Protein transport into canine pancreatic microsomes A quantitative approach Silvia Guth*, Christian Vo¨ lzing*, Anika Mu¨ ller, Martin Jung and Richard Zimmermann Medizinische Biochemie und Molekularbiologie, Universita ¨ t des Saarlandes, Homburg, Germany Transport o f p resecretory proteins i nto t he mammalian rough e ndoplasmic reticulum involves a protein translocase that comprises the integral membrane proteins Sec61ap, Sec61bp, and Sec61cp as core components. Electron microscopic analysis of protein tr anslocase i n r ough m icro- somal membranes suggested that between three and four heterotrimeric Sec61p complexes form th e central unit of protein translocase. Here we analyzed the stoichiometry of heterotrimeric Sec61p complexes p resent in cotranslationally active protein translocases of canine p ancreatic microsomes and various other lumenal and membrane components believed to be subunits of protein translocase and to be involved in covalent modifications. Based on these numbers, the c apacity f or cotra nslational transport was estimated for the e ndoplasmic reticulum of the human pancreas. Keywords: endoplasmic reticulum; mammalian microsomes; protein s ecretion; protein transport; pancreas. Transport o f presecretory proteins into mammalian rough microsomes involves cleavable signal peptides at the N-terminus o f the p recursor proteins and a p rotein translo- case in the microsomal m embrane [1]. Typically, transport occurs as a sequence of three consecutive steps, namely (a) specific membrane association of the precursor protein (also termed t argetin g), ( b) m embran e inse rtion, a nd ( c) comple- tion of translocation. S pecific membrane association of precursors in cotranslational transport involves two ribonu- cleoparticles – the ribosome [2] and the signal recognition particle (SRP) [3] – as well as their receptors on the endoplasmic reticulum (ER) surface (the SRP and ribosome receptors) [4,5]. P rotein t ranslocase (a) m ediates both membrane insertion and completion of translocation, (b) comprises Sec61ap, Sec61bp, and Sec61cp [5] as core components, and ( c) operates c o- or post-translationally. I n addition, heterotrimeric Sec61p complexes serve as specific ribosome-binding sites in cotranslational transport [6], and as sign al peptide receptors in g eneral [7]. The concentration of heterotrimeric Sec61p complexes and specifically bound ribosomes in defined s uspensions of mammalian microsomes (absorbance ¼ 50 at 280 nm in 2 % SDS, c orresponding to 1 equivalent per lL) has b een determined as 1.67–2.12 l M [8,9] and 0.27–0.39 l M [6,8], respective ly. These data were taken as a first sugg estion that oligomers o f heterotrimeric S ec61p complexes may be associated with ribosomes that are simultaneously engaged in p rotein synthesis and transloca- tion. Subse quently, cryo- and freeze fracture electron micro- scopic analysis of the Sec61p complexes, as present i n intact membranes, derived from canine pancreatic or yeast endo- plasmic reticulum, suggested that between three and four Sec61p complexes form the central unit of the protein translocase [8,10–12]. In addition to the heterotrimeric Sec61p complexes, Hsp70 protein family members of the ER lumen (BiP/Grp78 and Grp170) are part of the protein translocase a nd facilitate insertion of presecretory protein s into the Sec61p complex, as well as completion of translo- cation [13,14]. These Hsp70 protein family members o f the mammalian ER may be recruited t o the Sec61p complex by the membrane-integrated Hsp40 protein family members, Sec63p [9,15] and/or Mtj1p [16]. Sec62p [9,15], TRAMp [17], and the TRAP complex [ 18] appear to be additional subunits of protein translocase. Many precursor proteins that enter the E R are processed by the signal peptidase c omplex [19] and the oligosaccharyl transferase complex [20]. Therefore, it is not s urprising that these complexes a re in close proximity to protein translocase [21,22]. When misfolding occurs, the polypeptides are exported to t he cytosol and degraded by the proteasome. Protein export from the ER lumen to the cytosol is also mediat ed by Sec61p complexes [ 23,24]. Here we addressed the stoichiometry of the mammalian Sec61p complexes that are present in cotranslationally active protein translocases, by quantitative analysis of protein transport into pancreatic microsomes in single turnover translocation experiments, and of various other components, believed to b e subunits of protein translocase and involved in c ovalent modifications, by semiquantitative immunoblot analysis. Experimental procedures Materials The luciferase assay reagent and anti-luciferase immuno- globulin were obtained from Promega. The Translation kit Correspondence to R. Zimmermann, Medizinische Biochemie und Molekularbiologie, Universita ¨ t des Saarlandes, D-66421 Homburg, Germany. Fax: +49 6841 1626288; Tel.: +49 6841 1626510; E-mail: bcrzim@med-rz.uni-saarland.de Abbreviations: BiP, immunoglobulin heavy chain binding protein; ECL, enhanced chemiluminescence; ER, endoplasmic reticulum; ppl, preprolactin; PVDF, poly(vinylidene difluoride); SRP, signal recognition particle; SR, SRP receptor. *These authors contributed equally to this work. (Received 2 February 2004, revised 11 May 2004, accepted 10 June 2004) Eur. J. Biochem. 271, 3200–3207 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04252.x type II and firefly luciferase were from Roche Diagnostics. The peroxidase conjugate of goat antirabbit IgG was f rom Sigma Chemical Company. [ 35 S]Methionine, X-ray films and enhanced chemiluminescence (ECL) reagents were from Amersham Biosciences; poly(vinylidene difluoride) (PVDF) membranes were from Millipore. In vitro translation/translocation Protein synthesis was carried out in ra bbit reticulocyte lysates in the presence of [ 35 S]methionine, following the supplier’s recommendations (Translation kit type II; Roche Diagnostics). Subsequently, the samples were subjected to SDS/PAGE. The dried gels were analyzed in a phosphor- imager (Molecular Dynamics, Sunnyvale, C A, USA) usin g IMAGEQUANT software (version 5.1, Molecular Dynamics). Alternatively, the proteins were t ransferred to P VDF membranes a nd incubated with specific antibodies. The antibodies were visualized by ECL and subsequent exposure to X-ray film. X-ray films were analyzed by densitometry (Molecular Dynamics) using IMAGEQUANT software (ver- sion 3.0; Molecular Dynamics). Images from phosphori- mager and densitometry analyses were transferred into PHOTOSHOP software [version 3.0.5, Adobe Systems, Inc., San Jose, CA, USA] for production of figures. Luciferase activity was determined as describe d previously [25]. Quantification of proteins synthesized in vitro Firefly luciferase was used as an endogenous reference for the quantification of radiolabeled proteins. This is possible because luciferase, newly synthesized in rabbit reticulocyte lysate, is f olded to its native state with a very high efficiency and reproducibility. In a first set of experiments, serial dilutions of purified luciferase in reticulocyte lysate were subjected to luciferase activity measurements as well as to immunoblot analyses. The blot was incubated with anti- luciferase immunoglobulin and a peroxidase conjugate of secondary antibodies. The antibodies were visualized by ECL and subsequent exposure to X-ray film. The films were analyzed by densitometry. Both data sets gave rise to standard curves that served as a reference for unknown quantities o f luciferase in subsequent experiments (data not shown). In a second set of experiments, l uciferase was synthesized in reticulocyte lysate in the presence of [ 35 S]methionine for 60 min. Subsequently, luciferase activity and radioactivity, present in the luciferase b and after SDS/ PAGE, w ere determined for different aliquots of the same translation reaction by luminometry and phosphorimager analysis, respectively. In addition, different aliquots of the luciferase translation reactions were subjected to SDS/ PAGE and subsequent blotting to a PVDF membrane, together with serial dilutions of purified luciferase. The blot was a nalyzed as described a bove. Based on the two above-mentioned standard curves, the quantity of newly synthesized luciferase was determined. In both analyses, calculations were based on data points that lay in the linear range of the standard curves. The results from both analyses led to the conclusion that the concentration of de novo- synthesized luciferase in the reticulocyte lysate is  2n M after in vitro translation f or 60 min, i.e. for the b atches used of reticulocyte lysate and [ 35 S]methionine. Subsequently, the quantity of a given protein, after synthesis in the same batch of an in vitro system, was determined, with reasonable accuracy, by phosphorimager analysis of the r espective gel band and by its comparison with simultaneously synthes- ized firefly luciferase (assayed by both luminometry and phosphorimaging). In these experiments, the luciferase activity and the radioactivity analysis allowed the calcula- tion th at the concentration of a nascent p reprolactin polypeptide chain (ppl-86mer, four methionines) and full- length preprolactin (ppl, eight methionines) in the in vitro system were  250 n M and 100 n M , respectively, after translation for 20 and 45 min, respectively. The different methionine contents of luciferase (13 methionines) and the other proteins was taken into account. Quantification of microsomal proteins Dog pancreas microsomes were prepared and treated with nuclease and EDTA, as described previously [26]. The absorbance at 280 nm, in 2% SDS, of the final microsomal suspension was 50, corresponding to 1 equivalent per lL, or a protein concentration of  15 mgÆmL )1 . The Sec61p com- plex, SRP receptor, TRAMp, TRAP c omplex, s ignal peptidase complex and o ligosaccharyl transferase complex were purified according to previously published procedures [5,17,18,20] and used for quantification according to our previously published procedure [9]. B riefly, the quantity of protein, present in the respective band of a gel of a certain protein preparation, was determined by comparison with protein standards that were run on the same gel and stained simultaneously with Coomassie Brilliant Blue. Subsequently, an aliquot of the same sample of purified protein was run on the same gel together with increasing amounts of micro- somes, and the known quantity o f purified protein served as a standard for the Western blot signals, as determined by luminescence and densitometry of the X-ray films. In both analyses, calculations were based on data points that l ay in the linear range of the densitometry signals. We note that the calculations are based on the assumption that staining with Coomassie Brilliant Blue is uniform for all proteins and that this may not be absolutely true for every single protein. Therefore, the values that are given in Table 1 should be interpreted with this c aveat in mind. Results Capacity of canine pancreatic microsomes for SRP-dependent protein transport The ability to quantify n ewly synthesized proteins based on radioactivity analyses and co mparison with simultaneously synthesized firefly luciferase (as described in the Experi- mental procedures) allowed us to analyze the e fficiency of protein transport into canine pancreatic microsomes at different ratios of the precursor a nd Sec61ap. Single SRP– ribosome–nascent preprolactin chain (ppl-86mer) com- plexes were produced in the presence of increasing concen- trations of pancreatic microsomes. This experimental strategy is defined here as a single turnover experiment and was first described by Connolly & Gilmore [27]. Subsequently, the tran slation reactions were divided into four aliquots. One aliquot was untreated (Fig. 1A) and used Ó FEBS 2004 Protein transport into mammalian microsomes (Eur. J. Biochem. 271) 3201 for determining the total quantity o f ppl-86mer in the translation reaction. Microsomes were reisolated from the other three aliquots and untreated (Fig. 1B), or subjected to puromycin-induced translocation of the nascent presecre- tory protein (termed chase) (Fig. 1C), or subjected to chemical cross-linking (Fig. 1D). Luciferase was synthesized in parallel and quantified on the basis of its enzymatic activity. After SDS/PAGE and phosphorimager analysis, the precursor, mature protein and cross-linked precursor were quantified in comparison to luciferase analyzed simultaneously (as described in the Experimental proce- dures). The quantities (a) of total ppl-86mer (Fig. 1A,E), synthesized in the t ranslation reaction (b) o f m icrosome- bound ppl-86mer (Fig. 1B,F), (c) of chased, i.e. specifically bound ppl-86mer (Fig. 1C,G), and (d) of Sec61ap-associ- ated, i.e. cross-linked, ppl-86mer (Fig. 1D,H), were com- pared with t he quantities of Sec61p complexes [9] present in the various translation r eactions. According to t he efficien- cies of ppl-86mer synthesis, the ratios between precursor and Sec61ap varied b etween 12.5 and 0.4 (Fig. 1 E, e.g. the Table 1. Concentration of components in pancreatic rough microsomes (RM). In the case of complexes, the prote ins, shown in pa rentheses, were quantified. Note that the concentrations refer to a suspension of RM with an absorbance at 280 nm of 50, as measured in 2% SDS, corresponding to 1 eq uivalent per lL, or a protein concentration of  15 mgÆmL )1 or 0.33 m M (average molecular mass 50 000 kDa). BiP, immunoglobulin heavy chain binding protein; OST, oligosaccharyl transferase; SPase , signal peptida se; SRP, signal recognition particle. Component Concentration Reference High-affinity ribosome-binding sites a 0.27–0.39 l M [6,8] Cotranslationally operating translocases b 0.40–0.62 l M SRP receptor (SRap) 0.24 l M SRP receptor (SRbp) 0.47 l M Sec61p complex (Sec61ap) 2.12 l M [9] TRAMp 1.74 l M TRAP complex 1.30 l M [31] Sec62p 1.96 l M [9] Sec63p 1.98 l M [9] Mtj1p 0.36 l M [16] BiP 5.00 l M [30] Grp170 0.60 l M [30] SPase complex (SPC23-su) 0.52 l M OST complex (Ost48p) 1.60 l M a High salt resistant ribosome-binding sites at a concentration of Sec61 ap of 1.67 l M . b Productive binding sites for SRP–ribosome– nascent chain complexes; as deduced from Figs 1 and 3, respect- ively; average values. Fig. 1. Quantification of specific binding of nascent presecretory pro- teins to microsomes and Sec61ap in single turnover experiments. Nas- cent preprolactin (ppl-86mer) was synthesized in reticulocyte lysates in the presence o f [ 35 S]methionine and dog pancreas m icrosomes at the indicated concentrations [rough microsomes (RM), %, v/v]. After incubation for 20 min at 30 °C, the translation reactions were divided into four aliquots (A–D). One aliquot was untreated (A), and aliquots 2–4 w ere subjected to centrifugation (20 min, 15 0 00 g,2°C) (B–D). The pellet from the second aliquot was untreated thereafter (B). The pellet from the third aliquot was resuspended in buffer and incubated for 15 min at 30 °C in the presence of puromycin (1.25 m M )(C).The pellet from the fourth aliquot was resusp ended in buff er and incubate d with 335 l M succinimidyl 4-(N-maleimidometh yl)cycloh exane -1-ca rb- oxylate (SMCC) for 20 min at 0 °C, as described previously [13] (D). Firefly luciferase was synthesized in parallel and luciferase activity was assayed. Different aliquots of all sample s (including th e luciferase translation reaction) were subjectedtoSDS/PAGE(19.4%acrylamide + urea) and phosphorimager analysis. The q uantification of ppl- 86mer was carried out as described in the Experimental procedures (E–H). The data for 1 lL(A)and5lL (B–D) aliquots are s hown. Th e dotted line in (F) re presents th e sum of ppl-86m er a nd ppl-86mer cross- linkedtoSec61ap of (C), i.e. allows an estimation of the protein recovery after cross-linking. Note that the different electrophoretic mobility of ppl-86mer in lane 4 of (B) is caused by a gel artifact. ppl 86 xSec61ap, ppl-86mer cross-linked to Sec61ap; ppl 86 ,ppl-86mer; pl 56 ,pl-56mer. 3202 S. Guth et al. (Eur. J. Biochem. 271) Ó FEBS 2004 first data point: 250 nmol of ppl-86mer divided by 20 nmol of Sec61ap corresponds to a ratio of 12.5 : 1.0). The binding of ppl-86mer to microsomes (Fig. 1F) and the specific binding of ppl-86mer to microsomes (Fig. 1G), measured as chase t o p l-56mer, increased with increasing concentration of microsomes. A s e xpected, at t he low est concentration of microsomes, the total binding exceeded by far the specific binding [28] (Fig. 1F vs. 1G). However, the specific bindin g was significant at higher concentrations of microsomes (Fig. 1 F vs. 1G). There was a g ood corre lation between specific binding and cross-linking to Sec61ap (Fig. 1G vs. 1H) and cross-linking of specifically bound ppl-86mer was quite efficient (up to 80%; Fig. 1 G vs. 1H). For the two intermediate concentrations of microsomes (4 and 8% RM, respectively, i.e. at 25- and 12.5-fold dilutions) the results, shown in F ig. 1, indicate concentrations of specific binding sites for ribosome–nascent chain complexes o f ppl-86mer in the microsomal suspen sion between 0.34 (Fig. 1H, e.g. the second data point/80 nmol Sec61ap: 15 nmol of ppl- 86mer · 25 ¼ 375 nmol) and 0.46 l M (Fig. 1 G, e.g. the third data point/160 nmol Sec61ap: 38 nmol ppl- 86mer · 12.5 ¼ 475 nmol) (average: 0.4 l M ;Table 1)(note that the c oncentration of S ec61ap in the microsomal suspension is  2 l M ; Table 1). As cross-linking typically does not occur at efficiencies of 100%, the latter number seems to be more reliable. Thus, at t hese intermediate concentrations of microsomes, the average ratio between specifically bound ppl-86mer and Sec61 ap was 1.0 : 4.3 (Fig. 1 G; 0.46 /2 l M corresponds to a ratio of 1.0 : 4.3), i.e. about one in four Sec61ap molecules was able to bind the nascent precursor protein. Similar experiments we re carried out employing rough microsomes that had been pretreated with puromycin plus high salt and yielded similar results (data not shown). This must be caused by read-out synthesis on ribosomes that were attached to microsomes in the mammalian translation system. In order t o c onfirm the notion that the conditions of the single turnover experiment allow saturation of specific binding, a two -stage experiment was carried out. The first stage of this experiment was similar to the single turnover experiment, but was carried out in the absence of [ 35 S]methionine and at an intermediate concentration of microsomes (5% RM, v/v). Subsequently, microsomes were reisolated and subjected to a second translation reaction in the presence of [ 35 S]methionine, i.e. with or without prior EDTA treatment. In parallel, microsomes were subjected to a first mock tr anslation reaction in the absence of transcript and [ 35 S]methionine and, after reisolation and treatment with or without EDTA, t o a s econd translation reaction i n the p resence of [ 35 S]methionine. T he precursor preprolactin was synthesized in the second stage of the experiment, and the various microsomes were present in t hese translation reactions at two different final concentrations (5 or 10% RM, v/v; Fig. 2A). A control experiment for this second Fig. 2. Specific binding of nas cent presecretory p roteins to microsomes in a single turnover experiment prevents cotranslational transport of other presecretory proteins i n a s ub sequen t transport reaction. Nascent preprolactin (p pl-86mer) was synthesizedinreticulocytelysateinthe presence o f dog pancreas microsomes [5% rough microsomes (RM), v/v) (+mRNA)]. A mock translation minus transcript was carried out in parallel (–mRNA). After incubation for 20 m in at 30 °C, the translation r eactions were centrifuged (20 min, 15 000 g,2°C). The pellets were resuspended in buffer, divided into two aliquots, and incubated in the absence (–) or presence (+) of EDTA (5 m M )for 30 m in a t 3 0 °C. MgCl 2 (7.5 m M )wasaddedtoallsampleswhichwere then adjusted to the same final concentration of EDTA. Subsequently, preprolactin w as synthesized in reticulocyte lysates in the presence of [ 35 S]methionine and in the simultaneous pre sence of the different microsomes at the indicated concentrations (5 or 10% RM, v/v; stip- pled vs. solid bars in C) (A,C). In parallel, preprolactin was synthesized in rabbit reticulocyte lysates in the p resence of [ 35 S]methionine and in the simultan eous p resence o f dog pancreas micros omes at the indicated concentrations (RM, %, v/v) (B). After incubation for 45 min at 30 °C, the t ranslation reactions were s upplemented with puromycin and incubated further for 15 min at 30 °C. All samples were subjected to SDS/PAGE (19.4% acrylamide+urea) and phosphorimager ana- lysis ( A–C). Note that radiolabeled p pl-86mer was synthesized in the second translation rea ction as a result of the presence of mRNA which was carried over from the first translation reaction (A). The other presecretory proteins were analyzed in a similar manner (D,E) or under post-translational transport conditions (F), as described previously [29]. ppl, preprolactin; ppl 86 , ppl-86mer; pl, prolactin; pl 56 , pl-56mer; ppaf, prepro-a- factor; ppcecDHFR, preprocecropin - dihydrofolate reductase hybrid protein. Ó FEBS 2004 Protein transport into mammalian microsomes (Eur. J. Biochem. 271) 3203 translation confirmed that the concentrations of micro- somes that were used in t he secon d stage of the two-stage experiment allowed the detectio n of quantitative differences in transport e fficiencies (Fig. 2B). After SDS/PAGE, phos- phorimager analysis of precursor and mature protein was carried out (Fig. 2A,C). Transport of preprolactin was almost completely blocked when microsomes were analyzed which had previously been subjected to a single turnover experiment with ppl-86mer (Fig. 2A,C, lan es/bars 3 and 4). This effect is most obvious when the two concentrations of microsomes are compared (Fig. 2A,C, lanes/bars 3 vs. 4). However, transpo rt o f preprolactin was only minimally affected when microsomes were analyzed which had previ- ously been subjected to a single turnover experiment and, subsequently, to a chase of ppl-86mer with EDTA (Fig. 2 A,C, lanes/bars 1 and 2). Furthermore, microsomes were minimally affected by the first mock translation (Fig. 2 A,C, lanes/bars 5–8). Thus, the two-stage experiment demonstrated that the single turnover e xperiments had led to saturation of microsomes with respect to th eir transport capacity. This was confirmed by employing, in the two-stage experiment, a second precursor that is transported in an SRP-dependent manner and cotranslationally, yeast pre- pro-a-factor (Fig. 2D, bars 3 and 4 vs. 1 and 2), and a precursor that is transported predominantly in an SRP- dependent manner and cotranslationally when it is synthes- ized in the presence of microsomes, a preprocecropin– dihydrofolate reductase hybrid (Fig. 2E, bars 3 and 4 vs. 1 and 2) [ 29]. We note that yeast prepro-a-factor is t ranspor- ted into mammalian microsomes only cotranslationally and that the preprocecropin–dihydrofolate r eductase hybrid is transported into mammalian microsomes both co- and post-translationally under c otranslational conditions (Fig. 2 E) and, obviously, only post-translationally under post-translational conditions (Fig. 2F) [29]. SRP-independ- ent a nd post-tr anslational t ransport of the preprocecropin– dihydrofolate reductase hybrid was not affected by satura- tion of microsomes with respect to their cotranslational transport capacity (Fig. 2F). We note that the observation that the preprocecropin–dihydrofolate r eductase hybrid under cotranslational conditions was affected less than preprolactin and the prepro-a-factor (Fig. 2E vs. Fig. 2C and Fig. 2D, bars 3 and 4) is perfectly consistent with the fact that this precursor is transported into mammalian microsomes both co- and post-translationally under cotranslational conditions [29]. We reasoned that cross-linking of the ppl-86mer to Sec61ap should also be detectable at the level of Sec61ap and that quantification of cross-linking at the level of Sec61ap s hould directly demonstrate the relevance of the numbers stated above. Single SRP–ribosome–nascent pre- prolactin complexes were incubated with increasing con- centrations of pancreatic microsomes. Subsequently, the microsomes were reisolated and subjected to chemical cross-linking, or not cross-linked. SDS/PAGE of the samples, together with a serial dilutions of microsomes, was followed by blotting to P VDF. The b lot was i ncubated with anti-Sec61ap immunoglobulin and peroxidase conju- gate of secondary antibodies. The antibodies were visual- ized by ECL of the blots and subseq uent exposure to X-ray film (Fig. 3A,B). Indeed, an Sec61ap-related cross-linking product was detected which comprised the radioactive ppl- 86mer (Fig. 3C, lanes 6–9). This cross-linking product was specific as it depended on both transcript coding for ppl- 86mer and cross-linking reagent and because it was not detected after puromycin chase and subsequent cross- linking (Fig. 3D). Cross-linking was quantified after den- sitometry of the X-ray films. Between 27 and 35%, i.e. around one out of three to four Sec61ap molecules could be cross-linked to ppl-86mer under these conditions (Fig. 3 A,B, lanes 7 and 8). Thus, under conditions of saturation of microsomes with ppl-86mer, approximately every third or fourth Sec61ap molecule is in a position which allows cross-linking to ppl-86mer and chase to pl- 56mer, respectively. According to these results, the con- centrations of specific binding sites for ribosome-nascent chain complexes of ppl-86mer in the microsomal suspen- sion are between 0.54 and 0.7 l M (average: 0.62 l M ; Table 1 ) (note that the concentration of Sec61apinthe microsomal suspension is  2 l M ; t hus 27 and 35%, respectively, of cross-linked Sec61ap molecules correspond to concentrations of productive binding sites o f 0.54 and 0.7 l M ;Table1). Content of canine pancreatic microsomes of proteins involved in protein transport and covalent modifications Cotranslational membrane association of nascent precur- sor p roteins involves the SRP receptor (SR), comprising an a-subun it and a b-subunit. Heterotrimeric Sec61p com- plexes form the core unit of the protein translocase. In addition, protein translocase comprises the Hsp70 protein family member s of the ER lumen (BiP/Grp78 a nd Grp170) and their putative m embrane-integrated Hsp40 co-chaper- ones, Sec63p and Mtj1p. Furthermore, Sec62p, TRAMp, and the TRAP complex appear to be additional subunits of protein translocase. As many precursor proteins that enter the ER are cotranslationally processed by t he signal peptidase complex and the oligosaccharyl complex, these complexes must b e in close proximity to pro tein translo- case. Here w e d etermined t he concentration of these various components in the canine pancreatic microsomes, which had been used in the transport experiments discussed above, by semiquantitative immunoblot analysis (as d es- cribed in the Experimental procedures). The r esults are summarized, together w ith some p reviously published data, in Table 1. Discussion Quantitative aspects of cell-free systems for the analysis of protein transport into mammalian microsomes The transport data of this study suggest that the concen- tration of cotranslationally active protein translocases in defined suspensions of dog pancreas microsomes is  0.4–0.6 l M (average 0.5 l M ) (Table 1). This agrees reasonably well with the previously observed concentration of high-affinity ribosome-binding sites ( 0.3–0.4 l M at an Sec61apconcentrationof1.67l M ; Table 1) [6,8]. Consid- ering that active protein translocase contains three to fou r heterotrimeric complexes [8,10–12], these functionally defined concentrations correspond to 1.5 or 2 l M hetero- trimeric Sec61p complexes, as present in cotranslationally 3204 S. Guth et al. (Eur. J. Biochem. 271) Ó FEBS 2004 active protein translocases. Here we found that the saturation of cotranslationally active protein translocases with SRP–ribosome–nascent chain complexes inhibits co- translational t ransport o f other precursor polypeptides, but allows post-translational transport. This is consistent with the numbers discussed above and the t otal concentration of heterotrimeric Sec61p complexes ( 2 l M ; Table 1). Recently, we showed that co- and post-translational transport involves the Sec61p complex [32]. Thus, there are at least two populations of Sec61p complexes p resent in pancreatic microsomes; one class that provides the capacity for cotranslational protein transport and one class that provides the c apacity for post-translational transport. This is some- what reminiscent of t he situation in yeast [33]. However, it seems to us that in these mammalian microsomes the concentration of SR, rather than the concentrations of the translocase subunits Sec62p and Sec63p, may be the decisive factor for the ratio between the two different populations of Sec61p complexes (Table 1). We note that the concentration of SRbp may be a more reliable indicator of the concen- tration of SR because SRap has been shown to be rather sensitive t owards proteolytic attack during the isolation of pancreatic microsomes. Typically, protein translocation is accompanied by processing, by signal peptidase, of precursor proteins in transit. Furthermore, transient i nteraction of Sec61p com- plexes with signal peptidase was observed during protein translocation [21]. Therefore, we argued that signal pepti- dase should be present in microsomes at a similar concen- tration a s protein tra nslocase. Here we de termined a concentration of 0.52 l M for signal peptidase (Table 1). Thus, it seems that a single signal peptidase complex is associated with protein t ranslocase. We take this as further substantiation of the concentrations discussed above and circumstantial evidence for the oligomeric character of protein translocase. In c ontrast, most of the other subunits of protein translocase, as well as the oligosaccharyl transf- erase complex, are present in p ancreatic microsomes a t similar concentrations as heterotrimeric Sec 61p complexes. Thus, multiple copies of these proteins and complexes may be associated with oligomers o f heterotrimeric Sec61p complexes in intact membranes of the mammalian ER. We note that our results are consistent with the idea that the protein translocase of the ER cont ains t hree o r four heterotrimeric Sec61p complexes, but do not prove this. However, it should be equally clear that the fact that the homologous complexes from bacteria or archaea, termed SecYEG [34] or SecY complex [35], respectively, were crystallized as dimers and monomers, doe s not necessarily mean that these complexes are active as monomers. On the contrary, both electron microscopic [10–12] a nd, in particular, biophysical characterization [32,36] of active Sec61p complexes are consistent with an oligomeric state. Fig. 3. Quantification of the association of Sec61ap with nascent pre- secretory proteins in single turnover experiments. (A–C) Nascent pre- prolactin (ppl-86mer) was synthesized in reticulocyte lysates (20 lL) in the presence o f [ 35 S]methionine and dog pancreas microsomes at the indicated concentrations [rough microsom es (RM), % , v/v]. A fter incubation fo r 2 0 m in at 30 °C, the translation reactions were sub - jected to centrifugation (20 min, 15 000 g,2°C). The pellets were resuspended in buffer, divided into two aliquots, and incubated in the absence (– XL) or presence (+ XL) of succinimidyl 4-(N-maleimido- methyl)cyclohexane-1-carboxylate (SMCC, 335 l M )for20minat 0 °C. The proteins were subjected to SDS/PAGE (15% acrylamide) and subsequent blo tting to p oly(vinylide ne difluoride) (PVDF) mem - brane. A threefold serial dilution series of microsomes was analyzed on the same g el and blot (corresponding to 0.03, 0 .1, 0.3, 1 , and 3 lLof RM; lanes 5 through 1). The blot was incubated with rabbit anti- Sec61ap immunoglobulin and peroxidase c onjugate of goat anti-rabbit IgG. The antibodies were visualized by EC L analysis of the blots and subsequent exposure t o X-ray film (15 and 30 s exposures are shown in A and B, respectively). Subsequently, the blots were washed, dried and subjected to autoradiography (C). (D) Nascent preprolactin (ppl- 86mer) was synthesiz ed in rabbit reticulocyte lysate in the presence of dog pancreas microsome s (7.5% RM , v/v). A mock translation minus transcript was analyzed in parallel (– mRNA). After incubation for 20 m in at 30 °C, the t ranslation reactions were centrifuged (20 min, 15 000 g,2°C). The pellets were resuspen ded in b uffer. One aliquo t was incubated for 15 min a t 30 °C in the presen ce of puromycin (+ puromycin). The aliquots were incubated in the absence (– XL ) or presence (+ X L) of SMCC (335 l M )for20minat0°C, as indicated. The prote ins were separated by SDS/PAGE (1 5% acrylamide) an d blotted to a PVDF membrane. The blot was analyzed as de scribed above. pp l 86 xSec61ap, pp l-86mer cross-linked to Sec61 ap; ppl 86 , ppl-86mer. Ó FEBS 2004 Protein transport into mammalian microsomes (Eur. J. Biochem. 271) 3205 Quantitative considerations for the pancreatic ER Typically, a microsomal preparation from a canine pan- creas with an average weight of 40 g yields 40 mL of the defined microsomal suspension. The concentration in defined suspensions of dog pancreas microsomes, of 0.4– 0.6 l M , corresponds to a total of 20 nmol of cotransla- tionally active protein translocases p er canine pancreas. Thus, we calculate that 12 · 10 15 molecules of cotransla- tionally active protein translocases are present per canine pancreas, or about twice that number for a typical human pancreas. An average human produces  700 mL of pancreatic juice per day. The protein concentration of this body fluid is  70 0 mgÆmL )1 , thus the daily production of secretory proteins in the human pancreas amounts to  5 g. These 5 g correspond to  100 lmol, or 60 · 10 18 molecules, of secretory proteins per day (average molecular mass ¼ 50 000 kDa). Therefore, one can estimate that 2500 molecules of presecretory proteins are handled per cotranslationally active protein translocase in the human pancreatic ER per day, or  100 molecules per hour. Assuming that the yield of microsomes in the course of a microsomal preparation is never 100%, this number seems at least to be in the correct order of magnitude when the average speed of translation in a mamm alian cell is considered (up t o 300 amino acid residues per min). Acknowledgements We wish to thank R. Gilmore (DepartmentofBiochemistryand Molecular Biology, University of Massachusetts, Worchester, USA ) for a gift of anti-Ost48p serum. This work was supported by the DFG (grant C1/SFB 530) and by the Fonds der Che mischen Industrie. References 1. Blobel, G. & Dobberstein, B. (1975) Transfer o f protei ns across membranes. I. Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membrane- bound ribosomes of murine m yelom a. J. Cell Biol. 67, 835–851. 2. Perara, E., Rothman, R.E. & Lingappa, V.R. (1986) Uncoupling translocation f rom translation: implications for t ransport of pro- teins a cross membranes. Science 232, 348–352. 3. Walter, P. & Blobel, G. (1981) T ranslocation of proteins across the endoplasmic reticulum. II. Signal recognition protein, SRP, mediates the selective binding to microsomal membranes of in-vitro-assembled polysomes synthesizing secretory protein. J. Cell Biol. 91, 551–556. 4. Meyer, D.I., Krause, E. & Dobberstein, B. (1982) Secretory pro- tein translocation across membranes – the role of the Ôdocking proteinÕ. Nature 297, 647–650. 5. Go ¨ rlich, D. & Rapoport, T.A. (1993) Protein tran slocation into proteoliposomes rec onstit uted fro m purifie d com ponents o f th e endoplasmic reticulum membrane. Cell 75 , 615–630. 6. Kalies, K U., Go ¨ rlich, D. & R apoport, T.A. (1994) Binding of ribosomes to t he rough endoplasmic reticulu m mediated by the Sec61p-complex. J. Cell Biol. 126, 925 –934. 7. Jungnickel, B. & Rapoport, T.A. (1995) A posttargeting signal sequence recognitio n event in the end oplasmic reticulu m mem- brane. Cell 82 , 261–270. 8. Hanein, D., Matlack, K.E.S., Jungnickel, B., Plath, K., Kalies, K U., Miller, K.R., Rapoport, T.A. & Akey, C.W. (1996) Oligomeric rings of the Sec61p complex induced by ligands required for protein tran slo cation. Cell 87, 721–732. 9. Tyedmers, J., Lerner, M., Bies, C., D udek, J., Skowrone k, M., Haas, I., Heim, N., Nastainczyk, W., Volkmer, J. & Zimmer- mann, R. (2000) Homologs of the yeast Sec complex subunits Sec62p and Sec63p are abundant proteins in dog pancreas microsomes. Proc. Natl Acad. Sci. USA 97, 7214–7219. 10. Beckmann, R., Bubeck, D., Grassucci, R., Penczek, P., Verschoor, A., Blobel, G. & Frank, J. (1997) Alignment of conduits for the nascent polypeptide chain in the ribosome-Sec61 co mplex. Science 278, 2123–2126. 11. Beckmann, R., Spahn, C.M., Eswar, N., Helmers, J., Penczek, P.A.,Sali,A.,Frank,J.&Blobel,G.(2001)Architectureofthe protein-conducting channel associated with the translating 80S ribosome. Cell 107, 361–372. 12. Menetret, J F., Neuhof, A., Morgan, D.G., Plath, K., Rader- macher,M.,Rapoport,T.A.&Akey,C.W.(2000)Thestructure of ribosome-channel complexes engaged in protein transloc ation. Mol. Cell 6, 1219–1232. 13. Dierks, T., Volkmer, J., Schlenstedt, G., Jung, C., Sandholzer, U., Zachmann, K., Schlotterhose, P., Neifer, K., Schmid t, B. & Zimmermann, R. (1996) A microsomal ATP-binding protein involved in efficient protein transport into the mammalian endoplasmic reticulum. EMBO J. 15, 6932–6942. 14. Tyedmers, J., Lerner, M., Wiedmann, M., Volkmer, J. & Zimmermann, R. (2003) Polypeptide chain binding prote ins medi- ate complet ion of cotranslational protein translocation into the mammalian endoplas mic reticulum. EMBO Reports 4, 505–510. 15. Meyer, H A., G rau, H., Kraft, R., K ostka, S., Prehn, S., Kalies, K U. & Hartmann, E. (2000) Mammalian Sec61 is associated with Sec62 and Sec63. J. Biol. Chem. 275, 1 4550–14557. 16. Dudek,J.,Volkmer,J.,Bies,C.,Guth,S.,Mu ¨ ller, A., Lerner, M., Feick, P., Scha ¨ fer, K H., Morgenstern, E., H ennessy, F., Bl atch, G.L., Janoscheck, K., Heim, N., Frien, M., Nastainczyk, W. & Zimmermann, R. (2002) A novel type of cochaperone mediates transmembrane recruitment of Dn aK-like chaperones to r ibo- somes. EMBO J. 21, 2958–2967. 17. Voigt, S., Jungnickel, B., Hartmann, E. & Rapoport, T.A. (1996) Signal-sequence dependent function of the TRAM protein during early phases of protein transport across the ER membrane. J. Cell Biol. 134, 25–35. 18. Fons, R.D., Bogert, B.A. & Hegde, R.S. (2003) Substrate-specific function of the translocon associated protein complex during translocation across the ER membrane. J. Cell Biol. 160, 529–539. 19. Evans, E.A., Gilmore, R. & Blobel, G. (1986) Purification of microsomal signal peptidase as a complex. Proc.NatlAcad.Sci. USA 83, 581–585. 20. Kelleher, D.J., Kreibich, G. & Gilmore, R. (1992) Oligosacchary- ltransferase activity is associated with a protein complex com- posed of ribophorins I and II and a 48 kd protein. Cell 69, 55–65. 21. Kalies, K U., Rapoport, T.A. & Hartmann, E. (1998) The b subunit of the Sec61 complex facilitates cotranslational protein transport and interacts with the si gnal peptidase during translo- cation. J. Cell Biol. 141, 887–894. 22. Scheper, W., Thaminy, S ., Kais, S., Stagljar, I. & Ro ¨ misch, K. (2003) Coordination of N-glycosylation and protein translocation across the e ndoplasmic reticulum membrane by Sss1 protein. J. Biol. Chem. 278, 37998–38003. 23. Wiertz, E.J.H., Tortoralla, D., Bogyo, M., Yu, J., Mothes, W., Jones, T.R., R apoport, T.A. & Ploegh, H.L. (1996) Sec61-medi- ated transfer of a membrane prote in from the endoplasmic reticulum to the proteasome for destruction. Nature 384, 432–438. 24. Plemper, R.K., Bo ¨ mler,S.,Bordallo,J.,Sommer,T.&Wolf,D.H. (1997) Mutant analysis links the translocon and BiP to retrograde protein transport for E R degradation. Nature 388, 891–895. 25. Tyedmers, J., Brunke, M., Lechte, M., Sandholzer, U., Dierks, T., Schlotterhose, P., Schmidt, B. & Zimmermann, R. (1996) Efficien t folding of firefly luciferase after transport into mammalian 3206 S. Guth et al. (Eur. J. Biochem. 271) Ó FEBS 2004 microsomes in the absence of luminal chaperones and folding catalysts. J. Biol. Chem. 271, 1 9509–19513. 26. Watts, C ., Wickner, W. & Zimmermann, R. (1 983) M13 procoat and pre-immunoglobulin share processing specificity b ut use dif- ferent membrane receptor m echanisms. Proc. N atl Acad. Sci. USA 80, 2809–2813. 27. Connolly, T. & Gilmore, R. (1986) Formation of a functional ribosome-membrane junction during translocation requires the participation of a GTP-binding protein. J. Cell Biol. 103, 2253– 2261. 28. Murphy, E.C., III, Zheng, T. & Nicchitta, C.V. (1997) Identifi- cation of a novel stage of ribosome/nascent chain association with the endoplasmic reticulum membrane. J. Cell Biol. 136, 1213– 1226. 29. Schlenstedt, G ., Gudmundsson, G.H., Boman, H.G . & Zimmer- mann, R. (1990) A large presecretory p rotein translocates both cotranslationally, using signal recognition particle and ribosome, and p osttranslationally, without these ribonucleoparticles, when synthesized in the presence of mammalian microso mes. J. Bi ol. Chem. 265, 13960–13968. 30. Bies, C., Guth, S., Janoschek, K., Nastainczyk, W., Volkmer, J. & Zimmermann, R. (1999) A Scj1p homolog and folding catalysts present in dog pancreas microsomes. Biol. Chem. 380, 1175–1182. 31. Hartmann, E., Wiedmann, M. & Rapoport, T.A. (1989) A membrane component of the endoplasmic reticulum that may b e essential for protein t ranslocation. EMBO J. 8, 2225–2229. 32. Wirth, A., Jung, M., Bies, C., F rien, M ., Tyedmers, J., Zimmer- mann, R. & Wagner, R. (2003) The Sec61p complex is a dynamic precursor activated channel. Mol. Cell 12, 261–268. 33. Panzner, S ., Dreier, L., Hartmann, E., Kostka, S. & Rapoport, T.A. (1995) Posttranslation protein transport in yeast recon- stituted with a purified complex o f S ec proteins and Kar2p. Cell 81, 561–570. 34. Breyton, C., Haase, W., Rapoport, T.A., Ku ¨ hlbrandt, W. & Collinson, I. (2002) Three-dimensional structure of the bacterial protein translocation complex, S ecY EG. Nature 418 , 662–665. 35. van den Berg, B., Clemons, W.M., Collinson, I., Modis, Y., Hartmann, E., Harrison, S.C. & Rapoport, T.A. (2004) X-ray structure of a protein-conducting channel. Nature 427, 3 6–44. 36. Hamman, B.D., Chen, J.C., Johnson, E.E. & Johnso n, A.E. (1997) The aqueous pore through the translocon has a d iameter of 40–60 A during co-translational protein translocation at t he ER membrane. Cell 89, 535–544. Ó FEBS 2004 Protein transport into mammalian microsomes (Eur. J. Biochem. 271) 3207 . calculate that 12 · 10 15 molecules of cotransla- tionally active protein translocases are present per canine pancreas, or about twice that number for a typical human pancreas. An average human. complexes that are present in cotranslationally active protein translocases, by quantitative analysis of protein transport into pancreatic microsomes in single turnover translocation experiments, and. Protein transport into canine pancreatic microsomes A quantitative approach Silvia Guth*, Christian Vo¨ lzing*, Anika Mu¨ ller, Martin Jung and Richard Zimmermann Medizinische