Portal de Periódicos da CAPES

A Cryptic Rab1-binding Site in the p115 TetheringProtein

2005; Elsevier BV; Volume: 280; Issue: 27 Linguagem: Inglês

10.1074/jbc.m503925200

1083-351X

Matthew Beard, Ayano Satoh, James Shorter, Graham Warren,

Lipid Membrane Structure and Behavior

Small GTPases and coiled-coil proteins of the golgin family help to tetherCOPI vesicles to Golgi membranes. At the cis-side of the Golgi, theRab1 GTPase binds directly to each of three coiled-coil proteins: p115, GM130,and as now shown, Giantin. Rab1 binds to a coiled-coil region within the taildomain of p115 and this binding is inhibited by the C-terminal, acidic domainof p115. Furthermore, GM130 and Giantin bind to the acidic domain of p115 andstimulate p115 binding to Rab1, suggesting that p115 binding to Rab1 isregulated. Regulation of this interaction by proteins such as GM130 andGiantin may control the membrane recruitment of p115 by Rab1. Small GTPases and coiled-coil proteins of the golgin family help to tetherCOPI vesicles to Golgi membranes. At the cis-side of the Golgi, theRab1 GTPase binds directly to each of three coiled-coil proteins: p115, GM130,and as now shown, Giantin. Rab1 binds to a coiled-coil region within the taildomain of p115 and this binding is inhibited by the C-terminal, acidic domainof p115. Furthermore, GM130 and Giantin bind to the acidic domain of p115 andstimulate p115 binding to Rab1, suggesting that p115 binding to Rab1 isregulated. Regulation of this interaction by proteins such as GM130 andGiantin may control the membrane recruitment of p115 by Rab1. Targeting of transport vesicles to the correct membrane compartment is amultilayered process consisting of tethering, docking, and fusion.SNARE 1The abbreviations used are: SNARE, soluble NSF attachment protein receptors(where NSF is N-ethylmaleimide-sensitive factor); ER, endoplasmicreticulum; MBP, maltose-binding protein; GST, glutathioneS-transferase; DTT, dithiothreitol; PI, protease inhibitor;GTPγS, guanosine 5′-O-(thiotriphosphate).1The abbreviations used are: SNARE, soluble NSF attachment protein receptors(where NSF is N-ethylmaleimide-sensitive factor); ER, endoplasmicreticulum; MBP, maltose-binding protein; GST, glutathioneS-transferase; DTT, dithiothreitol; PI, protease inhibitor;GTPγS, guanosine 5′-O-(thiotriphosphate). proteins arethe best characterized components of the docking and fusion machinery. CognateSNARE pairs are thought to provide the core specificity for membrane fusion(1Rothman J.E. Nature. 1994; 372: 55-63Crossref PubMed Scopus (1995) Google Scholar, 2Mellman I. Warren G. Cell. 2000; 100: 99-112Abstract Full Text Full Text PDF PubMed Scopus (355) Google Scholar, 3Mayer A. Trends Biochem.Sci. 2001; 26: 717-723Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar).Tethering components act before SNAREs and are thought to provide an initialinteraction between a vesicle and target membrane(4Waters M.G. Hughson F.M. Traffic. 2000; 1: 588-597Crossref PubMed Scopus (91) Google Scholar, 5Lowe M. Curr. Biol. 2000; 10: R407-R409Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar, 6Pfeffer S. Mol. Cell. 2001; 8: 729-730Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar).Tethering is essential for transport and is mediated by a diverse array ofproteins including: GTPases of the Ypt/Rab and Arl families, coiled-coilproteins that can link membranes together, and large multiprotein assembliesrecently termed quatrefoil tethering complexes(7Zerial M. McBride H. Nat.Rev. Mol. Cell. Biol. 2001; 2: 107-117Crossref PubMed Scopus (2681) Google Scholar, 8Barr F.A. Short B. Curr.Opin. Cell. Biol. 2003; 15: 405-413Crossref PubMed Scopus (213) Google Scholar, 9Gillingham A.K. Munro S. Biochim. Biophys. Acta. 2003; 1641: 71-85Crossref PubMed Scopus (179) Google Scholar, 10Jackson C.L. Curr.Biol. 2003; 13: R174-R176Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar).At the entry face of the Golgi apparatus, multiprotein complexes include TRAPPand COG (11Sacher M. Jiang Y. Barrowman J. Scarpa A. Burston J. Zhang L. Schieltz D. Yates III, J.R. Abeliovich H. Ferro-Novick S. EMBO J. 1998; 17: 2494-2503Crossref PubMed Scopus (234) Google Scholar,12Whyte J.R. Munro S. J. CellSci. 2002; 115: 2627-2637Google Scholar), the Rab family GTPasesare Rabs 1 and 2 (13Tisdale E.J. Bourne J.R. Khosravi-Far R. Der C.J. Balch W.E. J. Cell Biol. 1992; 119: 749-761Crossref PubMed Scopus (417) Google Scholar), and thecoiled-coil proteins are p115, Giantin, and GM130. The latter are members ofthe golgin protein family, initially identified as antigens in certainautoimmune diseases (14Seelig H.P. Schranz P. Schroter H. Wiemann C. Renz M. J. Autoimmun. 1994; 7: 67-91Crossref PubMed Scopus (59) Google Scholar). Theprecise roles of tethering proteins, their mechanisms of action, and how theyinteract, however, remain unclear.The best characterized Golgi tethering proteins are mammalian p115 and itsyeast homologue Uso1p. These proteins are myosin-shaped, homodimericmolecules, each polypeptide of which comprises an N-terminal globular head, acoiled-coil tail, and a short C-terminal acidic domain(15Waters M.G. Clary D.O. Rothman J.E. J. Cell Biol. 1992; 118: 1015-1026Crossref PubMed Scopus (202) Google Scholar, 16Sztul E. Colombo M. Stahl P. Samanta R. J. Biol. Chem. 1993; 268: 1876-1885Abstract Full Text PDF PubMed Google Scholar, 17Sapperstein S.K. Walter D.M. Grosvenor A.R. Heuser J.E. Waters M.G. Proc. Natl. Acad. Sci. U. S.A. 1995; 92: 522-526Crossref PubMed Scopus (171) Google Scholar, 18Yamakawa H. Seog D.H. Yoda K. Yamasaki M. Wakabayashi T. J. Struct. Biol. 1996; 116: 356-365Crossref PubMed Scopus (64) Google Scholar).Uso1p is essential for exocytic transport(19Nakajima H. Hirata A. Ogawa Y. Yonehara T. Yoda K. Yamasaki M. J. Cell Biol. 1991; 113: 245-260Crossref PubMed Scopus (137) Google Scholar) and tethers COPIIvesicles to Golgi membranes in yeast(20Sapperstein S.K. Lupashin V.V. Schmitt H.D. Waters M.G. J. Cell Biol. 1996; 132: 755-767Crossref PubMed Scopus (158) Google Scholar, 21Barlowe C. J. CellBiol. 1997; 139: 1097-1108Crossref PubMed Scopus (155) Google Scholar, 22Cao X. Ballew N. Barlowe C. EMBO J. 1998; 17: 2156-2165Crossref PubMed Scopus (291) Google Scholar).p115 is essential for both exocytic transport and maintenance of the stackedstructure of mammalian Golgi membranes(23Puthenveedu M.A. Linstedt A.D. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 1253-1256Crossref PubMed Scopus (85) Google Scholar). It acts during ER toGolgi and intra-Golgi transport, as well as post-mitotic Golgi reassembly(15Waters M.G. Clary D.O. Rothman J.E. J. Cell Biol. 1992; 118: 1015-1026Crossref PubMed Scopus (202) Google Scholar,24Alvarez C. Garcia-Mata R. Hauri H.P. Sztul E. J. Biol. Chem. 2001; 276: 2693-2700Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar). p115 tethers COPIvesicles to Golgi membranes(25Sönnichsen B. Lowe M. Levine T. Jamsa E. Dirac-Svejstrup B. Warren G. J. CellBiol. 1998; 140: 1013-1021Crossref PubMed Scopus (251) Google Scholar). These data suggest thatp115/Uso1p functions to tether vesicles, although the molecular mechanism isstill unclear.Our working model for the mechanism by which p115 tethers COPI vesicles toGolgi membranes has been that it forms a “bridge,” simultaneouslybinding and linking Giantin in COPI vesicle membranes to GM130 on the Golgi(25Sönnichsen B. Lowe M. Levine T. Jamsa E. Dirac-Svejstrup B. Warren G. J. CellBiol. 1998; 140: 1013-1021Crossref PubMed Scopus (251) Google Scholar). This model for acis-golgin tethering complex arose from two important ideas: 1) GM130acts as the Golgi membrane anchor for p115; 2) p115 (anchored by GM130)tethers by simultaneously binding to Giantin in vesicle membranes.The first idea originated from studies into the mitotic disassembly ofmammalian Golgi stacks. GM130 is a mitotically regulated, p115-binding proteinpresent in highly purified Golgi membranes(26Nakamura N. Rabouille C. Watson R. Nilsson T. Hui N. Slusarewicz P. Kreis T.E. Warren G. J. Cell Biol. 1995; 131: 1715-1726Crossref PubMed Scopus (666) Google Scholar, 27Nakamura N. Lowe M. Levine T.P. Rabouille C. Warren G. Cell. 1997; 89: 445-455Abstract Full Text Full Text PDF PubMed Scopus (341) Google Scholar, 28Lowe M. Rabouille C. Nakamura N. Watson R. Jackman M. Jamsa E. Rahman D. Pappin D.J. Warren G. Cell. 1998; 94: 783-793Abstract Full Text Full Text PDF PubMed Scopus (255) Google Scholar, 29Lowe M. Gonatas N.K. Warren G. J. Cell Biol. 2000; 149: 341-356Crossref PubMed Scopus (126) Google Scholar).The N-terminal domain of GM130 binds to the acidic C-terminal domain of p115(27Nakamura N. Lowe M. Levine T.P. Rabouille C. Warren G. Cell. 1997; 89: 445-455Abstract Full Text Full Text PDF PubMed Scopus (341) Google Scholar,30Lesa G.M. Seemann J. Shorter J. Vandekerckhove J. Warren G. J. Biol. Chem. 2000; 275: 2831-2836Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar, 31Linstedt A.D. Jesch S.A. Mehta A. Lee T.H. Garcia-Mata R. Nelson D.S. Sztul E. J. Biol.Chem. 2000; 275: 10196-10201Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar, 32Dirac-Svejstrup A.B. Shorter J. Waters M.G. Warren G. J. Cell Biol. 2000; 150: 475-488Crossref PubMed Scopus (63) Google Scholar).GM130 is regulated by the mitotic kinase, CDK1/cyclin B(28Lowe M. Rabouille C. Nakamura N. Watson R. Jackman M. Jamsa E. Rahman D. Pappin D.J. Warren G. Cell. 1998; 94: 783-793Abstract Full Text Full Text PDF PubMed Scopus (255) Google Scholar). CDK1/cyclin B-mediatedphosphorylation of GM130 inhibits GM130 binding to p115(27Nakamura N. Lowe M. Levine T.P. Rabouille C. Warren G. Cell. 1997; 89: 445-455Abstract Full Text Full Text PDF PubMed Scopus (341) Google Scholar), and this correlates withthe inhibition of p115 binding to Golgi membranes in vivo(33Shima D.T. Haldar K. Pepperkok R. Watson R. Warren G. J. Cell Biol. 1997; 137: 1211-1228Crossref PubMed Scopus (194) Google Scholar) and in vitro(34Levine T.P. Rabouille C. Kieckbusch R.H. Warren G. J. Biol. Chem. 1996; 271: 17304-17311Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). Furthermore, p115localization to the Golgi region of cells is disrupted either bymicroinjection of a GM130 N-terminal peptide (N73), which inhibits p115binding to GM130, or by overexpression of a truncated GM130 that lacks thep115-binding domain (35Seemann J. Jokitalo E.J. Warren G. Mol. Biol. Cell. 2000; 11: 635-645Crossref PubMed Scopus (153) Google Scholar).The second idea, that p115 links GM130 to Giantin, arose from theobservation that GM130 and Giantin are asymmetrically distributed betweenGolgi membranes and vesicles. Giantin (but not GM130) is incorporated intoCOPI vesicles, during in vitro budding reactions. p115 stimulatesbinding of these vesicles to Golgi membranes(25Sönnichsen B. Lowe M. Levine T. Jamsa E. Dirac-Svejstrup B. Warren G. J. CellBiol. 1998; 140: 1013-1021Crossref PubMed Scopus (251) Google Scholar). Giantin, like GM130, isa major p115-binding protein in Golgi extracts(25Sönnichsen B. Lowe M. Levine T. Jamsa E. Dirac-Svejstrup B. Warren G. J. CellBiol. 1998; 140: 1013-1021Crossref PubMed Scopus (251) Google Scholar). However, whereasanti-GM130 antibodies (or the peptide N73) prevent p115 binding to Golgimembranes, antibodies against Giantin do not(25Sönnichsen B. Lowe M. Levine T. Jamsa E. Dirac-Svejstrup B. Warren G. J. CellBiol. 1998; 140: 1013-1021Crossref PubMed Scopus (251) Google Scholar). In contrast,anti-Giantin antibodies do prevent p115 binding to COPI vesicles, butanti-GM130 antibodies do not. Inhibition of p115 binding to either GM130 onGolgi membranes or Giantin in vesicles is sufficient to abolish the p115effect on tethering (25Sönnichsen B. Lowe M. Levine T. Jamsa E. Dirac-Svejstrup B. Warren G. J. CellBiol. 1998; 140: 1013-1021Crossref PubMed Scopus (251) Google Scholar,36Shorter J. Beard M.B. Seemann J. Dirac-Svejstrup A.B. Warren G. J. Cell Biol. 2002; 157: 45-62Crossref PubMed Scopus (167) Google Scholar). p115 is also necessary tolink Giantin and GM130 during co-immunoprecipitation experiments from Golgiextracts (32Dirac-Svejstrup A.B. Shorter J. Waters M.G. Warren G. J. Cell Biol. 2000; 150: 475-488Crossref PubMed Scopus (63) Google Scholar). The functionalimportance of these interactions is apparent because agents that inhibit p115binding to Giantin or to GM130 block tethering and fusion during an invitro assay for Golgi reassembly after mitosis(36Shorter J. Beard M.B. Seemann J. Dirac-Svejstrup A.B. Warren G. J. Cell Biol. 2002; 157: 45-62Crossref PubMed Scopus (167) Google Scholar). Furthermore, GM130 andGiantin function in vivo, because microinjected antibodies againstthese proteins inhibit exocytic transport(24Alvarez C. Garcia-Mata R. Hauri H.P. Sztul E. J. Biol. Chem. 2001; 276: 2693-2700Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar). The most parsimoniousexplanation for these results is that p115 tethers by linking GM130 on onemembrane to Giantin in the other.However, several lines of evidence are inconsistent with theseinterpretations. First, antibodies against p115, GM130, and Giantin causedifferent phenotypes when microinjected. Although antibodies to GM130 orGiantin inhibit transport, they do so at a later stage than antibodies againstp115 (24Alvarez C. Garcia-Mata R. Hauri H.P. Sztul E. J. Biol. Chem. 2001; 276: 2693-2700Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar). In otherexperiments, microinjection of anti-Giantin antibodies led to degradation, butthis did not inhibit progression through mitosis or Golgi reassembly in thedaughter cells, at least as assessed by immunofluorescence microscopy. Both ofthese processes are believed to depend on tethering and fusion. Similarly,microinjected anti-GM130 antibodies did not affect mitosis or Golgireassembly. In contrast, microinjected anti-p115 antibodies led to p115degradation and collapse of the Golgi structure(37Puthenveedu M.A. Linstedt A.D. J. Cell Biol. 2001; 155: 227-238Crossref PubMed Scopus (97) Google Scholar). These results suggest afunction for p115 that is independent of its interactions with GM130 andGiantin and were presented as arguing against the working model for tetheringby p115.Second, a prediction of the tethering model is that p115 constructs withoutthe GM130 and Giantin-binding sites should neither localize to the Golgiapparatus nor function to tether membranes. Indeed, inhibition of GM130binding to p115 (by microinjection of the N73 peptide or by truncation ofGM130) does block p115 localization to the Golgi apparatus(35Seemann J. Jokitalo E.J. Warren G. Mol. Biol. Cell. 2000; 11: 635-645Crossref PubMed Scopus (153) Google Scholar). Confusingly, however, aninitial study showed that p115 truncations lacking the GM130 andGiantin-binding domain do localize to the Golgi region of transfected cells(38Nelson D.S. Alvarez C. Gao Y.S. Garcia-Mata R. Fialkowski E. Sztul E. J. CellBiol. 1998; 143: 319-331Crossref PubMed Scopus (116) Google Scholar). This finding hasrecently been confirmed and extended by a gene replacement approach showingthat truncated p115, without the binding domain for GM130 or Giantin, issufficient to rescue Golgi morphology and transport in cells where endogenousp115 has been knocked down by RNA interference(23Puthenveedu M.A. Linstedt A.D. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 1253-1256Crossref PubMed Scopus (85) Google Scholar). It seems, therefore,that experiments in which p115 can no longer bind to GM130 (and Giantin)contradict those in which GM130 can no longer bind to p115. Furthermore, theacidic C-terminal domain of p115 is absent in Drosophila p115 and ismuch shorter in C. elegans(39Kondylis V. Rabouille C. J.Cell Biol. 2003; 162: 185-198Crossref PubMed Scopus (104) Google Scholar). Furthermore, aconditional lethal CHO cell line that contains no detectable GM130immunoreactivity has been described that has no apparent defect in Golgistructure or transport when grown at the permissive temperature(40Vasile E. Perez T. Nakamura N. Krieger M. Traffic. 2003; 4: 254-272Crossref PubMed Scopus (48) Google Scholar). Together these data aredifficult to reconcile with the model that p115 tethers by directly linkingGM130 to Giantin.A resolution for these discrepancies might be related to the function ofRab1, which also acts during ER to Golgi and intra-Golgi transport and bindsdirectly to p115 and golgin tethering proteins(41Allan B.B. Moyer B.D. Balch W.E. Science. 2000; 289: 444-448Crossref PubMed Scopus (382) Google Scholar, 42Moyer B.D. Allan B.B. Balch W.E. Traffic. 2001; 2: 268-276Crossref PubMed Scopus (208) Google Scholar, 43Weide T. Bayer M. Koster M. Siebrasse J.P. Peters R. Barnekow A. EMBO Rep. 2001; 2: 336-341Crossref PubMed Scopus (103) Google Scholar).Rab family GTPases function throughout the exocytic pathway and several familymembers are implicated in tethering(4Waters M.G. Hughson F.M. Traffic. 2000; 1: 588-597Crossref PubMed Scopus (91) Google Scholar,6Pfeffer S. Mol. Cell. 2001; 8: 729-730Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar,7Zerial M. McBride H. Nat.Rev. Mol. Cell. Biol. 2001; 2: 107-117Crossref PubMed Scopus (2681) Google Scholar,44Barr F.A. Curr. Biol. 1999; 9: 381-384Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar). Rab proteins undergo acycle of GTP binding and hydrolysis that switches between their active andinactive states, respectively. In the active state they bind effectors thatare either membrane proteins(43Weide T. Bayer M. Koster M. Siebrasse J.P. Peters R. Barnekow A. EMBO Rep. 2001; 2: 336-341Crossref PubMed Scopus (103) Google Scholar,45Diao A. Rahman D. Pappin D.J. Lucocq J. Lowe M. J. Cell Biol. 2003; 160: 201-212Crossref PubMed Scopus (195) Google Scholar,46Satoh A. Wang Y. Malsam J. Beard M.B. Warren G. Traffic. 2003; 4: 153-161Crossref PubMed Scopus (103) Google Scholar) or that become recruitedto membranes by the Rab itself(41Allan B.B. Moyer B.D. Balch W.E. Science. 2000; 289: 444-448Crossref PubMed Scopus (382) Google Scholar,47Short B. Preisinger C. Schaletzky J. Kopajtich R. Barr F.A. Curr. Biol. 2002; 12: 1792-1795Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar). After GTP hydrolysis, Rabproteins are removed from the membrane by Rab-GDP dissociation inhibitor. Thisallows recycling of GDP bound Rab for further rounds of transport(48Soldati T. Riederer M.A. Pfeffer S.R. Mol. Biol. Cell. 1993; 4: 425-434Crossref PubMed Scopus (123) Google Scholar).Rab1 recruits p115 to COPII vesicles during in vitro assays for ERto Golgi transport (41Allan B.B. Moyer B.D. Balch W.E. Science. 2000; 289: 444-448Crossref PubMed Scopus (382) Google Scholar). Italso binds GM130 and acts on the Golgi membrane(42Moyer B.D. Allan B.B. Balch W.E. Traffic. 2001; 2: 268-276Crossref PubMed Scopus (208) Google Scholar). The yeast homologue,Ypt1p, recruits Uso1p (p115) to membranes and acts in tethering during ER toGolgi and intra-Golgi transport(21Barlowe C. J. CellBiol. 1997; 139: 1097-1108Crossref PubMed Scopus (155) Google Scholar,22Cao X. Ballew N. Barlowe C. EMBO J. 1998; 17: 2156-2165Crossref PubMed Scopus (291) Google Scholar,49Lupashin V.V. Hamamoto S. Schekman R.W. J. Cell Biol. 1996; 132: 277-289Crossref PubMed Scopus (73) Google Scholar,50Cao X. Barlowe C. J. CellBiol. 2000; 149: 55-66Crossref PubMed Scopus (116) Google Scholar).While mapping the Rab1-binding site on p115 we noticed a dramaticenhancement in the apparent affinity of interaction when the p115 C-terminaldomain was removed. Further experiments showed that the p115 C-terminal domainbound to and competed for the Rab1-binding site on p115. The inhibition ofRab1 binding was relieved by either GM130 or Giantin, both of which binddirectly to p115 and to Rab1. This raises the possibility that binding ofthese golgins serves a regulatory role instead of, or in addition to, astructural role in tethering.EXPERIMENTAL PROCEDURESAntibodies—For anti-Giantin, antibodies (against 1–448and 1125–1695) were raised in rabbits, against hexahistidineHis6-tagged immunogens. Sera were concentrated (40% ammoniumsulfate precipitation) and dialyzed against 25 mm Tris (pH 8.0),150 mm KCl. Anti 1125–1695 was affinity-purified againstimmunogen covalently linked to cyanogen bromide-activated Sepharose (GEHealthcare). Other antibodies were as follows: GM130, monoclonal (TransductionLaboratories); p115, monoclonal(15Waters M.G. Clary D.O. Rothman J.E. J. Cell Biol. 1992; 118: 1015-1026Crossref PubMed Scopus (202) Google Scholar); GRASP65 (7E10),monoclonal (Francis Barr, Max Planck Institute of Biochemistry, Martinsried,Germany); GRASP65, polyclonal(51Wang Y. Seemann J. Pypaert M. Shorter J. Warren G. EMBO J. 2003; 22: 3279-3290Crossref PubMed Scopus (149) Google Scholar).Plasmids—Rab1a/pGEX4T3, Rab2/pGEX, Rab6/pGEX, and Rab11/pGEXwere gifts from Tommy Nilsson (Göteborg University, Göteborg,Sweden), Francis Barr (Max Planck Institute of Biochemistry), and David Sheff(University of Iowa), respectively.Rab1a(S25N), Rab1a(Q70L), Giantin 1–500/pET23a, Giantin121–500-maltose-binding protein (MBP)/pET23a, his-CC1,2,3 (p115652–812)/pQE9, his-CC1,2 (p115 652–776)/pQE9, his-CC1 (p115652–701)/pQE9, and his-CC2,3,4 (p115 704–933)/pQE9 were made byQuikChange® mutagenesis (Stratagene). Giantin 1–1197/pET23a, Giantin1–500-MBP/pET23a, GM130/pET23a, p115/pGEX6P1, and p115-A/pGEX6P1 weremade by PCR and subcloning (see supplemental Table S1 for primer sequences).Constructs were verified by sequencing.Other plasmids used in this study were as follows: pGCP364/pSG5, pGL88,pGL108, pGL141, pGL147, pGL101(30Lesa G.M. Seemann J. Shorter J. Vandekerckhove J. Warren G. J. Biol. Chem. 2000; 275: 2831-2836Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar), pBSGM130(26Nakamura N. Rabouille C. Watson R. Nilsson T. Hui N. Slusarewicz P. Kreis T.E. Warren G. J. Cell Biol. 1995; 131: 1715-1726Crossref PubMed Scopus (666) Google Scholar), GRASP65-his/pET30a(51Wang Y. Seemann J. Pypaert M. Shorter J. Warren G. EMBO J. 2003; 22: 3279-3290Crossref PubMed Scopus (149) Google Scholar), his-H (p1151–650)/pQE9, his-TA (p115 651–961)/pQE9, and his-T (p115651–933)/pQE9 (36Shorter J. Beard M.B. Seemann J. Dirac-Svejstrup A.B. Warren G. J. Cell Biol. 2002; 157: 45-62Crossref PubMed Scopus (167) Google Scholar).Recombinant Proteins—Recombinant proteins were purified onglutathione-Sepharose 4B (Amersham Biosciences) or nickel-nitrilotriaceticacid-agarose (Qiagen). GM130-his was further purified by ion-exchangechromatography (Hi-Trap SP column (pH 7.4)) (Amersham Biosciences) and gelfiltration (Hi-Prep 16/60 Sephacryl S-300 column) (Amersham Biosciences). TheGST moiety was removed from GST-p115 using PreScission protease (AmershamBiosciences).Rabs were loaded as described previously(52Christoforidis S. Zerial M. Methods. 2000; 20: 403-410Crossref PubMed Scopus (87) Google Scholar). Briefly, Rab was washedwith exchange buffer (20 mm HEPES (pH 7.4), 100 mm NaCl,10 mm EDTA, 5 mm MgCl2, 1 μmDTT, 1 μm guanosine nucleotide) and then incubated (three times,30 min, room temperature) in exchange incubation buffer (20 mmHEPES (pH 7.4), 100 mm NaCl, 10 mm EDTA, 5 mmMgCl2, 1 μm DTT, 1 mm guanosinenucleotide). After washing in stabilization buffer (20 mm HEPES (pH7.4), 100 mm NaCl, 5 mm MgCl2, 1μm DTT, 1 μm guanosine nucleotide) the Rab wasthen incubated (30 min, room temperature) in stabilization incubation buffer(20 mm HEPES (pH 7.4), 100 mm NaCl, 5 mmMgCl2, 1 μm DTT, 1 mm guanosinenucleotide).Superose 6 Chromatography—Recombinant p115, his-TA, or his-Twere filtered (0.45 μm) to remove any particulate matter and thengel-filtrated on a Superose 6 HR 10/30 column (Amersham Biosciences),equilibrated in column buffer (25 mm HEPES (pH 7.4), 200mm KCl, 1 mm DTT, 10% glycerol), at 0.2 ml/min. Half-mlfractions were collected, and aliquots were analyzed by SDS-PAGE withCoomassie staining. Thyroglobulin (669 kDa, Stokes radius 85.0 Å),ferritin (440 kDa, Stokes radius 61.0 Å), catalase (232 kDa, Stokesradius 52.2 Å), and aldolase (158 kDa, Stokes radius 48.1 Å) wererun as standards.Velocity Sedementation—Recombinant p115, his-TA, or his-Twere layered onto a linear glycerol gradient (10–30% (w/v)) in gradientbuffer (25 mm HEPES (pH 7.4), 200 mm KCl, 1mm DTT). Thyroglobulin (20.0 S), catalase (11.4 S), and bovineserum albumin (4.6 S) were run as standards. The gradients were centrifugedfor 5 h in an SW55 rotor (Beckman Coulter), at 4 °C and then fractionatedfrom the top into ∼0.4-ml fractions. Aliquots were analyzed by SDS-PAGEwith Coomassie staining.Golgi Membrane Extracts—Rat liver Golgi was purified fromrat liver as described (53Hui N. Nakamura N. Slusarewicz P. Warren G. Celis J. Cell Biology: A LaboratoryHandbook. Academic Press, New York1998: 46-55Google Scholar)and then extracted at 0.5 mg/ml with extraction buffer (20 mm HEPES(pH 7.4), 100 mm KCl, 5 mm MgCl2, 1% TritonX-100, 1 mg/ml soybean trypsin inhibitor, 1 μm DTT, EDTA freeprotease inhibitor (PI) mixture (Roche Applied Science)) for 30 min on ice andclarified by centrifugation (20 min, 16,000 × g, 4 °C). Forcarbonate stripping, rat liver Golgi (0.2 mg/ml) was incubated in carbonatebuffer (0.2 m sodium carbonate (pH 11) and PI) for 1 h on ice.Stripped rat liver Golgi was concentrated (30 min, 11,700 × g,4 °C, on a swing out microcentrifuge) onto a 2 μl sucrose cushion (2m), resuspended (0.2 mg/ml) in sucrose buffer (20 mmHEPES (pH 7.4), 100 mm KCl, 5 mm MgCl2, 200mm sucrose, 1 μm DTT, PI), reconcentrated, and thenextracted as above.In Vitro Transcription/Translation—Twenty-five μlreactions were performed using the TNT T7 coupled reticulocyte lysate system(Promega) using 0.5 μg of plasmid DNA and 2 μl of[35S]methionine (10 mCi/ml) per reaction. Two μl of eachreaction was analyzed by SDS-PAGE. The remainder was frozen in liquid nitrogenand stored at -80 °C until use.Binding Assays—Glutathione S-transferase(GST)-Rab1, immobilized on glutathione-Sepharose 4B, was incubated withputative binding proteins in binding buffer (20 mm HEPES (pH 7.4),100 mm KCl, 5 mm MgCl2, 1 mg/ml soybeantrypsin inhibitor, 1% Triton X-100, 1 μm DTT, PI supplementedwith guanosine nucleotide (1 mm)) for 1 h, rotating at 4 °C.Beads were washed twice in the same buffer and then once in nucleotide-freebinding buffer. Proteins were eluted (three incubations, 10 min, rotating,room temperature) in elution buffer (20 mm HEPES (pH 7.4), 100mm KCl, 20 mm EDTA, 1% Triton X-100, 1 mmDTT, 5 mm GDP), pooled, concentrated by trichloroacetic acidprecipitation, and then analyzed by SDS-PAGE.GST-p115CT binding assays were performed as for Rab1 except: 10 μg ofGST-p115CT (p115 886–961) was used. Beads were washed three times in thesame buffer then resuspended in 2× sample buffer (100 mm Tris(pH 6.8), 3% SDS, 15% glycerol, 5% DTT, 0.01% bromphenol blue)Quantification—Blots were quantified by scanning (EpsonExpression 1680 scanner) or by directly measuring the ECL signal (Kodak ImageStation 440CF). Signal intensity was compared with at least three standardsamples of the same protein, loaded as a series of 2-fold dilutions, anddetected on the same immunoblot. Signals were only quantified if theirmeasured intensity lay within a range encompassed by the standards and overwhich there was a linear relationship between the amount of material loadedand signal intensity.RESULTSRab1 Binds to the Coiled-coil Tail Region of p115—TheRab1-binding site on p115 had not previously been mapped. A set of truncationscorresponding to structural and functional domains in p115 was thereforeconstructed and tested for binding to GST-Rab1(Fig. 1A). p115 isthought to be a parallel homodimer of two polypeptide chains. Each is composedof an N-terminal globular head (H); a rod-like tail (T)containing four regions of predicted coiled-coil structure (CC1, CC2, CC3,CC4), and a short (28 amino acid), C-terminal acidic domain (A)(Fig. 1A)(17Sapperstein S.K. Walter D.M. Grosvenor A.R. Heuser J.E. Waters M.G. Proc. Natl. Acad. Sci. U. S.A. 1995; 92: 522-526Crossref PubMed Scopus (171) Google Scholar,18Yamakawa H. Seog D.H. Yoda K. Yamasaki M. Wakabayashi T. J. Struct. Biol. 1996; 116: 356-365Crossref PubMed Scopus (64) Google Scholar).Full-length p115 or truncations corresponding to the head (his-H), tail(his-T), or tail plus acidic (his-TA) domains (50 nm each) wereincubated with immobilized GST or GST-Rab1 fusion proteins (1 nmol/200 μlreaction volume). Fusion proteins of either wild type Rab1a (preloaded with anappropriate guanosine nucleotide) or point mutations stabilized in the GTPbound (GST-Rab1(QL)) or the GDP bound (GST-Rab1(GDP)) conformations weretested. Specifically bound proteins were eluted by incubation with GDP/EDTAbuffer. This technique has been used previously to measure effector binding toRab1(41Allan B.B. Moyer B.D. Balch W.E. Science. 2000; 289: 444-448Crossref PubMed Scopus (382) Google Scholar, 42Moyer B.D. Allan B.B. Balch W.E. Traffic. 2001; 2: 268-276Crossref PubMed Scopus (208) Google Scholar, 43Weide T. Bayer M. Koster M. Siebrasse J.P. Peters R. Barnekow A. EMBO Rep. 2001; 2: 336-341Crossref PubMed Scopus (103) Google Scholar,45Diao A. Rahman D. Pappin D.J. Lucocq J. Lowe M. J. Cell Biol. 2003; 160: 201-212Crossref PubMed Scopus (195) Google Scholar,46Satoh A. Wang Y. Malsam J. Beard M.B. Warren G. Traffic. 2003; 4: 153-161Crossref PubMed Scopus (103) Google Scholar) and other Rab familymembers(54Christoforidis S. Zerial M. Methods Enzymol. 2001; 329: 120-132Crossref PubMed Scopus (8) Google Scholar, 55Short B. Preisinger C. Korner R. Kopajtich R. Byron O. Barr F.A. J. Cell Biol. 2001; 155: 877-883Crossref PubMed Scopus (180) Google Scholar, 56Valsdottir R. Hashimoto H. Ashman K. Koda T. Storrie B. Nilsson T. FEBS Lett. 2001; 508: 201-209Crossref PubMed Scopus (94) Google Scholar, 57Cullis D.N. Philip B. Baleja J.D. Feig L.A. J. Biol. Chem. 2002; 277: 49158-49166Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar, 58Nagashima K. Torii S. Yi Z. Igarashi M. Okamoto K. Takeuchi T. Izumi T. FEBSLett. 2002; 517: 233-238Crossref PubMed Scopus (117) Google Scholar).In these assays, ∼5% of the full-length p115 bound GST-Rab1(QL),whereas ≤0.5%