Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

The Charge Distribution in the Cytoplasmic Loop of Subunit C of the F1F0 ATPase Is a Determinant for YidC Targeting

2008; Elsevier BV; Volume: 283; Issue: 15 Linguagem: Inglês

10.1074/jbc.m709408200

ISSN

1083-351X

Autores

Stefan Kol, Nico Nouwen, Arnold J. M. Driessen,

Tópico(s)

Photosynthetic Processes and Mechanisms

Resumo

YidC is a member of the Oxa1 family of proteins that facilitates the membrane insertion of a subset of inner membrane proteins in Escherichia coli. YidC acts as an insertase for membrane insertion of subunit c of the F1F0 ATP synthase (F0c), but the requirements for substrate recognition have remained unclear. Here, we have analyzed the role of the charged aminoacyl residues in F0c in YidC targeting and membrane insertion. Binding experiments demonstrate that F0c is targeted directly to YidC without the presence of a stable lipid surface-bound intermediate. Positive charges in the cytoplasmic loop of F0c are important determinants for YidC binding and subsequent membrane insertion. These data support a model in which F0c binds directly to YidC prior to its membrane insertion. YidC is a member of the Oxa1 family of proteins that facilitates the membrane insertion of a subset of inner membrane proteins in Escherichia coli. YidC acts as an insertase for membrane insertion of subunit c of the F1F0 ATP synthase (F0c), but the requirements for substrate recognition have remained unclear. Here, we have analyzed the role of the charged aminoacyl residues in F0c in YidC targeting and membrane insertion. Binding experiments demonstrate that F0c is targeted directly to YidC without the presence of a stable lipid surface-bound intermediate. Positive charges in the cytoplasmic loop of F0c are important determinants for YidC binding and subsequent membrane insertion. These data support a model in which F0c binds directly to YidC prior to its membrane insertion. The Escherichia coli inner membrane protein, YidC, is a member of the cytochrome oxidase biogenesis (Oxa) membrane protein family, which functions in promoting membrane insertion in chloroplasts, mitochondria, and bacteria (1Luirink J. von Heijne G. Houben E. de Gier J.W. Annu. Rev. Microbiol. 2005; : 329-355Crossref PubMed Scopus (155) Google Scholar). YidC is an essential protein in E. coli and other bacteria. A fraction of the YidC is associated with the Sec-translocase, where it interacts with transmembrane segments (TMSs) 3The abbreviations used are:TMStransmembrane segmentAMdiS4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acidF0cF1F0 ATP synthase subunit cIMVinner membrane vesicleIMPinner membrane proteinSRPsignal recognition particlePMFproton motive forceTricineN-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine.3The abbreviations used are:TMStransmembrane segmentAMdiS4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acidF0cF1F0 ATP synthase subunit cIMVinner membrane vesicleIMPinner membrane proteinSRPsignal recognition particlePMFproton motive forceTricineN-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine. of nascent inner membrane proteins (IMPs) during membrane integration (2Scotti P.A. Urbanus M.L. Brunner J. de Gier J.W. von Heijne G. van der Does C. Driessen A.J. M. Oudega B. Luirink J. EMBO J. 2000; 19: 542-549Crossref PubMed Scopus (298) Google Scholar). In addition, YidC also functions as an independent entity in the insertion of IMPs. The major coat proteins of bacteriophage M13 and Pf3, which were initially thought to insert spontaneously (3Kuhn A. FEMS Microbiol. Rev. 1995; 17: 185-190Crossref PubMed Google Scholar), were the first identified substrates of this pathway (4Samuelson J.C. Chen M. Jiang F. Moller I. Wiedmann M. Kuhn A. Phillips G.J. Dalbey R.E. Nature. 2000; 406: 637-641Crossref PubMed Scopus (417) Google Scholar, 5Chen M. Samuelson J.C. Jiang F. Muller M. Kuhn A. Dalbey R.E. J. Biol. Chem. 2002; 277: 7670-7675Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar). The first authentic E. coli YidC substrate described was subunit c of the F1F0 ATP synthase (F0c) (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar), whereas recently, the mechanosensitive channel of large conductance, MscL, was proposed to be another substrate of the YidC pathway (7Facey S.J. Neugebauer S.A. Krauss S. Kuhn A. J. Mol. Biol. 2007; 365: 995-1004Crossref PubMed Scopus (90) Google Scholar).The YidC dependence of F0c is of particular interest as it concerns a critical step in the assembly of the F1F0 ATPase. F0c is a component of the membrane-embedded F0 sector that associates with the cytosolic F1 domain (α3β3γδϵ) to assemble into a functional F1F0 ATPase. In E. coli, multiple F0c subunits assemble into a decameric ring structure (8Fillingame R.H. Angevine C.M. Dmitriev O.Y. Biochim. Biophys. Acta. 2002; 1555: 29-36Crossref PubMed Scopus (73) Google Scholar), which interacts with two b subunits and one a subunit (F0b and F0a, respectively). The F1F0 ATPase plays a central role in the energy metabolism of the cell, converting the energy stored in a transmembrane electrochemical gradient of protons into the synthesis of ATP from ADP and inorganic phosphate. During the catalytic cycle, F0c becomes protonated at an aspartic acid at position 61 in the second TMS. This residue is located at the subunit a/c interface, and protonation induces a rotation of the c-ring whereupon another F0c can be protonated by F0a. This process drives the rotation of subunit γ within the α3β3 hexamer of F1, and by conformational changes in the catalytic nucleotide binding sites, causes the synthesis and release of ATP (9Capaldi R.A. Aggeler R. Trends Biochem. Sci. 2002; 27: 154-160Abstract Full Text Full Text PDF PubMed Scopus (285) Google Scholar, 10Weber J. Senior A.E. FEBS Lett. 2003; 545: 61-70Crossref PubMed Scopus (235) Google Scholar). The biogenesis of the F1F0 ATPase has been studied in some detail. The current model is that upon the YidC-mediated insertion, F0c assembles into a decameric ring structure (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar, 8Fillingame R.H. Angevine C.M. Dmitriev O.Y. Biochim. Biophys. Acta. 2002; 1555: 29-36Crossref PubMed Scopus (73) Google Scholar, 12Kol S. Turrell B.R. de Keyzer J. van der Laan M. Nouwen N. Driessen A.J. M. J. Biol. Chem. 2006; 281: 29762-29768Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar) that subsequently interacts with the F0a/(F0b)2 subcomplex (13Stalz W.D. Greie J.C. Deckers-Hebestreit G. Altendorf K. J. Biol. Chem. 2003; 278: 27068-27071Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). It is commonly believed that, in the next stage, the preassembled F1 sector binds to the membrane-embedded F0 sector to form the functional ATPase. In vitro, mixing of the separately purified sectors results in the restoration of coupled ATPase activity (14Steffens K. Schneider E. Herkenhoff B. Schmid R. Altendorf K. Eur. J. Biochem. 1984; 138: 617-622Crossref PubMed Scopus (17) Google Scholar).The mechanism by which YidC inserts IMPs into the membrane is poorly understood. Analysis of the physicochemical properties of YidC-only substrates may provide further insights in this process. Although the currently identified substrate range of the YidC-only pathway is relatively small, a common feature of these membrane proteins is their small size, their hydrophobicity, and the presence of short hydrophilic periplasmic regions. M13 phage procoat protein (73 amino acids) and the E. coli MscL (136 amino acids) and F0c (79 amino acids) are similar in structure and contain two TMSs (or a signal sequence and TMS). The Pseudomonas aeruginosa Pf3 major coat protein (44 amino acids) contains only one TMS. Although in vivo depletion of the bacterial SRP component Ffh interferes with membrane assembly of a tagged F0c variant (15van Bloois E. Haan G.J. de Gier J.W. Oudega B. Luirink J. FEBS Lett. 2004; 576: 97-100Crossref PubMed Scopus (74) Google Scholar), reconstitution studies have demonstrated that M13, Pf3, and F0c are all targeted to YidC in a SRP-independent manner (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar, 16de Gier J.W. Scotti P.A. Saaf A. Valent Q.A. Kuhn A. Luirink J. von Heijne G. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 14646-14651Crossref PubMed Scopus (112) Google Scholar, 17Chen M. Xie K. Nouwen N. Driessen A.J.M. Dalbey R.E. J. Biol. Chem. 2003; 278: 23295-23300Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). MscL, on the other hand, seems to utilize SRP (7Facey S.J. Neugebauer S.A. Krauss S. Kuhn A. J. Mol. Biol. 2007; 365: 995-1004Crossref PubMed Scopus (90) Google Scholar). Currently, it is unclear how ribosome targeting by SRP is achieved as YidC lacks the C-terminal extension of the mitochondrial Oxa1p needed for ribosome binding (18Jia L. Dienhart M. Schramp M. McCauley M. Hell K. Stuart R.A. EMBO J. 2003; 22: 6438-6447Crossref PubMed Scopus (173) Google Scholar, 19Szyrach G. Ott M. Bonnefoy N. Neupert W. Herrmann J.M. EMBO J. 2003; 22: 6448-6457Crossref PubMed Scopus (191) Google Scholar). Moreover, it is not evident how SRP can discriminate between the YidC and Sectranslocases, i.e. the assumed default pathway for SRP-targeted nascent IMPs (20Valent Q.A. Scotti P.A. High S. de Gier J.W. von Heijne G. Lentzen G. Wintermeyer W. Oudega B. Luirink J. EMBO J. 1998; 17: 2504-2512Crossref PubMed Scopus (244) Google Scholar). Therefore, a major unresolved question is how these IMPs are targeted and recognized by YidC. Obviously, the hydrophobic character of the TMSs is important for YidC recognition in line with observations that TMSs of substrate proteins can be cross-linked to YidC (2Scotti P.A. Urbanus M.L. Brunner J. de Gier J.W. von Heijne G. van der Does C. Driessen A.J. M. Oudega B. Luirink J. EMBO J. 2000; 19: 542-549Crossref PubMed Scopus (298) Google Scholar). In addition, it has been suggested that the charge distribution in these IMPs may be critical for membrane binding and may possibly allow the electrophoretic translocation of the periplasmic regions and/or the formation of the final membrane topology (21van der Laan M. Nouwen N.P. Driessen A.J. M. Curr. Opin. Microbiol. 2005; 8: 182-187Crossref PubMed Scopus (53) Google Scholar). Before YidC was identified as an essential component for insertion, M13 procoat was believed to insert into the membrane spontaneously. Earlier studies with model systems suggested that the positive charges at the N and C termini of M13 procoat are important for an electrostatic binding to the membrane surface (22Kuhn A. Wickner W. Kreil G. Nature. 1986; 322: 335-339Crossref PubMed Scopus (55) Google Scholar, 23Gallusser A. Kuhn A. EMBO J. 1990; 9: 2723-2729Crossref PubMed Scopus (73) Google Scholar), whereas the negative charges in the connecting loop were proposed to be needed for the proton motive force (PMF)-dependent insertion, allowing the protein to adopt its final membrane topology (24Cao G. Cheng S. Whitley P. von Heijne G. Kuhn A. Dalbey R.E. J. Biol. Chem. 1994; 269: 26898-26903Abstract Full Text PDF PubMed Google Scholar). Interestingly, the negative charges appear not to be necessary for insertion per se (25Kuhn A. Zhu H.Y. Dalbey R.E. EMBO J. 1990; 9: 2385-2389Crossref PubMed Scopus (46) Google Scholar). Although those studies demonstrated that M13 procoat has a strong tendency to interact electrostatically with the membrane surface, it is unknown whether a membrane surface-bound form represents an authentic intermediate in YidC-mediated targeting and insertion. In contrast, membrane insertion of F0c occurs independently of the PMF, whereas its membrane topology is reverse of that of M13 procoat (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar). With F0c, the short N and C termini are translocated, whereas the loop that connects the two TMSs remains in the cytosol.Here, we have determined the role of the charged aminoacyl residues in the N terminus and cytosolic loop region of F0c in YidC targeting and membrane insertion using in vitro assays. Our data demonstrate that positive charges in the cytosolic loop region of F0c are important determinants for YidC recognition, and consequently, also for subsequent YidC-mediated membrane insertion. The data further suggest that membrane insertion does not occur via a stable membrane surface-bound intermediate of F0c, consistent with a model wherein newly synthesized F0c is targeted directly to YidC.EXPERIMENTAL PROCEDURESStrains and Plasmids—E. coli strain A19 was used to obtain the S135 lysate (26Clark A.J. Genetics. 1963; 48: 105-120Crossref PubMed Google Scholar). Strain SF100 (27Baneyx F. Georgiou G. J. Bacteriol. 1990; 172: 491-494Crossref PubMed Google Scholar) was used for topology determination and overexpression of YidC (28van der Laan M. Houben E.N. Nouwen N. Luirink J. Driessen A.J. M. EMBO Rep. 2001; 2: 519-523Crossref PubMed Scopus (95) Google Scholar). Strain JS7131 (4Samuelson J.C. Chen M. Jiang F. Moller I. Wiedmann M. Kuhn A. Phillips G.J. Dalbey R.E. Nature. 2000; 406: 637-641Crossref PubMed Scopus (417) Google Scholar), in which the yidC gene is under control of the araBAD promoter, was grown overnight in Luria-Bertani (LB) medium at 37 °C supplemented with 0.2% arabinose and 25 μg/ml spectinomycin. Cells were harvested, washed with warm LB, diluted to an OD660 of 0.4, and further grown in the presence of 0.2% glucose to deplete YidC or with 0.2% arabinose to generate non-depleted control cells. After every generation, the cultures were diluted with 1 volume of the same medium until the YidC-depleted cells stopped growing. Depletion was verified by immunoblotting with antibodies directed against YidC. All mutants were constructed in vector pET20AtpE-A79C (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar) (see Table 1).TABLE 1Site-directed F0c mutants with altered charge distribution in the N terminus and the middle loop regionWild-type F0cF0c mutantaThe indicated mutations were introduced into plasmid pET20AtpE-A79C. Only the changed residues with the corresponding substitutions are indicated. Codons are indicated between brackets. The nomenclature of the F0c mutants is explained under “Results”N00N++M+0+0+M00000M0–0–0M0000+M00+00M+0000M0–0–+M0–+–0M+–0–0M+–+–0M+–0–+M0–+–+N++M–––––M–––––Glu-2 (GAA)Gln (CAA)Lys (AAA)Lys (AAA)Asp-7 (GAT)Asn (AAT)Lys (AAA)Lys (AAA)Lys-34 (AAA)Leu (CTA)Leu (CTA)Leu (CTA)Leu (CTA)Leu (CTA)Leu (CTA)Leu (CTA)Glu (GAA)Glu (GAA)Glu-37 (GAA)Gln (CAA)Gln (CAA)Gln (CAA)Gln (CAA)Gln (CAA)Arg-41 (CGT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Glu (GAA)Glu (GAA)Asp-44 (GAT)Asn (AAT)Asn (AAT)Asn (AAT)Asn (AAT)Asn (AAT)Arg-50 (CGT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Leu (CTT)Glu (GAG)Glu (GAG)a The indicated mutations were introduced into plasmid pET20AtpE-A79C. Only the changed residues with the corresponding substitutions are indicated. Codons are indicated between brackets. The nomenclature of the F0c mutants is explained under “Results” Open table in a new tab In Vitro Insertion, Binding, and Association Assays—The Ribo-Max transcription kit (Promega) was used for the synthesis of mRNA in a coupled transcription/translation system. Reactions were carried out for 20 min at 37 °C in the absence or in the presence of YidC-depleted or non-depleted inner membrane vesicles (IMVs) as described previously. A small sample of the reaction mixture was used as a synthesis control, and the remainder was treated with 0.4 mg/ml proteinase K for 30 min on ice in the presence or absence of 1% Triton X-100. Samples were trichloroacetic acid-precipitated and analyzed by 18% SDS-PAGE and phosphoimaging. Binding assays were performed in the same manner, but the remainder was loaded on top of a sucrose cushion consisting of 10 mm HEPES-KOH, pH 7.5, 100 mm KCl, 0.5 mg/ml bovine serum albumin, 20% sucrose (v/v) sucrose, 1 mm dithiothreitol and spun at 25 p.s.i. for 20 min. Pellets were resuspended in 50 mm HEPES-KOH, pH 7.5, trichloroacetic acid-precipitated and analyzed by 18% SDS-PAGE and phosphoimaging.Labeling of F0c A79C—To determine the topology of membrane-inserted wild-type F0c and F0c N++M-----, in vitro labeling reactions were carried out in the presence of IMVs prepared from strain SF100. A small part of the reaction was removed as a synthesis control. The IMVs were subsequently reisolated by centrifugation through a sucrose cushion consisting of 10 mm HEPES-KOH, pH 7.5, 100 mm KCl, 0.5 mg/ml bovine serum albumin, and 20% (v/v) sucrose and resuspended in 50 mm HEPES-KOH, pH 7.5, 20% (v/v) glycerol. The membranes were then incubated for 20 min on ice with 0.15 mg/ml protease K. Digestion was stopped by the addition of 5 mg/ml phenylmethylsulfonyl fluoride. Labeling reactions were performed for 30 min at room temperature with 1 mm 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid (AMdiS) (Invitrogen) with and without Triton X-100 present. The labeling reaction was quenched with dithiothreitol, trichloroacetic acid-precipitated, and analyzed by Tricine-SDS-PAGE and phosphoimaging.RESULTSF0 Subunit c Charge Mutagenesis—In the cytoplasmic membrane, F0c forms a helical hairpin of two TMS. The short N and C termini are translocated into the periplasm, whereas the connecting loop region remains cytosolic (Fig. 1A). Except for the catalytic aspartic acid at position 61 in the second TMS, charged aminoacyl residues are found only at the N terminus (Glu-2 and Asp-7) and in the cytosolic loop (Lys-34, Glu-37, Arg-41, Asp-44, and Arg-50) connecting the two TMSs (M-region). To investigate the contribution of these charges in the targeting of F0c to YidC and the subsequent membrane insertion, we constructed a series of mutants in which the charge distribution and the net charge of the individual regions were systematically altered. To minimize changes in steric interactions with YidC, substitutions of amino acid residues were chosen such that they minimally affected the size of the side chain. To facilitate the determination of the membrane topology of the F0c mutants, mutations were made in an F0c derivative that contains a unique cysteine residue at position 79 and that, like wild-type F0c, inserts into membranes in a YidC-dependent manner (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar). The nomenclature used to describe the mutants first indicates the position in the protein, i.e. N or M, referring to the N terminus or the middle loop region, respectively, followed by the charges present in the region. For instance, the mutant that has the two N-terminal charges replaced by uncharged residues will be referred to as F0c N00. Table 1 summarizes all the mutants used in this study.Recognition of F0c by YidC Is Dependent on the Charge Distribution in the M-region—To study the YidC dependence of F0c binding and insertion, IMVs were isolated from E. coli strain JS7131 (4Samuelson J.C. Chen M. Jiang F. Moller I. Wiedmann M. Kuhn A. Phillips G.J. Dalbey R.E. Nature. 2000; 406: 637-641Crossref PubMed Scopus (417) Google Scholar) grown in the presence of arabinose (wild-type cells, YidC+ IMVs) or glucose (YidC-depleted cells, YidC- IMVs). Analysis of these IMVs by SDS-PAGE and immunoblotting showed that under YidC depleting conditions, the ∼25-kDa phage shock protein A was induced (29van der Laan M. Urbanus M.L. ten Hagen-Jongman C.M. Nouwen N. Oudega B. Harms N. Driessen A.J. M. Luirink J. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 5801-5806Crossref PubMed Scopus (125) Google Scholar) (Fig. 1B) and that the IMVs of cells grown in the presence of glucose contain only a very small amount of YidC (Fig. 1C). These IMVs were used in subsequent assays, unless indicated otherwise.To determine the role of the charged aminoacyl residues in the M-region of F0c in targeting to YidC, a combined binding/insertion assay was developed based on an in vitro transcriptional/translation reaction of F0c in the presence of IMVs followed by isolation of the IMVs by centrifugation through a sucrose cushion. Since targeting, binding, and insertion are tightly coupled processes, the sedimentation assay monitors both the amount of membrane-associated F0c and the amount of inserted F0c. In vitro synthesized wild-type full-length F0c co-sediments with YidC+ IMVs (Fig. 2A, lane 5), whereas only low amounts associate with YidC- IMVs (lane 6). The latter levels are only slightly higher than the sedimentation in the absence of IMVs (lane 4). This suggests that the residual binding is due to the small residual amounts of YidC present in the YidC- IMVs. The efficiency of the in vitro synthesis of F0c was not affected by the presence of IMVs (lanes 1–3). Increased levels of F0c were found to co-sediment with IMVs containing overproduced levels of YidC (Fig. 2E). These data show that the sedimentation of F0c with IMVs is strictly YidC-dependent and that newly synthesized F0c has little tendency to associate with lipids or other membrane proteins present in the IMVs.FIGURE 2Co-sedimentation of F0c with YidC is strongly decreased upon removal of two or more positive charges in the M-region. Wild-type (WT) F0c (A), F0c M+-+-0 (B), F0c M0000+ (C), and F0c M00000 (D) were synthesized in the absence of IMVs (lanes 1 and 4) or in the presence of 5 μg of YidC+ IMVs (lanes 2 and 5) or YidC- IMVs (lanes 3 and 6). E, wild-type F0c was synthesized in the absence of IMVs (lanes 1 and 4) or in the presence of 5 μg of SF100 IMVs (YidC+, lanes 2 and 5) or YidC-overexpressed SF100 IMVs (YidC2+, lanes 3 and 6). After translation, samples were directly analyzed on SDS-PAGE as translation controls (lanes 1–3) or spun through a sucrose gradient, and the collected membrane material was resuspended and trichloroacetic acid-precipitated (lanes 4–6). F, quantification of the co-sedimentation of charge mutants with YidC expressed as the ratio of F0c bound to YidC+ and YidC- IMVs. F0c mutants defective in membrane association showed a ratio of 1 or less. n.d., not determined.View Large Image Figure ViewerDownload Hi-res image Download (PPT)The sedimentation assay was further used to study the interaction of the remaining mutant F0c proteins with YidC. Removal of only one of the positive charges in the M-region (Fig. 2B, F0c M+-+-0) had little effect on the amount of F0c co-sedimented with the YidC+ IMVs. On the other hand, when only a single positive charge (F0c M0000+) (Fig. 2C) or no charge (F0c M00000) (Fig. 2D) is present in the M-region, sedimentation is no longer YidC-dependent, and only very low levels of F0c associate with the IMVs. Analysis of all M-region mutants in the YidC-dependent sedimentation assay (Fig. 2F) showed that the presence of two positive charges in the M-region suffices for efficient YidC-binding (F0c M0-+-+, F0c M+-0-+, and F0c M+-+-0), whereas a further reduction to either one or no positive charge entirely abolishes the YidC-dependent sedimentation. Neutralization of the two negative charged amino acids in the M-region (F0c M+0 + 0+) had little effect on the YidC-dependent sedimentation. With charge mutants of the N terminus of F0c, sedimentation of F0c N00 and F0c N++ is somewhat reduced as compared with the wild-type (F0c N--) but remained YidC-dependent. Taken together, the above data indicate that the positively charged amino acids in the M-region of F0c are important determinants for YidC recognition.YidC-mediated Membrane Insertion of F0c Is Dependent on the Presence of Positive Charges in the M-region—We next analyzed the set of mutants for membrane insertion using the previously described protease-protection assay. In the presence of YidC+ IMVs, membrane-inserted wild-type F0c becomes protease-resistant (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar) (Fig. 3A, lane 5), whereas with YidC- IMVs, only a low level of protease-protected F0c is observed (lane 6) that is slightly above the background levels observed in the absence of membranes (lane 4). Like wild-type F0c, a substantial amount of protease-protected F0c was observed when the two negatively charged aminoacyl residues at the N terminus were mutated (Fig. 3, B and C, lane 5; F0c N00 and F0c N++), although the YidC dependence of F0c N00 was somewhat reduced. These data show that F0c N00 and N++ insert into the membrane in a YidC-dependent manner and indicate that the charge distribution of the N terminus only marginally affects membrane insertion of F0c.FIGURE 3Membrane insertion of F0c is dependent on the presence of at least two positive charges in the M-region. Wild-type (WT) F0c (A), F0c N00 (B), F0c N++ (C), F0c M+-+-0(D), F0c M0000+ (E), and F0c M00000 (F) were synthesized in the absence of IMVs (lanes 1 and 4) or in the presence of 5 μg of YidC+ IMVs (lanes 2 and 5) or YidC- IMVs (lanes 3 and 6). After translation, samples were applied directly on SDS-PAGE as translation controls (lanes 1–3) or treated with 0.4 mg/ml proteinase K (PK) (lanes 4–6). G, quantification of the YidC dependence of membrane insertion of all charge mutants expressed as the ratio of F0c insertion to YidC+ and YidC- IMVs. F0c mutants defective in membrane insertion showed a ratio of 1 or less.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Next, the membrane insertion of the M-region mutants of F0c was analyzed (Fig. 3, D–G). Mutagenesis of only one of the three positively charged residues into a neutral or negatively charged aminoacyl residue hardly affected the YidC-dependent membrane insertion of F0c (Fig. 3, D and G; F0c M0-+-+, M+-+-0, and M+-0-+). However, a marked decrease in the YidC-dependent membrane insertion was observed upon the mutagenesis of at least two positive charges (Fig. 3, E and G;F0c M0-0-+, M0-+-0, and M+-0-0). Strikingly, the removal of all positive charges (Fig. 3G;F0c M0-0-0), neutralization of the M-region (Fig. 3, F and G;F0c M00000), or a complete replacement of the positively charged residues for glutamate (Fig. 3G, F0c M-----) completely abolishes the YidC-dependent membrane insertion. The latter mutant also excludes the possibility that the insertion defect is due to polarity effects of the introduced leucine residues. Neutralization of the two negatively charged amino acids (Fig. 3G; F0c M+0 + 0+) barely affected the insertion. It should be noted that in particular with F0c M00000, substantial levels of YidC-independent membrane insertion are observed. The F0c mutant with the reversed charge (F0c N++M-----) only weakly inserts into the membrane in a YidC-dependent manner (Figs. 3G and 4A). Overall, it is concluded that at least two positive charges in the M-region of F0c are required for YidC-mediated membrane insertion. The exact position of these charges in the cytosolic loop region is of only minor importance.FIGURE 4Membrane insertion, co-sedimentation and topology analysis of F0c N++M-----. A, reactions were carried out as described in the legend of Fig. 3. PK, proteinase K. B, reactions were carried out as described in the legend of Fig. 3, A and B. Wild-type (WT) F0c (C) and F0c N++M----- (D) were synthesized in the presence of 5 μg of IMVs prepared from strain SF100. A small sample of the reaction was taken as a synthesis control (lane 1), and the remainder was spun through a sucrose cushion and treated with proteinase K. Digestion was stopped by the addition of phenylmethylsulfonyl fluoride, and the reaction was aliquoted and trichloroacetic acid precipitated (lane 2) or treated with AMdiS in the absence (lane 3) or presence of 1% (v/v) Triton X-100 (TX100) (lane 4). These samples were then trichloroacetic acid-precipitated and analyzed by Tricine-SDS-PAGE and phosphoimaging. When F0c N++M----- is treated with proteinase K, a proteolytic fragment is formed (*).View Large Image Figure ViewerDownload Hi-res image Download (PPT)A Reversed Charge Mutant of F0c Shows Strong Membrane Binding but Poor Membrane Insertion—The charge balance of mutant F0c N++M----- is reminiscent of M13 procoat (3Kuhn A. FEMS Microbiol. Rev. 1995; 17: 185-190Crossref PubMed Google Scholar), a protein that has a membrane topology inverse of that of wild-type F0c. As the F0c N++M----- also behaved differently in the binding and insertion assays as compared with wild-type F0c, we decided to analyze its membrane topology. The F0c derivatives used in this study contain a C-terminal cysteine residue (Cys-79) that in wild-type F0c is translocated into the vesicle lumen. With the membrane-inserted wild-type F0c, the cysteine residue is inaccessible for labeling with the membrane-impermeable reagent AMdiS (6van der Laan M. Bechtluft P. Kol S. Nouwen N. Driessen A.J. M. J. Cell Biol. 2004; 165: 213-222Crossref PubMed Scopus (174) Google Scholar) (Fig. 4C, lane 3) but becomes accessible upon membrane solubilization with the detergent Triton X-100 (lane 4), as evidenced by a mobility change in SDS-PAGE due to the added mass of the AMdiS molecule. Incubation of membrane-inserted F0c N++M----- with AMdiS yields an altered labeling pattern as compared with wild-type F0c with a larger fraction of the protease-protected full-length F0c shifted toward the derivatized form (Fig. 4D, lane 3). This suggests that a major fract

Referência(s)