Directed mutagenesis of the beta-subunit of F1-ATPase from Escherichia coli

Oligonucleotide-directed mutagenesis was used to generate six mutant strains of Escherichia coli which had the following specific amino acid substitutions in the beta-subunit of F1-ATPase: (i) Lys-155----Gln; (ii) Lys-155----Glu; (iii) Gly-149----Ile; (iv) Gly-154----Ile; (v) Tyr-297----Phe;(vi) Tyr-354----Phe. The effects of each mutation on growth of cells on succinate plates or limiting (3 mM) glucose and on cell membrane ATPase activity and ATP-driven pH gradient formation were studied. The results showed Lys-155 to be essential for catalysis, as has been predicted previously from sequence homology and structural considerations; however, the results appear to contradict the hypothesis that Lys-155 interacts with one of the substrate phosphate groups because the Lys-155----Glu mutation was less detrimental than Lys-155----Gln. Gly-149 and Gly-154 have been predicted to be involved in essential conformational changes in F1-ATPase by virtue of their position in a putative glycine-rich flexible loop structure. The mutation of Gly-154----Ile caused strong impairment of catalysis, but the Gly-149----Ile mutation produced only moderate impairment. The two tyrosine residues chosen for mutation were residues which have previously received much attention due to their being the sites of reaction of the inactivating chemical modification reagents 4-chloro-7-nitrobenzofurazan (Tyr-297) and p-fluorosulfonylbenzoyl-5'-adenosine (Tyr-354). We found that mutation of Tyr-297----Phe caused only minor impairment of catalysis, and mutation of Tyr-354----Phe produced no impairment. Therefore, a direct role for either of these tyrosine residues in catalysis is unlikely.     

F0F1-ATPase gamma subunit mutations perturb the coupling between catalysis and transport.

We introduced mutations to test the function of the conserved amino-terminal region of the gamma subunit from the Escherichia coli ATP synthase (F0F1-ATPase). Plasmid-borne mutant genes were expressed in an uncG strain which is deficient for the gamma subunit (gamma Gln-14-->end). Most of the changes, which were between gamma Ile-19 and gamma Lys-33, gamma Asp-83 and gamma Cys-87, or at gamma Asp-165, had little effect on growth by oxidative phosphorylation, membrane ATPase activity, or H+ pumping. Notable exceptions were gamma Met-23-->Arg or Lys mutations. Strains carrying these mutations grew only very slowly by oxidative phosphorylation. Membranes prepared from the strains had substantial levels of ATPase activity, 100% compared with wild type for gamma Arg-23 and 65% for gamma Lys-23, but formed only 32 and 17%, respectively, of the electrochemical gradient of protons. In contrast, other mutant enzymes with similar ATPase activities (including gamma Met-23-->Asp or Glu) formed H+ gradients like the wild type. Membranes from the gamma Arg-23 and gamma Lys-23 mutants were not passively leaky to protons and had functional F0 sectors. These results suggested that substitution by positively charged side chains at position 23 perturbed the energy coupling. The catalytic sites of the mutant enzymes were still regulated by the electrochemical H+ gradient but were inefficiently coupled to H+ translocation in both ATP-dependent H+ pumping and delta mu H+ driven ATP synthesis.     

Analysis of structure and function of the B subunit of cholera toxin by the use of site-directed mutagenesis

Oligonucleotide-directed mutagenesis of ctxB was used to produce mutants of cholera toxin B subunit (CT-B) altered at residues Cys-9, Gly-33, Lys-34, Arg-35, Cys-86 and Trp-88. Mutants were identified phenotypically by radial passive immune haemolysis assays and genotypically by colony hybridization with specific oligonucleotide probes. Mutant CT-B polypeptides were characterized for immunoreactivity, binding to ganglioside GM1, ability to associate with the A subunit, ability to form holotoxin, and biological activity. Amino acid substitutions that caused decreased binding of mutant CT-B to ganglioside GM1 and abolished toxicity included negatively charged or large hydrophobic residues for Gly-33 and negatively or positively charged residues for Trp-88. Substitution of lysine or arginine for Gly-33 did not affect immunoreactivity or GM1-binding activity of CT-B but abolished or reduced toxicity of the mutant holotoxins, respectively. Substitutions of Glu or Asp for Arg-35 interfered with formation of holotoxin, but none of the observed substitutions for Lys-34 or Arg-35 affected binding of CT-B to GM1. The Cys-9, Cys-86 and Trp-88 residues were important for establishing or maintaining the native conformation of CT-B or protecting the CT-B polypeptide from rapid degradation in vivo.     

The proton pore in the Escherichia coli F0F1-ATPase: substitution of glutamate by glutamine at position 219 of the alpha-subunit prevents F0-mediated proton permeability

Three mutations in the uncB gene encoding the a-subunit of the F0 portion of the F0F1-ATPase of Escherichia coli were produced by site-directed mutagenesis. These mutations directed the substitution of Glu-219 by Gln, or of Lys-203 by Ile, or of Glu-196 by Ala. Strains carrying either the Lys-203 or Glu-196 substitutions showed growth characteristics indistinguishable from the coupled control strain. Properties of membrane preparations from these strains were also similar to those from the coupled control strain. The substitution of Glu-219 by Gln resulted in a strain which was unable to utilise succinate as sole carbon source and had a growth-yield characteristic of an uncoupled strain. Membrane preparations of the Glu-219 mutant were proton impermeable and the F1-ATPase activity was inhibited by about 50% when membrane-bound. The results are discussed with reference to a previously proposed intramembranous proton pore involving subunits a and c.     

Assembly of the adenosine triphosphatase complex in Escherichia coli: assembly of F0 is dependent on the formation of specific F1 subunits.

A strain of Escherichia coli (AN1007) carrying the polar uncD436 allele which affects the operon coding for the F1-F0 adenosine triphosphatase (ATPase) complex was isolated and characterized. The uncD436 allele affected the two genes most distal to the operon promoter, i.e., uncD and uncC. Although the genes coding for the F0 portion of the ATPase complex were not affected in strains carrying this mutant allele, the lack of reconstitution of washed membranes by normal F1 ATPase suggested that a functional F0 might not be formed. This conclusion was supported by the observation that the 18,000-molecular-weight F0 subunit, coded for by the uncF gene, was absent from the membranes. Plasmid pAN36 (uncD+C+), when inserted into a strain carrying the uncD436 allele, resulted in the incorporation of the 18,000-molecular-weight F0 subunit into the membrane. A further series of experiments with Mu-induced polarity mutants, with and without plasmid pAN36, showed that the formation of both the alpha- and beta-subunits of F1 ATPase was an essential prerequisite to the incorporation into the membrane of the 18,000-molecular-weight F0 subunit and to the formation of a functional F0. Examination of the polypeptide composition of membranes from various unc mutants allowed a sequence for the normal assembly of the F1-F0 ATPase complex to be proposed.     

Site-directed mutageneses of rat liver cytochrome P-450d: axial ligand and heme incorporation

By oligonucleotide-directed mutageneses, 13 substitutions of amino acids at the carboxy-terminal region of rat liver cytochrome P-450d were done as follows: (A) Phe-449----Tyr; (B) Gly-450----Ser; (C) Leu-451----Ser; (D) Gly-452----Glu; (E) Lys-453----Glu; (F) Arg-454----Leu; (G) Arg-455----Gly; (H) Cys-456----Tyr; (I) Cys-456----His; (J) Ile-457----Ser; (K) Gly-458----Glu; (L) Glu-459----Ala; (M) Ile-460----Ser. The CO-bound reduced forms of the wild type and mutants B-G, J, L, and M gave Soret peaks at 448 nm. The CO complex of mutant A gave a Soret peak at 445 nm. The intensities of the CO-bound forms of mutants A, C, D, and J were very small compared with that of the wild-type complex. The CO-reduced forms of mutants H, I, and K did not give a Soret peak around 450 nm at all. The 448-nm peak of mutant F was unstable and quickly disappeared with the concomitant appearance of a peak at 420 nm.(ABSTRACT TRUNCATED AT 250 WORDS)     

Analysis of fibrinogen A alpha-fusion proteins. Mutants which inhibit thrombin equivalently are not equally good substrates

We have examined the interaction of thrombin with fibrinogen A alpha chain residues 7-16. Using genetically engineered constructions, we have synthesized in Escherichia coli a fibrinogen A alpha 1-50 fusion protein and seven mutant proteins with single amino acid substitutions. These are: Asp7----Ala, Phe8----Tyr, Glu11----Ala, Gly12----Val, Gly13----Val, Gly14----Val, and Arg16----Leu. Competitive immunoassay of cell lysates showed that all the mutations but one, Arg16----Leu, altered the structure of the protein such that cross-reactivity with the A alpha-specific monoclonal antibody, Y18, was significantly reduced. The fusion proteins were purified and analyzed as thrombin inhibitors and substrates. All the fusion proteins are competitive inhibitors of the amidolytic hydrolysis of Spectrozyme TH, a thrombin-specific chromogenic substrate, with inhibition constants corresponding to that for fibrinogen. We conclude that these 7 amino acid substitutions do not alter thrombin binding to the fusion proteins. The fusion proteins were tested as substrates by monitoring thrombin-dependent peptide release. The natural sequence and three mutants, Asp7----Ala, Glu11----Ala, and Gly14----Val, are good substrates. The other mutants are either poor substrates or are not cleaved by thrombin within A alpha 1-50. These results indicate that residues between Asp7 and Arg16 are critical to efficient peptide hydrolysis, whereas residues outside this region are critical to thrombin binding.     

Amino acid substitutions in the epsilon-subunit of the F1F0-ATPase of Escherichia coli

A mutant strain of Escherichia coli was isolated in which Gly-48 of the mature epsilon-subunit of the energy-transducing adenosine triphosphatase was replaced by Asp. This amino acid substitution caused inhibition of ATPase activity (about 70%), loss of ATP-dependent proton translocation and lowered oxidative phosphorylation, but did not affect proton translocation through the F0. Purified F1-ATPase from the mutant strain bound to stripped membranes with the same affinity as the normal F1-ATPase. Partial revertant strains were isolated in which Pro-47 of the epsilon-subunit was replaced by Ser or Thr. Pro-47 and Gly-48 are predicted to be residues 2 and 3 in a Type II beta-turn and the Gly-48 to Asp substitution is predicted to cause a change from a Type II to a Type I or III beta-turn. Space-filling models of the beta-turn (residues 46-49) in the normal, mutant and partial revertant epsilon-subunits indicate that the peptide oxygen between Pro-47 and Gly-48 is in a different position to the peptide oxygen between Pro-47 and Asp-48 and that the substitution of Pro-47 by either Ser or Thr restores an oxygen close to the original position. It is suggested that the peptide oxygen between Pro-47 and Gly-48 of the epsilon-subunit is involved either structurally in inter-subunit H-bonding or directly in proton movements through the F1-ATPase.     

Essential residues in the polar loop region of subunit c of Escherichia coli F1F0 ATP synthase defined by random oligonucleotide-primed mutagenesis.

The conserved, polar loop region of subunit c of the Escherichia coli F1F0 ATP synthase is postulated to function in the coupling of proton translocation through F0 to ATP synthesis in F1. We have used a random mutagenesis procedure to define the essential residues in the region. Oligonucleotide-directed mutagenesis was carried out with a random mixture of mutant oligonucleotides, the oligonucleotide mixture being generated by chemical synthesis by using phosphoramidite nucleotide stocks that were contaminated with the other three nucleotides. Thirty mutant genes coding single-amino-acid substitutions in the region between Glu-37 and Leu-45 of subunit c were tested for function by analyzing the capacity of plasmids carrying the mutant genes to complement a Leu-4----amber subunit c mutant. All substitutions at the conserved Arg-41 residue resulted in loss of oxidative phosphorylation, i.e., transformants could not grow on a succinate carbon source. The other conserved residues were more tolerant to substitution, although most substitutions did result in impaired growth on succinate. We conclude that Arg-41 is essential in the function of the polar loop and that the ensemble of other conserved residues collectively maintain an optimal environment required for that function.     

The proton pore in the Escherichia coli F0F1-ATPase: a requirement for arginine at position 210 of the a-subunit

Site-directed mutagenesis was used to generate three mutations in the uncB gene encoding the a-subunit of the F0 portion of the F0F1-ATPase of Escherichia coli. These mutations directed the substitution of Arg-210 by Gln, or of His-245 by Leu, or of both Lys-167 and Lys-169 by Gln. The mutations were incorporated into plasmids carrying all the structural genes encoding the F0F1-ATPase complex and these plasmids were used to transform strain AN727 (uncB402). Strains carrying either the Arg-210 or His-245 substitutions were unable to grow on succinate as sole carbon source and had uncoupled growth yields. The substitution of Lys-167 and Lys-169 by Gln resulted in a strain with growth characteristics indistinguishable from a normal strain. The properties of the membranes from the Arg-210 or His-245 mutants were essentially identical, both being proton impermeable and both having ATPase activities resistant to the inhibitor DCCD. Furthermore, in both mutants, the F1-ATPase activities were inhibited by about 50% when bound to the membranes. The membrane activities of the mutant with the double lysine change were the same as for a normal strain. The results are discussed in relation to a previously proposed model for the F0 (Cox, G.B., Fimmel, A.L., Gibson, F. and Hatch, L. (1986) Biochim. Biophys. Acta 849, 62-69).     

Genetic evidence for interaction between the a and b subunits of the F0 portion of the Escherichia coli proton translocating ATPase

A mutation of the b subunit of the Escherichia coli proton translocating ATPase was previously described (Porter, A. C. G., Kumamoto, C., Aldape, K., and Simoni, R. D. (1985) J. Biol. Chem. 260, 8182-8187). This mutation, which causes substitution of aspartic acid for glycine at position 9 (basp9), results in loss of function of the ATPase complex. In this paper we describe the isolation and characterization of two mutations that partially suppress the effects of the basp9 alteration. The suppressor mutations cause amino acid substitutions at position 240 of the a subunit. Membranes derived from strains carrying a suppressor mutation and the basp9 mutation exhibited ATP-dependent proton translocating activity.     

Role of the delta subunit in enhancing proton conduction through the F0 of the Escherichia coli F1F0 ATPase.

We studied the effect of the delta subunit of the Escherichia coli F1 ATPase on the proton permeability of the F0 proton channel synthesized and assembled in vivo. Membranes isolated from an unc deletion strain carrying a plasmid containing the genes for the F0 subunits and the delta subunit were significantly more permeable to protons than membranes isolated from the same strain carrying a plasmid containing the genes for the F0 subunits alone. This increased proton permeability could be blocked by treatment with either dicyclohexyl-carbodiimide or purified F1, both of which block proton conduction through the F0. After reconstitution with purified F1 in vitro, both membrane preparations could couple proton pumping to ATP hydrolysis. These results demonstrate that an interaction between the delta subunit and the F0 during synthesis and assembly produces a significant change in the proton permeability of the F0 proton channel.     

Mutational replacements of conserved amino acid residues in the beta subunit resulted in defective assembly of H+-translocating ATPase (F0F1) in Escherichia coli

Mutant genes for the beta subunit of H+-translocating ATPase (F0F1) were cloned from Escherichia coli strains isolated in this laboratory. Determination of their nucleotide sequence revealed four missense mutations (strain KF39, Glu-41----Lys; strain KF16 and KF42, Glu-185----Lys; strain KF48, Gly-223----Asp; KF26 and 4 other strains, Ser-292----Phe). Two nonsense mutants (strain KF40, Gln-361----end; strain KF20, Gln-397----end) were also identified. Glu-41, Glu-185, and Ser-292 are conserved in the amino acid sequences of the beta subunits so far studied, and Gly-223, Gln-361, and Gln-397 are conserved in beta subunits from bacteria and mitochondria, but not in those from chloroplasts. The amounts of F1 subunits in the membranes of these strains were studied by immunochemical assay and two-dimensional gel electrophoresis. In the mutants studied, the amounts of alpha and beta subunits in the membranes were 69-21 and 46-2%, respectively, of the amounts in wild-type membranes, the amount depending on the strain. No delta and epsilon subunits were detected in membranes of a missense mutant KF16, although reduced amounts of alpha and beta subunits could be detected, suggesting that the F1 portion may not be connected to F0 through the delta and epsilon subunits. The altered residues in missense mutants or missing domains in nonsense mutants may be important for the subunit-subunit interactions or assembly of the entire complex. Genetic experiments on introduction of suppressor tRNA into strains KF40 and KF20 suggested that F1 could be active even when residue 361 or 397 was replaced by a Ser, Leu, or Tyr residue.     

Effect of the delta subunit on assembly and proton permeability of the F0 proton channel of Escherichia coli F1F0 ATPase.

During the assembly of the Escherichia coli proton-translocating ATPase, the subunits of F1 interact with F0 to increase the proton permeability of the transmembrane proton channel. We tested the involvement of the delta subunit in this process by partially and completely deleting uncH (delta subunit) from a plasmid carrying the genes for the F0 subunits and delta and testing the effects of those F0 plasmids on the growth of unc+ and unc mutant E. coli strains. We found that the delta subunit was required for inhibition of growth of unc+ cells. We also tested membranes isolated from unc-deleted cells containing F0 plasmids for F1-binding ability. In unc-deleted cells, these plasmids produced F0 in amounts comparable to those found in normal unc+ E. coli cells, while having only small effects on cell growth. These studies demonstrate that the delta subunit plays an important role in opening the F0 proton channel but that it does not serve as a temporary plug of F0 during assembly, as had been previously speculated (S. Pati and W. S. A. Brusilow, J. Biol. Chem. 264:2640-2644, 1989).     

A mutation in the alpha-subunit of F1-ATPase from Escherichia coli affects the binding of F1 to the membrane

The mutation Gly-29----Asp in the alpha-subunit of the F1-ATPase from Escherichia coli was characterized and shown to cause the following effects. 1) Oxidative phosphorylation was markedly impaired in vivo 2) Membrane ATPase and ATP-driven proton-pumping activities were decreased markedly. 3) Membranes were proton-permeable, and membrane-bound ATPase was dicyclohexylcarbodiimide-insensitive. Therefore, it appeared that integration between F1 and F0 was abnormal. This was confirmed directly by the demonstration that the mutant F1 bound poorly to stripped membranes from a normal strain. Purified, soluble mutant F1 had normal ATPase activity. These results suggest that residue Gly-29, which is strongly conserved in alpha-subunits of F1-ATPases, lies in a region of the alpha-subunit important for membrane binding. Thus, three regions of the F1-alpha-subunit have now been recognized, specialized for membrane binding, nucleotide binding, and alpha/beta intersubunit signal transmission, respectively. The approximate locations of the three regions are described.     

Three-dimensional structures of H-ras p21 mutants: molecular basis for their inability to function as signal switch molecules

The X-ray structures of the guanine nucleotide binding domains (amino acids 1-166) of five mutants of the H-ras oncogene product p21 were determined. The mutations described are Gly-12----Arg, Gly-12----Val, Gln-61----His, Gln-61----Leu, which are all oncogenic, and the effector region mutant Asp-38----Glu. The resolutions of the crystal structures range from 2.0 to 2.6 A. Cellular and mutant p21 proteins are almost identical, and the only significant differences are seen in loop L4 and in the vicinity of the gamma-phosphate. For the Gly-12 mutants the larger side chains interfere with GTP binding and/or hydrolysis. Gln-61 in cellular p21 adopts a conformation where it is able to catalyze GTP hydrolysis. This conformation has not been found for the mutants of Gln-61. Furthermore, Leu-61 cannot activate the nucleophilic water because of the chemical nature of its side chain. The D38E mutation preserves its ability to bind GAP.     

Suppression mutations in the defective beta subunit of F1-ATPase from Escherichia coli

The Escherichia coli mutant of the proton-translocating ATPase KF11 (Kanazawa, H., Horiuchi, Y., Takagi, M., Ishino, Y., and Futai, M. (1980) J. Biochem. (Tokyo) 88, 695-703) has a defective beta subunit with serine being replaced by phenylalanine at codon 174. Four suppression mutants (RE10, RE17, RE18, and RE20) from this strain capable of growth on minimal plate agar supplemented by succinate were isolated. The original point mutation at codon 174 was intact in these strains. Additional point mutations, Ala-295 to Thr, Gly-149 to Ser, Leu-400 to Gln, Ala-295 to Pro, for RE10, RE17, RE18, and RE20, respectively, were identified by the polymerase chain reaction and sequencing. These mutations, except for RE10, were confirmed as a single mutation conferring a suppressive phenotype by genetic suppression assay using KF11 as the host cells. The results indicated that Ser-174 has functional interaction with Gly-149, Ala-295, and Leu-400. The residues are located within the previously estimated catalytic domain of the beta subunit, indicating that this domain is indeed folded for the active site of catalytic function. Growth rates of the revertants in the minimal medium with succinate increased compared with that of KF11, showing that ATP synthesis recovered to some extent. The ATP hydrolytic activity in the revertant membranes increased in RE17 and RE20 but did not in RE10 and RE18, suggesting that synthesis and hydrolysis are not necessarily reversible in the proton-translocating ATPase (F1F0).     

Mutations in the conserved proline 43 residue of the uncE protein (subunit c) of Escherichia coli F1F0-ATPase alter the coupling of F1 to F0

The conserved Pro43 residue of the uncE protein (subunit c) of the Escherichia coli F1F0-ATPase was changed to Ser or Ala by oligonucleotide-directed mutagenesis, and the mutations were incorporated into the chromosome. The resultant mutant strains were capable of oxidative phosphorylation as indicated by their ability to grow on succinate and had growth yields on glucose that were 80-90% of wild type. Membrane vesicles from the mutants were slightly less efficient than wild type vesicles in ATP-driven proton pumping as indicated by ATP-dependent quenching of quinacrine fluorescence. The decreased quenching response was not due to increased H+ leakiness of the mutant membranes or to loss of F1-ATPase activity from the membrane. These results indicate that the mutant F1F0-ATPases are defective in coupling ATP hydrolysis to H+ translocation. The membrane ATPase activity of the mutants was inhibited less by dicyclohexylcarbodiimide than that of wild type. The decrease in sensitivity to inhibition by dicyclohexylcarbodiimide was caused primarily by dissociation of the F1-ATPase from the mutant F0 in the ATPase assay mixture. These results support the idea that Pro43, and neighboring conserved polar residues play an important role in the binding and functional coupling of F1 to F0. Although a Pro residue is found at position 43 in all species of subunit c studied, surprisingly, it is not absolutely essential to function.     

Mutations at Glu-32 and His-39 in the epsilon subunit of the Escherichia coli F1F0 ATP synthase affect its inhibitory properties.

A collection of amino acid substitutions at residues Glu-32 and His-39 in the epsilon subunit of the Escherichia coli F1F0 ATP synthase has been constructed by cassette mutagenesis. Substitutions for residue Glu-32 appeared to cause abnormal inhibition of membrane-bound F1 ATPase activity, and replacement of His-39 by Arg, Val, and Pro affected F1F0 interactions.     

The distribution of heavy-chain isoforms of myosin in airways smooth muscle from adult and neonate humans.

ATP synthesis by oxidative phosphorylation in Escherichia coli occurs in catalytic sites on the beta-subunits of F1-ATPase. Random mutagenesis of the beta-subunit combined with phenotypic screening is potentially important for studies of the catalytic mechanism. However, when applied to haploid strains, this approach is hampered by a preponderance of mutants in which assembly of F1-ATPase in vivo is defective, precluding enzyme purification. Here we mutagenized plasmids carrying the uncD (beta-subunit) gene with hydroxylamine or N-methyl-N'-nitro-N-nitrosoguanidine and isolated, by phenotypic screening and complementation tests, six plasmids carrying mutant uncD alleles. When the mutant plasmids were used to transform a suitable uncD- strain, assembly of F1-ATPase in vivo occurred in each case. Moreover, in one case (beta Gly-223----Asp) F1-ATPase assembly proceeded although it had previously been reported that this mutation, when present on the chromosome of a haploid strain, prevented assembly of the enzyme in vivo. Therefore, this work demonstrates an improved approach for random mutagenesis of the F1-beta-subunit. Six new mutant uncD alleles were identified: beta Cys-137----Tyr; beta Gly-142----Asp; beta Gly-146----Ser; beta Gly-207----Asp; beta-Gly-223----Asp; and a double mutant beta Pro-403----Ser,Gly-415----Asp which we could not separate. The first five of these lie within or very close to the predicted catalytic nucleotide-binding domain of the beta-subunit. The double mutant lies outside this domain; we speculate that the region around residues beta 403-415 is part of an alpha-beta intersubunit contact surface. Membrane ATPase and ATP-driven proton pumping activities were impaired by all six mutations. Purified F1-ATPase was obtained from each mutant and shown to have impaired specific ATPase activity.