A Conserved Membrane-spanning Amino Acid Motif Drives Homomeric and Supports Heteromeric Assembly of Presynaptic SNARE Proteins
2000; Elsevier BV; Volume: 275; Issue: 23 Linguagem: Inglês
10.1074/jbc.m910092199
ISSN1083-351X
AutoresRico Laage, Jan Rohde, Bettina Brosig, Dieter Langosch,
Tópico(s)Neuroscience and Neuropharmacology Research
ResumoAssembly of the SNARE proteins synaptobrevin/VAMP, syntaxin, and SNAP-25 to binary and ternary complexes is important for docking and/or fusion of presynaptic vesicles to the neuronal plasma membrane prior to regulated neurotransmitter release. Despite the well characterized structure of their cytoplasmic assembly domains, little is known about the role of the transmembrane segments in SNARE protein assembly and function. Here, we identified conserved amino acid motifs within the transmembrane segments that are required for homodimerization of synaptobrevin II and syntaxin 1A. Minimal motifs of 6–8 residues grafted onto an otherwise monomeric oligoalanine host sequence were sufficient for self-interaction of both transmembrane segments in detergent solution or membranes. These motifs constitute contiguous areas of interfacial residues assuming α-helical secondary structures. Since the motifs are conserved, they also contributed to heterodimerization of synaptobrevin II and syntaxin 1A and therefore appear to constitute interaction domains independent of the cytoplasmic coiled coil regions. Interactions between the transmembrane segments may stabilize the SNARE complex, cause its multimerization to previously observed multimeric superstructures, and/or be required for the fusogenic activity of SNARE proteins. Assembly of the SNARE proteins synaptobrevin/VAMP, syntaxin, and SNAP-25 to binary and ternary complexes is important for docking and/or fusion of presynaptic vesicles to the neuronal plasma membrane prior to regulated neurotransmitter release. Despite the well characterized structure of their cytoplasmic assembly domains, little is known about the role of the transmembrane segments in SNARE protein assembly and function. Here, we identified conserved amino acid motifs within the transmembrane segments that are required for homodimerization of synaptobrevin II and syntaxin 1A. Minimal motifs of 6–8 residues grafted onto an otherwise monomeric oligoalanine host sequence were sufficient for self-interaction of both transmembrane segments in detergent solution or membranes. These motifs constitute contiguous areas of interfacial residues assuming α-helical secondary structures. Since the motifs are conserved, they also contributed to heterodimerization of synaptobrevin II and syntaxin 1A and therefore appear to constitute interaction domains independent of the cytoplasmic coiled coil regions. Interactions between the transmembrane segments may stabilize the SNARE complex, cause its multimerization to previously observed multimeric superstructures, and/or be required for the fusogenic activity of SNARE proteins. SNAP (soluble NSF (N-ethylmaleimide-sensitive factor) attachment protein) receptor 3-[(3-cholamidopropyl)dimethylammonio]-1-propane sulfonate disuccinimidyl suberate hemagglutinin maltose-binding protein polyacrylamide gel electrophoresis synaptosomal associated protein of 25 kDa transmembrane segment vesicle associated membrane protein wild type monoclonal antibody Intracellular membrane fusion events, e.g. constitutive organelle traffic or Ca2+-regulated neurotransmitter release, require conserved sets of membrane proteins, designated SNAREs.1 The best characterized SNAREs are those mediating exocytosis of synaptic vesicles in neurons (reviewed in Refs. 1.Hanson P.I. Heuser J.E. Jahn R. Curr. Opin. Neurobiol. 1997; 7: 310-315Crossref PubMed Scopus (334) Google Scholar, 2.Linial M. J. Neurochem. 1997; 69: 1781-1792Crossref PubMed Scopus (81) Google Scholar, 3.Jahn R. Hanson P.I. Nature. 1998; 393: 14-15Crossref PubMed Scopus (79) Google Scholar, 4.Rizo J. Südhof T.C. Nat. Struct. Biol. 1998; 5: 839-842Crossref PubMed Scopus (66) Google Scholar). In detergent extracts from presynaptic nerve terminals, the single-span integral membrane SNAREs synaptobrevin (also referred to as VAMP) and syntaxin together with the peripheral membrane SNARE protein SNAP-25 form a stable ternary complex that is disassembled in vitro after binding of soluble α-SNAP by the ATPase NSF (5.Söllner T. Bennett M.K. Whiteheard S.W. Scheller R.H. Rothman J.E. Cell. 1993; 75: 409-418Abstract Full Text PDF PubMed Scopus (1582) Google Scholar, 6.Söllner T. Whiteheart S.W. Brunner M. Erdjument-Bromage H. Geromanas S. Tempst P. Rothman J.E. Nature. 1993; 362: 318-324Crossref PubMed Scopus (2628) Google Scholar). According to the original SNARE hypothesis (7.Rothman J.E. Nature. 1994; 372: 55-63Crossref PubMed Scopus (2007) Google Scholar), interaction between these SNARE partners bridges opposing vesicular and plasma membranes. Therefore, assembly and disassembly of ternary SNARE complexes would proceed in atrans-configuration that is regarded essential for vesicle docking, priming and/or fusion (8.Fasshauer D. Bruns D. Shen B. Jahn R. Brünger A.T. J. Biol. Chem. 1997; 272: 4582-4590Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar, 9.Fiebig K.M. Rice L.M. Pollock E. Brünger A.T. Nat. Struct. Biol. 1999; 6: 117-123Crossref PubMed Scopus (236) Google Scholar, 10.Weber T. Zemelman B.V. McNew J.A. Westermann B. Gmachl M. Parlati F. Söllner T.H. Rothman J.E. Cell. 1998; 92: 759-772Abstract Full Text Full Text PDF PubMed Scopus (2015) Google Scholar, 11.Ungermann C. Sato K. Wickner W. Nature. 1998; 396: 543-548Crossref PubMed Scopus (279) Google Scholar, 12.Chen Y.A. Scales S.J. Patel S.M. Doung Y.-C. Scheller R.H. Cell. 1999; 97: 165-174Abstract Full Text Full Text PDF PubMed Scopus (359) Google Scholar). On the other hand, SNARE complexes are also found on the vesicular (13.Otto H. Hanson P.I. Jahn R. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 6197-6201Crossref PubMed Scopus (231) Google Scholar, 14.Nichols B.J. Ungermann C. Pelham H.R.B. Wickner W.T. Haas A. Nature. 1997; 387: 199-202Crossref PubMed Scopus (380) Google Scholar) as well as the plasma (15.Taubenblatt P. Dedieu J.C. Gulik-Krzywicki T. Morel N. J. Cell Sci. 1999; 112: 3559-3567Crossref PubMed Google Scholar) membrane in a cis-configuration, i.e.side by side. Protein domains involved in the binary and ternary interactions leading to SNARE complex formation have been originally identified by in vitro binding studies using recombinant soluble fragments of synaptobrevin II, syntaxin 1A, and SNAP-25 as follows: (i) the cytoplasmic domain of synaptobrevin II; (ii) a carboxyl-terminal, membrane-proximal region of syntaxin 1A; and (iii) carboxyl-terminal plus amino-terminal regions of SNAP-25 (16.Hayashi T. McMahon H. Yamasaki S. Binz T. Hata Y. Südhof T.C. Niemann H. EMBO J. 1994; 13: 5051-5061Crossref PubMed Scopus (662) Google Scholar, 17.Calakos N. Bennett M.K. Peterson K.E. Scheller R.H. Science. 1994; 263: 1146-1149Crossref PubMed Scopus (366) Google Scholar, 18.Kee Y. Lin R.C. Hsu S.-C. Scheller R.H. Neuron. 1995; 14: 991-998Abstract Full Text PDF PubMed Scopus (191) Google Scholar, 19.Hao J.C. Salem N. Peng X.R. Kelly R.B. Bennett M.K. J. Neurosci. 1997; 17: 1596-1603Crossref PubMed Google Scholar, 20.Poirier M.A. Hao J.C. Malkus P.N. Chan C. Moore M.F. King D.S. Bennett M.K. J. Biol. Chem. 1998; 273: 11370-11377Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar, 21.Fasshauer D. Eliason W.K. Brunger A.T. Jahn R. Biochemistry. 1998; 37: 10354-10362Crossref PubMed Scopus (208) Google Scholar). More recently, structural studies confirmed that previously predicted cytoplasmic coiled coil domains of synaptobrevin II (H1, H2) and syntaxin 1A (H3) form a tightly packed parallel four-helical bundle together with two helices derived from SNAP-25 (HA and HB) in the SNARE complex (22.Sutton R.B. Fasshauer D. Jahn R. Brunger A.T. Nature. 1998; 395: 347-353Crossref PubMed Scopus (1916) Google Scholar, 23.Poirier M.A. Xiao W.Z. Macosko J.C. Chan C. Shin Y.K. Bennett M.K. Nat. Struct. Biol. 1998; 5: 765-769Crossref PubMed Scopus (418) Google Scholar). In contrast, the role of the carboxyl-terminal TMSs located at one side of the SNARE complex (24.Hanson P.I. Roth R. Morisaki H. Jahn R. Heuser J.E. Cell. 1997; 90: 523-535Abstract Full Text Full Text PDF PubMed Scopus (670) Google Scholar, 25.Lin R.C. Scheller R.H. Neuron. 1997; 19: 1087-1094Abstract Full Text Full Text PDF PubMed Scopus (274) Google Scholar, 26.Hohl T.M. Parlati F. Wimmer C. Rothman J.E. Sollner T.H. Engelhardt H. Mol. Cell. 1998; 2: 539-548Abstract Full Text Full Text PDF PubMed Scopus (107) Google Scholar) in these interactions has only been characterized in part. Transmembrane domains are known to participate in oligomerization of many different integral membrane proteins. It is thought that TMS self-assembly is driven by a close packing of their characteristically shaped surfaces that are defined by their primary structures (reviewed in Ref. 27.Lemmon M.A. Engelman D.M. Q. Rev. Biophys. 1994; 27: 157-218Crossref PubMed Scopus (177) Google Scholar). We have previously shown that the homodimeric structure originally observed for native synaptobrevin (28.Washbourne P. Schiavo G. Montecucco C. Biochem. J. 1995; 305: 721-724Crossref PubMed Scopus (110) Google Scholar, 29.Calakos N. Scheller R.H. J. Biol. Chem. 1994; 269: 24534-24537Abstract Full Text PDF PubMed Google Scholar, 30.Edelman L. Hanson P.I. Chapman E.R. Jahn R. EMBO J. 1995; 14: 224-231Crossref PubMed Scopus (390) Google Scholar) is preserved with recombinant synaptobrevin II, where it depends on a specific amino acid motif within the TMS (31.Laage R. Langosch D. Eur. J. Biochem. 1997; 249: 540-546Crossref PubMed Scopus (73) Google Scholar). Furthermore, a direct interaction of syntaxin and synaptobrevin TMSs in synthetic proteoliposomes was recently reported (32.Margittai M. Otto H. Jahn R. FEBS Lett. 1999; 446: 40-44Crossref PubMed Scopus (76) Google Scholar). Surprisingly, we found that the synaptobrevin II homodimerization motif is conserved within the TMS of syntaxin 1A. This predicted the involvement of this motif in a homotypic interaction of syntaxin 1A as well as in its heterophilic binding to synaptobrevin II. Here, we (i) identify the minimal TMS amino acid motif required for synaptobrevin II homodimerization, (ii) show that a similar motif mediates syntaxin 1A homodimerization, and (iii) demonstrate that TMSs and cytoplasmic coiled coil domains cooperate in synaptobrevin/syntaxin heterodimerization. Rat syntaxin 1A cDNA was amplified by polymerase chain reaction with VENT polymerase (Biolabs) from a plasmid template as described previously for rat synaptobrevin II (31.Laage R. Langosch D. Eur. J. Biochem. 1997; 249: 540-546Crossref PubMed Scopus (73) Google Scholar) with primers containing NheI (sense primer) andBglII (antisense primer) restriction sites. The amplified fragment was cut with NheI and BglII and ligated into the pET 21d (Novagen)-based plasmid pSNiR, previously cut withNheI and BamHI. In the resulting constructs, the amino termini of synaptobrevin or syntaxin are fused in frame to the carboxyl terminus of a tripartite fusion moiety consisting of the coding sequence of the HA marker epitope in case of synaptobrevin or the c-myc epitope for syntaxin, Staphylococcus aureusnuclease A, and a linker-region coding for 9 amino acids. Construction of plasmid pToxR-A16 was described previously (33.Gurezka R. Laage R. Brosig B. Langosch D. J. Biol. Chem. 1999; 274: 9265-9270Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar). All other pToxR constructs were made by ligating synthetic oligonucleotide cassettes encoding the desired sequences into plasmid pHKToxR(TMIl4)MalE (34.Kolmar H. Hennecke F. Götze K. Janzer B. Vogt B. Mayer F. Fritz H.-J. EMBO J. 1995; 14: 3895-3904Crossref PubMed Scopus (64) Google Scholar) previously cut with NheI and BamHI. All constructs were verified by dideoxy sequencing. All mutants were made by oligonucleotide-directed mutagenesis performed according to Kunkelet al. (35.Kunkel T.A. Roberts J.D. Zakour R.A. Methods Enzymol. 1987; 154: 367-382Crossref PubMed Scopus (4558) Google Scholar) on single-stranded templates (T7 Mutagene-kit, Bio-Rad). All mutations were verified by dideoxy sequencing. ToxR activities were determined in quadruplicates in 3–11 independent experiments as described (33.Gurezka R. Laage R. Brosig B. Langosch D. J. Biol. Chem. 1999; 274: 9265-9270Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar) and are given in Miller units (mean ± S.E.). Western blot analysis was done as described (33.Gurezka R. Laage R. Brosig B. Langosch D. J. Biol. Chem. 1999; 274: 9265-9270Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar). PD-28 cell growth assays were done as described (36.Brosig B. Langosch D. Protein Sci. 1998; 7: 1052-1056Crossref PubMed Scopus (196) Google Scholar), and the absorbance was read at 650 nm at a pathlength of 6 mm. To correct for slightly different membrane insertion efficiencies, the β-galactosidase activities elicited by the mutant sequences were normalized to the A 650 values obtained after 48 h of cell growth. Proteins, encoded by pSNiR vectors, were expressed inEscherichia coli BL21(DE3)pLsyS (Novagen) as described (31.Laage R. Langosch D. Eur. J. Biochem. 1997; 249: 540-546Crossref PubMed Scopus (73) Google Scholar), solubilized with 2% (v/v) polyoxyethylene 9 lauryl ether (Sigma) for mild SDS-PAGE, or with 2% (w/v) CHAPS (Applichem, Darmstadt) in solubilization buffer (50 mm Hepes, pH 7.9, 1 mNaCl, 1 mm EDTA, 1 mm phenylmethylsulfonyl fluoride) for all other applications. The usual concentration of recombinant protein after solubilization was ∼1 mg/ml. Biosynthetic labeling was performed by addition of 7.5 μCi of [35S]methionine/cysteine (2:1) (Amersham Pharmacia Biotech) per ml of bacterial culture 15 min after induction of expression with isopropyl-β-d-thiogalactopyranoside. CHAPS-solubilized proteins (see above) were treated with DSS (Pierce) dissolved in dimethyl sulfoxide. Dimethyl sulfoxide never exceeded 2% (v/v) of the reaction volume. The samples were shaken for 5 min at room temperature, and reactions were then quenched with 100 mm Tris, pH 7.4, for 10 min. The samples were precipitated with methanol (37.Wessel D. Flügge U.I. Anal. Biochem. 1984; 138: 141-143Crossref PubMed Scopus (3170) Google Scholar) prior to SDS-PAGE. CHAPS-solubilized proteins were separated by SDS-PAGE (∼20 μg/lane) and subsequently blotted onto nitrocellulose membrane (Amersham Pharmacia Biotech Hybond-C pure). The membrane was blocked for 30 min in TBB (50 mm Tris-Cl, pH 8.0, 150 mm NaCl, 0.1% (v/v) Triton X-100, 1 mm EDTA, 4% (w/v) nonfat dry milk powder). The blocking solution was replaced by 15 μg/ml 35S-labeled wt protein of the cognate SNARE partner in TBB while shaking was continued for 2 h at room temperature. The blots were washed twice for 10 min in TBS (50 mm Tris-Cl, pH 8.0, 150 mm NaCl) and once in TWB (TBS + 0.1% (v/v) Triton X-100). Blots were air-dried and exposed to a BAS-MP imaging plate (Fuji, Japan) overnight, and bound35S-labeled protein was quantified using a BAS-1000 bio-imaging analyzer (Fuji, Japan). CHAPS-solubilized HA-tagged synaptobrevin II derivatives were incubated with wt35S-labeled syntaxin (final concentration 250 ng/μl) in solubilization buffer at 4 °C for 2 h. The monoclonal anti-HA antibody 12CA5 (a kind gift of Dr. Lerner, Scripps Institute, San Diego, CA) was added at 20 ng/μl, and incubation was continued with shaking overnight at 4 °C. Protein A-Sepharose CL-4B (Amersham Pharmacia Biotech) was added, pelleted (Hereaus Biofuge 4000 rpm) after 2 h, washed twice with buffer A (10 mm Tris-Cl, pH 7.5, 150 mm NaCl, 1 mm EDTA, 0.2% (v/v) Triton X-100), twice with buffer B (10 mm Tris-Cl, pH 7.5, 500 mm NaCl, 1 mm EDTA, 0.2% (v/v) Triton X-100), and once with buffer C (10 mm Tris-Cl, pH 7.5, 0.2% (v/v) Triton X-100). After washing, bound proteins were eluted with SDS sample buffer and separated by SDS-PAGE. Gels were dried and exposed to a BAS-MP imaging plate (Fuji, Japan), and the amount of co-precipitated radiolabeled syntaxin was quantified using a BAS-1000 bio-imaging analyzer (Fuji, Japan). Proteins solubilized with 2% polyoxyethylene 9 lauryl ether were precipitated with methanol (37.Wessel D. Flügge U.I. Anal. Biochem. 1984; 138: 141-143Crossref PubMed Scopus (3170) Google Scholar), redissolved in SDS sample buffer (50 mm Tris-HCl, pH 6.8, 1% (w/v) SDS, 6 m urea, 50 mmdithiothreitol, 20% (w/v) sucrose), separated by SDS-PAGE (5 μg/lane), and visualized by Western blotting using the HA-mAb 12CA5 for HA-tagged synaptobrevin or the myc-mAb 9E10 for myc-tagged syntaxin 1A as described (31.Laage R. Langosch D. Eur. J. Biochem. 1997; 249: 540-546Crossref PubMed Scopus (73) Google Scholar). Minigels were run at 4 °C, and samples were not boiled prior to electrophoresis. We previously described an amino acid motif in the synaptobrevin II TMS that is essential for its homodimerization. Mutational analysis had identified 6 amino acid residues (Leu-99, Ile-102, Cys-103, Leu-107, Ile-110, and Ile-111) whose exchange to alanine significantly reduced self-interaction of the recombinant proteins expressed in E. coli as analyzed by SDS-PAGE under mild conditions. Simultaneous mutation of three of these residues (L99A/C103A/I111A = syb-mult) abrogated homodimerization almost completely (31.Laage R. Langosch D. Eur. J. Biochem. 1997; 249: 540-546Crossref PubMed Scopus (73) Google Scholar) (Figs. 1 A and2 A). To examine whether this motif is sufficient for synaptobrevin II TMS-TMS interaction, we compared a set of synaptobrevin II mutants for self-interaction by mild SDS-PAGE. Replacing the TMS by an oligoalanine sequence (syb-A15), which does not self-interact (33.Gurezka R. Laage R. Brosig B. Langosch D. J. Biol. Chem. 1999; 274: 9265-9270Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar), reduced homodimerization as efficiently as the previously characterized "mult" mutations (Fig.2 A). Grafting the critical 6 residues onto this oligoalanine background (syb-A8) completely restored homodimerization thus identifying them as minimal amino acid motifs responsible for synaptobrevin II TMS-TMS interaction. As mutation of alanine 104 previously did not influence homodimerization (31.Laage R. Langosch D. Eur. J. Biochem. 1997; 249: 540-546Crossref PubMed Scopus (73) Google Scholar), this residue is not considered as part of this motif.Figure 2Homodimerization of synaptobrevin II and syntaxin 1A. Wild type and mutated full-length proteins were separated by SDS-PAGE and visualized by Western blotting. A,the 66-kDa homodimer of syb-wt (synaptobrevin/nuclease A fusion = 33-kDa monomer) is disrupted upon different TMS mutations (syb-A15 and syb-mult) but preserved with a 6-residue TMS motif (syb-A8).B, the 100-kDa homodimer of syx-wt (syntaxin/nuclease A fusion = 50-kDa monomer) is disrupted upon TMS-deletion (syx-cyt) or mutation (syx-A15 and syx-mult) but appears even stronger with a conserved 6-residue TMS motif (syx-A8).View Large Image Figure ViewerDownload Hi-res image Download (PPT) A sequence comparison identified 5 (Ile-270, Cys-271, Leu-275, Ile-278, and Ile-279) out of the 6 positions to be conserved in the TMS of syntaxin 1A, the natural binding partner of synaptobrevin II (Fig.1 A). Therefore, we examined the potential role of this motif in syntaxin 1A homodimerization. Analysis of recombinant syntaxin 1A by SDS-PAGE under mild conditions indeed revealed its partial homodimerization. Dimer formation was almost completely abrogated when the TMS had been deleted (syx-cyt); its central 15 hydrophobic amino acids were replaced by the oligoalanine sequence (syx-A15) or mutated in three positions (syx-mult = M267A/C271A/I279A). Grafting the motif homologous to the synaptobrevin II homodimerization motif (Met-267, Ile-270, Cys-271, Leu-275, Ile-278, and Ile-279) onto the oligoalanine sequence (syx-A8) resulted in syntaxin homodimerization that was even stronger than that of the wt protein (Fig.2 B). Thus, syntaxin 1A also forms homodimers by sequence-specific self-interaction of its TMS based on an amino acid motif almost identical to that previously identified in synaptobrevin II. To examine whether the cytoplasmic coiled coil domains of synaptobrevin and syntaxin are involved in homodimerization, multiple mutations were made in positions relevant for binary and ternary SNARE protein interactions (Fig. 1 C) (Refs. 18.Kee Y. Lin R.C. Hsu S.-C. Scheller R.H. Neuron. 1995; 14: 991-998Abstract Full Text PDF PubMed Scopus (191) Google Scholar, 19.Hao J.C. Salem N. Peng X.R. Kelly R.B. Bennett M.K. J. Neurosci. 1997; 17: 1596-1603Crossref PubMed Google Scholar, 23.Poirier M.A. Xiao W.Z. Macosko J.C. Chan C. Shin Y.K. Bennett M.K. Nat. Struct. Biol. 1998; 5: 765-769Crossref PubMed Scopus (418) Google Scholar, and see below). These mutations had no detectable effect on homodimerization of both proteins, indicating that the cytoplasmic coiled coil domains do not self-interact (results not shown, but see Fig. 4). To examine self-interaction of the SNARE TMSs in the absence of the cytoplasmic domains and incorporated into membranes, we used the ToxR transcription activator system. We previously established this system as a sensitive tool to study TMS-TMS interactions using the structurally well characterized glycophorin A TMS dimer for reference (36.Brosig B. Langosch D. Protein Sci. 1998; 7: 1052-1056Crossref PubMed Scopus (196) Google Scholar, 38.Langosch D.L. Brosig B. Kolmar H. Fritz H.-J. J. Mol. Biol. 1996; 263: 525-530Crossref PubMed Scopus (220) Google Scholar). The ToxR protein is anchored by a single TMS of choice within the inner membrane of expressing E. coli cells where it is thought to exist in a monomer/dimer equilibrium. The dimeric form binds to the cholera toxin promoter thus activating expression of a downstreamlacZ gene in a reporter strain (Fig.3 A). β-Galactosidase expression is therefore diagnostic of ToxR self-assembly in the membrane. Here, we replaced the ToxR TMS by the synaptobrevin II or syntaxin 1A TMSs to study their self-interaction. The transcriptional activities of these chimeric ToxR proteins indicated that both SNARE TMSs self-interacted similarly well in the membrane. The degrees of interaction were comparable to that of a previously characterized membrane-spanning leucine zipper (33.Gurezka R. Laage R. Brosig B. Langosch D. J. Biol. Chem. 1999; 274: 9265-9270Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar) and below that of the glycophorin A TMS (36.Brosig B. Langosch D. Protein Sci. 1998; 7: 1052-1056Crossref PubMed Scopus (196) Google Scholar, 38.Langosch D.L. Brosig B. Kolmar H. Fritz H.-J. J. Mol. Biol. 1996; 263: 525-530Crossref PubMed Scopus (220) Google Scholar). Since the signals were reduced by the mult mutations to statistically significant degrees (Fig. 3 B; two-tailed Student's t test, p < 0.05), the interactions are sequence-specific and involve the same faces of the transmembrane helices as determined for full-length proteins in detergent solution. The weaker effects of the mult mutations found here as compared with detergent solution is assumed to result from higher protein concentrations and/or preorientation of the interacting domains in the membrane (39.Grasberger B. Minton A.P. DeLisi C. Metzger H. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6258-6262Crossref PubMed Scopus (191) Google Scholar). The minimal interaction motifs were determined using an oligoalanine host sequence (A16) previously shown to partially partition into the membrane where it stays largely monomeric (33.Gurezka R. Laage R. Brosig B. Langosch D. J. Biol. Chem. 1999; 274: 9265-9270Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar). The 6-residue motifs, which were sufficient for homodimerization of full-length SNAREs in detergent (Fig. 2), partially restored self-interaction in the membrane. To obtain wild-type levels of homodimerization, however, these motifs had to be expanded by two additional conserved isoleucine residues completing the contiguous areas of interfacial residues (Figs. 3 B and 7). To exclude that different concentrations of the ToxR chimeric proteins in the membranes distorted the signals, we ascertained their similar expression levels by Western blot analysis (Fig. 3 C). Furthermore, we assessed the efficiency of the ToxR constructs to integrate into the inner bacterial membrane by testing their ability to functionally complement the maltose-binding protein (MalE) deficiency of PD28 cells. Due to a deletion in MalE, this strain is unable to grow in minimal medium with maltose as the only carbon source (40.Bedouelle H. Duplay P. Eur. J. Biochem. 1988; 171: 541-549Crossref PubMed Scopus (113) Google Scholar). In cells expressing correctly inserted ToxR membrane proteins, however, the MalE domain allows maltose uptake and thus cell growth (36.Brosig B. Langosch D. Protein Sci. 1998; 7: 1052-1056Crossref PubMed Scopus (196) Google Scholar). All constructs complemented MalE deficiency to comparable degrees, thus confirming their similarly efficient membrane integration; a ToxR protein with deleted TMS (ΔTM) did not support cell growth (Fig. 3 D) as expected (36.Brosig B. Langosch D. Protein Sci. 1998; 7: 1052-1056Crossref PubMed Scopus (196) Google Scholar). In sum, both SNARE TMSs are capable of self-assembling in membranes in the absence of the cytoplasmic domains. Conserved motifs of 8 amino acids are sufficient to mimic these homotypic interactions. Conservation of the self-interacting TMS motifs suggested that TMS-TMS interactions may also be important for heterodimerization of both proteins. This is in line with a recently reported contribution of the TMSs to synaptobrevin/syntaxin interaction (32.Margittai M. Otto H. Jahn R. FEBS Lett. 1999; 446: 40-44Crossref PubMed Scopus (76) Google Scholar). On the other hand, synaptobrevin II and syntaxin 1A are known to interact via cytoplasmic coiled coil domains in binary as well as in ternary complexes including SNAP-25 (16.Hayashi T. McMahon H. Yamasaki S. Binz T. Hata Y. Südhof T.C. Niemann H. EMBO J. 1994; 13: 5051-5061Crossref PubMed Scopus (662) Google Scholar, 18.Kee Y. Lin R.C. Hsu S.-C. Scheller R.H. Neuron. 1995; 14: 991-998Abstract Full Text PDF PubMed Scopus (191) Google Scholar, 19.Hao J.C. Salem N. Peng X.R. Kelly R.B. Bennett M.K. J. Neurosci. 1997; 17: 1596-1603Crossref PubMed Google Scholar, 20.Poirier M.A. Hao J.C. Malkus P.N. Chan C. Moore M.F. King D.S. Bennett M.K. J. Biol. Chem. 1998; 273: 11370-11377Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar, 22.Sutton R.B. Fasshauer D. Jahn R. Brunger A.T. Nature. 1998; 395: 347-353Crossref PubMed Scopus (1916) Google Scholar, 23.Poirier M.A. Xiao W.Z. Macosko J.C. Chan C. Shin Y.K. Bennett M.K. Nat. Struct. Biol. 1998; 5: 765-769Crossref PubMed Scopus (418) Google Scholar). The hallmark of coiled coil structures is that the a and d positions within repeatedabcdefg motifs form the hydrophobic core of the helix-helix interfaces. Mutating a and d positions of the H3 region of syntaxin previously resulted in loss of ternary as well as binary interactions (18.Kee Y. Lin R.C. Hsu S.-C. Scheller R.H. Neuron. 1995; 14: 991-998Abstract Full Text PDF PubMed Scopus (191) Google Scholar). To compare the contribution of the cytoplasmic coiled coil domains and the TMSs to heterodimerization, we generated mutations in either part of synaptobrevin II and syntaxin 1A. To test these proteins (tagged with HA or myc epitopes, respectively) for their ability to form homo- and heterodimers, they were co-incubated, cross-linked with DSS, and analyzed by SDS-PAGE followed by immunoblotting. Wild-type proteins formed homodimers plus an additional band of intermediate apparent molecular weight that was detected with both antibodies and thus identified as the heterodimer. This result suggests that homodimerization competes with heterodimerization. Mutating fivea and d positions in the coiled coil domains of synaptobrevin (syb-60/84) or syntaxin (syx-230/251) (Figs.1 C and 7) strongly reduced heterodimerization, as expected (Fig. 4). In comparison, the TMS mult mutations had a detectable but somewhat less pronounced effect on heterodimerization as assayed by DSS cross-linking or SDS-PAGE analysis under mild conditions (data not shown). To evaluate quantitatively the roles of coiled coil domains and TMSs in heterophilic interaction, we developed an overlay assay. Equal amounts of wt and mutant proteins were separated by SDS-PAGE, blotted onto nitrocellulose, and probed with 35S-labeled wt-binding partners. By quantifying bound radioactivity, different degrees of heterodimerization were determined for wt and mutants. The results confirmed the importance of both coiled coil domains for heterophilic binding (Fig.5 A). Importantly, the multiple TMS mutations (syb-mult and syx-mult) or replacement of the TMSs by the oligoalanine sequence (syb-A15 and syx-A15) also significantly decreased binding to their respective wt SNARE partner (Fig. 5). To identify the minimal amino acid motifs responsible for heterophilic TMS-TMS interaction, we tested the proteins with the minimal homodimerization motifs. In case of syntaxin 1A, the motif of 6 residues (syx-A8) was sufficient for wt level of heterodimerization with wt synaptobrevin II, whereas the 8-residue motif that completely restored homodimerization in the ToxR system (syb-A6) was required in the reverse configuration. By using an independent experimental approach, we investigated the influence of the mutations on the ability of synaptobrevin to co-precipitate 35S-labeled wt syntaxin from detergent solution. Upon co-incubation, we immunoprecipitated the synaptobrevin proteins and quantitated co-precipitated syntaxin upon SDS-PAGE. In agreement with the overlay assay (Fig. 5), the coiled coil mutant syb-60/84 as well
Referência(s)