Nucleotide sequence of the genes coding for alpha, beta and gamma subunits of the proton-translocating ATPase of Escherichia coli

A nucleotide sequence of 2328 base pairs comprising a portion of the gene cluster for the proton-translocating ATPase of $\text{E. coli}$ was determined. The sequence covers most of the gene for α subunit, the entire gene for γ subunit and the amino terminal portion of the gene for β subunit, along with the flanking regions of these genes. The amino acid sequences of these subunits deduced from the DNA sequences indicate that the α and γ subunits have 513 and 287 amino acid residues, respectively. A possible secondary structure for each subunit was estimated from the inferred primary structure. The intercistronic regions between the genes for α and γ and between γ and β are 49 and 26 base pairs, respectively. The significance of codon usage in these genes is discussed in correlation with their expression.     

Subunit specific antisera to the Escherichia coli ATP synthase: effects on ATPase activity, energy transduction, and enzyme assembly

We obtained antisera to each of the five subunits (α, β, γ, δ, and ?) of the F₁ portion of the proton-translocating ATPase from Escherichia coli (ECF₁). No cross-reaction between the antiserum to a given subunit and any of the other four subunits was observed by Ouchterlony immunodiffusion. The α antiserum reacted only with the denatured α chain. Antibodies to either subunit β or subunit γ inhibited the ATPase activity of the enzyme. The ATPase activity of the holoenzyme in the everted membrane vesicles was just as sensitive as purified ECF₁ to inhibition by the anti-β or anti-γ serum. A prolonged digestion of ECF₁ with trypsin removed intact γ from ECF₁, but did not alter the sensitivity of the ATPase to inhibition by the anti-γ serum. Proteolytic fragments were isolated from the trypsinized enzyme. They gave an immunoprecipitation band with the anti-γ serum, but none of the other subunit antisera. The antiδ serum detached ECF₁ from everted membrane vesicles and completely blocked both the ATP- and respiration-dependent pyridine nucleotide transhydrogenase, an energylinked membrane function. The δ antiserum had no effect on the ATPase activity of the ECF₁. The e antiserum stimulated the ATPase activity of purified ECF₁ as shown previously (P. P. Laget and J. B. Smith, Arch. Biochem. Biophys.197, 83, 1979), but strongly inhibited the holoenzyme in membrane vesicles. The α antiserum completely blocked the ATP-driven transhydrogenase. The same antiserum maximally inhibited the respiratory chain-driven reaction by only 35%. These observations indicate that the antiserum selectively affected energy transduction mediated by the ATPase. The protonmotive force generated by substrate oxidation was probably not dissipated by the ? antiserum. Adsorbing the δ or ? antiserum with everted membrane vesicles selectively removed those antibodies that reacted with membrane-bound ATPase. The adsorbed sera still reacted strongly with purified ECF₁, and prevented it from restoring ATP-dependent proton translocation in ECF₁-depleted vesicles. Therefore, it appears that more of the δ and the ? subunit is exposed in the purified ECF₁ molecule than in the membrane-bound enzyme.     

Subunit-subunit interactions in TF1 as revealed by ligand binding to isolated and integrated α and β subunits

The binding of 3′-O-(1-naphthoyl)adenosinetriphosphate (1-naphthoyl-ATP), ATP and ADP to TF₁ and to the isolated α and β subunits was investigated by measuring changes of intrinsic protein fluorescence and of fluorescence anisotropy of 1-naphthoyl-ATP upon binding. The following results were obtained. (1) The isolated α and β subunits bind 1 mol 1-naphthoyl-ATP with a dissociation constant (K_D(1-naphthoyl-ATP)) of 4.6 μM and 1.9 μM, respectively. (2) The K_D(ATP) for α and β subunits is 8 μM and 11 μM, respectively. (3) The K_D(ADP) for α and β subunits is 38 μM μM and 7 μM, respectively. (4) TF₁ binds 2 mol 1-naphthoyl-ATP per mol enzyme with K_D = 170 nM. (5) The rate constant for 1-naphthoyl-ATP binding to α and β subunit is more than 5 · 10⁴ M⁻¹s⁻¹. (6) The rate constant for 1-naphthoyl-ATP binding to TF₁ is 6.6 · 10³ M⁻¹ · s⁻¹ (monophasic reaction); the rate constant for its dissociation in the presence of ATP is biphasic with a fast first phase (k^A₋₁ = 3 · 10⁻³s⁻¹) and a slower second phase (k^A₋₂ < 0.2 · 10⁻³s⁻¹). From the appearance of a second peak in the fluorescence emission spectrum of 1-naphthoyl-ATP upon binding it is concluded that the binding sites in TF₁ are located in an environment more hydrophobic than the binding sites on isolated α and β subunits. The differences in kinetic and thermodynamic parameters for ligand binding to isolated versus integrated α and β subunits, respectively, are explained by interactions between these subunits in the enzyme complex.     

Specific labelling of the (Ca2+ + Mg2+)-ATPase of Escherichia coli with 8-azido-ATP and 4-chloro-7-nitrobenzofurazan

1. 8-Azido-ATP is a substrate for Escherichia coli (Ca²⁺ + Mg²⁺)-ATPase (E. coli F₁).2. Illumination of E. coli F₁ in the presence of 8-azido-ATP causes inhibition of ATPase activity. The presence of ATP during illumination prevents inhibition.3. 8-Azido-ATP and 4-chloro-7-nitrobenzofurazan (NbfCl) bind predominantly to the α subunit of the enzyme, but also significantly to the β subunit.4. The α subunit of E. coli F₁ seems to have some properties that in other F₁-ATPases are associated with the β subunit.     

Biosynthetic approach to the structure of F1-ATPase (BF1 factor) from Micrococcus lysodeikticus membranes: in vivo labeling of the polypeptides and study of their relationship

The F₁-ATPase or BF₁ factor was purified from Micrococcus lysodeikticus substrain B grown in a synthetic medium in the presence of tritiated amino acids. When analyzed in sodium dodecyl sulfate-7% polyacrylamide gels, the fresh purified preparation contained α, β, γ subunits (referred as the intrinsic subunits) and two other polypeptides (designated as X and component of relative mobility 1.0) whose status as subunits remains to be established. This overall polypeptide composition was similar to that of the F₁-ATPase isolated from the same strain grown in complex medium (J. Carreira, J. M. Andreu, M. Nieto, and E. Muñoz., 1976 Mol. Cell. Biochem.10, 67–76). The distribution of ³H-labeled amino acids into purified F₁-ATPase and its constituent polypeptides under different stages of growth was used to investigate the biosynthetic relationship between the different polypeptides. The incorporation of amino acids into purified BF₁ factor was slower than that of cytoplasmic and other membrane proteins. In isotope-dilution and chase experiments, F₁-ATPase showed one of the slowest rates of decay of the incorporated label. These results point out that F₁-ATPase of M. lysodeikticus undergoes slower turnover than the overall cytoplasmic and membrane proteins. Pulse and chase experiments allowed us to conclude that the α, β, γ subunits and the components of relative mobility 1.0 are independent with differences in their turnover and therefore do not bear any apparent relation as precursors-products. The two major subunits represent seemingly the “core” of ATPase, the β subunit behaving like the most stable component. On the other hand, the γ subunit appears to be synthesized independently from this α + β complex.     

Subunit composition, function, and spatial arrangement in the Ca2+-and Mg2+-activated adenosine triphosphatases of Escherichia coli and Salmonella typhimurium.

The Ca²⁺- and Mg²⁺-activated ATPases of Escherichia coli NRC 482 and Salmonella typhimurium LT2 were purified to homogeneity. Both enzymes consisted of five polypeptides (α-?). The molecular weights of the α, β, and ? polypeptides were 56,800, 51,800 and 13,200 for both enzymes. The molecular weights of the γ and δ polypeptides of the E. coli and S. typhimurium ATPases were 32,000 and 20,700, and 30,900 and 21,500, respectively. In both ATPases the stoichiometry of the subunits was α₃β₃γδ? as determined with the ¹⁴C-labeled enzymes. The ATPases of either organism reacted with equal effectiveness with ATPase-deficient particles of the other organism to reconstitute energy-dependent transhydrogenase activity. Treatment of the homogeneous ATPases of both organisms with TPCK-trypsin stimulated ATPase activity but resulted in destruction of coupling factor activity. Trypsin treatment completely digested the δ and ? polypeptides, and removed up to 70% of the γ polypeptide. In the presence of the bifunctional cross-linking reagent dithiobis(succinimidyl propionate) ATPase activity was lost and cross-linking of α to β polypeptides occurred. Crosslinking of α to α or β to β polypeptides was not detected. The function of the individual polypeptides of the ATPase is discussed and a model for their spatial arrangement in the enzyme is presented.     

Normal‐mode‐based modeling of allosteric couplings that underlie cyclic conformational transition in F1 ATPase

F₁ ATPase, a rotary motor comprised of a central stalk ( γ subunit) enclosed by three α and β subunits alternately arranged in a hexamer, features highly cooperative binding and hydrolysis of ATP. Despite steady progress in biophysical, biochemical, and computational studies of this fascinating motor, the structural basis for cooperative ATPase involving its three catalytic sites remains not fully understood. To illuminate this key mechanistic puzzle, we have employed a coarse‐grained elastic network model to probe the allosteric couplings underlying the cyclic conformational transition in F₁ ATPase at a residue level of detail. We will elucidate how ATP binding and product (ADP and phosphate) release at two catalytic sites are coupled with the rotation of γ subunit via various domain motions in α ₃ β ₃ hexamer (including intrasubunit hinge‐bending motions in β subunits and intersubunit rigid‐body rotations between adjacent α and β subunits). To this end, we have used a normal‐mode‐based correlation analysis to quantify the allosteric couplings of these domain motions to local motions at catalytic sites and the rotation of γ subunit. We have then identified key amino acid residues involved in the above couplings, some of which have been validated against past studies of mutated and γ ‐truncated F₁ ATPase. Our finding strongly supports a binding change mechanism where ATP binding to the empty catalytic site triggers a series of intra‐ and intersubunit domain motions leading to ATP hydrolysis and product release at the other two closed catalytic sites. Proteins 2009. © 2009 Wiley‐Liss, Inc.     

Partial purification of active delta and epsilon subunits of the membrane ATPase from escherichia coli

We have partially purified active delta and epsilon subunits of the E. coli membranebound Mg²⁺ -ATPase (ECF₁). Treating purified ECF₁ with 50% pyridine precipitates the major subunits (α, β, and γ) of the enzyme, but the two minor subunits (δ and ϵ), which are present in relatively small amounts, remain in solution. The delta and epsilon subunits were then resolved from one another by anion exchange chromatography. The partially purified epsilon strongly inhibits the hydrolytic activity of ECF₁. The epsilon fraction inhibits both the highly purified five-subunit ATPase and the enzyme deficient in the δ subunit. The latter result indicates that the delta subunit is not required for inhibition by epsilon. By contrast, two-subunit enzyme, consisting chiefly of the α and β subunits, was insensitive to the ATPase inhibitor, suggesting that the γ subunit may be required for inhibition by epsilon. The partially purified delta subunit restored the capacity of ATPase deficient in delta to recombine with ATPase-depleted membranes and to reconstitute ATP-dependent transhydrogenase. Previously we reported (Biochem. Biophys. Res. Commun. 62:764 [1975]) that a fraction containing both the delta and epsilon subunits of ECF₁ restored the capacity of ATPase missing delta to recombine with depleted membranes and to function as a coupling factor in oxidative phosphorylation and for the energized transhydrogenase. These reconstitution experiments using isolated subunits provide rather substantial evidence that the delta subunit is essential for attaching the ATPase to the membrane and that the epsilon subunit has a regulatory function as an inhibitor of the ATPase activity of ECF₁.     

Analysis of the Open and Closed Conformations of the β Subunits in Thermophilic F1-ATPase by Solution NMR

F₁-ATPase, composed of α, β, γ, δ, and ? subunits, is a unique enzyme in terms of its rotational catalytic activity. The smallest unit showing this function is the α₃β₃γ complex. We have investigated the α₃β₃γ?^ΔC (?^ΔC, truncated ?) complex from thermophilic Bacillus PS3 (TF₁′, 360 kDa) in the solution state by using the combination of extensive deuteration, segmental-labeling, and CRINEPT (cross-correlated relaxation-enhanced polarization transfer) NMR. Well-resolved CRINEPT-HMQC (heteronuclear multiple-quantum correlation) spectra of partially ¹⁵N-labeled TF₁′ were obtained for this huge and asymmetric protein complex. The spectrum of the C-terminal domain of the β subunit revealed that the open form of the β subunit in the TF₁′ complex is similar to that of the free β monomer. The open β subunit in the TF₁′ complex does not exhibit high affinity for nucleotides unlike the monomer, but this is in agreement with the results of single-molecule analysis of TF₁α₃β₃γ. On the other hand, the closed form of the β subunit in the TF₁′ complex was shown to be distinct from that of the nucleotide-bound β monomer. This is consistent with a previous report that the closed form of the TF₁β monomer could be a catalytically activated state. The loop between the N-terminal β-barrel and the central domain is highly flexible in the TF₁′ complex, in contrast to that in the α₃β₃ hexamer, suggesting that it is affected by the presence of the γ subunit in this area.     

Adenine nucleotide binding sites in normal and mutant adenosine triphosphatases of Escherichia coli

Three types of assays were used to characterize adenine nucleotide binding sites on the Ca²⁺, Mg²⁺-activated ATPase of normal Escherichia coli and its unc A 401 and unc D 412 mutants. ADP was bound mainly at a single site in normal and mutant ATPase. In the absence of divalent cations ATP was bound at a single high-affinity and three low-affinity sites in normal and unc D ATPases. The 2′,3′-dialdehyde (oADP) obtained by periodate oxidation of ADP reacted with both low- and high-affinity sites whereas oATP was bound primarily at a low-affinity site. Two types of adenine nucleotide binding sites, a high-affinity site reacting with ATP and ADP and a low-affinity site for ATP, were detected by the effects of these nucleotides on the fluorescence of the aurovertin D-ATPase complex. This high-affinity site(s) was present in normal and mutant ATPases. However, the fluorescence response at both high- and low-affinity sites was modified in the unc D ATPase as a consequence of the abnormal β subunit in this enzyme. Normal fluorescence responses were not induced by the binding of oADP or oATP to the ATPases. ATP was bound at a single site on isolated α subunits of the enzyme. Since this site was not detected in the unc A ATPase, it is unlikely to be the high-affinity site detected in the intact enzyme or the binding site for the endogenous tightly bound adenine nucleotides found in the purified ATPase. It is more probable that the site detected on the isolated α subunit from the normal enzyme is that which binds oADP since this site was absent in the unc A ATPase. Pretreatment of the normal ATPase with either N, N′-dicyclohexyl-carbodiimide (DCCD) or with 4-chloro-7-nitrobenzofurazan (NbfCl), reagents which inhibit ATPase activity by reacting with a β subunit, affected binding of oADP to α subunit(s) but had less effect with oATP. Inhibition of oADP binding could be due to conformational changes induced in the α subunit by the reaction of DCCD and NbfCl with a β subunit, or to steric reasons. If the latter hypothesis is correct, the active site of the ATPase would be at the interface between α and β subunits of the enzyme.     

Characterization of subunit structural alterations which occur during purification of nitrate reductase from Escherichia coli

Nitrate reductase from Escherichia coli, purified to homogeneity after release from membranes by deoxycholate treatment, was composed of two subunits of 155,000 (α) and 58,000 (β) daltons and contained no cytochrome b₁. Analysis of fractions at different stages of purification by gel electrophoresis and immunoprecipitation revealed that during the early steps of the purification cytochrome b₁ dissociated from the enzyme and the β subunit was altered in size as determined by sodium dodecyl sulfate-gel electrophoresis. Analysis of the peptide patterns obtained by partial proteolysis of isolated α and β subunits established that these subunits are composed of distinct sequences and ruled out a precursor-product relationship between the two subunits. The β subunit was altered during the purification by loss of a 2000-dalton fragment, apparently from its carboxyl terminus. The protease inhibitor tosyllysine chloromethylketone protected nitrate reductase from more extensive degradation by endogenous proteases during the purification but did not prevent the removal of the 2000-dalton fragment. This carboxyl terminal fragment was part of a 15,000-dalton sequence which was removed by trypsin and which was required for the self-associating character of the unmodified enzyme monomers. From the structural changes which occurred during the purification procedure, it is proposed that the carboxyl terminal segment of the β subunit is involved in the binding of nitrate reductase to cytochrome b₁ and its association with the membrane.     

Reconstitution of ATPase activity from the isolated alpha, beta, and gamma subunits of the coupling factor, F1, of Escherichia coli.

The three major subunits (α, β and γ) of the coupling factor, F₁ ATPase, of $\text{Escherichia coli}$ were separated and purified by hydrophobic column chromatography after the enzyme was dissociated by cold inactivation. The ability to hydrolyze ATP was reconstituted by dialyzing the mixture of subunits against 0.05 M Tris-succinate, pH 6.0, containing 2 mM ATP and 2 mM MgCl₂. A mixture containing α, β and γ regained ATP hydrolyzing activity. Individual subunits alone or mixtures of any two subunits did not develop ATPase activity, except for a low but significant activity with α plus β. The reconstituted ATPase had a K_m of 0.23 mM for ATP and a molecular weight by sucrose gradient density centrifugation of about 280,000.     

Identification of the altered subunit in the inactive F1ATPase of an Escherichia coli uncA mutant.

ATPase activity was restored to the inactive coupling factor, F₁ATPase, of $\text{Escherichia coli}$ strain AN120 ($\text{uncA401}$) by reconstitution of the dissociated complex with an excess of wild-type α subunit. Large excesses of α gave the highest levels of activity. The other subunits which are required for the reconstitution of ATPase activity, β and γ, did not complement the mutant enzyme. These results indicate that the α polypeptide of the AN120 ATPase is defective.     

Quantitative analysis on inactivation and reactivation of recombinant glycerol dehydratase from Klebsiella pneumoniae XJPD-Li

The genes encoding glycerol dehydratase were cloned and characterized by genomic DNA from Klebsiella pneumoniae XJPD-Li, and the assigned accession number EF634063 was available from the GenBank database. The DNA sequence analysis showed that the clone included three ORFs (dhaB, dhaC and dhaE, encoding α, β and γ subunit of glycerol dehydratase, respectively). Among three subunits of glycerol dehydratase, amino acid residues H₁₃, S₁₉₃, N₃₅₉, E₄₀₇, and M₅₁₅ of α subunit, N₄₇, L₁₅₀, V₁₈₉ of β subunit are different with what had been reported. Subsequently, the expression vector was constructed and transformed into E. coli BL21, and the colony carried genes of glycerol dehydratase were selected. SDS-PAGE examination showed that the three subunits were well expressed. The specific activity of recombined glycerol dehydratase reached to 0.299 U mg^?1, which was about 3 times comparing with that of the wild strain. The research also displayed that both glycerol and O₂ could inactive the glycerol dehydratase expressed in E. coli quickly in 10 min. The inactivated glycerol dehydratase could be effectively reactivated under the system as follows: the concentration of ATP, Mg²⁺ and coenzyme B₁₂ were 50 mM, 10 mM and 3 μM, respectively, when the ratio (W/W) of glycerol dehydratase to reactivation factor was 4:1. The O₂-inactivated and glycerol-inactivated dehydratase could be reactivated to 97.3% and 98.9% of initial activity in 10 min in above-mentioned conditions, respectively. The reactivation factor together with ATP was considered as the “ON/OFF” reactivating condition.     

Characterization of CF1 from the Diatom Odontella sinensis

The CF₁ moiety of the chloroplast ATPase of the diatom Odontella sinensis was solubilized from isolated thylakoids by chloroform extraction. Further purification was achieved by HPLC on a Superose-6 column. The resulting four-subunit complex was identified as CF₁ lacking subunit δ. The larger two subunits, α and β, showed cross-reactivity with antisera raised against the homologous subunits of spinach-CF₁. Western blot analysis further revealed that — contrary to other ATPases — migration in SDS-PAGE of α was faster than migration of β, suggesting a deletion of 40 to 50 amino acids in subunit α of Odontella. The assumed deletion does not involve the N-terminal side of the protein, as was established by protein sequencing. The N-terminal sequences of subunits α and γ showed highest homologies with the equivalent subunits of blue-green algae. According to SDS-PAGE, the apparent molecular weights of the four Odontella subunits were 53.2 (β), 51.2 (α), 39.3 (γ) and 16.2 (ε) kD. ATPase activity of isolated Odontella-CF₁ could be induced by trypsin or octylglucoside, and to a lesser extent by sulfite or by alcohols such as methanol or ethanol.     

Cloning and expression of the gene encoding catalytic subunit of thermostable glucose dehydrogenase from Burkholderia cepacia in Escherichia coli

We have cloned a 1620-nucleotide gene encoding the catalytic subunit (α subunit) of a thermostable glucose dehydrogenase (GDH) from Burkholderia cepacia. The FAD binding motif was found in the N-terminal region of the α subunit. The deduced primary structure of the α subunit showed about 48% identity to the catalytic subunits of sorbitol dehydrogenase (SDH) from Gluconobacter oxydans and 2-keto-d-gluconate dehydrogenases (2KGDH) from Erwinia herbicola and Pantoea citrea. The α subunit of B. cepacia was expressed in Escherichia coli in its active water-soluble form, showing maximum dye-mediated GDH activity at 70 °C, retaining high thermal stability. A putative open reading frame (ORF) of 507 nucleotides was also found upstream of the α subunit encoding an 18-kDa peptide, designated as γ subunit. The deduced primary structure of γ subunit showed about 30% identity to the small subunits of the SDH from G. oxydans and 2KGDHs from E. herbicola and P. citrea.     

Studies on pituitary follitropin. II. Isolation and characterization of the subunits of the ovine hormone.

The α and β subunits of highly potent ovine follitropin have been isolated by dissociation in 8 m urea, pH 7.5, and chromatography on DEAE-Sephadex A25. The isolated subunits display microheterogeneity on polyacrylamide gel electrophoresis and have very low activity in follitropin-specific radioreceptor and radioimmunoassays. The tryptophan fluorescence spectra of native follitropin and the isolated β subunit are different. The recombinant of follitropin α + β subunit had the same activity as the native hormone in the radioimmunoassay, but its activity in the radioreceptor and in vivo bioassay was about 65% of the intact hormone. Substitution of the follitropin α by ovine lutropin α subunit (prepared by a method not involving urea) to form the recombinant restored full activity in all the three assays investigated. The formation of recombined hormone proceeds at a rapid rate and is almost complete by 6 h. The α and β subunits of ovine follitropin differ from each other in amino acid composition. No significant differences were apparent in their carbohydrate composition. The amino acid composition of the ovine follitropin α and lutropin α subunits are very similar. The oxidized α subunit has phenylalanine at its NH₂-terminus while aspartic acid is present at this position in the oxidized β subunit.     

Conversion of stable ATPase to labile ATPase by acetylation, and the alpha beta and alpha gamma subunit complexes during its reconstitution

The differences between thermophilic ATPase (TF₁) and mesophilic ATPases (F₁'s) were examined by converting TF₁ to a labile F₁ by acetylation. The acetyl TF₁ was similar to mesophilic F₁'s in its low resistance to treatments with urea, high salt, heat, cold and acids. However, some TF₁ specific properties, such as aurovertin resistance, Cd²⁺ requirment and binding to TF_o were not impaired. Acetyl-TF₁ showed poor reconstitutability, mainly because its α subunit was acetylated. The reported ATPase activity of the βγ subunit complex was disproved and formation of the αβ and αγ subunit complexes was confirmed by an improved reconstitution method of the ATPase.     

α and β subunits of F1-ATPase are required for survival of petite mutants in Saccharomyces cerevisiae

Although Saccharomyces cerevisiae can form petite mutants with deletions in mitochondrial DNA (mtDNA) (ρ^?) and can survive complete loss of the organellar genome (ρ^o), the genetic factor(s) that permit(s) survival of ρ^? and ρ^o mutants remain(s) unknown. In this report we show that a function associated with the F₁-ATPase, which is distinct from its role in energy transduction, is required for the petite-positive phenotype of S. cerevisiae. Inactivation of either the α or β subunit, but not the γ, δ, or ? subunit of F₁, renders cells petite-negative. The F₁ complex, or a subcomplex composed of the α and β subunits only, is essential for survival of ρ^o cells and those impaired in electron transport. The activity of F₁ that suppresses ρ^o lethality is independent of the membrane F_o complex, but is associated with an intrinsic ATPase activity. A further demonstration of the ability of F₁ subunits to suppress ρ^o lethality has been achieved by simultaneous expression of S. cerevisiae F₁α and γ subunit genes in Kluyveromyces lactis– which allows this petite-negative yeast to survive the loss of its mtDNA. Consequently, ATP1 and ATP2, in addition to the previously identified AAC2, YME1 and PEL1/PGS1 genes, are required for establishment of ρ^? or ρ^o mutations in S. cerevisiae.