ABSTRACT

The signal recognition particle receptor consists of two subunits of 72 kDa (SR) and 30 kDa (SR). Assembly of SR on the endoplasmic reticulum membrane can occur independent of the signal recognition particle-mediated translocation pathway. To identify the sequences within SR necessary for membrane binding, a series of amino-terminal and internal deletion mutants was constructed and translated in a cell-free system. In addition, nascent SR polypeptides of varying lengths were generated by cycloheximide treatment of translation reactions. Microsome binding assays performed on these polypeptides revealed a membrane binding domain consisting of the amino-terminal 140 residues of SR. This domain includes the two hydrophobic sequences originally proposed to bind to membranes and a highly charged region not previously implicated in membrane assembly. Furthermore, the domain forms a protease-resistant folding unit that after proteolysis can target and anchor onto microsomes. Extraction of microsomal SR at high pH supplemented with 1 M NaSCN suggests that SR and the membrane binding domain are not integrated in the endoplasmic reticulum membrane. The membrane binding domain is also the major site of tight binding with SR, suggesting that SR plays a role in the membrane assembly of SR.

INTRODUCTION

In mammalian cells, secretory signal sequences of nascent polypeptide chains are bound by the ribonucleoprotein signal recognition particle (SRP)()as they emerge from the ribosome. Targeting to polypeptide translocation sites on the endoplasmic reticulum (ER) membrane then occurs via the interaction of SRP with the SRP receptor on the cytoplasmic face of the ER membrane (1, 2). The major components of this targeting pathway are conserved in eukaryotes and possibly in prokaryotes (for review, see Ref. 3).

The SRP receptor has been isolated as a heterodimer of two polypeptides that migrate in SDS-PAGE as 72-kDa (SR) and 30-kDa (SR) species (4). Both subunits are resistant to extraction from the membrane with urea or high salt and have been characterized as integral membrane proteins by resistance to extraction at high pH(2, 4, 5) . Protease dissection of SR on microsomes or purified by affinity chromatography revealed a translocation active cytoplasmic fragment of about 58 kDa and a fragment of about 14 kDa containing a putative membrane anchor(5, 6, 7) . The cDNA for SR encodes a 638-residue polypeptide containing two stretches of hydrophobic amino acids (residues 1-22 and 64-79) near the amino terminus that were proposed to serve as membrane anchors, as well as three clusters of charged (mostly basic) residues between residues 84 and 243(8) . The cytoplasmic elastase fragment of SR was shown to consist of the sequence from residue 152 to the carboxyl terminus and contains a GTP binding site(8, 9) . The cytoplasmic elastase fragment can assemble on trypsin-digested membranes to restore translocation activity, suggesting that it may bind SR directly(10) . SR is predicted from the primary amino acid sequence to have a single transmembrane domain near the amino terminus and a GTP binding site near the cytoplasmic carboxyl terminus(11) .

SR has previously been shown to target and anchor onto the ER membrane in vitro by a mechanism independent of the SRP-mediated pathway(10) . Membrane assembly and functional reconstitution of SR can occur post-translationally and in the absence of GTP or ATP. Cell-free synthesized SR can also restore SRP-mediated translocation activity to microsomes in which the endogenous SR has been inactivated by digestion with trypsin or by alkylation of free sulfhydryls. The binding of SR onto trypsin-digested microsomes is labile to urea, suggesting that the subunit is not assembled on the membrane by spontaneous insertion into the lipid bilayer(10) .

The exact mechanism by which SR assembles on the membrane is unknown. Furthermore, the sequences within SR required for interaction with SR have not been identified. To investigate these issues, we have assayed deletion mutants of SR translated in a cell-free system for salt-resistant binding onto ER microsomes. An amino-terminal domain of SR including amino acids 1-140 was found to be necessary for membrane binding. Immunoprecipitation experiments indicate that the domain is also responsible for binding to SR. The SR membrane binding domain appears to be an independent folding unit that is tightly bound to SR but not integrated into the ER membrane. A new model of SR membrane assembly is proposed in which both hydrophobic and hydrophilic regions ofSR anchor the protein to the membrane primarily by interacting with the transmembrane SR.

EXPERIMENTAL PROCEDURES

Materials and General Methods

Transcription reactions with SP6 polymerase were performed as described previously(14) . Cell-free translation reactions were performed in rabbit reticulocyte lysate (RRL) and labeled with [S]methionine as described previously(12) ; translation products were analyzed by SDS-PAGE (15, 16) followed by fluorography. Canine pancreatic rough microsomes were obtained as described, and either extracted with 0.5 M KOAc (KRMs) or washed by Sepharose CL-2B gel exclusion chromatography (CRMs)(13) .

Polyclonal antisera against SR and SR were raised in rabbits injected with purified bacterially overexpressed fusion proteins. Plasmid pMAC142 contained the sequence encoding amino acids 39-295 of SR inserted into pRIT-2T (Pharmacia) resulting in a fusion protein with S. aureus Protein A. Plasmid pMAC359 encoded amino acids 208-265 of canine SR fused to glutathione S-transferase in the vector pMAC241, a modification of pGEX-2T (Pharmacia) with an enhanced polylinker. The SR and SR fusion proteins were purified using IgG-Sepharose and glutathione-Sepharose columns, respectively. Other antisera were kind gifts of J. J. M. Bergeron, R. Gilmore, and T. Rapoport.

Plasmids

Plasmid pMAC3 encodes the polypeptide SRD1, containing amino acids 79 to the stop codon of SR and therefore having the two hydrophobic regions deleted from the amino terminus of SR. Plasmid pMAC456 encodes the polypeptide SRD3 in which residues 156-250 of SRN are deleted, removing part of the second and all of the third charged regions of SR. Plasmid pMAC55 encodes the polypeptide SRD4 containing an initial methionine followed by a glycine residue and residues 28 to the stop codon of SR, deleting the first hydrophobic region of SR.

Plasmid pMAC205 encodes the first 176 amino acids of SRN followed by Ser-Asn-Tyr-Ser-Arg-stop codon. This polypeptide, SRX2, includes the two hydrophobic regions and the first two charged regions of SR. Plasmid pMAC268 encodes SRX3, containing the polypeptide sequence Met-Gly-Ala-Pro followed by amino acids 28 to the stop codon of SRX2 and deleting the first hydrophobic region from SRX2. Plasmid pMAC135 encodes SRX6, containing residues 1-38, Asn-Ser and residues 79 to the end of SRX2, thereby deleting the second hydrophobic region of SRX2. Plasmid pMAC362 encodes SRX7, containing residues 1-79 and 103 to the stop codon of SRX2 and deleting the first charged region of SRX2.

Plasmid pMAC459 encodes SRD6, containing the sequence of SRN with amino acids 39-79 replaced by Asn-Ser and thus deleting the second hydrophobic region from SRN. Plasmid pMAC494 encodes the polypeptide SRD7, having the sequence of SRN with amino acids 79-103, and therefore the first charged region, deleted.

Plasmid pMAC455 codes for a chimaeric SR polypeptide (SR-MD), containing the first 29 residues of mouse SR followed by the predicted transmembrane and cytoplasmic domains of canine SR. The chimaeric polypeptide was used because the cDNA sequence of canine SR was incomplete and the encoded protein was missing the initiation site and an unknown number of amino-terminal residues. However, the missing residues were predicted to be in the ER lumen (11) and less likely to interact with SR. The lumenal domain of canine SR was therefore replaced with the complete amino-terminal lumenal domain of mouse SR, and the DNA sequence encoding this polypeptide was inserted into the vector pSPUTK. For immunoprecipitation experiments, plasmid pMAC690 was constructed encoding SR-MD with two copies of the influenza hemeagglutinin epitope tag at the amino terminus (HASR-MD). The sequence of the epitope tag was provided by inserting the DNA encoding SR-MD into the plasmid pG7SCTHA2(35) . The resulting coding sequence was inserted behind the SP6 promoter of plasmid pMAC334, a version of pGEM3 with the 5`-untranslated region of pSPUTK and the 3`-untranslated region of bovine preprolactin. Plasmid pMAC508 encoding the integral membrane protein SSgPA has been previously reported(38) .

Cell-free Translations and Membrane Targeting

Incomplete nascent SRN polypeptides of different lengths were generated by terminating cell-free translation reactions at various times with 1 mM cycloheximide. To assay membrane targeting of these polypeptides, a 20-µl aliquot of each reaction was incubated with 5 equivalents of KRMs for 5 min at 24 °C. An equal volume of buffer containing 1 M NaCl, 50 mM EDTA, and 20 mM Tris-Cl, pH 8.0, was added at 4 °C. The mixture was layered over a 100-µl sucrose step gradient containing 500 mM sucrose, 500 mM NaCl, 25 mM EDTA, and 20 mM Tris, pH 8.0, and the membranes were pelleted by centrifugation in an Airfuge at 20 psi (100,000 g) for 10 min. The top 75 µl (supernatant) was recovered, and peptidyl-tRNA was precipitated by adding 500 µl of 2% cetyltrimethylammonium bromide (CTAB) and 500 µl of 0.5 M NaOAc, pH 5.0(17) . Equivalent portions of the pellet and supernatant fractions were analyzed by SDS-PAGE.

Proteolytic Digestions

Proteinase K digestions of KRMs at 1 equivalent/µl were performed for 1 h at 4 °C with either 0 or 10 µg/ml proteinase K. The reactions were terminated with 1 mM phenylmethylsulfonyl fluoride and 2 µg/ml aprotinin. The microsomes were adjusted to 500 mM NaCl and pelleted in an Airfuge at 20 psi (100,000 g) for 10 min. Immunoblots were probed for SR and visualized using an alkaline phosphatase color reaction.

Membrane Extractions and Immunoprecipitations

Microsomes were extracted with high pH following a modified procedure based on the published assay(21) . 2 ml of KRMs at 1 equivalent/µl were loaded onto a 100-ml Sepharose CL-2B gel exclusion column equilibrated and eluted with 1 M NaSCN, 0.2 M NaCO, pH 11.5, and 10 mM dithiothreitol. 1.5-ml fractions were collected and concentrated by trichloroacetic acid precipitation for SDS-PAGE analysis. Immunoblots were visualized as above.

For immunoprecipitations of the SR mutants with SR-MD, 10-µl RRL translation reactions were mixed with 10-µl reactions of SR-MD after translation was complete and incubated at 24 °C for 30 min. The mixtures were then diluted in 500 µl of buffer (100 mM Tris-Cl, pH 8.0, 100 mM NaCl, 1% Triton X-100) at 4 °C, and the translation products were isolated using a monoclonal Sepharose affinity matrix. To prepare the affinity matrix, IgG against SR was purified from ascites fluid (4) and coupled to CNBr-activated Sepharose. As controls, 10-µl translation reactions of SRN, the deletion mutants and SR-MD were immunoprecipitated using the same monoclonal Sepharose.

To co-precipitate various SR mutants with HASR-MD, RRL translation reactions synthesizing HASR-MD were carried out in the presence of KRMs. A 5-µl aliquot of the HASR-MD reaction was incubated with a 30-µl translation reaction of each SR mutant at 24 °C for 30 min. The mixture was loaded onto a 0.8-ml Sepharose CL-2B column equilibrated and eluted with buffer containing 250 mM NaCl, 100 mM KOAc, 10 mM Tris-OAc, pH 7.5, 2.5 mM MgCl, and 1 mM dithiothreitol. Fractions containing the excluded volume of the column were pooled, adjusted to 350 mM NaCl, 5% glycerol, and 1% Triton X-100, and HASR-MD was recovered using monoclonal antibodies against the hemeagglutinin epitope and Protein G Affi-Gel (Pharmacia).

RESULTS

Sequences within SR Required for Membrane Binding

In a previous study of the membrane assembly of SR, anchored and loosely bound molecules could be separated by a simple pelleting assay in the presence of 2 M urea(10) . However, this assay could not clearly distinguish membrane-bound polypeptides from large insoluble aggregates. Therefore, to assay the deletion mutants for tight membrane binding, translation reactions containing microsomes were fractionated by Sepharose CL-2B gel exclusion chromatography at high ionic strength. Endogenous microsomal SR was identified by immunoblotting and found to elute solely in the excluded volume of the columns (fraction4, marked with arrowhead for all columns used in Fig. 2, data not shown). The included volume (fractions7-12 in all assays) was determined using the endogenous hemoglobin in the RRL lysate. Because the included volume eluted as a broad peak, fractions7, 9, and 11 are shown in Fig. 2as representative fractions.

The two hydrophobic stretches in SR were deleted in SRD1 and, as expected, this polypeptide did not fractionate with microsomes as it was recovered only in the included volume (Fig. 2b, compare lanes1-6 with lanes7-12). Constructs that removed only the first (SRD4) or second (SRD6) of the hydrophobic sequences were also assayed. Although a fraction of SRD6 aggregated in RRL and therefore is recovered in fractions 5 and 6, the aggregates were clearly resolved from fraction 4 containing membranes (Fig. 2d, compare lanes8 and 9 to lane7). Neither SRD4 nor SRD6 were able to bind efficiently onto microsomes (Fig. 2, c and d, compare lanes1 and 7). Surprisingly, a construct (SRD7) that left both the hydrophobic regions intact but deleted an adjacent section of strongly charged residues (amino acids 79-103) was also unable to bind efficiently onto microsomes (Fig. 2e, compare lanes1 and 7). Although analysis of this molecule was complicated by the presence of large aggregates (Fig. 2e, lane1), there was still a large portion of unaggregated polypeptide (Fig. 2e, lanes10-12) that was expected to be targetted to the membrane. As a control, a construct with a deletion in a region of the SR sequence containing numerous basic amino acids (residues 156-250) but containing an intact amino terminus (SRD3) was found to fractionate with microsomes as expected (Fig. 2f, compare lanes1 and 7).

These results suggested that sequences beyond the predicted membrane anchor of SR (8) may be required for membrane assembly. The carboxyl-terminal domain of SR (residues 152-638) has been shown to target to translocation sites on the ER but not anchor to the membrane in a manner resistant to high salt or urea concentrations(1, 10) . To directly examine the membrane assembly of the amino-terminal region of SR, a construct (SRX2) containing the first 176 amino acids of SRN was assayed. Although the putative carboxyl-terminal targeting domain was deleted from SRX2, the polypeptide bound onto microsomes ( Fig. 2g, compare lanes1 and 7). Therefore, there are at least two targeting sequences in SR, but only the amino-terminal sequence mediates tight binding onto membranes. To analyze the sequence of SRX2 further, plasmids were constructed with deletions within the SRX2 coding region (SRX3, SRX6, and SRX7, Fig. 1). However, cell-free synthesized SRX3, SRX6, and SRX7 were unable to clearly bind onto membranes and formed very large aggregates in the presence or absence of microsomes (data not shown).

The positively charged amino acid sequence deleted in SRD7 may be specifically required for membrane binding. However, it is also possible that the deleted sequence is not itself involved in membrane assembly but that deletion adversely affects protein folding around an adjacent membrane anchor sequence. To address this issue, cell-free translation reactions of SRN were terminated with cycloheximide at different times after initiation to generate a series of ribosome-bound peptidyl-tRNA translation intermediates with a range of lengths. Ribosome-bound nascent polypeptides prepared in this manner should be free of aggregates. The reactions were incubated with microsomes to allow targeting of the nascent chains and then adjusted to 500 mM NaCl and 25 mM EDTA. The membranes were separated by centrifugation and analyzed for the presence of bound polypeptides. Peptidyl-tRNA was precipitated from the supernatant with CTAB (17) to recover nascent chains not bound to the microsomes. It was expected that if the amino-terminal hydrophobic regions of SRN (up to around residue 80) were sufficient for membrane binding while attached to ribosomes, then polypeptides of molecular weight greater than or equal to 13 kDa (corresponding to about residue 120, presuming 40 amino acids at most are sequestered within the ribosome(22, 23) ) would be detected in the membrane fraction. On the other hand, if membrane binding required sequences beyond the hydrophobic regions, then only larger polypeptides (approximately 190 amino acids for a 150 residue membrane binding domain) would be recovered with the microsomes.

Nascent SRN polypeptides of discrete sizes from 10 kDa upward (estimated by migration in SDS-PAGE) could be detected after precipitation with CTAB (Fig. 3, lanes7-12). The CTAB-precipitated products reflected the polypeptides present in the total translation reaction (Fig. 3, lane13). However, no polypeptides smaller than a 23-kDa translation intermediate were recovered with microsomes (Fig. 3, lanes1-6). This 23-kDa product was chased to full-length SRN when the translation reactions were allowed to proceed for 1 h and is therefore a true translation intermediate (data not shown). The deletion mutant SRX2 containing 180 amino acids also migrates as a 23 kDa band, suggesting the 23-kDa nascent polypeptide contains a similar number of residues. This is too large to consist of the hydrophobic regions of SRN alone, but it is consistent with a membrane binding domain of approximately 140 amino acids. These data therefore support the hypothesis that sequences carboxyl-terminal to the hydrophobic regions of SR are necessary for membrane assembly.

Domain Structure of the SR Membrane Binding Sequence

The comparative resistance of the amino-terminal fragment of SR to proteinase K digestion whether or not the protein is attached to membranes suggests that the anchoring domain forms a folded unit. To test this directly, the deletion mutants SRX2, SRX3, SRX6, and SRX7 (see Fig. 1) were assayed for resistance to Proteinase K. SRX2 contains the complete amino-terminal domain of SRN, and the other polypeptides have deletions within the SRX2 sequence. Cell-free translation reactions of the polypeptides were digested with 10 µg/ml Proteinase K on ice for up to 1 h and analyzed at intermediate time points. The SRX3, SRX6, and SRX7 polypeptides were rapidly degraded under these conditions, with less than 30% of the initial populations remaining after 20 min (Fig. 4b). In contrast, more than 60% of the initial population of SRX2 polypeptide remained after 40 min of digestion (Fig. 4b). Interestingly, these data are reflected in the membrane binding behavior of SRX3, SRX6, and SRX7 reported above. In addition, the deletion mutants SRD4, SRD6, and SRD7 containing deletions in the SRX2 region of SRN cannot bind onto membranes (Fig. 1). Taken together, these data suggest that the deletions within SRX2 lead to misfolding, and that SRX2 forms an independently folded protein domain.

Membrane Binding of SR Correlates with Binding to SR

In studies of the translocation and membrane integration of proteins in cell-free systems, it has been observed that extraction with high pH alone was not always sufficient to distinguish peripherally bound proteins from integrated polypeptides(24) . To increase the stringency of the high pH extractions, microsomes were extracted in buffer containing 1 M NaSCN, 0.2 M NaCO, pH 11.5, and 10 mM dithiothreitol, and membranes were separated from extracted material by Sepharose CL-2B gel exclusion chromatography. Fractions were analyzed for the presence of SR and SR by immunoblot analysis. As controls, the immunoblots were also probed for the integral membrane proteins SSR and the 48-kDa subunit of oligosaccharyl transferase (OST48), the cytosolic protein actin, and the 54-kDa subunit of the peripheral membrane SRP (SRP54)(25, 26, 27, 28, 29, 30) .

Under these conditions, microsomal SR was detected solely in the predetermined excluded volume of the column, corresponding to fractions 4-6 (Fig. 5a, lanes1-3). While a fraction of microsomal SR eluted in fractions 4 and 5 (Fig. 5a, lanes1 and 2), the majority of the SR polypeptides eluted in a broad peak between fractions 24 and 32 (represented by fractions24, 28, and 32, Fig. 5a, lanes8-10). Visual inspection of this experiment and replicate trials indicated that approximately 20% or less of the SR population remained on membranes in the excluded volume. As expected, the integral membrane control proteins SSR and OST48 were observed almost exclusively in the membrane fractions (Fig. 5a, lanes1-3), while actin and peripherally bound SRP54 eluted in a broad peak centered around fraction 30 (Fig. 5a, lanes9-11). It appears that while perturbation of the membrane at pH 11 was not sufficient to extract SR(5) , most SR polypeptides can be clearly separated from integral membrane ER proteins under the conditions used here.

Surprisingly, both SR and SR were detected only in the aqueous supernatant (Fig. 5b, compare lanes1and 2). The partitioning of SR into the aqueous phase is consistent with its strongly hydrophilic primary sequence (8) and the apparently anomalous membrane interaction demonstrated in Fig. 5a. While SR appears to be integral membrane in high pH extractions supplemented with 1 M NaSCN (Fig. 5a), it is possible that the tight interaction between the receptor subunits (4) causes SR to partition in the aqueous phase with SR. We therefore digested microsomes with 5 µg/ml trypsin for 1 h at 4 °C to proteolyze SR while leaving SR unaffected (10) and then solubilized the membranes as above. After partitioning, tryptic fragments of SR (Fig. 5b, lane3), but no full-length protein, were detected in the aqueous phase, and SR was detected predominantly in the detergent phase (Fig. 5b, compare lanes3 and 4). This agrees with recent results indicating the integral membrane nature of SR(11) . As expected, in our solubilization conditions, OST48 was observed almost entirely in the detergent phase after cloud point separation (Fig. 5b, compare lanes5 and 6), and calreticulin partitioned solely into the aqueous phase (Fig. 5b, compare lanes7 and 8).

These results suggest that SR is anchored largely by binding to the transmembrane SR polypeptide. To determine if this interaction is mediated by the membrane binding domain of SR, we assayed the SR deletion mutants used to map the SR anchoring domain for the ability to bind SR in co-immunoprecipitations. A cDNA encoding canine SR was available but lacked a complete amino terminus(11) . However, a complete cDNA of mouse SR was available(11) , so a plasmid coding for a hybrid murine/canine SR (SR-MD) was constructed. The mouse and dog sequences are highly homologous (88% identity, 93% similarity), both having a single putative transmembrane domain and a carboxyl-terminal GTP-binding consensus sequence predicted to be on the cytoplasmic side of the ER(11) . The SR-MD hybrid was constructed to contain the amino-terminal lumenal domain of mouse SR and the transmembrane and carboxyl-terminal cytoplasmic domains of canine SR. The junction between the sequences was selected because the binding site for SR was expected to be in the transmembrane or cytoplasmic domain of SR.

RRL translation reactions of SR-MD were mixed with translation reactions of various SR deletion mutants and immunoprecipitated using monoclonal antibodies against SR(4) . A fraction of SR-MD was observed to co-precipitate with SRN (Fig. 6a, lanes1 and 4) and SRD3 polypeptides (Fig. 6a, lane3), but not with SR-EF ( Fig. 6a, lane 2). Furthermore, SR-MD did not co-precipitate with SRD4 or SRD1 (Fig. 6a, lanes5 and 6) nor with SRD6 or SRD7 (data not shown). Control experiments (data not shown) suggest that the relatively poor co-precipitation of SR-MD with SRN is likely due to inefficient formation of dimers in the absence of membranes. As expected in control immunoprecipitations of translation reactions containing only SRN, the deletion mutants or SR-MD, no protein bands corresponding to SR-MD were observed (data not shown). These results suggest that the polypep-tide sequences involved in the membrane anchoring of SR are sufficient for interaction with SR in the absence of membranes.

DISCUSSION

We have shown here that a folded amino-terminal membrane binding domain of SR containing hydrophobic and charged amino acids is required for tight binding to SR. The SR membrane binding sequence contains approximately 140 residues and forms an independently folded protein domain (Fig. 4b). Membrane binding is observed when this region of SR is generated by proteolysis of intact molecules either before or after targeting to microsomes (Fig. 4a) or by cell-free synthesis as an isolated polypeptide (Fig. 2g and 3). Furthermore, the SR membrane binding domain binds directly to SR in the absence of other membrane proteins or lipids (Fig. 6a). Despite the presence of a membrane targeting signal within the carboxyl-terminal domain of SR (10), deletion of either hydrophobic or charged sequences from the amino-terminal domain abolishes tight binding to the membrane (Fig. 2) and to SR (Fig. 6b). Our results therefore suggest that the membrane binding domain of SR is not inserted into the membrane (Fig. 5), but the entire domain is involved in binding to SR. The remarkably strong interaction between the subunits is resistant to 1% nonionic detergent and high ionic strength (Fig. 6b), pH 11, and 2 M urea (2, 4, 5, 10) and most likely requires both hydrophobic and nonhydrophobic interactions. While the exact sequences within the SR membrane binding domain that are in direct contact with SR remain to be determined, we expect they will include polar and nonpolar amino acids.

This model is not contradicted by the primary sequence of SR, as the two hydrophobic regions within the anchoring domain are of comparatively low hydrophobicity and are both broken by lysine residues (8). Although the data cannot entirely discount the possibility of interactions between the membrane lipids and SR, the relative extractibility of membrane bound SR in Fig. 5a suggests that these interactions are not typical of a membrane protein with even a single transmembrane domain. Our results are more consistent with the hydrophobic regions of SR contributing to intersubunit contacts.

Hydrophobic interactions alone are not sufficient for receptor dimer assembly, as the deletion mutant SRD7 that has the same hydrophobic sequences as full-length SR was unable to bind SR (Fig. 6b). The importance of nonhydrophobic protein-protein interactions is further demonstrated by the cosegregation of both subunits as a complex in the aqueous phase after cloud point extraction (Fig. 5b). Moreover, SR could be dissociated from membrane-bound SR by the combined disruption of polar and hydrophobic interactions with pH 11.5 and 1 M NaSCN, without solubilizing the microsomal lipid bilayer (Fig. 5a).

A slight molar excess of SR over SR on the ER membrane has been reported (1.1 mol of SR/mol of SR)(4) . In our model, novel SR polypeptides targetted to the ER membrane would be anchored via these unpaired SR molecules. Anchoring of novel SR is saturable at a concentration similar to that of excess SR on the membrane(10) . The identity of the trypsin-sensitive membrane component required for anchoring of novel SR (10) is still unresolved. However, despite the apparent resistance of SR to protease digestion(10, 11) , our results suggest that SR is the required factor. The SR membrane binding domain within SRX2 is necessary and sufficient for co-precipitation of SR and SR (Fig. 6), and, similar to full-length SR, the binding of SRX2 onto trypsin-treated microsomes is labile to urea (data not shown).

The domain of SR required for membrane anchoring and tight binding to SR has been demonstrated to be unnecessary for functional assembly of the receptor on the ER membrane(10) . Therefore, tight binding between the receptor subunits is not required for receptor activity. This suggests a dual role for SR, as a membrane anchor for SR and as a part of the translocation machinery. A specific role for SR in translocation has not been directly demonstrated, but the GTPase activity of SRP54 requires binding to the SRP receptor(31) , and SR has been shown to be labeled in vitro with GTP(11) .

The two-domain structure of SR is likely evolutionarily conserved. The sequence of a homologue of SR has been obtained from yeast and contains a complete amino-terminal sequence(32) . A yeast homologue of SR has now been identified(11) , and we predict a similar pattern of interactions between these proteins. The E. coli homolog of SR, FtsY, begins at residue 126 of the canine sequence (32, 33) and thus corresponds closely to the carboxyl-terminal domain of mammalian SR. Interestingly, a bacterial homologue of SR has not been identified. Since the mammalian SR anchoring domain that mediates binding to the subunit is absent in FtsY, there may not be a homologue of SR in E. coli. However, FtsY has been reported to be resistant to high pH extraction despite the absence of hydrophobic domains(34) , suggesting that it may also bind to an integral membrane protein.

As demonstrated in Fig. 3, nascent SR polypeptides can assemble on microsomes while still attached to ribosomes. Since membrane binding appears to require a folded amino-terminal domain to interact with SR, targeting in this manner would still be essentially post-translational. However, this suggests that folding of SR and receptor dimer assembly can occur cotranslationally, at least in vitro. While post-translational targeting of SR molecules has been demonstrated in vitro(10) , the subunit may assemble co-translationally in vivo. We have therefore begun to investigate the possibility that SR assembles onto the membrane during a pause in translation.

FOOTNOTES

*

This work was supported by a Medical Research Council of Canada grant. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

§

To whom correspondence should be addressed. Tel.: 905-525-9140 (ext. 2075); Fax: 905-522-9033.

The abbreviations used are: SRP, signal recognition particle; ER, endoplasmic reticulum; PAGE, polyacrylamide gel electrophoresis; RRL, rabbit reticulocyte lysate; SR and SR, SRP receptor and subunits; KRMs, canine pancreatic rough microsomes extracted with 0.5 M KOAc; CRMs, CL-2B column washed canine pancreatic rough microsomes; CTAB, cetyltrimethylammonium bromide; OST48, the 48 kD subunit of oligosaccharyl transferase.

REFERENCES

1. Gilmore, R., Blobel, G., and Walter, P. (1982) J. Cell Biol.95, 463-469 [Abstract/Free Full Text]
2. Gilmore, R., Walter, P., and Blobel, G. (1982) J. Cell Biol.95, 470-477 [Abstract/Free Full Text]
3. Wolin, S. L. (1994) Cell77, 787-790 [CrossRef][Medline] [Order article via Infotrieve]
4. Tajima, S., Lauffer, L., Rath, V. L., and Walter, P. (1986) J. Cell Biol.103, 1167-1178 [Abstract/Free Full Text]
5. Hortsch, M., Avossa, D., and Meyer, D. I. (1985) J. Biol. Chem.260, 9137-9145 [Abstract/Free Full Text]
6. Meyer, D. I., and Dobberstein, B. (1980) J. Cell Biol.87, 498-502 [Abstract/Free Full Text]
7. Meyer, D. I., and Dobberstein, B. (1980) J. Cell Biol.87, 503-508 [Abstract/Free Full Text]
8. Lauffer, L., Garcia, P. D., Harkins, R. N., Coussens, L., Ullrich, A., and Walter, P. (1985) Nature318, 334-338 [CrossRef][Medline] [Order article via Infotrieve]
9. Römisch, K., Webb, J., Herz, J., Prehn, S., Frank, R., Vingron, M., and Dobberstein, B. (1989) Nature340, 478-482 [CrossRef][Medline] [Order article via Infotrieve]
10. Andrews, D. W., Lauffer, L., Walter, P., and Lingappa, V. R. (1989) J. Cell Biol.108, 797-810 [Abstract/Free Full Text]
11. Miller, J. D., Tajima, S., Lauffer, L., and Walter, P. (1995) J. Cell Biol.128, 273-282 [Abstract/Free Full Text]
12. Jackson, R. J., and Hunt, T. (1983) Methods Enzymol.96, 50-74 [Medline] [Order article via Infotrieve]
13. Walter, P., and Blobel, G. (1983) Methods Enzymol.96, 84-93 [Medline] [Order article via Infotrieve]
14. Gurevich, V. V., Pokrovskaya, I. D., Obukhova, T. A., and Zozulya, S. A. (1991) Anal. Biochem.195, 207-213 [CrossRef][Medline] [Order article via Infotrieve]
15. Laemmli, U. K. (1970) Nature227, 680-685 [CrossRef][Medline] [Order article via Infotrieve]
16. Schagger, H., and von Jagow, G. (1987) Anal. Biochem.166, 368-379 [CrossRef][Medline] [Order article via Infotrieve]
17. Hobden, A. N., and Cundliffe, E. (1978) Biochem. J.170, 57-61 [Medline] [Order article via Infotrieve]
18. Bordier, C. (1981) J. Biol. Chem.256, 1604-1607 [Abstract/Free Full Text]
19. Andrews, D. W., Gupta, J., and Abisdris, G. (1993) Mol. Cell. Biol.13, 4703-4713 [Abstract/Free Full Text]
20. Theisen, N., Schulze Lohoff, E., von Figura, K., and Hasilik, A. (1986) Anal. Biochem.152, 211-214 [CrossRef][Medline] [Order article via Infotrieve]
21. Fujiki, Y., Hubbard, A. D., Fowler, S., and Lazarow, P. B. (1982) J. Cell Biol.93, 97-102 [Abstract/Free Full Text]
22. Malkin, L. I., and Rich, A. (1967) J. Mol. Biol.26, 329-346 [CrossRef][Medline] [Order article via Infotrieve]
23. Blobel, G., and Sabatini, D. D. (1970) J. Cell Biol.45, 130-145 [Abstract/Free Full Text]
24. Andrews, D. W., Young, J. C., Mirels, L. F., and Czarnota, G. J. (1992) J. Biol. Chem.267, 7761-7769 [Abstract/Free Full Text]
25. Wiedmann, M., Görlich, D., Hartmann, E., Kurzchalia, T. V., and Rapoport, T. A. (1989) FEBS Lett.257, 263-268 [CrossRef][Medline] [Order article via Infotrieve]
26. Görlich, D., Prehn, S., Hartmann, E., Herz, J., Otto, A., Kraft, R., Wiedmann, M., Knespel, S., Dobberstein, B., and Rapoport, T. A. (1990) J. Cell Biol.111, 2283-2294 [Abstract/Free Full Text]
27. Wada, I., Rindress, D., Cameron, P. H., Ou, W.-J., Doherty, J. J., II, Louvard, D., Bell, A. W., Dignard, D., Thomas, D. Y., and Bergeron, J. J. M. (1991) J. Biol. Chem.266, 19599-19610 [Abstract/Free Full Text]
28. Kelleher, D. J., Kreibich, G., and Gilmore, R. (1992) Cell69, 55-65 [CrossRef][Medline] [Order article via Infotrieve]
29. Silberstein, S., Kelleher, D. J., and Gilmore, R. (1992) J. Biol. Chem.267, 23658-23663 [Abstract/Free Full Text]
30. Walter, P., and Blobel, G. (1982) Nature99, 691-698
31. Miller, J. D., Wilhelm, H., Gierasch, L., Gilmore, R., and Walter, P. (1993) Nature366, 351-354 [CrossRef][Medline] [Order article via Infotrieve]
32. Ogg, S. C., Poritz, M. A., and Walter, P. (1992) Mol. Biol. Cell3, 895-911 [Abstract]
33. Gill, D. R., Hatfull, G. F., and Salmond, G. P. C. (1986) Mol. & Gen. Genet.205, 134-145
34. Luirink, J., ten Hagen-Jongman, C. M., van der Weijden, C. C., Oudega, B., High, S., Dobberstein, B., and Kusters, R. (1994) EMBO J.13, 2289-2296 [Medline] [Order article via Infotrieve]
35. Glover, J. R., Andrews, D. W., and Rachubinski, R. A. (1994) Proc. Natl. Acad. Sci. U. S. A.91, 10541-10545 [Abstract/Free Full Text]
36. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory, NY
37. Falcone, D., and Andrews, D. W. (1991) Mol. Cell. Biol.11, 2656-2664 [Abstract/Free Full Text]
38. Janiak, F., Glover, J. R., Leber, B., Rachubinski, R. A., and Andrews, D. W. (1994) Biochem. J.300, 191-199
39. Michalak, M., Milner, R. E., Burns, K., and Opas, M. (1992) Biochem. J.285, 681-692

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles: