8088 Correction Proc. Natl. Acad. Sci. USA 92 (1995) Cell Biology. In the article "Nascent polypeptide-associated complex protein prevents mistargeting of nascent chains to the endoplasmic reticulum" by Brett Lauring, Hideaki Sakai, Gert Kreibich, and Martin Wiedmann, which appeared in number 12, June 6, 1995 ofProc. Natl. Acad. Sci. USA (92, 5411-5415), the following correction should be noted. On p. 5413, right column, Fig. 2A, the sample shown in lane 6 had SRP present and not missing as indicated. Downloaded by guest on September 25, 2021 Proc. Natl. Acad. Sci. USA Vol. 92, pp. 5411-5415, June 1995 Cell Biology

Nascent polypeptide-associated complex protein prevents mistargeting of nascent chains to the endoplasmic reticulum BRETT LAURING*, HIDEAKI SAKAItt, GERT KREIBICH*, AND MARTIN WIEDMANNt§ *Department of Cell Biology, New York University School of Medicine, New York, NY 10016; and tCellular Biochemistry and Biophysics Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021 Communicated by David D. Sabatini, New York University Medical Center, New York NY, March 10, 1995

ABSTRACT We show that, after removal of the nascent chains. Both the inappropriate targeting and translocation of polypeptide-associated complex (NAC) from ribosome- signalless polypeptides (in the absence of NAC) are SRP associated nascent chains, ribosomes synthesizing proteins independent. lacking signal peptides are efficiently targeted to the endo- plasmic reticulum (ER) membrane. After this mistargeting, translocation across the ER membrane occurs, albeit less MATERIALS AND METHODS effi'ciently than for a nascent secretory polypeptide, perhaps In Vitro Transcription and Translation and Isolation of because the signal peptide is needed to catalyze the opening of Nascent Chain Complexes. In vitro transcription and transla- the translocation pore. The mistargeting was prevented by the tion of truncated mRNAs were as described (7). Before use, addition of purified NAC and was shown not to be mediated rabbit reticulocyte lysates (8) were centrifuged at 14,000 x g by the signal recognition particle and its receptor. Instead, it for 3 min. Truncated mRNAs were translated at 26°C for 20 appears to be a consequence of the intrinsic affinity of min, a temperature that best preserves the ribosome-nascent ribosomes for membrane binding sites, since it can be blocked chain complexes. After translation, 9 vol of dilution buffer by competing ribosomes that lack associated nascent polypep- [DB; 40 mM Hepes/0.5 M KOAc/5 mM Mg(OAc)2/2 mM tides. We propose that, when bound to a signalless ribosome- dithiothreitol (DTT), pH 7.5] was added and the ribosome- associated nascent polypeptide, NAC sterically blocks the site nascent chain complexes were recovered by centrifugation in the ribosome for membrane binding. (100,000 rpm, 40 min, 4°C, TLA 100.4 rotor; Beckman) through a 1.5-ml high-salt-containing sucrose cushion [HSS; The signal recognition particle (SRP) selects signal sequence- 0.5 M sucrose in DB supplemented with protease inhibitors (9) bearing ribosomes for targeting to the endoplasmic reticulum and 0.8 unit of RNasin per ,pl (Promega)]. The complexes were (ER) (1, 2). Even though SRP acts positively to select such resuspended in translation blank buffer (TBB), as described ribosomes for targeting, it may also be necessary for cells to (6), using 0.5 vol of the buffer unit volume of the original possess a mechanism that prevents nascent chains lacking translation mixture. Insoluble material was then removed by signal peptides from being mistargeted to, and consequently centrifugation at 14,000 x g for 10 min at 4°C. Recovery of the mistranslocated across, the ER membrane. In particular, it nascent chains was typically 50-75%. These complexes were seems likely that such mistargeting would otherwise occur free of NAC as assessed by Western blotting (data not shown) given the high affinity of ribosomes for binding sites in ER or by a photocrosslinking approach (6). membranes (3, 4), which are located near the translocon (5). Preparation ofRibosomes. Canine pancreas ribosomes were A heterodimeric protein, the nascent polypeptide- prepared by puromycin-high-salt stripping (10) of rough mi- associated complex (NAC), was recently purified on the basis crosomes (11). Yeast ribosomes were prepared from a trans- of its ability to bind to ribosome-associated nascent polypep- lation lysate (12). After pelleting through a 2.0 M sucrose tide chains as they are synthesized and is likely to be one of the cushion, ribosomes were resuspended in RBB [50 mM Hepes/ first cytosolic factors that contacts them (6). NAC was found 100 mM KOAc/5 mM Mg(OAc)2/2 mM DTT/0.8 unit of to bind to all regions or domains of ribosome-associated RNasin per ,lI/protease inhibitors] containing 1 mM puromy- nascent chains tested, with the notable exception of fully cin and 500 mM KOAc and then sedimented through HSS exposed signal peptides (6). NAC can therefore serve to ensure cushions. The ribosomal pellets were resuspended in RBB that only signal peptides remain available for SRP binding and containing 0.25 M sucrose. All ribosome preparations were thus could contribute to the fidelity of translocation. free of NAC as judged by Western blotting. Prior to use, We now present evidence for an additional and surprising ribosomal resuspensions were homogenized and insoluble role of the NAC protein. This is that its binding to non-signal material was removed by centrifugation at 14,000 x g for 10 regions of nascent chains serves to prevent the mistargeting of min. Molar ratios of competing ribosomes and nascent chains ribosomes containing nascent chains to the ER. We found that (see Fig. 3) were calculated based onA26o readings made in 1% in the absence of NAC, ribosomes bearing any nascent chain SDS. bind to the ER membrane regardless of whether or not the Nascent Chain Targeting Assay. Truncated ribosome- nascent chain contains a signal peptide or bound SRP. This nascent chain complexes in 0.5 vol of TBB lacking nucleotides targeting results from the intrinsic affinity of the ribosomes for and the energy-generating system were incubated with 200 nM binding sites at the ER membrane. A fraction of the nascent NAC or NAC buffer (6) for 2 min at 26°C prior to addition of chains lacking signal peptides brought to the membrane in this manner could subsequently be completely translocated. Puri- Abbreviations: ER, endoplasmic reticulum; SRP, signal recognition fied NAC was able to prevent this inappropriate targeting and particle; NAC, nascent polypeptide-associated complex; ffLuc, firefly translocation, most likely by sterically blocking the membrane luciferase; pPL, preprolactin; CAT, chloramphenicol acetyltrans- attachment site in the ribosome, after it binds to the nascent ferase. tPresent address: Department of Pharmacology, Nagasaki University, School of Dentistry, 1-7-1 Sakamoto, Nagasaki 852 Japan. The publication costs of this article were defrayed in part by page charge §To whom reprint requests should be addressed at: Cellular Biochem- payment. This article must therefore be hereby marked "advertisement" in istry and Biophysics Program, Memorial Sloan-Kettering Cancer accordance with 18 U.S.C. §1734 solely to indicate this fact. Center, 1275 York Avenue, New York, NY 10021. 5411 5412 Cell Biology: Lauring et at Proc. Natl. Acad Sci. USA 92 (1995) 0.3 equivalent of EDTA/KOAc-stripped rough microsomes 86 amino acids of the signal peptide-containing preprolactin per ,ul (11). After 3 min at 26°C and 5 min on ice, 20-,l samples (86aapPL), were translated in the rabbit reticulocyte lysate were mixed with 2.3 M sucrose in RBB to give a final sucrose translation system. After translation, ribosome-nascent chain concentration of 2.1 M. Samples were transferred to 750-,l complexes were isolated by centrifugation through high-salt- tubes and overlaid with 360 Al of 1.9 M sucrose in RBB. Tubes containing sucrose cushions, which removes the majority of were filled with RBB and then centrifuged (45,000 rpm, 1 hr, associated cytosolic factors, including NAC (6). These com- 4°C, SW 55 rotor; Beckman). Gradients were then frozen in plexes were used as substrates for binding to microsomal liquid nitrogen and cut into three fractions with a sharp Rambo membranes that had been stripped of their ribosomes with knife. The nascent chain content of each fraction was analyzed EDTA and KOAc (11). Importantly, no SRP was added to the by SDS/PAGE and fluorography. system. Samples were fractionated in sucrose gradients (3, 4), which were collected in three fractions (Fig. 1A). Top fractions contained the ER membranes that floated up together with RESULTS any targeted ribosome-nascent chain complexes, bottom frac- NAC Prevents Default Targeting to the ER of Ribosomes tions contained free, untargeted ribosome-nascent chain com- Containing Signalless Nascent Chains. Isolated nontranslat- plexes, and middle fractions were devoid of both ribosomes ing ribosomes bind to sites on ER membranes with an affinity and membranes. Fig. 1A (Left) shows that an approximately constant of -10-8 M. We modified the standard ribosome equal fraction of each type of nascent chain was targeted to the binding assay (3, 4) to enable us to measure the contributions ER membrane. It is not surprising that the signal peptide- of nascent chains and NAC to the ribosome-membrane inter- containing 86aapPL was targeted because it is known that, action. The 3' truncated mRNAs lacking stop codons were during in vitro translation in the rabbit reticulocyte lysate translated in vitro to generate truncated nascent polypeptide system, endogenous SRP binds to the signal peptide and this chains that remain stably associated with the ribosomes as binding is resistant to high-salt extraction (14). It is surprising, peptidyl-tRNAs (7). After purification, these ribosome- however, that the ribosome-nascent chain complexes lacking nascent chain complexes were used as substrates in the ribo- signal peptides (77aaffLuc and 76aaCAT) bound to the mem- some binding assay. brane to the same extent as those containing the truncated Truncated mRNAs encoding signalless polypeptides corre- secretory protein, because when they are not subjected to the sponding to the N-terminal 76 and 77 amino acids of the high-salt treatment only the latter bound to the membrane bacterial chloramphenicol acetyltransferase (76aaCAT) and (Fig. 1B). Since NAC, which binds to non-signal peptide the peroxisomal firefly luciferase (77aaffLuc), respectively, as domains of ribosome-nascent chain complexes, is removed by well as the mRNA for a polypeptide corresponding to the first treatment with the high-salt medium (6), we hypothesized that, A -NAC +WN T M B T M c 86aapPL D TMB TMB TMB TMB 77aaffLuc & g77aaffLuc _E B * 4W4No Sa ,- ql g77aaffLuc 76aaCAT - TMB TMB v TEj# 56aaPL 77aaffLuc A&~~~~~ta0 N- 0 . aTRAP - - X L S HSLS HS. 1 2 345 1 23 45 56 1 2 3 4 5 6 77aaffLuc 86aapPL NAC - - + - + 1 2 FIG. 1. NAC regulates ribosome binding to the ER membrane. (A) NAC prevents the membrane association of ribosomes containing signal-free nascent chains but not of those containing a bona fide nascent secretory protein. High-salt-stripped 86aapPL, 77aaffLuc, and 76aaCAT ribosome-nascent chain complexes were incubated either without (lanes 1-3) or with (lanes 4-6) 200 nM NAC for 3 min at 26°C and 5 min on ice prior to addition of EDTA/KOAc-stripped rough microsomes (0.3 equivalent/,A) (11). After incubation for 3 min at 26°C and 5 min on ice, 20-,ul samples were underlaid in discontinuous sucrose gradients and centrifuged. The gradients were fractionated and, after trichloroacetic acid precipitation, the nascent chain content of each fraction was analyzed by SDS/PAGE (15% acrylamide gels), followed by fluorography. Top fractions (T) contained the membranes with bound ribosome-nascent chain complexes; bottom fractions (B) contained free, untargeted ribosomes. The distribution of aTRAP, a marker for rough ER membranes (26), assessed by Western blotting is shown. (B) 77aaffLuc ribosome-nascent chain complexes are not targeted to the ER membrane if they are prepared under conditions that do not strip NAC from the nascent chains. 77aaffLuc (lanes 1-3) and 86aapPL (lanes 4-6) ribosome-nascent chain complexes were prepared as usual except that the dilution buffer and the sucrose cushions contained 100 mM KOAc (low salt) rather than 500 mM KOAc. The complexes were incubated with membranes, fractionated, and analyzed exactly as in A. T, top fraction; M, middle fraction; B, bottom fraction. (C) 77aaffLuc ribosome-nascent chain complexes targeted to the ER membrane in the absence of NAC become bound -by a high-salt-resistant linkage. After incubation of 77aaffLuc or 86aapPL containing ribosome-nascent chain complexes with membranes as inA, samples were adjusted, as indicated, to high salt (HS; 0.5 M KOAc) prior to flotation or kept as controls at the usual low salt (LS; 100 mM KOAc) concentration. The bottom layers in the gradients contained the same salt concentrations as the samples. After centrifugation for 80 min, the gradients were fractionated as inA. The elevated salt concentration slightly altered the flotation of the membranes (data not shown). Lanes represent the top (T), middle (M), or bottom (B) layers of the gradients. (D) NAC prevents translocation of nascent chains from salt-stripped ribosome-nascent chain complexes that lack signal sequences in their nascent polypeptides but not of those that contain them. NAC was added back (200 nM) to near its original concentration. Aliquots of salt-stripped ribosome-nascent chain complexes were preincubated, as indicated, with or without NAC (200 nM), as inA, or with puromycin (1 mM; lane 1) before the addition of SRP-depleted, KOAc-washed rough microsomes (14) at a concentration of 0.1 equivalent/pul and incubation for 3 min at 26°C and 5 min on ice. After this targeting, all samples were adjusted to 1 mM puromycin and 0.5 M KOAc and incubated for 20 min at 37°C and then 15 min on ice to release chains from the ribosomes. NAC blocked translocation of 77aaffLuc (compare lanes 2 and 3) but not of 86aapPL (lanes 4 and 5). g77aaffLuc and 56aaPL indicate positions of the translocated chains. Phosphorlmager analysis indicates that the translocation efficiency of 86aapPL is 10-fold more efficient than that of 77aaffLuc. (E) Translocation of nascent polypeptides lacking signal peptides occurs in the presence of a cytosol depleted of NAC. Bovine brain cytosol was depleted of NAC by passage over a heparin-agarose column, as described (6). This procedure removes >90% of NAC but <10% of total protein from the cytosol. Stripped 77aaffLuc ribosome-nascent chain complexes were incubated with either cytosol (lane 1) or NAC-depleted cytosol (lane 2) for 5 min at 26°C prior to addition of EDTA/KOAc-washed rough microsomes (0.1 equivalent/pl). After 5 min at 26°C, 1 mM puromycin was added and samples were incubated at 37°C for 20 min and on ice for 15 min to release polypeptides from the ribosomes. Samples were analyzed as in D. Cell Biology: Lauring et al Proc. Natl. Acad Sci USA 92 (1995) 5413 when present, NAC may block the interaction of ribosomes A synthesizing proteins lacking signal peptides with the ER 0L (9D G) membrane. Fig. 1A (Right) shows that this is, indeed, the case. ._ oL i+ ~-~ I-HgC B The addition of NAC prevented targeting to the membrane of X F-n (. 0« Z ribosome-nascent chain complexes that lack signal peptides -L C'D +o (76aaCAT, 77aaffLuc) but not of those complexes with chains M that do contain a signal (86aapPL). g77aaffLuc -OF-C 77aaffLuc Whereas nonprogrammed ribosomes can readily be ex- 86aapPL tracted from microsomal membranes with high salt, the bind- ing of ribosomes containing nascent chains of secretory pro- 56aaPL teins is known to be resistant to salt extraction (13, 15). Fig. 1 C apyrase --: - - + 1 2 34 shows that for both salt-stripped 77aaffLuc- and 86aapPL- FIG. 2. Translocation of 77aaffLuc can proceed independently of containing complexes, significant and approximately equal SRP. (A) High-salt-stripped ribosome-nascent chain complexes were fractions of targeted nascent chains were resistant to high-salt resuspended in TBB lacking nucleotides and the energy-generating extraction. Therefore, by this criterion, in the absence of NAC, system. The nucleotides (1 mM ATP, 1 mM GTP, 2 mM GMP-PNP) the strength of the association of the two types of ribosome- were then added (14) as indicated. All samples received 0.1 equivalent nascent chain complexes with the membrane appears to be of SRP-depleted rough microsomes per ,ul that were prepared by equivalent. high-salt and puromycin stripping (10). Lane 5, sample received 20 nM We next determined to what extent the nascent chains SRP (18) and was then incubated for 5 min at 26°C prior to addition targeted in the absence of NAC could be translocated across of the stripped rough microsomes. Samples were incubated for 5 min the ER membrane. Translocation of the signal-containing at 26°C followed by puromycin treatment as described in the legend to Fig. 1D. After trichloroacetic acid precipitation, samples were ana- 86aapPL was assessed by the appearance of a lower molecular lyzed by SDS/PAGE and fluorography. To rule out that the mem- weight band that results from signal peptide cleavage, whereas branes were contaminated with residual levels of nucleotides, one translocation of the signalless 77aaffLuc was assessed by the sample contained stripped rough microsomes that had been pretreated appearance of a band of lower electrophoretic mobility that with apyrase (lane 6) as described (14). g77aaffLuc indicates the results from N-glycosylation (6). Fig. 1D demonstrates that, as glycosylated and translocated product. (B) Ribosome-nascent chain previously reported (6), upon release from the ribosomes with complexes containing 86aapPL either lacking the various nucleotides puromycin, both types of nascent chains targeted in the or supplemented with them, as indicated, were all incubated with 20 absence of NAC were translocated. It should be noted, how- nM SRP for 5 min at 26°C prior to addition of puromycin-high-salt- ever, that although in the absence of NAC roughly equal stripped rough microsomes and incubation for 5 min at 26°C. Samples proportions of 86aapPL and 77aaffLuc ribosome-nascent were treated with puromycin and processed as in A. 56aaPL indicates chain complexes became associated with the ER membrane the signal cleaved and translocated product. (see Fig. 1A), the extent of translocation across the membrane proteins in our cytosol-depleted conditions. We exploited the was considerably more efficient (- 10-fold) when a signal fact, known from previous studies with secretory proteins (19), peptide was present, a fact consistent with the idea that the that guanine nucleotide binding to SRP is necessary for it to signal peptide functions in gating the translocation pore (16, be released from the nascent polypeptide, which in turn is a 17). The translocation of the 77aaffLuc depended on the requirement for the subsequent translocation (19). Thus, as ribosome bringing the nascent chain to the ER, for if the shown in Fig. 2B, in the presence of SRP, translocation of the nascent chain was released from the ribosome with puromycin signal peptide-containing 86aapPL nascent chains required prior to addition of the microsomes, no translocation occurred GTP (lanes 2 and 4) or a nonhydrolyzable analog of it (lane 1). (Fig. 1D, lane 1). As expected from the effect of NAC on In effect, when nucleotides were omitted (lane 3), SRP binding ribosome binding, restoration of NAC to the ribosome- to the nascent chain became a trap that blocked further steps nascent chain complexes before addition of the membranes in its translocation 2A prevented translocation of the 77aaffLuc but not of the (14, 19). Fig. demonstrates that, on the 86aapPL (lanes 3 and 5). other hand, translocation of 77aaffLuc nascent chains was not Thus, NAC is sufficient to prevent mistargeting ofribosomes blocked when guanine nucleotides were omitted. Further- and the resultant mistranslocation. This function is likely to be more, addition of GTP with or without ATP or addition of the specific for NAC, as the translocation of 77aaffLuc still nonhydrolyzable GTP analogue GMP-PNP did not increase occurred in the presence of a cytosol that was depleted of NAC the extent of 77aaffLuc translocation (Fig. 2A, compare lanes by heparin chromatography (Fig. 1E) as described (6). 2-4 to lane 1). Finally, pretreatment of the membranes with Translocation of Proteins Lacking Signal Peptides Can apyrase to remove any residual nucleotide did not abolish Proceed Independently of SRP. The targeting and resultant transport (Fig. 2A, lane 6). Thus, although SRP has been translocation of nascent chains lacking signal peptides, just shown to be capable of associating with ribosome-nascent described, occurred in the absence of added SRP. In fact, Fig. chain complexes lacking signal peptides (6), the inability of 2A shows that the addition of SRP to a reaction mixture SRP to "trap" 77aaffLuc complexes in the absence of guanine containing salt-stripped ribosome-nascent chain complexes nucleotides demonstrates both that SRP associated differently and SRP-depleted microsomes-i.e., rough microsomes with the two types of ribosome-nascent chain complexes and treated with puromycin and high salt (10) that contain no that its function is not necessary for translocation of the 54-kDa SRP detectable by Western blotting (data not mistargeted signalless nascent chains. shown)-did not result in a significant increase in the trans- The Intrinsic Affinity of Ribosomes for Binding Sites in the location of 77aaffLuc nascent chains [compare lanes 1-4 (no ER Membrane Determines the Targeting and Translocation of SRP) with lane 5 (20 nM SRP)]. Signalless Polypeptides. Since after depletion of NAC, chains Since it has been previously shown that, in the absence of lacking signals can be translocated in vitro independently of NAC, SRP can inappropriately interact with non-signal pep- SRP, it seemed likely that targeting under these conditions is tide regions of nascent chains (6) and since it remains formally mediated directly by the intrinsic affinity of ribosomes for ER possible that small, undetected levels of contaminating SRP membranes. To assess this possibility, a competition assay was (either on the SRP-depleted microsomes or with the isolated performed using as competitors nonprogrammed ribosomes ribosome-nascent chain complexes) were responsible for the obtained from dog pancreas rough microsomes by puromycin/ mistargeting, we sought to determine whether blocking the high-salt treatment, which also strips them of nascent chains. function of SRP affects the translocation of nonsecretory These were mixed in increasing concentrations with a fixed 5414 Cell Biology: Lauring et aL Proc. Natl. Acad Sci. USA 92 (1995) amount of salt-extracted 77aaffLuc ribosome-nascent chain in the cytosol but is bound to the ER membrane (20). After it complexes. All ribosome preparations were found to be free of was demonstrated in vitro that rough microsomal fractions NAC, as assessed by immunoblotting (data not shown). Be- could be disassembled into both ribosomal and membranous cause the luciferase nascent chains lack a signal peptide and subfractions, it was possible to reconstitute the ribosome- both GTP and SRP were omitted from these assays, we could membrane junction by binding nontranslating ribosomes to focus on the contribution of the ribosome itself to targeting. ER-derived membranes (3, 15) that had been stripped of their The nontranslating ribosomes prepared from canine pancreas ribosomes. blocked the transport of the 77aaffLuc nascent chains (Fig. 3, lanes 1-5). Rat liver and rabbit reticulocyte ribosomes were The signal hypothesis maintains that information necessary also able to compete (data not shown). The unlikely possibil- for sorting or targeting ribosomes to the ER membrane is not ities that the competing ribosomes acted by either merely related to ribosomal components but instead is contained in displacing the truncated nascent chains from the ribosome- the particular product being synthesized on each ribosome- nascent chain complexes or by releasing the polypeptides from namely, in its signal sequence (21, 22). Positive selection of the tRNAs were eliminated by the findings that addition of the nascent chains by SRP is clearly critical to explain how proteins competing ribosomes did not affect the number of nascent containing signal peptides are targeted but does not explain in chains that could be recovered with sedimentable ribosomes any demonstrated way how, given the affinity of ribosomes for (lanes 15-18) or the amount of peptidyl-tRNAs that could be ER binding sites, the cell could prevent the majority of precipitated (7) by the cationic detergent cetyltrimethylam- ribosome binding sites from being occupied by ribosomes monium bromide (lanes 11-14). Furthermore, when the non- synthesizing proteins lacking signal peptides (which in nonse- programmed ribosomes were added only after the ribosome- cretory cell types might be a large fraction of the active nascent chain complexes had been preincubated with the ribosomes). The NAC protein now can be seen to resolve this membranes, and hence had been given the opportunity to be dilemma. targeted to their binding sites and have the nascent chain The present results make it clear that NAC critically con- complexes become engaged with the translocation apparatus, tributes to maintaining fidelity in targeting not only, as shown no competition was observed (lane 6). These results indicate that nontranslating ribosomes can block targeting of salt- in earlier work (6), by preventing inappropriate binding of SRP stripped ribosome-nascent chain complexes by competing for to signalless polypeptides, but also by preventing the binding of ribosome binding sites in the ER membrane and that they do ribosomes bearing those polypeptides to the ER membrane. not impair the translocation reaction per se. The competition Thus, NAC and SRP appear to play complementary roles in is specific for ribosomes, as neither 10 ,tM albumin (data not ensuring the specificity of translocation. shown) nor a NAC-depleted cytosol (6) (see Fig. 1E) blocked It is significant that NAC can interact with very short chains translocation. Moreover, yeast ribosomes were much less (<35 amino acids; see ref. 6), for this implies that NAC is likely effective than mammalian ribosomes in blocking translocation to be one of the first nonribosomal factors that an emerging (Fig. 3, lanes 7-10). polypeptide encounters. NACwould therefore be in a position, with respect to both the ribosome and the nascent chain, to prevent inappropriate interactions with the translocon. DISCUSSION The SRP and signal peptide-independent targeting ob- The elucidation of the secretory pathway was initially linked to served in this work, as defined by ribosome binding and the observation that a subpopulation of ribosomes is not free resistance to salt extraction, is an efficient process. Although 1 2 3 4 5 6 7 8 9 10 1112131415 161718 the translocation of the signalless polypeptide was much less efficient than that of pPL, we believe that both are effected by the same cellular machinery and that the lower efficiency of * translocation of 77aaffLuc may reflect a requirement for a signal peptide to gate the translocation channel (16, 17). Thus, we have previously shown (6) that the translocation of the 77aaffLuc that occurs in the absence of NAC cannot be carried 0 2 4 816(16)1 3 6 11 0 4 8 16 0 4 8 16 Canine pancreas Yeast CTABr Sedimented a. FIG. 3. Ribosome binding can target nascent chains independently of SRP. Mammalian ribosomes lacking nascent chains can prevent the translocation of signalless polypeptides by competing with the corre- sponding ribosome-nascent chain complexes for ribosome binding NAC depletion n1 sites on the ER membrane. For 10-,ul assays, 3 ,ul of salt-stripped 77aaffLuc ribosome-nascent chain complexes in TBB lacking nucle- b. c.a otides was mixed with unprogrammed dog pancreas ribosomes at the molar ratios indicated below each lane prior to addition of 1 equivalent of KOAc-washed rough microsomes. After 5 min at 26°C, 1 mM puromycin and 0.5 M KOAc were added and the incubation continued at 37°C for 20 min to induce translocation (lanes 1-5). In one sample (lane 6), the competing ribosomes were added only after incubation of the ribosome-nascent chain complexes with the KOAc-washed rough microsomes for 5 min at 26°C. Asterisk indicates glycosylated and translocated product. Competing ribosomes do not release the nascent Translocation chains from the tRNA or displace the truncated nascent chains from the ribosomes. After components were mixed and incubated for 5 min FIG. 4. A model for the role of NAC in blocking mistargeting of at 26°C, as in lanes 1-10, but without puromycin, nascent chains ribosomes containing signalless nascent chains. It is proposed that the present as peptidyl-tRNAs were recovered by precipitation with the ribosome contains a membrane attachment site (M) near the site of cationic detergent cetyltrimethylammonium bromide (CTABr) (lanes exit of the nascent chain from the large ribosomal subunit. When 11-14). Alternatively, after the incubation, ribosome-associated ribosomes are synthesizing nascent chains lacking signal peptides (a), polypeptide chains were recovered by sedimentation of the ribosomes NAC binds to the nascent chain and therefore covers the M site, which through high-salt sucrose cushions (lanes 15-18). Samples were ana- in turn prevents mistargeting from occurring. Removal of NAC by lyzed by SDS/PAGE and fluorography. Molar excess of competing high-salt stripping (b) results in exposure of the M site. Mistargeting ribosomes is indicated below each lane. and the resultant translocation (c) can then occur. Cell Biology: Lauring et aL Proc. Natl. Acad Sci USA 92 (1995) 5415 out if the membranes were pretreated with N-ethylmaleimide, Scientist Training Program (B.L., Grant ST 32 GM 07308), the an agent that is known to block the translocation of secretory American Cancer Society (G.K., Grant CB-111A), and the Sloan- polypeptides. We have also been able (unpublished observa- Kettering Institute (M.W.). tions) to crosslink, using a photoactivatable reagent, the 77aaffLuc to Sec 61p, a core component of the translocon (10). 1. Walter, P. & Blobel, G. (1981) J. Cell Biol. 91, 551-558. 2. Gilmore, R. (1993) Cell 75, 589-592. Whereas SRP positively selects for ribosomal targeting by 3. Borgese, N., Mok, W., Kreibich, G. & Sabatini, D. D. (1974) J. binding to signal peptides as they emerge from the ribosome, Mol. Biol. 88, 559-580. NAC by binding only to non-signal peptide regions prevents 4. Kreibich, G., Marcantonio, E. E. & Sabatini, D. D. (1983) Meth- the targeting that the ribosome otherwise would mediate ods Enzymol. 96, 520-530. through its direct interaction with the ER membrane ribosome 5. Yu, Y. H., Sabatini, D. D. & Kreibich, G. (1990) J. Cell Bio. 111, receptors. It should be noted that we have not yet been able to 1335-1342. remove the SRP from the 86aapPL nascent chain complexes 6. Wiedmann, B., Sakai, H., Davis, T. A. & Wiedmann, M. (1994) and thus to determine whether NAC, by binding to portions of Nature (London) 370, 434-440. the nascent chain following the signal, would also be able to 7. Gilmore, R., Collins, P., Johnson, J., Kellaris, K. & Rapiejko, P. prevent the binding of such SRP-depleted ribosomes to the ER (1991) Methods Cell Biol. 34, 223-239. membrane. 8. Jackson, R. J. & Hunt, T. (1983) Methods Enzymol. 96, 50-74. An attractive possibility for the mechanism of action ofNAC 9. Erickson, A. H. & Blobel, G. (1983) Methods Enzymol. 96,38-50. 10. Gorlich, D., Prehn, S., Hartmann, E., Kalies, K. U. & Rapoport, would be that when bound to a nascent chain, it sterically T. A. (1992) Cell 71, 489-503. blocks a ribosomal membrane binding site (M site) located 11. Walter, P. & Blobel, G. (1983) Methods Enzymol. 96, 84-93. near the nascent chain exit site in the large ribosomal subunit, 12. Waters, M. G. & Blobel, G. (1986) J. Cell Biol. 102, 1543-1550. which interacts directly with a ribosome receptor(s) (23-25) at 13. Gilmore, R. & Blobel, G. (1985) Cell 42, 497-505. the ER membrane (Fig. 4). Under other circumstances, when 14. High, S., Flint, N. & Dobberstein, B. (1991) J. Cell Bio. 113, a signal peptide is synthesized, SRP rather than NAC binds, 25-34. halting translational elongation and leading to targeting via 15. Adelman, M. R., Blobel, G. & Sabatini, D. D. (1973) J. Cell Biol. interaction with the SRP receptor. 56, 206-229. 16. Simon, S. M. & Blobel, G. (1992) Cell 69, 677-684. Note Added in Proof. Consistent with the data presented in Fig. 2 17. Crowley, K. S., Liao, S., Worrell, V. E., Reinhart, G. D. & showing that SRP interacts differently with signal peptide-containing, Johnson, A. E. (1994) Cell 78, 461-471. as opposed to signalless, nascent chains, Ted Powers and Peter Walter 18. Walter, P. & Blobel, G. (1983) Methods Enzymol. 96, 682-691. (personal communication) and Ines Moller and M.W. (unpublished 19. Connolly, T. & Gilmore, R. (1989) Cell 57, 599-610. observations) have shown that more SRP cofractionates with ribo- 20. Palade, G. E. (1975) Science 189, 347-358. somes synthesizing proteins containing signals than with those syn- 21. Blobel, G. & Sabatini, D. D. (1971) in Biomembranes, ed. Man- thesizing signalless polypeptides. son, L. A. (Plenum, New York), Vol. 2, pp. 193-195. 22. Blobel, G. & Dobberstein, B. (1975) J. Cell Biol. 67, 835-851. We thank Drs. Y. Yu and B. Wiedmann for helpful suggestions and 23. Lake, J. A. (1985) Annu. Rev. Biochem. 54, 507-530. comments. We also thank Drs. H. Lodish and J. Rothman and 24. Kreibich, G. & Sabatini, D. D. (1992) Curr. Biol. 2, 90-92. members of the Wiedmann, Kreibich, and Rothman laboratories for 25. Kalies, K.-U., Gorlich, D. & Rapoport, T. A. (1994) J. Cell Biol. critically commenting on the manuscript. Special thanks also to Drs. 126, 925-934. M. Adesnik and D. Sabatini for their time and effort in greatly 26. Hartmann, E., Gorlich, D., Kostka, S., Otto, A., Kraft, R., improving the clarity of the manuscript. This work was supported by Knespel, S., Burger, E., Rapoport, T. A. & Prehn, S. (1993) Eur. Heidi Chen and the New York University Medical Center Medical J. Biochem. 214, 375-381.