Vol. 12, No. 7

MOLECULAR AND CELLULAR BIOLOGY, JUlY 1992, p. 3288-3296 0270-7306/92/073288-09$02.00/0 Copyright ©3 1992, American Society for Microbiology

Topology and Functional Domains of Sec63p, an Endoplasmic Reticulum Membrane Protein Required for Secretory Protein Translocation DAVID FELDHEIM, JONATHAN ROTHBLAT,t AND RANDY SCHEKMAN*

Howard Hughes Medical Research Institute, Division of Biochemistry and Molecular Biology, Barker Hall, University of California, Berkeley, California 94720 Received 27 March 1992/Accepted 23 April 1992

SEC63 encodes a protein required for secretory protein translocation into the endoplasmic reticulum (ER) of Saccharomyces cerevisiae (J. A. Rothblatt, R. J. Deshaies, S. L. Sanders, G. Daum, and R. Schekman, J. Cell Biol. 109:2641-2652, 1989). Antibody directed against a recombinant form of the protein detects a 73-kDa polypeptide which, by immunofluorescence microscopy, is localized to the nuclear envelope-ER network. Cell fractionation and protease protection experiments confirm the prediction that Sec63p is an integral membrane protein. A series of SEC63-SUC2 fusion genes was created to assess the topology of Sec63p within the ER membrane. The largest hybrid proteins are unglycosylated, suggesting that the carboxyl terminus of Sec63p faces the cytosol. Invertase fusion to a loop in Sec63p that is flanked by two putative transmembrane domains produces an extensively glycosylated hybrid protein. This loop, which is homologous to the amino terminus of the Escherichia coli heat shock protein, DnaJ, is likely to face the ER lumen. By analogy to the interaction of the DnaJ and Hsp70-like DnaK proteins in E. coli, the DnaJ loop of Sec63p may recruit luminal Hsp7O (BiP/GRP78/Kar2p) to the translocation apparatus. Mutations in two highly conserved positions of the DnaJ loop and short deletions of the carboxyl terminus inactivate Sec63p activity. Sec63p associates with several other proteins, including Sec6lp, a 31.5-kDa glycoprotein, and a 23-kDa protein, and together with these proteins may constitute part of the polypeptide translocation apparatus. A nonfunctional DnaJ domain mutant allele does not interfere with the formation of the Sec63p/Sec6lp/gp3l.5/p23 complex. We have previously described a set of temperature-sensitive lethal mutants of Saccharomyces cerevisiae that fail to localize a signal peptide-bearing cytosolic enzyme chimera to the lumen of the endoplasmic reticulum (ER) (14, 38). Characterization of these mutants showed that the products of four genes, SEC61, SEC62, SEC63, and SEC65, are required for translocation of secretory precursor proteins across the ER membrane (14, 38, 45a). Additional alleles of SEC63 that inhibit protein import into the nucleus have also been described (39). In addition to the SEC genes described above, several other genes are required for the proper targeting and insertion of presecretory proteins into the ER. Hsp7O homologs appear to be required in the cytosol and ER lumen for translocation to occur. Cytosolic Hsp7O, encoded by the SSA genes, is required for efficient translocation of prepro-a-factor into the ER both in vivo and in vitro (8, 12). An ER luminal Hsp7O homolog, BiP (the KAR2 gene product), has been identified, and some alleles of this gene block translocation of presecretory proteins (33, 37, 50). The deduced amino acid sequence of Sec63p suggests it is an integral membrane protein with interesting structural features (39). The predicted sequence is of a 73-kDa protein with three transmembrane segments. The carboxyl terminus is highly charged; 31 of the last 47 amino acids are aspartate or glutamate. Sec63p contains an internal 70-amino-acid domain that has 43% sequence identity to the amino terminus of the Escherichia coli heat shock protein DnaJ. DnaJ works in concert with the E. coli Hsp7O homolog, DnaK, to promote bacteriophage lambda DNA replication. In a com*

pletely purified replication system, the DnaJ and DnaK proteins are required to mediate the disassembly of the A 0-some complex, which activates the helicase activity of DnaB protein to initiate X DNA replication (2, 27, 52). The homology of Sec63p to DnaJ and the requirement for Hsp7O homologs for efficient translocation suggest that one role of Sec63p may be to interact with an Hsp7O homolog to promote protein translocation. To gain a better understanding of Sec63p and its function, we have analyzed the intracellular distribution, topology, and essential features of the protein. In this report, we show that Sec63p is a 73-kDa integral ER membrane protein with three transmembrane domains which position the DnaJ homology loop in the ER lumen and the highly charged carboxyl terminus in the cytosol. Both the carboxyl terminus and the DnaJ-like domain are required for Sec63p function.

MATERIALS AND METHODS Strains, materials, plasmids, and general methods. The following strains were used in this study: RSY255 (leu23,-112, ura3-52 MA Tot), RSY151 (leu2-3,-112 ura3-52pep4-3 sec63-1 MATao), YPH500 (ura3-52 lys2-801 ade2-101 his3A200 leu2-Al MATTa), YT455 (ura3-52 ade2-100 Asuc2 A4 Ta), RSY800 (sec63: :HIS3 ura3-52/ura3-52 lys2-801!lys2801 ade2-101/ade2-101 trplA63/trplA63 his3-A200/his3-A200 leu2-A1!leu2-A1), and RSY587 (sec63::HIS3 ura3-52 lys2-801 ade2-101 trpJA63 his3-A200 leu2-AJ, MAIToa) containing pDF41. Escherichia coli DH5a harboring plasmid pUV5GlS was used to isolate lytic ,1,3 glucanase (43). Yeast cells were grown in rich or minimal medium as described previously (15). Reagents and their sources were as follows: phenylmeth-

Corresponding author.

t Present address: Department of Biological Sciences, Dartmouth

College, Hanover, NH 03755.

3288

VOL. 12, 1992

ylsulfonyl fluoride, pepstatin, leupeptin, 1,10-phenanthroline, tunicamycin, and Freund's complete and incomplete adjuvants (Sigma Chemical Co., St. Louis, Mo.); CNBractivated Sepharose 4B and protein A-Sepharose CL-4B (Pharmacia-LKB Biotechnology, Piscataway, N.J.); Tran'S-label (ICN, Costa Mesa, Calif.); alkaline phosphatase immunoblotting reagents (Bio-Rad, Richmond, Calif.); Citifluor (Citifluor Ltd., London, U.K.); and Geneclean (BiolOl, La Jolla, Calif.). All restriction enzymes and molecular biology reagents were purchased from Boehringer Mannheim (Indianapolis, Ind.) or Bethesda Research Laboratories (Gaithersburg, Md.). Affinity-purified Sec62p antibody and Sec23p antibody have been described elsewhere (11, 21). Kar2p antibody was a gift of Mark Rose (Department of Molecular Biology, Princeton University, Princeton, N.J.). Ascites fluid from the hybridoma myc-9E10 was provided by Karl Kuchler (this department) (17). Plasmids pRIT2T (32) and pEX3 (45) were used to generate SEC63 gene fusions for production and purification of anti-Sec63p sera. The multicopy yeast plasmids YEp351 and YEp352 and the centromeric plasmid pRS315 have been described previously (22, 44). Plasmids pDF14, pDF11, and pDF1 contain a 3.5-kb HindIlI fragment of pTK1 (39) inserted into YEp351, YEp352, and pUC119, respectively (49). pDF3 was created by inserting a 2.5-kb Hindll fragment from pTK1, which contains the minimum complementing subclone of SEC63, into the SmaI site in pUC119. SEC63SUC2 fusions were constructed in the SUC2 fusion vectors pSEY304 (2,um, URA3) (3), pSEYC306 (CEN4-ARS1, URA3) (25), pCS29, and pDF21. Each of these vectors contains the coding sequence for 511 amino acids of mature invertase preceded by a polylinker (46). pCS29 and pDF21 were constructed by replacing the polylinker of pSEY304 and pSEY306 with that of pUC119. pDF4 was made by inserting the SacI-BamHI fragment of pDF3 into pCS29 that was cleaved with Sacl and BamHI. Plasmid pDF41 was constructed by digesting pDF1 with EcoRV and inserting the

oligonucleotide 5'-CTCGAGGAGCAGAAATTAATCAGC GAAGAGGACCTCCTCAGGAAGAGGTAA- 3' annealed with its complement. This reading frame encodes amino acids corresponding to the epitope of human c-myc which is recognized by the monoclonal antibody myc-9E10 (17). The peptide EEQKLISEEDLLRK (single-letter amino acid code) corresponding to amino acid residues 409 to 423 of human c-myc (prepared by David King, this department) was synthesized on an ABI 431A peptide synthesizer, using 9-fluorenylmethoxycarbonyl amino acids activated with 2-(lH-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate in the presence of 1-hydroxybenzotriazole and diisopropylethylamine. The crude product (-90% purity) was purified by reverse-phase high-pressure liquid chromatography (10 pl/300 A Cl8-silica column) eluted with a gradient of acetonitrile in water, both containing 0.1% trifluoroacetic acid, to yield a product of >98% purity. The calculated molecular weight of the peptide was 1,730.9 and was experimentally found to be 1,730.7 + 0.6. This peptide was used as competitor in immunoprecipitation experiments. Common recombinant DNA techniques were performed essentially as described by Maniatis et al. (29). Unidirectional exonuclease III digestion was used to generate 3' deletions of SEC63 (20). Transfer of proteins from sodium dodecyl sulfate (SDS)polyacrylamide gels to nitrocellulose filters was performed as described previously (19). Filters were blocked, and all antibody incubations and washes were conducted with 2%

TOPOLOGY AND FUNCTIONAL DOMAINS OF Sec63p

3289

nonfat dry milk in Tris-buffered saline (25 mM Tris, 150 mM NaCI) plus 0.1% Nonidet P-40. Detection of filter-bound antibodies with alkaline phosphatase-conjugated goat antirabbit immunoglobulin G (IgG) was performed according to the supplier's instructions or by the enhanced chemiluminescence method (Amersham, Arlington Heights, Ill.). Preparation, fractionation, and proteolysis of cell extracts. Extracts of wild-type (RSY255) cells or sec63::HIS3 cells containing pDF41 (RSY587) were prepared as described elsewhere (11) except that spheroplasts were prepared with lyticase contained in the shock fluid of E. coli DH5a(pUV5G1S) (34, 43). To determine the nature of the membrane association of Sec63p, spheroplasts were broken by osmotic shock in lysis buffer (2,000 OD6. [optical density at 600 nm] unit cell equivalents per ml in 200 mM mannitol-100 mM NaCI-25 mM sodium phosphate-[pH 7.4], 1 mM MgCl2-1 mM phenylmethylsulfonyl fluoride) and centrifuged at 370 x g as described elsewhere (11). The supernatant fraction was diluted with 0.1 volume of 10% Triton X-100, 0.1 M Na2CO3 (pH 11), or 5 M NaCl or with 0.2 volume of 8 M urea, incubated for 20 min at 0WC, and centrifuged at 96,500 x g in a TLA 100 rotor (Beckman Instruments, Palo Alto, Calif.). Pellet fractions were resuspended in lysis buffer to the same volume, and samples were prepared for SDS-polyacrylamide gel electrophoresis (PAGE); 1 OD6. cell equivalent of lysate was applied in each gel lane. Protease treatments were performed as follows. Spheroplasts prepared from RSY587 were lysed by Dounce homogenization in buffer 88 (20 mM N-2-hydroxyethylprperazine-N'-2-ethanesulfonic acid [HEPES; pH 6.8], 150 mM potassium acetate, 250 mM sorbitol, 1 mM magnesium acetate) at 100 OD600 cell equivalents per ml. The 370 x g supernatant fraction was centrifuged at 12,000 x g in an SS34 rotor, and the particulate fraction was used as the starting material. Samples were separated into aliquots and treated at 0°C with 500 pug of trypsin per ml plus 500 ,g of proteinase K (Sigma) per ml in the presence or absence of 1% (vol/vol) Triton X-100 in buffer 88. After various times, the reactions were terminated by the addition of an equal volume of 40% trichloroacetic acid (TCA). TCA precipitates were washed with 1 ml of ice-cold acetone and solubilized with sample buffer. One OD6. cell equivalent of lysate was applied in each SDSPAGE lane. Production and affinity purification of anti-Sec63p antisera. Polyclonal antisera recognizing Sec63p epitopes were obtained by immunizing rabbits with protein A-Sec63p hybrid proteins expressed in E. coli. Protein A fusions were constructed by inserting the 2.2-kb StuI-HindII fragment of pDF1 into pRIT2T. Likewise, a lacZ-SEC63 fusion was created by inserting the same fragment into pEX3. The StuI-HindII fusion encoded a hybrid protein fused in frame, 13 amino acids upstream of the putative initiator methionine of Sec63p and extending the full length (603 amino acids) of the structural gene. The 5' and 3' termini were converted to blunt ends by using T4 DNA polymerase. The fragment was inserted into the SmaI cloning site in the pRIT and pEX vectors, and E. coli CD14 [cI857 rpsL F' (lacIql lacZAM15 Pro' TnlO)], a generous gift from C. d'Enfert (this department), was transformed to ampicillin resistance. Transformants containing SEC63 DNA inserts in the correct orientation, as judged by restriction analysis, were tested for fusion protein expression. Protein A fusions were expressed by induction at 42°C or by induction with 1 mM isopropylthiogalactopyranoside (IPTG). Subsequent large-scale expression of the StuI-HindII gene fusion was performed by

3290

FELDHEIM ET AL.

IPTG induction. lacZ-SEC63 fusions were expressed by heat induction. The expressed protein A-Sec63p fusions were recovered by sonically lysing bacteria in 10% glycerol-5 mM EDTA100 mM sodium phosphate (pH 7) containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 5 mM 1,10phenanthroline, and 10 ,uM leupeptin) and then subjected to ultracentrifugation (Beckman TLA 100.3 rotor at 100,000 x g for 15 min at 4°C). The hybrid resulting from the fusion of the Stu-HindII fragment to protein A sedimented with the particulate fraction, which was solubilized in 0.2% SDS-50 mM Tris-HCl (pH 7.5) containing protease inhibitors by heating to 85°C for 10 min. The solubilized material was then diluted fivefold with immunoprecipitation (IP) buffer (1.25% Triton X-100, 190 mM NaCl, 6 mM EDTA, 0.2% SDS, 60 mM Tris-HCl [pH 7.5]), resulting in a final detergent concentration of 1% Triton X-100-0.2% SDS. The fusion proteins were affinity purified on IgG-Sepharose Fast Flow 6 (Pharmacia-LKB Biotechnology), prepared according to the manufacturer's instructions. The bound fraction was washed extensively with IP buffer and IP buffer containing 2 M urea, and then fractions were eluted with 0.2 M glycine (pH 2.5) into neutralizing aliquots of 1.5 M Tris-HCl (pH 8.8). The elution profile of the protein A-Sec63p hybrids was determined by SDS-PAGE and immunoblotting with alkaline phosphatase-conjugated goat IgG (Bio-Rad Laboratories, Richmond, Calif.). The enriched fractions were pooled and dialyzed against phosphate-buffered saline (PBS; 20 mM sodium phosphate [pH 7.4], 150 mM NaCl). Rabbits were injected subcutaneously with 150 pg of protein in an emulsion containing Freund's incomplete adjuvant. The LacZ-Sec63 protein fusion was somewhat unstable in E. coli. Inclusion bodies recovered by centrifugation at 12,000 x g from sonically lysed, heat-induced E. coli contained immunoreactive material that displayed heterogeneous electrophoretic mobility. Therefore, to affinity purify antiserum, 10 mg of an E. coli inclusion body pellet was solubilized in PBS plus 0.2% SDS, and the resultant protein was coupled to 5 ml of CNBr-activated Sepharose 4B according to the manufacturer's instructions. Ten milliliters of anti-Sec63p crude serum raised against the protein A-Sec63 fusion protein was diluted fivefold in PBS and applied to the Sec63p-LacZ affinity column. The column was washed extensively with PBS, and antibodies were eluted with 0.2 M glycine (pH 2.2). The eluate was rapidly neutralized with 2 M Tris (pH 8.8). Antibodies in the peak fractions, measured by A280, were concentrated on a 200-pl protein A-Sepharose column (Pharmacia-LKB Biotechnology) and eluted with 0.2 M glycine (pH 2.2). Peak fractions as determined by A280 were dialyzed against PBS and stored in PBS plus 5 mM sodium azide and 1 mg of bovine serum albumin per ml. A 1:3,000 dilution of affinity-purified antibody was used to detect Sec63p on immunoblots. Immunofluorescence. Immunofluorescence microscopy was performed essentially as described by Pringle et al. (36). Anti-Sec63p antibodies were used at dilutions of 1/10 to 1/50. Bound primary antibodies were decorated with fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG, and slides were mounted in Citifluor after being supplemented with 1 mg of 4',6-diamino-2-phenylindole (DAPI) per ml. Cells were examined at x 400 magnification on a Nikon Optiphot microscope. Images were recorded on Kodak TMAX P3200 black-and-white film with a Nikon FX-35WA camera attached to a AFX-IIA shutter controller. Construction and analysis of SEC63-SUC2 gene fusions. A set of SEC63-SUC2 gene fusions was created by construct-

MOL. CELL. BIOL.

ing unidirectional exonuclease III 3' deletions from the starting plasmid, pDF4. pDF4, which contains the entire SUC2 coding sequence downstream from the SEC63 gene, was digested with SstI and BamHI. Exonuclease III digestions were performed as described previously (20), using 15-s time intervals at 37°C. The resultant single strand termini were degraded with Si nuclease and then treated with Klenow fragment to generate blunt ends. The plasmids were ligated overnight at 17°C and transformed into E. coli. Deletion plasmids were introduced by transformation into YT455, and cell extracts were immunoblotted with invertase antiserum to detect hybrid proteins. Clones that produced hybrid proteins were sequenced and selected for further study. SEC63A523-SUC2 was constructed by directly fusing the SUC2 gene in frame to the Spel site in SEC63. All fusions initially were constructed in pCS29 and were transferred as needed to pDF21, a centromeric plasmid. The number of amino acids deleted from the C terminus of Sec63p is indicated in each fusion name. Glycosylation of hybrid proteins was assessed with various SEC63-SUC2 fusion plasmids in strain YT455. Cells were grown to mid-logarithmic phase in minimal medium lacking uracil and treated with or without tunicamycin (10 ,ug/ml) for 5 min. Radiolabeling was performed with Tran35S-label (30 jiCi/OD600 cells). Cells were lysed and invertase immunoprecipitated as described elsewhere (15). Site-specific mutagenesis of SEC63. Site-specific mutagenesis was performed by using the polymerase chain reaction (PCR) essentially as described previously (24). Briefly, a degenerate 35-mer that corresponds to nucleotides 449 to 484 of SEC63 [5'-AAITITC(A/T)TC(C/G)AG(C/A)TAAATTAG CAAAGGGCCTAACA-3'] and a second primer corresponding to nucleotides 263 to 283 (5'-AGCAGGAGAAATATTA TAAT-3') were used to amplify plasmid pDF14. The resultant PCR product was excised by digestion with SpeI and Drall and then religated into pDF19 from which the SpeIDrall fragment had been excised. The PCR products were sequenced to identify mutant SEC63 alleles. Mutant genes were subcloned into either the centromeric plasmid pRS315 (44) or the multicopy plasmid YEp351 to create plasmids pDF25 and pDF32 (Asp-157--*Ala, CEN, and 2,um, respectively), pDF28 and pDF33 (Pro-156-->Asn, CEN, and 2,um, respectively), and pDF24 (Pro-156-*Asn, Asp-157--Ala, CEN). Plasmid pDF50 was constructed by digesting pDF24 with SphI and NheI and ligating it into SphI-NheI-digested pDF41. Mutations were tested for their ability to complement a chromosomal deletion of SEC63 by transforming the plasmids into strain RSY587, which contained the uracilbased plasmid pDF14. The ability to lose the Ura+ plasmid in such transformants was determined with 5-fluoro-orotic acid as described previously (6). Immunoprecipitation of the Sec63p complex. Cells were grown to an OD60 of 0.2 to 0.8 in yeast nitrogen base (Difco, Detroit, Mich.) supplemented with required amino acids and 100 ,uM (NH4)2SO4. After centrifugation, cells were resuspended at 5 OD60 units per ml, and 6 ml was radiolabeled with 900 puCi of Tran-35S-label (ICN). Labeling was terminated by the addition of 10 mM sodium azide, and cells were centrifuged, chilled, converted to spheroplasts, and lysed by agitation with glass beads in 0.6 ml of 200 mM mannitol-100 mM NaCl-25 mM NaPO4 (pH 7.4)-i mM MgCl2 (Iysis buffer). Unbroken cells were removed by centrifugation at 200 x g for 4 min in an HB4 rotor (Dupont/Sorvall). The supernatant fraction was centrifuged at 37,000 x g for 12 min in a TLA 100.3 rotor of the TL-100 ultracentrifuge (Beckman). Sedimented membranes were resuspended in lysis

VOL. 12, 1992

TOPOLOGY AND FUNCTIONAL DOMAINS OF Sec63p

12

3291

ken by heating the precipitates at 65°C for 20 min in Laemmli sample buffer containing 2% ,-mercaptoethanol. RESULTS

150

em 68FIG. 1. Evidence that affinity-purified anti-Sec63p antibodies specifically recognize a 73-kDa protein in yeast cells. Wild-type cells (lane 1) or wild-type cells transformed with SEC63 on a multicopy vector (pDF14; lane 2) were pulse-labeled for 15 min with Tran-35Slabel, and glass bead extracts were immunoprecipitated with affinity-purified anti-Sec63p antiserum. Immunoprecipitates were subjected to SDS-PAGE and fluorography. Sizes are indicated in kilodaltons.

buffer containing 10% glycerol to a final concentration of 50 OD600 cell equivalents per ml. Aliquots (75 pul) of the membrane fraction were treated with the cleavable crosslinking reagent dithiobis succinimidylpropionate (DSP; Pierce, Rockford, Ill.) (0.2 mg/ml) for 15 min at 20°C. Alternatively, membranes were mixed with Triton X-100 (1%, final concentration) and samples were rotated for 20 min at 4°C. After clarification at 100,000 x g for 30 min in the TLA 100.3 rotor, solubilized fractions (50 pul) were treated with DSP as described above. Samples were quenched by the addition of an equal volume of 0.2 M ammonium acetate. SDS (1%) was added, and samples were heated at 65°C for 20 min. Samples (100 to 200 Al) were mixed with -6 ,ug of the anti-c-myc monoclonal IgG (myc-9E10) and 25 ,ul of protein A-Sepharose (prepared as described by the supplier). Precipitation mixtures were mechanically rotated for 8 h at 4°C. Where indicated, competitor peptide (final concentration of c-myc peptide, 50 ,ug/ml) was added along with the monoclonal antibody. Immunoprecipitates were collected and washed as described previously (38). Cross-links were bro-

Detection of Sec63p in yeast cell extracts and in cells. The SEC63 DNA sequence predicts a 73-kDa protein capable of spanning a lipid bilayer three times. To characterize the SEC63 gene product further, we raised antiserum to the full-length Sec63 protein fused in frame to the C terminus of Staphylococcus aureus protein A. The hybrid gene was transformed into E. coli and induced for expression. Hybrid proteins were purified by IgG affinity chromatography, and antigen was injected into rabbits. The resultant antiserum was affinity purified by absorption to immobilized LacZSec63p. Affinity-purified antibodies were tested for their specificity by immunoprecipitation of crude yeast cell extracts (Fig. 1). The anti-Sec63 antiserum detected a polypeptide of 73 kDa and a second smaller polypeptide (lane 1). Both immunoreactive species were overproduced when cells contained the SEC63 gene on a 2,um plasmid (lane 2). The smaller band was probably a proteolytic fragment of Sec63p because it varied in amount from experiment to experiment (data not shown). These results confirmed the size of Sec63p predicted from the sequence analysis. Immunofluorescence microscopy was performed to determine the intracellular localization of Sec63p in the wild type and in strains overproducing Sec63p. Fixed, permeabilized cells were probed with anti-Sec63p antibodies and then with FITC-conjugated goat anti-rabbit antibodies. Nuclei were labeled with the DNA binding dye DAPI. In wild-type cells, Sec63p immunoreactivity was localized in a perinuclear staining pattern, with some peripheral ER staining (Fig. 2). This localization was similar to that of Kar2p and Sec62p, both ER proteins (11, 37). Similar but more intense staining patterns were seen with cells containing the 2p.m plasmid pDF24, indicating that overproduction of Sec63p did not result in its mislocalization (data not shown). Sec63p is an integral membrane protein. To test whether Sec63p behaves as an integral membrane protein, as predicted by its sequence, we prepared membrane fractions and extracted them under conditions which strip nonintegral proteins from the membrane or solubilize integral membrane proteins. Figure 3 demonstrates that approximately 90% of the Sec63p remained in the particulate fraction after treat-

I

FIG. 2. Localization of Sec63p to the ER. Wild-type cells grown at 30°C were fixed with formaldehyde, probed with affinity-purified anti-Sec63p antibodies, and then subjected to secondary decoration with FITC-conjugated goat anti-rabbit IgG. Magnification, x400. (A) Cells stained with anti-Sec63p; (B) DAPI staining of nuclei.

3292

MOL. CELL. BIOL.

FELDHEIM ET AL.

O' Tx-100

Q

-

15' 30' + - + - +

C',

0 U .~z

HO

z

co

co

o

0

His -LsP P

co

-

Ct Sec63p

-l

x

sls

5FSPI

CYc-myc

Sec63p-

Sec62p Sec23p -

C Kar2p -

-

__

S-

t-_-

FIG. 3. Evidence that Sec63p is an integral membrane protein. Membrane fractions were prepared and treated with either 0.5 M NaCl, 0.1 M Na2CO3 (pH 11), 1.6 M urea, or 1% Triton X-100. After incubation on ice for 20 min, all samples were separated into supernatant (S) or pellet (P) fractions by centrifugation at 96,600 x g, subjected to SDS-PAGE, and immunoblotted with an anti-Sec63p (1:3,000), anti-Sec62p (1:1,000), or anti-Sec23p (1:500) antibody.

FIG. 4. Evidence that Sec63p is accessible to exogenous protease. Membranes prepared from RSY587 were either mock digested or treated with proteinase K (500 ,ug/ml) plus trypsin (500 p,g/ml) in the absence or presence of 1% Triton X-100. All digests were terminated by the addition of an equal volume of 40% TCA. The TCA precipitates were resuspended in sample buffer, subjected to SDS-PAGE, and immunoblotted with an anti-Sec63p (1:3,000), anti-Kar2p (1:3,000), or anti-human c-myc (1:5,000) antibody. Times are indicated in minutes.

ment with 0.5 M NaCl, 0.1 M Na2CO3 (pH 11), or 1.6 M urea but was released into the soluble fraction upon treatment with 1% Triton X-100. As a control, Sec62p fractionated exclusively as an integral ER membrane protein (11). In contrast, the peripheral membrane protein, Sec23p (21), was partially solubilized with all conditions. Topology of Sec63p in the ER membrane. Analysis of the SEC63 sequence predicts three transmembrane domains. The DnaJ homology region is between the putative second and third transmembrane domains, leaving a 424-amino-acid carboxyl terminus. This structure could be accommodated in two possible orientations of Sec63p: the DnaJ-like region could face the ER lumen with the carboxyl terminus in the cytosol; alternatively, the DnaJ-like region could be oriented toward the cytosol with the carboxyl terminus in the ER lumen. These two predicted topologies suggest distinct protease sensitivities of Sec63p in cell homogenates. If the carboxyl terminus of Sec63p is luminal, Sec63p may be resistant to exogenous protease unless the membrane is solubilized with detergent. If Sec63p is in the alternate conformation, Sec63p should be sensitive to proteolysis. To distinguish between these two orientations, we tagged Sec63p on its C terminus with a human c-myc epitope. This construct complemented a chromosomal Asec63::HIS3 strain (data not shown). Samples of a medium-speed pellet fraction (see Materials and Methods) were subjected to digestion by a mixture of proteinase K and trypsin in the presence and absence of Triton X-100 (Fig. 4). After quenching of the reaction with 20% TCA, samples were prepared for SDS-PAGE, transferred to nitrocellulose, and immunoblotted with antisera raised against the ER luminal protein Kar2p (37), Sec63p, or human c-myc. In the absence of detergent, Sec63p was sensitive to protease, resulting in a smaller protease-resistant fragment (-68 kDa). This fragment did not react with the c-myc antibody, indicating that the carboxyl terminus of Sec63p is sensitive to protease digestion in the absence of detergent. A c-myc immune reaction was seen only in the zero-time protease control samples. Kar2p was resistant to protease unless the ER was solubilized. In the presence of detergent, two forms of Sec63p were produced, the -68-kDa

species seen in the absence of detergent and a new -42-kDa species that could correspond to an additional cleavage in a luminally exposed domain of Sec63p. The sensitivity of Sec63p to protease implies that a segment of the protein is exposed to the cytosol and a portion is protected by the ER membrane. The C-terminal segment containing the c-myc epitope is exposed to the cytosol. To refine the topology of Sec63p, we constructed a series of Sec63p-invertase hybrid proteins that had various amounts of the Sec63p carboxyl terminus deleted. The fusion of invertase to Sec63p domains that normally are directed to the cytosol will result in hybrid proteins with invertase in the cytosol, while fusions of invertase to domains of Sec63p that are ER luminal will result in invertase exposed to the ER lumen. Because invertase receives extensive N-linked glycosylation (48) and Sec63p is not glycosylated (data not shown), a luminal disposition of invertase would be detected by the presence of N-linked carbohydrate on the Sec63p-invertase fusion protein. A series of Sec63pinvertase hybrids was created by unidirectional exonuclease III deletions or by directing gene fusions to unique restriction sites in SEC63. The glycosylation state of the hybrid proteins was monitored by pulse-radiolabeling strain YT455 (Asuc2) containing various gene fusions in the presence or absence of the glycosylation inhibitor, tunicamycin. Extracts from radiolabeled cells were prepared, and invertase was immunoprecipitated with anti-invertase antiserum. The immunoprecipitates were subjected to SDS-PAGE and autoradiography (Fig. SB). Cells containing a plasmid with a termination codon between SEC63 and SUC2(pDF4) had a small amount of immunoreactive material that was not glycosylated, presumably because of an internal transcription promoter created by the gene fusion (Fig. SB, lanes 9 and 10). Cells expressing the Sec63pA28-, Sec63pA608-, or Sec63pA610-invertase hybrids contained immunoreactive material of the size expected for unglycosylated hybrid proteins (Fig. 5). In these cases, tunicamycin treatment had no effect on the size of the newly synthesized hybrid protein. A Sec63pA529-invertase hybrid protein, corresponding to a fusion junction in the DnaJ homology region, was exten-

VOL. 12, 1992

TOPOLOGY AND FUNCTIONAL DOMAINS OF Sec63p --I

A529 /D153A

A608

P152N

L--

I_

Complementation Null Sec63

-

-

Sec63- TS

B Deletion A28 A529 A608 Tuni F- + - + - +

A610 Vector _

1 _

97468 -

1

2 3

4

5 6

7

8

9

10

FIG. 5. Analysis of the topology of Sec63p, using Sec63p-invertase fusion proteins. (A) Schematic diagram of Sec63p. Boxed regions indicate predicted transmembrane domains based on the sequence of Sadler et al. (39). Endpoints of carboxy-terminal deletions used for invertase fusions are denoted. The DnaJ homology domain is indicated between the second and third transmembrane domains. (B) Wild-type cells transformed with the SUC2 vector pDF4 or with the indicated Sec63p-invertase fusion construct were pulse-labeled for 25 min with Tran-35S-label in the presence or absence of 10 ,ug of tunicamycin (tuni) per ml, as indicated. Glass bead extracts were treated with anti-invertase serum, and the immunoprecipitates were subjected to SDS-PAGE and autoradiography. A610 has two more amino acids deleted more than does A608. Sizes are indicated in kilodaltons.

sively glycosylated (Fig. 5; compare lanes 3 and 4). These data suggest that the carboxyl terminus of Sec63p faces the cytosol and that the DnaJ-like domain faces the ER lumen. The carboxyl terminus and DnaJ-like domains of Sec63p are essential. Carboxyl-terminal deletions of increasing length were evaluated for retention of function by complementation of sec63 mutations. A Sec63A14-invertase hybrid protein that had a deletion of eight negatively charged amino acids complemented the thermosensitivity of sec63-1 and the inviability of ASEC63. Deletion of just 14 more amino acids (7 more charged residues) of Sec63p resulted in the complete inability to complement sec63-1 or ASEC63 (Fig. SA). These results indicated that most of the carboxyl terminus of Sec63p was essential for Sec63p function. The DnaJ-like region of SEC63 shows 43% protein se-

quence identity to the N-terminal -70 amino acids of the bacterial protein (4, 35). Five other dnaJ homologs have been sequenced recently, including three yeast genes, SCJ1 (5), YDJJ, (7), and SIS1 (28), and two bacterial genes, the dnaJ genes from Caulobacter crescentus (18) and Mycobacteria tuberculosis (26). The alignments of the N termini of these proteins with Sec63p are shown in Fig. 6. The amino acids corresponding to proline 156 and aspartate 157 are identical in all DnaJ-like proteins. To test whether these residues were essential for Sec63p function, we changed proline 156 to asparagine and aspartate 157 to alanine by site-specific mutagenesis and introduced the mutant genes into sec63-1 or ASEC63. Neither mutant complemented the growth defect of the sec63 mutants when the genes were carried on either a centromere-based or 2,um plasmid (Fig. 5A), although both mutant proteins were expressed at approximately wild-type levels (data not shown). These results show that two invariant positions within the DnaJ homology region are required for Sec63p function. The Sec63p complex is not affected by DnaJ domain mutations. Sec63p can be cross-linked to four other proteins (Sec62p, Sec6lp, gp3l.5, and p23) that are precipitated from detergent-solubilized yeast membrane fractions by antibody against Sec62p (13). Sec62p is much less abundant than Sec63p; consequently, precipitation with Sec63p antibody detects a complex with only three of the subunits (Sec6lp, gp3l.5, and p23) (13). To test whether the integrity of this complex is affected by a nonfunctional dnaJ point mutant allele, we tagged mutant Sec63p (Pro-156--Asn Asp157- Ala double mutant) with a human c-myc epitope to allow the isolation of a complex from cells carrying mutant and wild-type copies of SEC63. Wild-type strains transformed with either a c-myc-tagged wild-type Sec63p or a tagged dnaJ mutant Sec63p were radiolabeled with Tran-355label and lysed, and membrane fractions were prepared and treated with the cleavable cross-linking reagent DSP either before or after detergent treatment. Samples were quenched, treated with SDS, and then mixed with an anti-human c-myc monoclonal antibody in the presence or absence of excess synthetic c-myc peptide. Immunoprecipitates were solubilized and resolved by SDS-PAGE, and radioactive proteins were visualized on a Molecular Dynamics PhosphoImager. Four prominent proteins corresponding to Sec63p, Sec6lp, gp3l.5, and p23 (13) were detected in samples precipitated from cross-linked membranes (Fig. 7, lanes 1 and 3). These species were detected in dnaJ mutant or wild-type fractions and failed to be precipitated in immune reactions that

IEI GISTSASDRDIKSAYRKLSVKF

SEC63 ScSi YDJ1 SIsi DnaJ (E.coli) DnaJ (M. tuberculosis) DnaJ (C. crecent us)

DI GVPVTATDVEIKKAYRKCALK H GVSPSANEQELKKGYRKAALK YEI GVSKTAEEREIRKAYKRLAMKY

SEC63 ScSi YDJ1

LTPDEKSVMEETQVQIT -NAGSEEA-HQKFIEVG

S151 DnaJ (E.coli) DnaJ (M.tuberculosis) DnaJ (C. crecentus)

3293

AI EIDKDATEKEIKSAYRQLSKK QE EI

GVSSDASPEEIKRAYRKLARDL W GVTNTIDEAGLKSNVNKLAMEH- .W YEST YDVIS

|ELVRQNXLKYGHPDGPQ PEKKKI DQFGADAVKN

-EAAEKFKEASA------ YEI S PEKRDI QTE--KFKEISE----- EI PQKREI RNQGDKEA-EAKFKEIK YE ISQK ANPGNPAAG-ERKFAVS HNS PEKRK RN-GGCENAAGRFKEIN YSVS SQK

DQFGEDGLSG DQYGLEAARS DQYGHAAFEQ DETRRIFAGG DRFGHAAGQR

FIG. 6. Comparison of Sec63p with DnaJ proteins in S. cerevisiae yeast, E. coli, M. tuberculosis, and C. crescentus. Shaded areas indicate identity in all seven proteins. Asterisks indicate the amino acid residues mutated and shown to be essential for Sec63p function.

3294

MOL. CELL. BIOL.

FELDHEIM ET AL. Whole membranes

Solubilized

membranes

mutant W.T.

mutant

+

+-

Peptide

+ L;

Sec63pSec6

W.T_

+

_

1p

gp3 1.5p23

1

2

3

4

5

6 7

8

FIG. 7. Sec63p complex with and without dnaJ domain mutation. DnaJ domain mutant forms of Sec63p assemble into the Sec63p complex. 35S-labeled membranes were prepared (see Materials and Methods), and aliquots were cross-linked with DSP before (lanes 1 to 4) or after (lanes 5 to 8) solubilization with 1% Triton X-100. Samples were immunoprecipitated with the human c-myc monoclonal antibody myc-9E10 in the absence (lanes 1, 3, 5, and 7) or presence (lanes 2, 4, 6, and 8) of competitor peptide. Samples were electrophoresed on an SDS-12.5% polyacrylamide gel, and exposed on a Molecular Dynamics PhosphoImager. Lanes: 1, 2, 5, and 6, RSY255 containing pDF50 (Sec63-Myc, P155N, D156A); 3, 4, 7, and 8, RSY255 containing pDF41 (Sec63-Myc). Species labeled Sec63p, Sec6lp, gp31.5, and p23 were confirmed by coincident electrophoresis with the cross-linked Sec protein complex precipitated with anti Sec63p antibody (13). W.T., wild type.

contained excess synthetic c-myc peptide (Fig. 7, lanes 1 to 4). Control experiments showed that antibody against Sec63p precipitated the complex in the presence or absence of synthetic c-myc peptide (not shown). Sec6lp association with the complex was labile. Cross-links formed in detergent-solubilized membrane fractions yielded a subcomplex consisting of Sec63p, gp3l.5, and p23 (Fig. 7, lanes 5 and 7). Again, no difference was seen in extracts of the dnaJ mutant and wild-type Sec63p, and all three proteins failed to be precipitated in the presence of competitor peptide. Thus, the dnaJ mutations affect some aspect of Sec63p orientation or function other than assembly into the multisubunit Sec protein complex. Since the dnaJ mutant proteins were expressed in a wild-type strain, it is possible that mutant and wild-type Sec63p coassemble into a functional complex. DISCUSSION SEC63 is an essential gene that is required for the translocation of a range of secretory proteins into the ER lumen. Sequence analysis of SEC63 predicts a 73-kDa protein with a highly negatively charged carboxyl terminus and an internal domain homologous to the E. coli DnaJ protein. Using antisera prepared against Sec63p, we have shown that Sec63p is a 73-kDa protein localized to the ER membrane. Characterization of Sec63p by using Sec63p-invertase fusion proteins predicts that Sec63p has three transmembrane domains with the DnaJ homologous region in the ER lumen and the carboxyl terminus in the cytosol. The carboxylterminal 28 amino acids are essential for Sec63p function, as are conserved residues in the DnaJ homology domain. Hybrid proteins have been used to map the topology of many E. coli inner membrane proteins (30, 40) as well as the yeast ER membrane proteins 3-hydroxyl-3-methylglutaryl coenzyme A reductase (42), Sec62p (11), and Secl2p (10).

This strategy appears to be applicable for many membrane proteins regardless of orientation in the membrane. Analysis of alkaline phosphatase fusions to the E. coli leader peptidase and the aspartate receptor correctly predict topology even though the first hydrophobic domains are in opposite orientations (31). Although our data do not address this point, an ER luminal orientation of the amino terminus of Sec63p is most likely. This unusual topology has been observed in only a few cases. Examples include E. coli leader peptidase and the ER cytochrome P-450 (8). In leader peptidase, the region immediately carboxyl terminal to the first transmembrane domain is highly charged: 15 of 39 amino acids are charged. This region is thought to act as a poison sequence prohibiting membrane assembly when placed on either side of an otherwise functional signal peptide (51). It is noteworthy that Sec63p has a similar charged region downstream of its first transmembrane domain: 23 of 52 amino acids downstream of the first hydrophobic domain are charged. Although Sec63p has no N-terminal signal peptide, the charged region may serve to force the insertion of the amino terminus and the DnaJ-like domain of Sec63p into the ER lumen. The method of using hybrid proteins to predict topology may not be applicable for all proteins. Analysis of the gene encoding 3-hydroxyl-3-methylglutaryl coenzyme A reductase, HMG1, with HIS4 fusions suggested that the third and fourth hydrophobic segments act in concert during assembly of the protein into the ER membrane (42). Proteins that span a membrane in a 3 sheet, such as the mitochondrial and E. coli porins, may assemble in a cooperative manner (16). Fusion protein analysis would not work in these instances. Several independent results support the conclusions concerning Sec63p topology derived from the hybrid protein analysis. A hybrid consisting of the entire Sec63p linked to invertase complements a sec63 null mutation, indicating that the invertase moiety does not interfere with Sec63p folding. The glycosylation state of several invertase fusions in the carboxyl-terminal domain of Sec63p produced a consistent pattern of unglycosylated hybrid proteins, whereas fusion to the DnaJ-like domain produced highly glycosylated species. Thus, the location of the fusion junction with respect to putative membrane anchor segments, rather the sequence of the junction, dictates the topological fate of the invertase reporter moiety. Finally, both glycosylated and nonglycosylated hybrid proteins are localized to the same nuclear envelope-ER network seen with authentic Sec63p (data not shown). All except the last few residues of the acidic carboxyl terminus of Sec63p are essential for translocation activity. The cytosolic location of this acidic domain of Sec63p suggests a possible means of interaction with Sec62p, which contains a basic carboxyl terminus situated similarly. However, as we have shown previously, the potentially electrostatically interacting domains are not the only determinants of Sec62p-Sec63p association. A nonfunctional Sec62-Suc2p hybrid that lacks the charged C terminus of Sec62p nevertheless interferes with the growth of a sec62-1 mutant strain in a manner that is partially compensated by overproduction of Sec63p (11). Thus, some region proximal to the C terminus of Sec62p interacts with and may be titrated by Sec63p. The DnaJ-like domain of Sec63p projects into the ER lumen. In analogy to the interactions of DnaJ and DnaK in E. coli, it is possible that the DnaJ domain of Sec63p physically interacts with Kar2p (yeast homolog of BiP/GRP78), a DnaK homolog involved in the translocation of presecretory proteins across the ER membrane (50). This proposal is sup-

VOL. 12, 1992

TOPOLOGY AND FUNCTIONAL DOMAINS OF Sec63p

DnaJ domain

N

FIG. 8. Model of topologies of Sec62p, Sec63p, and Kar2p. The DnaJ homology region of Sec63p is in bold type.

ported by two genetic experiments. First, thermosensitive alleles of sec63 and kar2 grow well at 25°C, yet the combination of these two mutations produces complete inviability (36a). The sec63-1 mutation results in a substitution at an invariant alanine residue (A181T) within the DnaJ domain (44a). Synthetic lethal combinations have been shown to occur with proteins known to interact, such as a- and

,B-tubulins (23) and the

two subunits of the mitochondrial

processing protease, Maslp and Mas2p (57a). Second, certain mutations in kar2 suppress the growth defect in strains

carrying the sec63-1 mutation in an allele-specific manner (36a). Allele-specific suppression is believed to be an indication of biological interaction (23). One example of allelespecific suppression of genes whose proteins have been shown to interact are the yeast genesACTl and SAC6. ACTI encodes a structural gene for actin, and SAC6 has been shown to encode an actin-binding protein (1). The predicted topology of Sec63p is consistent with the DnaJ domain being in the proper compartment to interact with Kar2p. A model combining predicted orientations of Sec62p (11) and Sec63p and the potential interaction with the Kar2p is shown in Fig. 8. How might the DnaJ-like domain of Sec63p serve to promote the activity of Kar2p? Sec63p comprises one cog in a multisubunit complex required for ER translocation. Kar2p may be anchored in proximity to translocating chains by direct association with the DnaJ-like loop. In this position, Kar2p may act directly on translocating polypeptides to facilitate refolding within the ER, thereby driving penetration to completion. Alternatively, by analogy to the proposed function of DnaJ and DnaK from E. coli, Kar2p may act via Sec63p to promote the association and dissociation of membrane subunits of the translocation complex, providing an element of regulation to the activity of the translocation machinery. It is also possible that the DnaJ domain and BiP participate in secretory polypeptide folding during and after the translocation event. Recently, we reported the detection of Sec63p in a multiprotein complex with Sec62p, Sec6lp, and two other unidentified proteins (13). Considering the models described above, we examined the complex for association with endogenous Kar2p. No reproducible association between Kar2p and wild-type or dnaJ mutant Sec63p was detected. The observation that normal levels and constituents of the complex associate with a nonfunctional dnaJ mutant Sec63p suggests that Kar2p may not be involved in assembly of the complex. While cross-linking has failed to reveal an associ-

3295

ation between Kar2p and Sec63p, the mild conditions reported here to isolate the c-myc-tagged Sec63p complex may allow a more functional study of the influence of Sec63p on catalytic activities of Kar2p. ACKNOWLEDGMENTS We thank Robin Wright for invaluable help with immunofluorescence, Arend Sidow for assistance with PCR, Karl Kuchler and Stan Grell for providing monoclonal antibody myc-9E10, David King for synthesizing the c-myc peptide, Colin Stirling, Christophe d'Enfert, Ray Deshaies, and Sylvia Sanders for stimulating discussions, Mark Rose, Pam Silver, Kim Arndt, Lucille Shapiro, and Mike Yaffe for providing data prior to publication, and Linda Wuestehube, Sylvia Sanders, Nancy Pryer, and Christophe d'Enfert for critical reading of the manuscript. We also thank Peggy McCutchan-Smith for expert assistance in preparing the manuscript. REFERENCES 1. Adams, A. E. M., D. Botstein, and D. G. Drubin. 1989. A yeast actin-binding protein is encoded by SAC6, a gene found by suppression of an actin mutation. Science 243:231-233. 2. Alfano, C., and R. McMacken. 1989. Ordered assembly of nucleoprotein structures at the bacteriophage A replication origin during the initiation of DNA replication. J. Biol. Chem. 264:10699-10708. 3. Bankaitis, V. A., L. M. Johnson, and S. D. Emr. 1986. Isolation of yeast mutants defective in protein targeting to the vacuole. Proc. Natl. Acad. Sci. USA 81:9075-9079. 4. Bardwell, J. C. A., K. Tilly, E. Craig, J. King, M. Zylicz, and C. 5.

6. 7. 8.

9.

10.

11.

12.

13.

14.

15.

16.

Georgopoulos. 1986. The nucleotide sequence of the Escherichia coli K12 dnaJ+ gene. J. Biol. Chem. 261:1782-1785. Blumberg, H., and P. Silver. 1991. A homologue of the bacterial heat-shock gene DnaJ that alters protein sorting in yeast. Nature (London) 349:627-630. Boeke, J., J. Trueheart, G. Natsoulis, and G. R. Fink. 1987. 5-Fluoro-orotic-acid as a selective agent in yeast molecular genetics. Methods Enzymol. 154:164-175. Caplan, A. J., and M. G. Douglas. 1991. Characterization of YDJ1: a yeast homologue of the bacterial dnaJ protein. J. Cell Biol. 114:609-621. Chirico, W. J., M. G. Waters, and G. Blobel. 1988. 70K heat shock related proteins stimulate protein translocation into microsomes. Nature (London) 332:805-810. DeLemos-Chiarandini, C., A. B. Frey, D. D. Sabatini, and G. Kreibich. 1987. Determination of the membrane topology of the phenobarbital-inducible rat liver cytochrome P-450 isozyme PB-4 using site-specific antibodies. J. Cell Biol. 104:209-219. d'Enfert, C., C. Barlowe, S. Nishikawa, A. Nakano, and R. Schekman. 1991. Structural and functional dissection of a membrane glycoprotein required for vesicle budding from the endoplasmic reticulum. Mol. Cell. Biol. 11:5727-5734. Deshaies, R., and R. Schekman. 1990. Structural and functional dissection of Sec62p, a membrane-bound component of the yeast endoplasmic reticulum protein import machinery. Mol. Cell. Biol. 10:6024-6035. Deshaies, R. J., B. D. Koch, M. Werner-Washburne, E. A. Craig, and R. Schekman. 1988. A subfamily of stress proteins facilitates translocation of secretory and mitochondrial precursor polypeptides. Nature (London) 332:800-805. Deshaies, R. J., S. L. Sanders, D. A., Feldheim, and R. Schekman. 1991. Assembly of yeast Sec proteins involved in translocation into the endoplasmic reticulum into a membrane-bound multisubunit complex. Nature (London) 349:806-807. Deshaies, R. J., and R. Schekman. 1987. A yeast mutant defective at an early stage in import of secretory protein precursors into the endoplasmic reticulum. J. Cell Biol. 105: 633-645. Deshaies, R. J., and R. Schekman. 1989. SEC62 encodes a putative membrane protein required for protein translocation into the yeast endoplasmic reticulum. J. Cell Biol. 109:26532664. Dihanich, M. 1990. The biogenesis and function of eukaryotic

3296

FELDHEIM ET AL.

porins. Experientia 46:146-153. 17. Evan, G. I., G. K. Lewis, G. Ramsay, and J. Michael Bishop. 1985. Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product. Mol. Cell. Biol. 5:3610-3616. 18. Gomes, S. L., J. W. Gober, and L. Shapiro. 1990. Expression of the Caulobacter heat shock gene dnaK is developmentally controlled during growth at normal temperatures. J. Bacteriol. 172:3051-3059. 19. Harlow, E., and D. Lane. 1988. Antibodies: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 20. Heinikoff, S. 1987. Unidirectional digestion with exonuclease III in DNA sequence analysis. Methods Enzymol. 155:156-165. 21. Hicke, L., and R. Schekman. 1989. Yeast Sec23p acts in the cytoplasm to promote protein transport from the endoplasmic reticulum to the Golgi complex in vivo and in vitro. EMBO J. 8:1677-1684. 22. Hill, J. E., A. M. Myers, T. J. Koerner, and A. Tzagoloff. 1986. Yeast/E. coli shuttle vectors with multiple unique restriction sites. Yeast 2:163-168. 23. Huffaker, T. C., M. A. Hoyt, and D. Botstein. 1987. Genetic analysis of the yeast cytoskeleton. Annu. Rev. Genet. 21:259284. 24. Innis, M. A., D. H. Gelfand, J. J. Sninsky, and T. S. White. 1990. PCR protocols: a guide to methods and applications, p. 177-183. Academic Press, Inc., San Diego, Calif. 25. Johnson, L. M., V. A. Bankaitis, and S. D. Emr. 1987. Distinct sequence determinants direct intracellular sorting and modification of a yeast vacuolar protease. Cell 48:875-885. 26. Lathigra, R. B., D. B. Young, D. Sweetser, and R. A. Young. 1988. A gene from Mycobacterium tuberculosis which is homologus to the DnaJ heat shock protein of E. coli. Nucleic Acids Res. 16:1636-1638. 27. Liberek, K., C. Georgopoulos, and M. Zylicz. 1988. Role of the Escherichia coli DnaK and DnaJ heat shock proteins in the initiation of bacteriophage X DNA replication. Proc. Natl. Acad. Sci. USA 85:6632-6636. 28. Luke, M. M., A. Sutton, and K. T. Arndt. 1991. Characterization of SIS1, a Saccharomyces cerevisiae homologue of bacterial dnaJ proteins. J. Cell Biol. 114:623-638. 29. Maniatis, T., E. F. Frisch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 30. Manoil, C., and J. Beckwith. 1986. A genetic approach to analyzing membrane protein topology. Science 233:1403-1408. 31. Manoil, C., J. J. Mekalanos, and J. Beckwith. 1990. Alkaline phosphatase fusions: sensors of subcellular location. J. Bacteriol. 172:515-518. 32. Nilsson, B., L. Abrahmsen, and M. Uhlen. 1985. Immobilization and purification of enzymes with staphylococcal protein A gene fusion vectors. EMBO J. 4:1075-1080. 33. Normington, K., K. Kohno, Y. Kozutsumi, M.-J. Gething, and J. Sambrook. 1989. S. cerevisiae encodes an essential protein homologous in sequence and function to mammalian BiP. Cell 57:1223-1236. 34. Nossal, N. G., and L. A. Heppel. 1966. The release of enzymes by osmotic shock from Escherichia coli in exponential phase. J. Biol. Chem. 241:3055-3062. 35. Ohki, M., F. Tamura, S. Nishimura, and H. Uchida. 1986. Nucleotide sequence of the Escherichia coli dnaJ gene and purification of the gene product. J. Biol. Chem. 261:1778-1781. 36. Pringle, J. R., R. A. Preston, A. E. M. Adams, T. Stearns, D. G. Drubin, B. K. Haarer, and E. W. Jones. 1989. Fluorescence microscopy methods for yeast. Methods Cell Biol. 31:358-435.

MOL. CELL. BIOL.

36a.Rose, M. Personal communication. 37. Rose, M. D., L. M. Misra, and J. P. Vogel. 1989. KAR2, a Karyogamy gene, is the yeast homolog of the mammalian BiPIGRP78 gene. Cell 57:1211-1221. 38. Rothblatt, J. A., R. J. Deshaies, S. L. Sanders, G. Daum, and R. Schekman. 1989. Multiple genes are required for proper insertion of secretory proteins into the endoplasmic reticulum in yeast. J. Cell Biol. 109:2641-2652. 39. Sadler, I., A. Chiang, T. Kurihara, J. Rothblatt, J. Way, and P. Silver. 1989. A yeast gene important for protein assembly into the endoplasmic reticulum and the nucleus has homology to DnaJ, an Escherichia coli heat shock protein. J. Cell. Biol. 109:2665-2675. 40. San Millan, J. L., D. Boyd, R. Dalbey, W. Wickner, and J. Beckwith. 1989. Use ofphoA fusions to study the topology of the Escherichia coli inner membrane protein leader peptidase. J. Bacteriol. 171:5536-5541. 41. Scott, J., and R. Schekman. 1980. Lyticase: endoglucanase and protease activities that act together in yeast cell lysis. J. Bacteriol. 142:414-423. 42. Sengstag, C., C. Stirling, R. Schekman, and J. Rine. 1990. Genetic and biochemical evaluation of eukaryotic membrane protein topology: multiple transmembrane domains of Saccharomyces cerevisiae 3-hydroxyl-3-methylglutaryl coenzyme A reductase. Mol. Cell. Biol. 10:672-680. 43. Shen, A., P. Chretien, L. Bastien, and S. Slilaty. 1991. Primary sequence of the glucanase gene from Oerskovia xanthineolytica. J. Biol. Chem. 266:1058-1063. 44. Sikorski, R. S., and P. Hieter. 1989. A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122:19-27. 44a.Silver, P. Personal communication. 45. Stanley, K. K., and J. P. Luzio. 1984. Construction of a new family of high efficiency bacterial expression vectors: identification of cDNA clones coding for human liver proteins. EMBO J. 3:1429-1434. 45a.Stirling, C. J., and E. W. Hewitt. 1992. The S. cerevisiae SEC65 gene encodes a component of yeast signal recognition particle with homology to human SRP19. Nature (London) 356:534-537. 46. Taussig, R., and M. Carlson. 1983. Nucleotide sequence of the yeast SUC2 gene for invertase. Nucleic Acids Res. 11:19431954. 47. Toyn, J., A. R. Hibbs, P. Sanz, J. Crowe, and D. I. Meyer. 1988. In vivo and in vitro analysis of ptll, a yeast ts mutant with a membrane-associated defect in protein translocation. EMBO J. 7:4347-4353. 48. Trimble, R., and F. Maley. 1977. Subunit structure of external invertase from Saccharomyces cerevisiae. J. Biol. Chem. 252: 4409-4412. 49. Vieira, J., and J. Messing. 1987. Production of single-stranded plasmid DNA. Methods Enzymol. 153:3-11. 50. Vogel, J. P., L. M. Misra, and M. D. Rose. 1990. Loss of BiPIGRP78 function blocks translocation of secretory proteins in yeast. J. Cell Biol. 110:1885-1896. 51. von HeiJne, G., W. Wickner, and R. E. Dalby. 1988. The cytoplasmic domain of Escherichia coli leader peptidase is a "translocation poison" sequence. Proc. Natl. Acad. Sci. USA 85:3363-3366. 51a.Yaffe, M. Personal communication. 52. Zylicz, M., 0. Ang, K. Liberek, and C. Georgopoulos. 1989. Initiation of X DNA replication with purified host- and bacteriophage encoded proteins: the role of the dnaK, dnaJ, and grpE heat shock proteins. EMBO J. 8:1601-1608.

Topology and functional domains of Sec63p, an endoplasmic reticulum membrane protein required for secretory protein translocation.

SEC63 encodes a protein required for secretory protein translocation into the endoplasmic reticulum (ER) of Saccharomyces cerevisiae (J. A. Rothblatt,...
2MB Sizes 0 Downloads 0 Views