.     

Novel α-Glucosidase from Aspergillus nidulans with Strong Transglycosylation Activity

Aspergillus nidulans possessed an α-glucosidase with strong transglycosylation activity. The enzyme, designated α-glucosidase B (AgdB), was purified and characterized. AgdB was a heterodimeric protein comprising 74- and 55-kDa subunits and catalyzed hydrolysis of maltose along with formation of isomaltose and panose. Approximately 50% of maltose was converted to isomaltose, panose, and other minor transglycosylation products by AgdB, even at low maltose concentrations. The agdB gene was cloned and sequenced. The gene comprised 3,055 bp, interrupted by three short introns, and encoded a polypeptide of 955 amino acids. The deduced amino acid sequence contained the chemically determined N-terminal and internal amino acid sequences of the 74- and 55-kDa subunits. This implies that AgdB is synthesized as a single polypeptide precursor. AgdB showed low but overall sequence homology to α-glucosidases of glycosyl hydrolase family 31. However, AgdB was phylogenetically distinct from any other α-glucosidases. We propose here that AgdB is a novel α-glucosidase with unusually strong transglycosylation activity.     

Characterization of a yeast nuclear gene (MST1) coding for the mitochondrial threonyl-tRNA1 synthetase

The wild-type yeast nuclear gene MST1 complements mutants defective in mitochondrial protein synthesis. The gene has been sequenced and shown to code for a protein of 54,030 kDa. The predicted product of MST1 is 36% identical over its 462 residues to the Escherichia coli threonyl-tRNA synthetase. Amino-acylation of wild-type mitochondrial tRNAs with a mitochondrial extract from mst1 mutants fail to acylate tRNAThr1 (anticodon: 3'-GAU-5') but show normal acylation of tRNAThr2 (anticodon: 3'-UGU-5'). These data suggest the presence of two separate threonyl-tRNA synthetases in yeast mitochondria. Antibodies were prepared against a trpE/MST1 fusion protein containing the 321 residues from the amino-terminal region of the E. coli anthranilate synthetase and 118 residues of the mitochondrial threonyl-tRNA synthetase. Antibodies to the fusion protein detect a 50-55-kDa protein in wild type yeast mitochondria but not in mitochondria of a strain in which the chromosomal MST1 gene was replaced by a copy of the same gene disrupted by insertion of the yeast LEU2 gene. The ability of the mutant with the inactive MST1 gene to charge tRNAThr2 argues strongly for the existence of a second threonyl-tRNA synthetase gene.     

The histidine tRNA genes of yeast

Yeast has at least seven nuclear histidine tRNA genes although there is a single tRNAHis. We have sequenced three of the histidine tRNA genes. The genes have identical coding sequences and the DNA anti-codon sequence GTG corresponds to the GUG anti-codon in tRNAHis. None of the three yeast histidine tRNA genes has an intervening sequence. Two of the three genes contain repeated DNA elements in the region adjacent to the 5' end of the histidine tRNA gene. One of the elements, sigma, is 18 base pairs (bp) from the 5' end of each of these genes, sigma elements are highly conserved and flanked by 5-bp repeats. The other element, delta, is at variable distances from the tRNA gene; one is 439 bp from a histidine tRNA gene and the other is 52 bp from a histidine tRNA gene. These solo delta elements are quite divergent when compared with delta s associated with transposon yeast elements and are not flanked by 5-bp repeats.     

Yeast ribosomal protein S33 is encoded by an unsplit gene.

The structure of the gene coding for ribosomal protein S33, - a protein which escapes the coordinate control of ribosomal protein synthesis in rna 2 mutant cells -, was determined by sequence analysis. The gene comprises an uninterrupted coding region of 204 nucleotides encoding a protein of 8.9 kD. Like for other yeast ribosomal protein genes that have been sequenced so far, a relatively strong codon bias was observed. By S1 nuclease mapping the 5' end of the S33 mRNA was shown to be located at 11 to 15 nucleotides upstream from the initiation codon.