Evidence for Assembly of Small Multidrug Resistance Proteins by a “Two-faced” Transmembrane Helix
2006; Elsevier BV; Volume: 281; Issue: 22 Linguagem: Inglês
10.1074/jbc.m600434200
ISSN1083-351X
AutoresArianna Rath, Roman A. Melnyk, Charles M. Deber,
Tópico(s)RNA and protein synthesis mechanisms
ResumoClinically significant bacterial resistance to drugs and cytotoxic compounds can be conferred by the energy-dependent efflux of toxicants, catalyzed by proteins embedded in the bacterial cell membrane. One such group of proteins, the small multidrug resistance family, are drug/proton antiporters that must oligomerize to function, a process that requires the assembly of at least two inactive monomers by intermolecular association of their four transmembrane helices. Here, we have used peptides that correspond to each of the four wild type transmembrane helices of the Halobacterium salinarum protein Hsmr and a corresponding library of mutant peptides to determine the interactive surfaces that likely contribute to protein oligomerization. Hsmr peptides were examined for strong (sodium dodecyl sulfate-resistant) and weaker (perfluorooctanoate-resistant) helix-helix interactions, in conjunction with circular dichroism, fluorescence energy transfer measurements, and molecular modeling. The results are compatible with a scheme in which two faces of helix four permit self-assembly via a higher affinity asymmetric pairing and a lower affinity symmetric interaction, resulting in a discrete tetramer. Our finding that two surfaces of helix four can contribute to the stability of small multidrug resistance protein assembly provides a molecular basis for the design of therapeutics that target this antibiotic resistance mechanism. Clinically significant bacterial resistance to drugs and cytotoxic compounds can be conferred by the energy-dependent efflux of toxicants, catalyzed by proteins embedded in the bacterial cell membrane. One such group of proteins, the small multidrug resistance family, are drug/proton antiporters that must oligomerize to function, a process that requires the assembly of at least two inactive monomers by intermolecular association of their four transmembrane helices. Here, we have used peptides that correspond to each of the four wild type transmembrane helices of the Halobacterium salinarum protein Hsmr and a corresponding library of mutant peptides to determine the interactive surfaces that likely contribute to protein oligomerization. Hsmr peptides were examined for strong (sodium dodecyl sulfate-resistant) and weaker (perfluorooctanoate-resistant) helix-helix interactions, in conjunction with circular dichroism, fluorescence energy transfer measurements, and molecular modeling. The results are compatible with a scheme in which two faces of helix four permit self-assembly via a higher affinity asymmetric pairing and a lower affinity symmetric interaction, resulting in a discrete tetramer. Our finding that two surfaces of helix four can contribute to the stability of small multidrug resistance protein assembly provides a molecular basis for the design of therapeutics that target this antibiotic resistance mechanism. The advances made in controlling and treating infectious diseases using antibiotics are threatened by the increase of multidrug resistance in pathogenic organisms (1Walsh C. Nature. 2000; 406: 775-781Crossref PubMed Scopus (1199) Google Scholar). Clinically significant resistance to therapeutic compounds can be conferred by proteins embedded in the bacterial cell membrane that use energy-dependent mechanisms to extrude a wide variety of toxicants from the cell (2Nikaido H. Science. 1994; 264: 382-388Crossref PubMed Scopus (1271) Google Scholar). Among these, the small multidrug resistance (SMR) 4The abbreviations used are: SMR, small multidrug resistance; TM, transmembrane; WT, wild type; FRET, fluorescence resonance energy transfer; PFO, sodium perfluorooctanoate; TAMRA, 5- (and 6-) carboxytetramethylrhodamine; dansyl, 5-dimethyl-amino-1-naphthalenesulfamoyl; dabcyl, 4-dimethylaminoazobenzene-4′-sulfamoyl; MES, 4-morpholineethanesulfonic acid. proteins are highly prevalent (>60 homologs in both Gram-positive and Gram-negative bacteria (3Ninio S. Rotem D. Schuldiner S. J. Biol. Chem. 2001; 276: 48250-48256Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar)) proton/drug antiporters of 100–110 amino acids (4Paulsen I.T. Skurray R.A. Tam R. Saier Jr., M.H. Turner R.J. Weiner J.H. Goldberg E.B. Grinius L.L. Mol. Microbiol. 1996; 19: 1167-1175Crossref PubMed Scopus (245) Google Scholar). SMRs catalyze the efflux of large (up to 10 Ä cross-section), structurally diverse hydrophobic cations such as acriflavine, ethidium, methyl viologen, tetracycline, and tetraphenylphosphonium (5Ubarretxena-Belandia I. Tate C.G. FEBS Lett. 2004; 564: 234-238Crossref PubMed Scopus (35) Google Scholar). The monomer of the best characterized family member, the Escherichia coli SMR protein EmrE, is a tightly packed bundle of four antiparallel transmembrane (TM) α-helices with short loops (6Arkin I.T. Russ W.P. Lebendiker M. Schuldiner S. Biochemistry. 1996; 35: 7233-7238Crossref PubMed Scopus (100) Google Scholar, 7Schwaiger M. Lebendiker M. Yerushalmi H. Coles M. Groger A. Schwarz C. Schuldiner S. Kessler H. Eur. J. Biochem. 1998; 254: 610-619Crossref PubMed Scopus (89) Google Scholar). The minimum structural unit of EmrE is a dimer that is unusual in that its component monomers have distinct conformations (8Tate C.G. Kunji E.R. Lebendiker M. Schuldiner S. EMBO J. 2001; 20: 77-81Crossref PubMed Scopus (103) Google Scholar, 9Ubarretxena-Belandia I. Baldwin J.M. Schuldiner S. Tate C.G. EMBO J. 2003; 22: 6175-6181Crossref PubMed Scopus (171) Google Scholar, 10Tate C.G. Ubarretxena-Belandia I. Baldwin J.M. J. Mol. Biol. 2003; 332: 229-242Crossref PubMed Scopus (71) Google Scholar, 11Ma C. Chang G. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 2852-2857Crossref PubMed Scopus (77) Google Scholar). The structural inequivalence of these subunits may relate to an antiparallel insertion of individual EmrE polypeptides (5Ubarretxena-Belandia I. Tate C.G. FEBS Lett. 2004; 564: 234-238Crossref PubMed Scopus (35) Google Scholar, 9Ubarretxena-Belandia I. Baldwin J.M. Schuldiner S. Tate C.G. EMBO J. 2003; 22: 6175-6181Crossref PubMed Scopus (171) Google Scholar, 11Ma C. Chang G. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 2852-2857Crossref PubMed Scopus (77) Google Scholar, 12Pornillos O. Chen Y.J. Chen A.P. Chang G. Science. 2005; 310: 1950-1953Crossref PubMed Scopus (71) Google Scholar), although evidence for antiparallel topology was not obtained in cysteine accessibility and labeling studies (13Ninio S. Elbaz Y. Schuldiner S. FEBS Lett. 2004; 562: 193-196Crossref PubMed Scopus (48) Google Scholar). It is also unresolved whether the pre-existence of alternate monomer folds is required for, or is the consequence of, multimer formation. Although the in vivo oligomeric state of EmrE may be a dimer of dimers (11Ma C. Chang G. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 2852-2857Crossref PubMed Scopus (77) Google Scholar, 14Elbaz Y. Steiner-Mordoch S. Danieli T. Schuldiner S. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 1519-1524Crossref PubMed Scopus (125) Google Scholar), drug efflux activity requires the assembly of two SMR molecules (15Yerushalmi H. Lebendiker M. Schuldiner S. J. Biol. Chem. 1996; 271: 31044-31048Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar). A detailed understanding of the TM helix-helix interactions that stabilize the self-assembly of SMR drug pumps is thus of considerable interest in the development of therapeutics that could target this resistance mechanism by disrupting SMR oligomerization. Like other multimeric membrane proteins, SMRs are stabilized by weak (van der Waals and electrostatic) intermolecular interactions between TM helix faces, defined as individual sets of residues on distinct surfaces of each α-helical TM segment (16Popot J.L. Engelman D.M. Biochemistry. 1990; 29: 4031-4037Crossref PubMed Scopus (823) Google Scholar). The intermolecular helix-helix interactions that assemble SMR oligomers vary in strength among family members (17Ninio S. Schuldiner S. J. Biol. Chem. 2003; 278: 12000-12005Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar) but are of sufficient stability in the Halobacterium salinarum homolog Hsmr to resist the strongly denaturing conditions of SDS-PAGE that disrupt oligomerization of EmrE (17Ninio S. Schuldiner S. J. Biol. Chem. 2003; 278: 12000-12005Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). The SDS-resistant Hsmr dimer thus allows for the detection of key SMR intermolecular interactions that could be denatured in its less stable counterparts. We have independently been studying Hsmr as a model system, using peptides that correspond to each of its four wild type (WT) TM helices, along with a corresponding library of mutant peptides. Here, we examine these Hsmr TM peptides for stronger (sodium dodecyl sulfate-resistant) and weaker (sodium perfluorooctanoate-resistant) TM-TM interactions, in conjunction with CD and fluorescence energy transfer (FRET) measurements and molecular modeling, to show that helix four has two interactive surfaces that may contribute to the stability of SMR protein oligomers. Peptide Synthesis and Purification—Prediction of Hsmr TM segments was performed using the program TM Finder (18Deber C.M. Wang C. Liu L.P. Prior A.S. Agrawal S. Muskat B.L. Cuticchia A.J. Protein Sci. 2001; 10: 212-219Crossref PubMed Scopus (111) Google Scholar) with all default parameters and with gap length set to zero. Lysine-tagged peptides corresponding to Hsmr residues 2–24 (TM-1), 30–51 (TM-2), 56–79 (TM-3), and 85–105 (TM-4) with the sequences K-HPYAYLAAAIAAEVAGTTALKLS-K, KK-PAPSVVVLVGYVSSFYFLGLVL-KKK, KKK-VGVVYGTWAAVGIVATALVGVVF-KKK, and KKK-VAGVVGLALIVAGVVVLNVAS-KK were synthesized with a Pioneer peptide synthesizer (Applied Biosystems) using standard Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry (19Amblard M. Fehrentz J.A. Martinez J. Subra G. Methods Mol. Biol. 2005; 298: 3-24PubMed Google Scholar) on a PAL-PEG-PS 4′-aminomethyl-3′, 5′-dimethoxyphenoxyvaleric acid-poly(ethylene glycol) polystyrene resin (Applied Biosystems) that produced an amidated C terminus upon peptide cleavage. The lysine residues added to each peptide sequence were not anticipated to affect peptide orientation or stoichiometry (20Melnyk R.A. Partridge A.W. Deber C.M. Biochemistry. 2001; 40: 11106-11113Crossref PubMed Scopus (84) Google Scholar, 21Therien A.G. Deber C.M. J. Biol. Chem. 2002; 277: 6067-6072Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar, 22Partridge A.W. Melnyk R.A. Yang D. Bowie J.U. Deber C.M. J. Biol. Chem. 2003; 278: 22056-22060Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar, 23Zouzoulas A. Therien A.G. Scanzano R. Deber C.M. Blostein R. J. Biol. Chem. 2003; 278: 40437-40441Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). Labeling with 4-dimethylaminoazobenzene -4′-sulfonyl chloride (dabcyl chloride), 5-dimethylamino-1-naphthalenesulfonyl chloride (dansyl chloride), or 5- (and 6-) carboxytetramethylrhodamine succinimidyl ester (TAMRA-SE) (Molecular Probes) was accomplished by incubating the resin-bound peptide with excess label under basic conditions overnight. Cleaved peptides were purified by reverse phase-high pressure liquid chromatography on a C4 preparative column (Phenomenex). Mass spectrometry was used to confirm the molecular weight of the purified peptides, and the Micro BCA assay (Pierce) was used to determine concentration. Peptide Electrophoresis—SDS-PAGE on precast NuPAGE gels containing 12% acrylamide in MES buffer (Invitrogen) was performed on 2 μg of peptide samples in homooligomerization experiments. One μgof unlabeled peptide was mixed with 1 μg of TAMRA-labeled peptide samples before loading gels in SDS-PAGE heterooligomerization experiments to keep total peptide concentration identical to that of unlabeled peptides. Sodium perfluorooctanoate (PFO)-PAGE was performed as described (24Ramjeesingh M. Huan L.J. Garami E. Bear C.E. Biochem. J. 1999; 342: 119-123Crossref PubMed Scopus (124) Google Scholar) on precast Tris-Glycine gels containing 18% acrylamide in Tris-Glycine buffer (Invitrogen). A total peptide concentration of 2 μg was used in PFO-PAGE experiments with unlabeled peptides; heterooligomerization of labeled peptides was performed as described above. Coomassie Blue staining was used on all gels to visualize peptides and was used in conjunction with UV transillumination to detect TAMRA-labeled peptides in heterooligomerization experiments. Densitometry measurements were performed using NIH Image. CD and Fluorescence Measurements—CD spectra in buffer and SDS were recorded in a 0.1-cm path length cuvette on a Jasco J-720 circular dichroism spectropolarimeter at peptide concentrations of 20–50 μm in SDS and in a 0.01-cm path length cuvette at peptide concentrations of 50–100 μm in PFO. The mean residue ellipticity expected for peptides in a 100% helical conformation was calculated using the formula max[θ]222=[(n−4.6)(−40,000)]/n where n is the number of residues in the peptide (25Chen Y.H. Yang J.T. Chau K.H. Biochemistry. 1974; 13: 3350-3359Crossref PubMed Scopus (1971) Google Scholar, 26Gans P.J. Lyu P.C. Manning M.C. Woody R.W. Kallenbach N.R. Biopolymers. 1991; 31: 1605-1614Crossref PubMed Scopus (211) Google Scholar). According to this relation, the mean residue ellipticities (deg cm2 dmol–1 res–1) at 222 nm corresponding to 100% helical conformations are –32,640 (TM-1); –33,185 (TM-2); –33,867 (TM-3); and –32,923 (TM-4). For FRET quenching experiments, a 1 μm solution of dansyl-labeled peptide (donor fluor) was mixed with 0–4μm of the corresponding dabcyl-labeled peptide (acceptor) in SDS or PFO micelles and allowed to equilibrate at room temperature overnight. Total peptide concentration was kept constant at 5 μm by the addition of the appropriate concentration of unlabeled peptide. Emission spectra were obtained in a 1-cm path length cuvette by exciting the dansyl fluor at an excitation wavelength of 341 nm with a 2 nm bandpass and monitoring emission from 450 to 650 nm with a 4 nm bandpass at each point in the titration on a Photon Technology International C-60 spectrofluorimeter. Spectra were integrated between 450 and 650 nm using the FELIX software provided by the manufacturer. Integrated total fluorescence intensities at each point in the titration (F) were normalized by the initial integrated total fluorescence intensity in the presence of 0 μm acceptor peptide (Fo). Mole fraction acceptor (Pa) was calculated according to the relation Pa=[acceptorpeptide]/[totalpeptide] Data were fit as described (27Adair B.D. Engelman D.M. Biochemistry. 1994; 33: 5539-5544Crossref PubMed Scopus (138) Google Scholar) using the program Kaleidagraph. In FRET competition experiments, a 1 μm dansyl-labeled peptide solution was mixed with: 1 μm unlabeled peptide; 1 μm dabcyl-labeled peptide; or 1 μm dabcyl and 2 μm unlabeled peptide. CD and fluorescence measurements were performed in buffer (10 mm Tris, pH 7.2, 10 mm NaCl), buffer with 20 mm SDS, or buffer with 50 mm PFO. Molecular Modeling—Energy-minimized models of the interaction between two helices corresponding to Hsmr residues 85–105 (TM-4) were produced as described (28Wang C. Deber C.M. J. Biol. Chem. 2000; 275: 16155-16159Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar). Design of Hsmr TM Peptides—We selected a peptide-based approach to evaluate helix-helix contacts in Hsmr because SMR proteins have short connecting loops between TM segments (5–9 residues in length based on NMR helix boundary assignments (7Schwaiger M. Lebendiker M. Yerushalmi H. Coles M. Groger A. Schwarz C. Schuldiner S. Kessler H. Eur. J. Biochem. 1998; 254: 610-619Crossref PubMed Scopus (89) Google Scholar)) and must therefore be predominantly stabilized by intermolecular TM helix-helix contacts. This technique permits parallel and antiparallel helix-helix interactions to be detected, an ideal methodology for SMR oligomers that may assemble with inverted topologies (5Ubarretxena-Belandia I. Tate C.G. FEBS Lett. 2004; 564: 234-238Crossref PubMed Scopus (35) Google Scholar, 9Ubarretxena-Belandia I. Baldwin J.M. Schuldiner S. Tate C.G. EMBO J. 2003; 22: 6175-6181Crossref PubMed Scopus (171) Google Scholar, 11Ma C. Chang G. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 2852-2857Crossref PubMed Scopus (77) Google Scholar, 12Pornillos O. Chen Y.J. Chen A.P. Chang G. Science. 2005; 310: 1950-1953Crossref PubMed Scopus (71) Google Scholar). Peptides corresponding to each WT TM segment of Hsmr were initially synthesized, with boundaries selected by referring to NMR α-helix assignments in EmrE (7Schwaiger M. Lebendiker M. Yerushalmi H. Coles M. Groger A. Schwarz C. Schuldiner S. Kessler H. Eur. J. Biochem. 1998; 254: 610-619Crossref PubMed Scopus (89) Google Scholar), the α-helices observed in EmrE x-ray structures (11Ma C. Chang G. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 2852-2857Crossref PubMed Scopus (77) Google Scholar, 12Pornillos O. Chen Y.J. Chen A.P. Chang G. Science. 2005; 310: 1950-1953Crossref PubMed Scopus (71) Google Scholar), and use of the TM-segment predicting program TM Finder (18Deber C.M. Wang C. Liu L.P. Prior A.S. Agrawal S. Muskat B.L. Cuticchia A.J. Protein Sci. 2001; 10: 212-219Crossref PubMed Scopus (111) Google Scholar). The peptides synthesized (TM-1, residues 2–24; TM-2, 36–55; TM-3, 56–79; TM-4, 85–105; see “Experimental Procedures” for sequences) incorporate multiple lysine residues at their N and C termini following the “Lys tag” technique (20Melnyk R.A. Partridge A.W. Deber C.M. Biochemistry. 2001; 40: 11106-11113Crossref PubMed Scopus (84) Google Scholar) that increases peptide solubility while preserving the in vivo oligomeric stoichiometry and native helix-helix interfaces of TM segments (20Melnyk R.A. Partridge A.W. Deber C.M. Biochemistry. 2001; 40: 11106-11113Crossref PubMed Scopus (84) Google Scholar, 21Therien A.G. Deber C.M. J. Biol. Chem. 2002; 277: 6067-6072Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar, 22Partridge A.W. Melnyk R.A. Yang D. Bowie J.U. Deber C.M. J. Biol. Chem. 2003; 278: 22056-22060Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar, 23Zouzoulas A. Therien A.G. Scanzano R. Deber C.M. Blostein R. J. Biol. Chem. 2003; 278: 40437-40441Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). Folding of Hsmr TM Peptides in Membrane-mimetic Environments—CD spectra of the four WT Hsmr peptides in the presence and absence of SDS or PFO micelles are shown in Fig. 1. With the exception of TM-2, which had a broad minimum centered near 210 nm, each Hsmr TM peptide displayed CD spectra consistent with a random-coil conformation in aqueous buffer (Fig. 1A). The TM-1, TM-3, and TM-4 peptides had α-helical CD spectra with minima at 208 and 222 nm in the presence of SDS (Fig. 1B) or PFO (Fig. 1C), secondary structure consistent with the conformation of native SMR proteins. The spectrum of TM-2, however, had a broad minimum centered at ∼217 nm in SDS and PFO micelles (Fig. 1, B and C), characteristic of β-strand conformation; its role in the self-assembly of Hsmr was therefore not investigated further. Homo- and Heterooligomerization of Hsmr TM Peptides in SDS—Oligomerization of the TM-1, TM-3, and/or TM-4 peptides was evaluated by SDS-PAGE (Fig. 2). TM-1 is monomeric and TM-4 is dimeric in the 12% NuPAGE system (Fig. 2A), suggesting that the TM-1/TM-1 contacts described in SMR proteins (29Soskine M. Steiner-Mordoch S. Schuldiner S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12043-12048Crossref PubMed Scopus (77) Google Scholar), although important for function (5Ubarretxena-Belandia I. Tate C.G. FEBS Lett. 2004; 564: 234-238Crossref PubMed Scopus (35) Google Scholar), may not stabilize the SDS-resistant Hsmr oligomer. Note that the molecular masses of the three constructs vary (TM-1 and TM-4, 2.56 kDa; TM-3, 3.12 kDa), and their migration rates on the gel are thus not comparable. Further, although TM-1 and TM-3 report oligomeric states >1.0 under some conditions, the TM-4 peptide is consistently dimeric (Fig. 2A) (or tetrameric (vide infra)) throughout our experiments and has been identified as an intermolecular contact site in EmrE (29Soskine M. Steiner-Mordoch S. Schuldiner S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12043-12048Crossref PubMed Scopus (77) Google Scholar). This intriguing ability of The TM-4 peptide to form discrete dimers and tetramers led us to explore the role of TM-4 in Hsmr stabilization in the remainder of this study. All combinations of TM-1, TM-3, and TM-4 were tested for their ability to form SDS-resistant heterooligomers. Unlabeled TM-1, TM-3, or TM-4 peptides were “mixed and matched” in equimolar ratios with peptides labeled with the UV-fluorescent dye TAMRA; here, co-migration of the labeled peptide with its unlabeled counterpart is indicative of heterooligomerization (20Melnyk R.A. Partridge A.W. Deber C.M. Biochemistry. 2001; 40: 11106-11113Crossref PubMed Scopus (84) Google Scholar). Each TAMRA-labeled peptide had a migration pattern similar to its unlabeled counterpart, and co-migration of unlabeled and TAMRA-labeled peptides was not observed for any combination of the TM-1, TM-3, and/or TM-4 peptides (compare Fig. 2B and 2C). We therefore concluded that, although intermolecular TM-1/TM-4 contacts have been described in EmrE (12Pornillos O. Chen Y.J. Chen A.P. Chang G. Science. 2005; 310: 1950-1953Crossref PubMed Scopus (71) Google Scholar, 29Soskine M. Steiner-Mordoch S. Schuldiner S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12043-12048Crossref PubMed Scopus (77) Google Scholar), heterooligomeric pairs of TM-1, TM-3, and/or TM-4 might not be mediating SDS-resistant dimer formation in Hsmr. Homo- and Heterooligomerization of Hsmr TM Peptides in PFO—The ability of the TM-1, TM-3, and TM-4 peptides to homo- and heterooligomerize in PFO micelles was also tested (Fig. 3). PFO is a “mild” detergent that preserves helix-helix interactions denatured by SDS (21Therien A.G. Deber C.M. J. Biol. Chem. 2002; 277: 6067-6072Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar), and PFO-PAGE can preserve the native quaternary structures of membrane proteins (24Ramjeesingh M. Huan L.J. Garami E. Bear C.E. Biochem. J. 1999; 342: 119-123Crossref PubMed Scopus (124) Google Scholar). We therefore expected PFO to preserve the SDS-resistant TM-4 dimer and also permit the formation of any lower affinity helix-helix interactions among the Hsmr peptides. Oligomerization of TM-1 and TM-3 in PFO-PAGE was unchanged when compared with SDS-PAGE (Fig. 3A). The TM-4 peptide, however, migrated as a tetramer (Fig. 3A). Since any interaction strong enough to resist SDS denaturation should persist in PFO (21Therien A.G. Deber C.M. J. Biol. Chem. 2002; 277: 6067-6072Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar, 24Ramjeesingh M. Huan L.J. Garami E. Bear C.E. Biochem. J. 1999; 342: 119-123Crossref PubMed Scopus (124) Google Scholar), we hypothesized that TM-4 self-assembles via at least two sets of intermolecular interactions, one that is SDS-sensitive, but permitted in PFO, and another that is resistant to denaturation by both SDS and PFO. “Mix and match” experiments for heterooligomerization on PFO-PAGE were performed on both unlabeled (Fig. 3B) and TAMRA-labeled and unlabeled TM peptides (Fig. 3, C and D). TAMRA labeling did not affect gel migration of TM-1, TM-3, and TM-4 when compared with unlabeled peptide (compare Fig. 3A and 3D). Migration of the unlabeled TM-1, TM-3, and TM-4 peptides was also unchanged when mixed and matched (compare Fig. 3A and 3B), and co-migration was not observed for all possible combinations of TAMRA-labeled and unlabeled peptides (compare Fig. 3C and 3D). These results are consistent with two possibilities: either the surface(s) of TM-4 that mediate homooligomerization do not bind TM-1 and/or TM-3 or any potential TM-1/TM-4 or TM-3/TM-4 interactions are of lower affinity than TM-4/TM-4 self-associations. TM-4 Stoichiometry Determined by Fluorescence Energy Transfer—The stoichiometry of TM-4 oligomerization was tested in FRET experiments in SDS or PFO micelles (Fig. 4). In these assays, co-localization of the peptides within 30 Ä, indicative of peptide oligomerization, results in energy transfer and a stoichiometry-dependent quenching of the dansyl group fluorescence by the dabcyl acceptor group (27Adair B.D. Engelman D.M. Biochemistry. 1994; 33: 5539-5544Crossref PubMed Scopus (138) Google Scholar). Little fluorescence quenching is expected for peptides that are monomeric, a linear relationship between dansyl fluorescence yield and acceptor peptide concentration is expected for dimers, and a non-linear relationship is indicative of higher order oligomers (27Adair B.D. Engelman D.M. Biochemistry. 1994; 33: 5539-5544Crossref PubMed Scopus (138) Google Scholar). Within experimental error, no change in the total fluorescence yield between 450 and 650 nm from the dansyl-TM-1 peptide after excitation at 341 nm was observed upon titration with an increasing proportion of dabcyl-TM-1 in SDS (Fig. 4A) or PFO (Fig. 4B), confirming that TM-1 is monomeric and acting as a negative control for oligomerization in the FRET assay. The linear decrease observed in dansyl-TM-4 peptide fluorescence upon titration with acceptor peptide in SDS is diagnostic of dimer formation of and fits to a stoichiometry of 2.3 ± 0.1 units/oligomer with an R-value of 0.996 (Fig. 4A), whereas the non-linear decrease in dansyl-TM-4 fluorescence yield observed in PFO micelles fits to a stoichiometry of 3.2 ± 0.2 units/oligomer with an R-value of 0.998 (Fig. 4B). These PFO data mirror the 3:1 protein:substrate stoichiometry described for ligand binding by SMR proteins in vivo (30Muth T.R. Schuldiner S. EMBO J. 2000; 19: 234-240Crossref PubMed Scopus (168) Google Scholar) and interpreted as representing an equilibrium between dimers and tetramers (14Elbaz Y. Steiner-Mordoch S. Danieli T. Schuldiner S. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 1519-1524Crossref PubMed Scopus (125) Google Scholar). Formation of Tetramers by TM-4 Requires the Interaction of Two Helix Faces—For the TM-4 peptide to form oligomers with greater than dimeric stoichiometry, it must have at least two helix surfaces capable of intermolecular interactions as oligomerization cannot proliferate beyond a dimer if a single interaction-competent face exists on each TM-4 helix (Fig. 5A). However, if two helix surfaces interact in a pair-wise or symmetric manner with their counterparts, the stoichiometry of the resulting multimer is difficult to control (Fig. 5B). Such unchecked oligomerization, which would be visible on gels as “laddering” of the TM-4 peptide to oligomers of undefined size, is not observed in our experiments. Two sets of conditions exist in which TM-4 stoichiometry can be restricted to a discrete tetramer: (i) if symmetric interactions are not possible, and each interaction-competent face participates in asymmetric contacts (Fig. 5C) or (ii) if a limited combination of symmetric and asymmetric contacts is permitted (Fig. 5D). Our search for the contact site(s) of TM-4 was guided by evidence for symmetric TM-4/TM-4 contacts observed in oligomeric EmrE (29Soskine M. Steiner-Mordoch S. Schuldiner S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12043-12048Crossref PubMed Scopus (77) Google Scholar) on a surface that we have termed the “Ala-Val face” (Fig. 6). The role of this surface in TM-4 self-assembly was assessed with mutant peptides in which two Ala-Val face residues, Ala-92 and Val-95, as well as an adjacent residue, Val-85, were substituted singly and/or in combination with their corresponding residues from EmrE. Because the EmrE dimer is denatured in SDS (13Ninio S. Elbaz Y. Schuldiner S. FEBS Lett. 2004; 562: 193-196Crossref PubMed Scopus (48) Google Scholar), these replacements were expected to disrupt TM-4 dimerization. The resulting mutants, A92M, V95S, A92M/V95S, and V85L, had CD spectra essentially identical to WT TM-4 (data not shown), and all but V85L reduced the affinity of TM-4 dimers on SDS-PAGE gels, evidenced by the appearance of a monomeric species (Fig. 7A). Relative band intensities determined by densitometry measurements show that the A92M, V95S, and A92M/V95S mutants reduce dimerization to ∼50% of WT levels (Table 1). Each of these mutants also exhibited decreased or eliminated fluorescence quenching when compared with WT in FRET experiments (Fig. 7B). Not every Ala-Val face residue capable of bridging monomers (29Soskine M. Steiner-Mordoch S. Schuldiner S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12043-12048Crossref PubMed Scopus (77) Google Scholar) contributes to TM-4/TM-4 affinity, however; a peptide containing an Asn-102 to Ala replacement (Fig. 6) remained dimeric on SDS-PAGE (Fig. 7A) and in FRET experiments (Fig. 7B). As Asn residues are typical mediators of SDS-resistant helix-helix association (31Gratkowski H. Lear J.D. DeGrado W.F. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 880-885Crossref PubMed Scopus (291) Google Scholar, 32Zhou F.X. Merianos H.J. Brunger A.T. Engelman D.M. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2250-2255Crossref PubMed Scopus (319) Google Scholar), we hypothesized that the TM-4 dimer was stabilized via strong van der Waals, rather than electrostatic, interactions (33Johnson R.M. Heslop C.L. Deber C.M. Biochemistry. 2004; 43: 14361-14369Crossref PubMed Scopus (33) Google Scholar).FIGURE 7Homooligomerization of Hsmr TM-4 mutants in SDS. A, SDS-PAGE of WT and mutant TM-4 peptides. Gold and green asterisks indicate residue substitutions on the Ala-Val face and Gly face, respectively. Oligomeric states are indicated to the right. Although the V95S mutant is loaded at a concentration equivalent to the other peptides, it consistently displays reduced staining by Coomassie Blue in SDS gels (data not shown) and therefore appears as a less intense band. B,
Referência(s)