F-ATPase: Forced Full Rotation of the Rotor Despite Covalent Cross-link with the Stator

Karin Gumbiowski, Martin Müller, Oliver Pänke, Parichart Promto, Stephanie Winkler, Wolfgang Junge, Siegfried Engelbrecht

Journal of Biological Chemistry · 2001 · 23 citations · 34 references

DOIFull text

Open access

Concepts

Abstract

In ATP synthase (FOF1-ATPase) ion flow through the membrane-intrinsic portion, FO, drives the central “rotor”, subunits c10εγ, relative to the “stator” ab2δ(αβ)3. This converts ADP and Pi into ATP. Vice versa, ATP hydrolysis drives the rotation backwards. Covalent cross-links between rotor and stator subunits have been shown to inhibit these activities. Aiming at the rotary compliance of subunit γ we introduced disulfide bridges between γ (rotor) and α or β (stator). We engineered cysteine residues into positions located roughly at the “top,” “center,” and “bottom” parts of the coiled-coil portion of γ and suitable residues on α or β. This part of γ is located at the center of the (αβ)3 domain with its C-terminal part at the top of F1 and the bottom part close to the FO complex. Disulfide bridge formation under oxidizing conditions was quantitative as shown by SDS-polyacrylamide gel electrophoresis and immunoblotting. As expected both the ATPase activities and the yield of rotating subunits γ dropped to zero when the cross-link was formed at the center (γL262C ↔ αA334C) and bottom (γCys87 ↔ βD380C) positions. But much to our surprise disulfide bridging impaired neither ATP hydrolysis activity nor the full rotation of γ and the enzyme-generated torque of oxidized F1, which had been engineered at the top position (γA285C ↔ αP280C). Apparently the high torque of this rotary engine uncoiled the α-helix and forced amino acids at the C-terminal portion of γ into full rotation around their dihedral (Ramachandran) angles. This conclusion was supported by molecular dynamics simulations: If γCys285-Val286 are attached covalently to (αβ)3 and γAla1-Ser281 is forced to rotate, γGly282-Ala284 can serve as cardan shaft. In ATP synthase (FOF1-ATPase) ion flow through the membrane-intrinsic portion, FO, drives the central “rotor”, subunits c10εγ, relative to the “stator” ab2δ(αβ)3. This converts ADP and Pi into ATP. Vice versa, ATP hydrolysis drives the rotation backwards. Covalent cross-links between rotor and stator subunits have been shown to inhibit these activities. Aiming at the rotary compliance of subunit γ we introduced disulfide bridges between γ (rotor) and α or β (stator). We engineered cysteine residues into positions located roughly at the “top,” “center,” and “bottom” parts of the coiled-coil portion of γ and suitable residues on α or β. This part of γ is located at the center of the (αβ)3 domain with its C-terminal part at the top of F1 and the bottom part close to the FO complex. Disulfide bridge formation under oxidizing conditions was quantitative as shown by SDS-polyacrylamide gel electrophoresis and immunoblotting. As expected both the ATPase activities and the yield of rotating subunits γ dropped to zero when the cross-link was formed at the center (γL262C ↔ αA334C) and bottom (γCys87 ↔ βD380C) positions. But much to our surprise disulfide bridging impaired neither ATP hydrolysis activity nor the full rotation of γ and the enzyme-generated torque of oxidized F1, which had been engineered at the top position (γA285C ↔ αP280C). Apparently the high torque of this rotary engine uncoiled the α-helix and forced amino acids at the C-terminal portion of γ into full rotation around their dihedral (Ramachandran) angles. This conclusion was supported by molecular dynamics simulations: If γCys285-Val286 are attached covalently to (αβ)3 and γAla1-Ser281 is forced to rotate, γGly282-Ala284 can serve as cardan shaft. Ni2+-nitrilotriacetic acid F1-ATPase from E. coli dithiothreitol 5,5′-dithiobis-(2-nitrobenzoic acid) 4-morpholinepropanesulfonic acid room temperature ATP synthases of bacteria, chloroplasts, and mitochondria utilize ion-motive force for the synthesis of ATP from ADP and phosphate (1Boyer P.D. Annu. Rev. Biochem. 1997; 66: 717-749Crossref PubMed Scopus (1595) Google Scholar, 2Junge W. Lill H. Engelbrecht S. Trends Biochem. Sci. 1997; 22: 420-423Abstract Full Text PDF PubMed Scopus (439) Google Scholar, 3Fillingame R.H. Nat. Struct. Biol. 2000; 7: 1002-1004Crossref PubMed Scopus (16) Google Scholar). When operating in reverse (F-ATPase) the enzyme hydrolyzes ATP and generates ion-motive force. The ATP synthase, in its simplest bacterial form, consists of eight different subunits, five in the F1 portion, (αβ)3γδε, and three in the Fo portion,ab2 c10 (4Jiang W. Hermolin J. Fillingame R.H. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 4966-4971Crossref PubMed Scopus (216) Google Scholar). The former subunits catalyze substrate conversion and the latter are responsible for ion translocation. ATP hydrolysis by both isolated F1(5Duncan T.M. Bulygin V.V. Zhou Y. Hutcheon M.L. Cross R.L. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 10964-10968Crossref PubMed Scopus (459) Google Scholar, 6Sabbert D. Engelbrecht S. Junge W. Nature. 1996; 381: 623-626Crossref PubMed Scopus (464) Google Scholar, 7Noji H. Yasuda R. Yoshida M. Kinosita K. Nature. 1997; 386: 299-302Crossref PubMed Scopus (1966) Google Scholar, 8Katoyamada Y. Noji H. Yasuda R. Kinosita K. Yoshida M. J. Biol. Chem. 1998; 273: 19375-19377Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar, 9Omote H. Sambonmatsu N. Saito K. Sambongi Y. Iwamoto-Kihara A. Yanagida T. Wada Y. Futai M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 7780-7784Crossref PubMed Scopus (125) Google Scholar, 10Noji H. Häsler K. Junge W. Kinosita Jr., K. Yoshida M. Engelbrecht S. Biochem. Biophys. Res. Commun. 1999; 260: 597-599Crossref PubMed Scopus (83) Google Scholar, 11Hisabori T. Kondoh A. Yoshida M. FEBS Lett. 1999; 463: 35-38Crossref PubMed Scopus (62) Google Scholar) and membrane-bound F1 (12Zhou Y.T. Duncan T.M. Bulygin V.V. Hutcheon M.L. Cross R.L. Biochim. Biophys. Acta. 1996; 1275: 96-100Crossref PubMed Scopus (69) Google Scholar) drives the rotation of γ(ε) relative to the (αβ)3 barrel. The counterpart of these rotor elements in F1 is the c-ring in Fo (13Sambongi Y. Iko Y. Tanabe M. Omote H. Iwamoto-Kihara A. Ueda I. Yanagida T. Wada Y. Futai M. Science. 1999; 286: 1722-1724Crossref PubMed Scopus (420) Google Scholar, 14Pänke O. Gumbiowski K. Junge W. Engelbrecht S. FEBS Lett. 2000; 472: 34-38Crossref PubMed Scopus (178) Google Scholar, 15Tsunoda S.P. Aggeler R. Yoshida M. Capaldi R.A. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 898-902Crossref PubMed Scopus (138) Google Scholar). Subunitsa–b2–δ-(αβ)3therefore form the stator, and subunitsc10–γ-ε form the rotor (15Tsunoda S.P. Aggeler R. Yoshida M. Capaldi R.A. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 898-902Crossref PubMed Scopus (138) Google Scholar, 16Engelbrecht S. Junge W. FEBS Lett. 1997; 414: 485-491Crossref PubMed Scopus (118) Google Scholar). Mechanistically, the tight coupling between subunits c10and γε poses the question how the symmetry “mismatch” between the presumed stepwise rotation of the proteolipid oligomer (36° per step) and γε (120° per step) is overcome. An elastic power transmission between FO and F1 has been shown (17Cherepanov D.A. Junge W. Biophys. J. 2001; 81: 1234-1244Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar, 18Pänke O. Cherepanov D.A. Gumbiowski K. Engelbrecht S. Junge W. Biophys. J. 2001; 81: 1220-1233Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar). Aiming at the internal elasticity of subunit γ, we introduced disulfide bridges at three different positions along the region where γ passes through the (αβ)3 cylinder (“top,” “center,” and “bottom,” cf. Fig. 1). By introducing a cross-link between the penultimate residue of γ and a suitable partner on α via engineered Cys residues in the region where γ snugly fits into the “hydrophobic bearing” of F1 (γA285C ↔ αP280C) (19Abrahams J.P. Leslie A.G.W. Lutter R. Walker J.E. Nature. 1994; 370: 621-628Crossref PubMed Scopus (2753) Google Scholar), we hoped to torsionally strain subunit γ upon ATP hydrolysis. Although the oxidized F1 carrying similarly engineered cross-links at the center (γL262C ↔ αA334C) or bottom (γCys87 ↔ βD380C) positions were hydrolytically inactive (as expected), quantitative disulfide bridge formation at the top position surprisingly did not impede the enzyme function at all, as apparent from both ATP turnover and the rotation assay with a fluorescent actin filament attached to γ (7Noji H. Yasuda R. Yoshida M. Kinosita K. Nature. 1997; 386: 299-302Crossref PubMed Scopus (1966) Google Scholar). All restriction enzymes were purchased from New England Biolabs (Frankfurt/Main, Germany) or MBI Fermentas (St. Leon-Rot, Germany). Oligonucleotide primers were synthesized by MWG Biotech (Ebersberg, Germany). Ni2+-nitrilotriacetic acid (Ni-NTA)1 horseradish peroxidase and Ni-NTA superflow were from Qiagen (Hilden, Germany). Biotin-PEAC5-maleimide was from Dojindo (via Gerbu Biotechnik, Gaiberg, Germany). Lumi-light Western blotting kit was obtained from Roche Molecular Biochemicals (Mannheim, Germany). Other reagents were of the highest grade commercially available. The plasmid pKH7 (all wild-type cysteines substituted by alanines (21Kuo P.H. Ketchum C.J. Nakamoto R.K. FEBS Lett. 1998; 426: 217-220Crossref PubMed Scopus (48) Google Scholar), His6 tag N-terminal at subunit β, γK109C (10Noji H. Häsler K. Junge W. Kinosita Jr., K. Yoshida M. Engelbrecht S. Biochem. Biophys. Res. Commun. 1999; 260: 597-599Crossref PubMed Scopus (83) Google Scholar)) was used as starting plasmid. For the αP280C, αA334C, γL262C, γA87C (because pKH7 had all endogenous Cys residues substituted for, the wild-type residue γCys87 had to be re-introduced), and βD380C mutations, pBluescript II SK (+/−) subclones were generated by insertion of the KpnI/XhoI, KpnI/SacI, and SacI/XbaI fragments of pKH7. Site-directed mutagenesis was carried out by polymerase chain reaction using the oligonucleotide 5′-GCTGCTCCGTCGTTGTCCAGGACGTGAAG-3′ and its complement 5′-TTATCGAAACTCAGTGTGGTGACGGTCGGTCG-3′ for αP280C, 5′-GAAACGTCACCACACTGAGTTTCGATAATCGG-3′ for αA334C, 5′-TTAAAGAGCTGCAGTGTGTATACAACAAAGATAG-3′ and 5′-CTTTGTATA-CACACTGCAGCTCTTTAATCAG-3′ for γL262C, 5′-CGACCGACCGTGGTTTGTGTGGTGGTTTGAAC-3′ and its complement for γA87C, and 5′-CCATCCTGGGTATGTGTGAACTGTCTGAAG-3′ and its complement for βD380C (22Weiner M.P. Costa G.L. Schoettlin W. Cline J. Mathur E. Bauer J.C. Gene (Amst.). 1994; 151: 119-123Crossref PubMed Scopus (415) Google Scholar). The KpnI/SacI and KpnI/XhoI fragments of pKH7 were substituted with the corresponding fragments carrying the γL262C or αP280C mutations by restriction and ligation (resulting in plasmids pBP4 and pMM6, respectively). The αA334C/γL262C double mutant was then generated by exchanging the KpnI/XhoI fragment of pBP4 with the corresponding mutated fragment. The γA285C mutant (pMM9) was generated by standard polymerase chain reaction with pKH7 as template and the oligonucleotide primers 5′-GGTACCAATGAGCTCCATCATGTTTA-3′, 5′-CTTCACGTCCTGGACAACGACG-GAGCAAGC-3′, and 5′-CTCGGGGGCCGCCTGTGTTTAAACAGG-3′ and its complement. The KpnI/SacI fragment of pMM6 was then exchanged against the corresponding fragment containing the γA285C mutation by standard restriction/ligation. The βD380C/γA87C double mutant was generated by substitution of the KpnI/XbaI fragment of pKH7 with the KpnI/SacI fragment carrying the γA87C mutation and SacI/XbaI fragment of pKG4 carrying the βD380C mutation by triple ligation. Successful cloning was checked by nucleotide sequencing. The resulting plasmids were pMM10 (αP280C/γA285C), pPP2 (αA334C/γL262C), and pSW3 (βD380C/γA87C). Escherichia colistrain DK8 (23Klionsky D.J. Brusilow W.S.A. Simoni R.D. J. Bacteriol. 1984; 160: 1055-1060Crossref PubMed Google Scholar) was transformed with the required plasmid, and cells were grown on minimal medium. Cells were collected at A600 = 0.8. Membranes were isolated and purified essentially according to Wise (24Wise J.G. J. Biol. Chem. 1990; 265: 10403-10409Abstract Full Text PDF PubMed Google Scholar) except that 5 mm DTT was added to all buffers in the case of PP2. EF1 was extracted by treatment with 1 mm EDTA in the presence of 1 (MM10, SW3) or 5 (PP2) mm DTT and applied to an anion-exchange column (Tosoh Fractogel TSK-DEAE 650(S), Toyo Soda, Darmstadt, Germany) equilibrated with buffer A (50 mmTris/H2SO4, 1 mm DTT, and 10% (v/v) methanol, pH 7.4). EF1 was eluted from the column using a stepwise gradient of 0.5 mNa2SO4 in buffer A. Enzyme-containing fractions (75–150 mm Na2SO4) were combined (yielding 20 mg of protein for MM10 and SW3, 10 mg for PP2 from 12-liter cultures). For further purification aliquots were gel-filtrated through PD-10 columns (Amersham Pharmacia Biotech), which were equilibrated with buffer B (50 mm Tris-HCl, 5 mm MgCl2, 50 mm KCl, 10% (v/v) glycerol, pH 7.5) and then applied to Ni-NTA affinity chromatography (5 mg of protein/ml Ni-NTA-agarose) equilibrated with the same buffer. After washing with buffer B (containing 20 mm imidazole and 1 mm DTT), pure EF1 was eluted with 150 mm imidazole and 1 mm DTT in buffer B. For cross-linking studies, the eluate was reapplied to a PD-10 gel filtration column equilibrated with buffer B. Protein determination was carried out according to Sedmak and Grossberg (25Sedmak J.J. Grossberg S.E. Anal. Biochem. 1994; 79: 544-552Crossref Scopus (2502) Google Scholar), resulting in 0.5–1 mg/ml protein, and purity was checked by SDS electrophoresis (Amersham Pharmacia Biotech Phast system, 8–25% gradient gels) without β-mercaptoethanol in the sample buffer. Staining was carried out with Coomassie I. H. In A 5 for Scholar). formation of the cross-link was by treatment of the enzyme in buffer B with or 50 at temperature for in the presence of mm ATP. The were with 10 mm and 5 The of the cross-link formation was shown by of the with 50 at All were to electrophoresis on a 8–25% SDS-polyacrylamide were with Coomassie by PubMed Scopus Google Scholar). The were and the gel were using were into flow cells of Germany) by Protein were in the per peroxidase in 20 mm 50 mm KCl, 5 mm 10 mg/ml in buffer EF1 in buffer with buffer 0.5 in buffer with buffer O. Gumbiowski K. Junge W. Engelbrecht S. FEBS Lett. 2000; 472: 34-38Crossref PubMed Scopus (178) Google Scholar) in buffer with buffer and 20 mm mg/ml 50 and 5 mm ATP in buffer D. of of the chain the of fluorescent as from the of fluorescent the flow This that the actin were attached to subunit γ in the to at and for the required we used This was then by with a of for 20 this the of the of of to the the of was An of was used for the of MM10 at at residue to of cysteines required for cross-link was then oxidized at a protein of 0.5–1 mg/ml in buffer B with in the presence of for by PD-10 gel filtration column in buffer B. of cross-link formation was as An was with a and to a the of the disulfide bridge in the oxidized the assay was carried out in the of This in of the actin was by the to by a and by the by a high this 5 in as on a A sample of EF1 was into the flow and with fluorescent actin The rotation of was for to rotation was to were and further by using the and A of subunit γ from E. coli was by the New Scholar) from of the enzyme Although the of γ is not the the of the parts is The portion of γ the residues with by and the part of the central and eight residues of the The protein was by J. J. Chem. 79: Scopus Google Scholar), which formed a cylinder with a of and of In the protein and The molecular dynamics were carried out with the R. M. R. A. N. J. A. K. K. J. 1999; 151: Scopus Google Scholar) using the force D.J. S. M. J. Chem. Scopus Google Scholar). A was used to A of the α-helix was by the was applied to the of of the residues and the of the and the in on the The was equilibrated and then the rotation of γ was forced by a torque applied to the of the The torque was by on the of and The and of the were at of the molecular dynamics by the A. K. 1997; Google Scholar). As a an torque of and the torque the torque upon γ under conditions the rotation of the α-helix as a the of and of the of were at positions that were obtained by the In with the of and H. O. PubMed Scopus Google Scholar), we that the of the α-helix was to the of The treatment of full by the where the force on is into a and a and J. A of the of and on in Scholar), the of the α-helix as with the at of of were In the case of forced the of the α-helix was the was the (αβ)3 the of the of γ, we did not this of (αβ)3 the of the and the of the as we used the to the the function in not and their are through and The of to the function the of the α-helix and its when a torque of was of the function the along the and the of the of ATPase activity was at protein of 10 in 50 mm Tris-HCl, pH mm MgCl2, and 10 mm After for 5 at the reaction was by the of and the Pi was D. Anal. Biochem. PubMed Scopus Google Scholar). Western blotting on was carried out in the Pharmacia Biotech against and were used at of and A against was used then for by EF1 mutant (10Noji H. Häsler K. Junge W. Kinosita Jr., K. Yoshida M. Engelbrecht S. Biochem. Biophys. Res. Commun. 1999; 260: 597-599Crossref PubMed Scopus (83) Google Scholar) was used as starting In this all wild-type cysteines are substituted by alanines (21Kuo P.H. Ketchum C.J. Nakamoto R.K. FEBS Lett. 1998; 426: 217-220Crossref PubMed Scopus (48) Google Scholar), cysteine was introduced in γ and carried an engineered His6 tag at the suitable for disulfide bridge formation between subunits γ and α or β, cysteine residues were introduced at positions or in and in γ cysteine is in wild-type EF1 was in and in β. The mutant was MM10 the αA334C/γL262C mutant was PP2 and the βD380C/γA87C mutant was The latter is to a mutant that was used by Cross and T.M. Bulygin V.V. Zhou Y. Hutcheon M.L. Cross R.L. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 10964-10968Crossref PubMed Scopus (459) Google Scholar). E. coli strain DK8 with the three plasmids as on as the the we or for the oxidized and of from these plasmids with a or the of the cysteines at the and After both and had activities of the activity of was under the same conditions and without of (50 or was used for the of engineered cysteine residues resulting in B. After a the γ had quantitative cross-link formed by Fig. the quantitative formation of the and by SDS-polyacrylamide gel electrophoresis and Western of the of with and for The did not formation under oxidizing conditions not The disulfide bridges were in all three by treatment with 50 mm DTT for with of the ATPase and we high molecular yield by the formation of an that the was not formed sample for SDS we oxidized by and SDS gel electrophoresis in the of of were obtained not This that the disulfide bridge formation in and not SDS sample and their ATP hydrolysis activities both in with and ↔ γA87C T.M. Bulygin V.V. Zhou Y. Hutcheon M.L. Cross R.L. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 10964-10968Crossref PubMed Scopus (459) Google to our the ATPase activity of at of the of the disulfide bridge between residues and did not impede ATP hydrolysis at of disulfide bridge formation on ATPase in a This was to obtained by and PubMed Scopus Google Scholar). that cross-linking in F1 the penultimate endogenous cysteine residue J. Biol. Chem. 1984; Full Text PDF PubMed Google Scholar)) had a on ATPase activity have been as an against the rotary of ATP Y. Annu. Rev. Biol. 2000; PubMed Google Scholar). For of rotation of the γ subunit of the assay of Noji (7Noji H. Yasuda R. Yoshida M. Kinosita K. Nature. 1997; 386: 299-302Crossref PubMed Scopus (1966) Google Scholar) was an of was the used cysteines and which were required for the disulfide these conditions the with in the presence of mm ATP in an quantitative formation of the cross-link After on rotating were under conditions of actin of are shown in Fig. B. to of the were the high yield and the generated torque were in the same as with the enzyme or the The of rotating attached to γK109C under conditions in which γA285C was to for internal at the C-terminal of γ, the ATPase activity under these conditions was from the activity with the Disulfide upon ATP by the generated be the of a disulfide at the ATP hydrolysis O. Cherepanov D.A. Gumbiowski K. Engelbrecht S. Junge W. Biophys. J. 2001; 81: 1220-1233Abstract Full Text Full Text PDF PubMed Scopus (126) Google the disulfide bridge in the oxidized enzyme of ATP as with that were then to the ATP hydrolysis and to SDS gel electrophoresis not The high torque generated by was to the α-helix and to The C-terminal portion of subunit γ three residues with which are to be for of the α-helix For the rotation of γAla1-Ser281 in the γGly282-Ala284 around the around the and is that the and are by of that a in the was by in the of PubMed Scopus Google Scholar). The torque by We used molecular dynamics to the of γ turnover of The dynamics of part of the C-terminal α-helix in subunit γ is shown in Fig. 5 by using the and internal and Fig. 5 A the of for residues in different positions of the The of the dynamics to the of the a torque of was applied to the a to a position by elastic of the The in this After that the of the C-terminal residues and was and the to The dynamics of the rotation were parts with a high of were by the of and the rotation was by the along the internal The dynamics of and for the five C-terminal residues are shown in Fig. B and that the internal of and essentially the α-helix is that the α-helix was by of at the dynamics were by the rotation of at the of the along the of by and PubMed Scopus Google Scholar). This have from the different of and have the of (all of were except In our case the was to an and the to the in the of the and are expected to be the in a In we the hydrolysis activity and rotation of γ in EF1 with engineered disulfide bridges between subunits α to the and γ (rotor) of F1 at three different positions along the of When the bridge was under both activities were in all three oxidizing conditions were the engineered disulfide bridge was at the center or bottom parts of In to the the of the disulfide bridge at the top position had on the of ATP hydrolysis and the rotation of subunit This was not by the forced of the disulfide by the rotary of ATP hydrolysis under our conditions nor was by a rotation around the disulfide out from the rotation Fig. The was to the and of the the penultimate and and are by of that a in the was by on a of that the high torque generated by ATP hydrolysis by EF1 is to an α-helix and to The activities of in with or ATP and rotation along with the of the molecular dynamics are of the in the wild-type enzyme as a In the top part of γ be to This is under by and is The used in this were a of of of Germany).

References

34