| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Cell Biology and Kaplan Cancer Center (V.H.B., A.S., G.K.), New York University School of Medicine, New York, New York 10016; Cellular Biochemistry and Biophysics Program (K.v.L.), Memorial Sloan-Kettering Cancer Center, New York, New York 10021; and Department of Pathology (B.L.), Columbia University, New York, New York 10027
Address all correspondence and requests for reprints to: Virginia H. Black, Department of Cell Biology, New York University School of Medicine, 550 First Avenue, New York, New York 10016. E-mail: blackv01{at}med.nyu.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The more complex forms also occur in hepatocytes and cultured cells in which 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGR), a key enzyme in cholesterol synthesis, is overexpressed as well as in cultured cells overexpressing other proteins characteristic of the SER, e.g. cytochromes P450 (8). HMGR has a high turnover rate (9), and its levels in adrenocortical cells are regulated in a sterol and ACTH-dependent manner (10, 11). Changes in SER volume and enzyme content with functional state indicate that there must be fluctuations in synthesis and degradation of HMGR and other enzymes in the cholesterol and steroid biosynthetic pathway. Yet morphologically identifiable rough endoplasmic reticulum (RER), presumably required for these functions, is sparse in these cells (less than 0.3% of the cytoplasmic volume) (6). Bound ribosomes occur on short endoplasmic reticulum (ER) cisternae and in patches scattered along predominantly smooth-surfaced, randomly arranged tubules or at the periphery of smooth cisternal and crystalloid arrays (Figs. 1 and 2, arrows). Although the RER increases 2- to 4-fold after ACTH treatment, this morphological appearance does not change significantly (6). Cisternae densely covered with ribosomes are rarely seen.
|
|
). Using microsomes prepared from these cells, a large number of proteins have been identified that are involved in targeting, translocation and processing of proteins synthesized on membrane bound ribosomes (12, 13, 14, 15) (for summary see Table 1). Some of these components have been shown to be largely confined to rough microsomes derived from the RER of protein-secreting cells: signal recognition particle (SRP) receptor (SR) (58), Sec61
(33), subunits of the translocon-associated protein (TRAP) complex (59) as well as ribophorins I and II (RI and RII) (60, 61, 62, 63).
|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell fractionation
Microsomal subfractions from guinea pig adrenals and liver as well as the adrenals of several other species obtained from Pel-Freez (Rogers, AR) were prepared as previously described (65, 66). Liver microsomes were prepared from rats fasted for 18 h using this method, a modification of that described by Adelman et al. (67), as well as by the modification of Adelman’s method described by Kruppa and Sabatini (68). Dog pancreas rough microsomes were prepared as previously described (61). In some cases, ribosomes were removed from the microsomal membranes after subfractionation, using high salt treatment and puromycin, as described previously (69). All fractions were stored at –70 C.
Antibodies
The proteins associated with protein translocation and processing that were investigated as a part of this study are listed in Table 1. Antibodies directed against RI and RII, OST48, Ig heavy chain-binding protein (BiP) [glucose regulated protein (GRP)78], and TRAP were as previously described (47, 70). Dr. Andrei Nikonov made the antipeptide antibody to defender against apoptotic death (DAD)1 (amino acids 76–91) in the laboratory of one of the investigators (G.K.). The antibody to the SR, raised against the 16 C-terminal amino acids of the -subunit, was also made in the laboratory (of G.K.). Antibodies for Sec61 and Sec61ß were raised in the laboratory of Dr. Martin Wiedmann (Memorial Sloan-Kettering Cancer Center, New York, NY) and affinity purified by Dr. Robert Levy while a student in one of the authors’ laboratories (G.K.). Antibodies against Sec62 and Sec63 were received from Dr. R. Zimmermann (Universität des Saarlandes, Sarbrucken, Germany). Antibody against GRP94 was obtained from Dr. Christopher Nicchitta (Duke University, Durham, NC). Antibody for signal peptidase SPC12 was from Dr. Enno Hartmann (University of Lubeck, Lubeck, Germany) (40). Antibody against the ribosomal S3 protein was obtained from Dr. Vicky Richon and was made by Dr. Xianbo Zhou in the laboratory of Drs. Paul Marks and Richard Rifkind (Memorial Sloan-Kettering Cancer Center). Antibodies to enzymes involved in sterol and steroid synthesis, HMGR, cytochromes P450 17-hydroxylase (CYP17), and 3ß-hydroxysteroid dehydrogenase (3ßHSD) as well as cytochromes P450 involved in xenobiotic metabolism, CYP1A and CYP2B, were as described previously (11, 64, 66, 71, 72). All of the antibodies were raised in rabbits, most against purified proteins, unless specified as antipeptides. All recognized proteins of the appropriate size in Western blots and most showed minimal cross-reactivity with other proteins under the conditions employed.
Immunoblotting
Microsomal proteins were separated by electrophoresis on 8% sodium dodecyl sulfate (SDS)-polyacrylamide gels and transferred electrophoretically to nitrocellulose sheets, as previously described (66, 71). Smaller proteins, DAD1 and Sec61ß, were separated electrophoretically on 20% SDS-polyacrylamide gels in a Tris-tricine buffer system, according to published protocols (73). In general, 12 µg of protein were used for each sample. However, protein concentrations were adjusted for optimal visualization in some cases, as indicated in the figure legends. Blots were stained with Ponceau S (Sigma Diagnostics, St. Louis, MO) to visualize corun standards and check protein loading and transfer, as previously described (66, 71). Tween 20 and/or 5% dry milk in PBS were used as blocking agents. Peroxidase-conjugated secondary antibodies (Capell, Durham, NC) were visualized using 3,3'-diaminobenzidine tetrahydrochloride (DAB; Polysciences, Warrington PA) or the enhanced chemiluminescence (ECL) Western blotting analysis system (Amersham Life Science, Buckinghamshire, UK). Once optimal conditions for each antibody had been established, subsequent blots were cut into pieces so that immunoreactions for proteins of differing sizes could be performed on each set of samples in overlapping combinations, e.g. calnexin, RI or RII, and Sec61; GRP94, BiP, RI or RII, CYPs, and Sec61 or TRAP. This permitted a more accurate assessment of the relative distribution of multiple proteins among the microsomal subfractions. Blots for each protein were run three to four times. Quantitation of immunoblots was performed using NIH Image.
Immunocytochemistry
Isolated adrenocortical cells were prepared and grown on coverslips, as previously described (74). After 3–5 d in culture, in the presence or absence of ACTH (100 mU/ml), the cells were rinsed briefly with PBS and fixed for 10 min on ice with methanol that had been stored at –20 C. Subsequent incubations were carried out at room temperature in 1% milk in PBS [blocking, 1 h; primary antibody (1:200), 1–5 h; three rinses; secondary antibody (fluorescein-conjugated AffiniPure F(ab')2 fragment donkey antirabbit IgG (H+L), Jackson ImmunoResearch Laboratories Inc, West Grove PA)(1:125), 1 h], followed by several rinses in plain PBS. Coverslips were adhered to glass slides with Citiflour mounting medium (Citifluor mountant medium #0, Ted Pella, Redding, CA). Corun controls were incubated in the secondary antibody alone. Confocal microscopy was performed with a LSM510 laser scanning microscope (Zeiss, Göttinger, Germany).
Analysis of OST activity
The octanoyl tripeptide (OTP), N-octanoyl-Asn-Tyr-Thr-amide, which contains the acceptor sequence for N-glycosylation, was received from Dr. Felix Wieland (University of Heidelberg, Heidelberg, Germany). [125I]OTP was prepared and used as a glycosyl acceptor to assay oligosaccharyltransferase activity, as previously described (75, 76, 77). After incubation with microsomal protein (30–300 µg) for 1 h at 37 C in the presence and absence of the glucosidase inhibitor, castanospermine (50 µM), the glycosylated tripeptide was bound to a conA Sepharose column and eluted with buffer containing methyl--D-mannopyranoside and Triton X-100. Activity in the eluate from the conA Sepharose column is expressed as counts per minute per microgram microsomal protein used in the assay.
Analysis of glycotripeptides by thin layer chromatography (TLC) was performed as previously described (75, 76, 77). To further define the pattern of glycosylation, the eluted samples were subjected to enzymatic digestion with -mannosidase and Endo H (77). Aliquots containing approximately equal counts per minute from each sample were incubated with Jack bean -mannosidase (6 U/ml; Glyko, Inc., Novato, CA) or Endo H (Roche, Mannheim, Germany) in the 1x buffer supplied with the -mannosidase, as described in the accompanying literature. The total incubation mixture was 5–9 µl. After incubation at 37 C, overnight (
16 h), the reaction mixture was spotted on TLC plates, without further processing and the components separated in the solvent system described above.
Ribosome targeting and binding
For the targeting assay, truncated mRNA lacking a stop codon encoding the first 86 amino acids of preprolactin (86pPL) was translated in a rabbit reticulocyte lysate supplemented with [35]S-Met and trifluoromethyldiazirinobenzoic acid-modified lys-tRNA as previously described (78) to create nascent chains modified for photocross-linking. After translation, cycloheximide was added to 0.5 mM and microsomes were added. After incubation for 2 min at room temperature and 10 min on ice, samples were irradiated to induce cross-linking. The samples were trichloroacetic acid (TCA) precipitated, resuspended in Laemmli sample buffer, and electrophoresed through 12% gels. The 62-kDa photoadduct representing the 86pPL nascent chain cross-linked to SRP (54 kDa) was visualized by fluorography.
For the binding assay, 86pPL mRNA was translated in a wheat germ lysate supplemented with [35]S-Met and high salt extracted ribosome/nascent chain complexes (RNCs) were prepared and isolated as previously described (79). RNCs (5 µl) were incubated with puromycin/KOAc washed dog pancreas microsomes (PKRMs, 10 mg/ml or 1 eq/µl) or adrenal smooth microsomes (12 mg/ml) for 2 min at room temperature and 10 min on ice and bound RNCs separated from unbound RNCs by flotation in discontinuous sucrose gradients, as previously described (79). The top gradient fractions containing membranes were collected and analyzed for RNC content by autoradiography after SDS-PAGE.
Assay of cotranslational translocation and processing
Two truncated mRNAs were prepared. The first, used for assay of signal peptide cleavage, encoded the C-terminal fragment (CTF; 99 amino acids) of amyloid precursor protein with bovine pPL signal peptide (SP) sequence added at the 5' end. The second, used for assay of N-glycosylation, was derived from plasmid pSF1, received from Dr. S. M. Simon (Rockefeller University, New York, NY). pSF1 contained the complete cDNA of opsin inserted in the Sac1 site of an SP6/4 vector (80). It was linearized with BsaH1 before runoff transcription with SP6 RNA polymerase (SP6 Cap-Scribe; Roche Molecular Biochemicals, Mannheim, Germany) to prepare truncated mRNA encoding the first 156 codons of bovine opsin (op-156). This region includes the first three transmembrane-spanning segments and a lumenally exposed N-terminal domain that possesses two N-glycosylation sites (81). In vitro translation/translocation reactions of the mRNAs were performed in a rabbit reticulocyte lysate system (nuclease-treated lysate; Promega, Madison, WI) supplemented with [35]S-Met at 30 C for 45 min in a total volume of 13 µl. In the case of SP-CTF, the reaction was diluted with buffer (20 mM HEPES, 100 mM NaCl) and an aliquot layered over a high salt sucrose cushion (150 µl, 100 mM KCl, 2 mM MgAc2, 0.5 M sucrose). After a brief spin (10 min at 60,000 rpm, 4 C) in a TL100 centrifuge (Beckman Coulter, Inc., Fullerton, CA), the resulting pellet, containing microsomes and translocated polypeptides, was resuspended in sample buffer [100 mM Tris (pH 6.8), 4% SDS, 20% glycerol]. The remainder of the translation/translocation reaction was precipitated with saturated ammonium sulfate and the TCA washed pellet resuspended in sample buffer containing additional Tris to buffer any residual TCA. In the case of op-156, the entire reaction was precipitated with saturated ammonium sulfate, washed with TCA, and resuspended in the same sample buffer. Aliquots of each preparation were analyzed for newly synthesized polypeptides by autoradiography after SDS-PAGE.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(listed in bold). We compared the distribution of the immunoreactive proteins in Western blots of adrenal microsomal subfractions with levels seen in similarly prepared microsomal subfractions from liver and in pancreatic rough microsomes.
Microsomal subfraction protein and ribosome content
As expected, based on the morphology of the cells (Figs. 1 and 2), smooth microsomes comprised a considerably greater percentage of the total microsomal fraction prepared from adrenals of control animals than from the liver, whereas rough microsomes were more abundant in liver (Fig. 3)1. However, when animals were treated with xenobiotics, which induce the SER in hepatocytes, the liver smooth microsomal fraction increased, reaching levels comparable with those in adrenal smooth microsomes.
|
). Little, if any, immunoreactive protein was present in the smooth microsomes, in agreement with the previous biochemical assays.
|
). As previously reported (62), RI was a prominent component of rough microsomes from the dog pancreas, a tissue devoted almost solely to the synthesis of secretory proteins. Among microsomal subfractions obtained from liver of both rat and guinea pig, RI was in highest concentration in the rough microsomes. Its concentration was lower in the intermediate microsomal fraction, which bore fewer ribosomes, and a small, somewhat variable amount was detectable in liver smooth microsomes.
In contrast, in adrenal microsomal subfractions, the concentration of RI in smooth microsomes was equal to or greater than that seen in the rough microsomes. Stripping ribosomes from adrenal microsomes after subfractionation increased the intensity of RI immunoreactivity, an index of its concentration per milligram microsomal protein, in rough microsomes by about 30%. However, the amount of this OST subunit in the smooth microsomes remained particularly striking. Given the fact that the smooth microsomes comprise the bulk of the total microsomal fraction from adrenal tissue (Fig. 3), the total amount of ribophorin I was considerably greater in adrenal smooth than in adrenal rough microsomal fractions (Fig. 4).
We then compared levels of other OST subunits, RII, OST48, and DAD1, in microsomal fractions from both tissues with those of RI (see Figs. 5 and 7). The results were similar to those obtained for RI. In microsomal fractions from liver, these proteins were all in highest concentration in the rough microsomes. Yet in those from the adrenal, all occurred in equal or greater concentration in nonribosome-bearing microsomal subfractions: smooth > intermediate > rough.
|
|
and -ß, core components of the translocation apparatus, were in highest concentration in rough microsomes from pancreas and liver (see Figs. 5 and 7). However, in adrenal microsomes Sec61 and -ß were detectable in similar concentrations in the smooth and rough microsomal subfractions (Fig. 5).
SR and the 12-kDa subunit of the signal peptidase complex (SPC12), representing complexes involved in targeting of nascent chains to the ER and cleavage of signal peptides, respectively, were more uniformly distributed among the subfractions from both liver and adrenal (see Figs. 5 and 7). However, TRAP, one of four subunits of the TRAP complex, which colocalizes with native ribosome-channel complexes, was localized primarily in the ribosome-bearing subfractions from both adrenal and liver (see Figs. 5 and 7). Little TRAP was detected in the smooth microsomes obtained from either organ.
Sec62 and Sec63, mammalian homologs of proteins essential for posttranslational translocation in yeast, were present in both the rough and smooth microsomal subfractions from liver and adrenal (data not shown).
Molecular chaperones
We examined several proteins involved in folding of newly synthesized polypeptides and ER quality control, generally considered to be distributed throughout the ER: two lumenal proteins, BiP and GRP94, and one membrane protein, calnexin (see Table 1). In liver microsomes, particularly those of the guinea pig, BiP and GRP94 were in higher concentration in the rough microsomes (see Fig. 7). In adrenal microsomal subfractions, their concentrations in smooth microsomes were equal to or slightly greater than in rough microsomes (Fig. 5). Calnexin was present in all microsomal subfractions from liver but did not show a consistent differential pattern of distribution (see Fig. 7). In adrenal microsomes, the levels of calnexin were higher in the smooth microsomes (Fig. 5).
Enzymes involved in sterol and steroid synthesis
Having found high levels in adrenal smooth microsomes of all translocon subunits and associated elements examined, except for ribosomal protein and TRAP, it seemed important to confirm that preferential localization in smooth microsomes of enzymes involved in sterol and steroid synthesis was retained in these subcellular fractions. Therefore, we examined the levels of HMGR and two enzymes involved in steroid synthesis, 3ßHSD and 17-hydroxylase, a cytochrome P450 (CYP17). All of these enzymes were in highest concentration in the adrenal smooth microsomes (Fig. 5). None were detectable in corun rough microsomes from liver or pancreas, although HMGR was present in liver smooth and intermediate microsomes (data not shown).
Immunocytochemistry
To determine whether the translocon components and steroidogenic enzymes have a similar distribution in situ, immunocytochemistry was performed on isolated adrenocortical cells. These cells retain their abundant SER in vitro, particularly in the presence of ACTH (74). Both CYP17 and Sec61were distributed fairly evenly throughout the tubular ER network of the isolated cells (Fig. 6, A and B, respectively). Corun controls showed minimal staining (Fig. 6C). Little change in the distribution of CYP17 and Sec61 was detected in ACTH-treated cells at this resolution (data not shown). However, in some cells, hotspots of intense labeling occurred for both proteins, which seemed to be more abundant in ACTH-treated cells (data not shown). More detailed analysis will be required to see whether these correspond to the organized arrays of SER observed in these cells with the electron microscope (4, 5, 6).
|
). Each xenobiotic agent induced the appropriate CYP: PB treatment resulted in increased levels of CYP2B and 3MC treatment resulted in increased levels of CYP1A, particularly in the smooth microsomes (Fig. 7).2
The increase in SER was relatively specific for enzymes involved in xenobiotic metabolism. There was a slight increase in the concentration of HMGR in all liver microsomal subfractions from treated animals and the concentration of the molecular chaperones, BiP and GRP94, did shift toward the smooth microsomal fraction in treated animals (Fig. 7). However, although there was some shift of the OST complex subunits, as well as Sec61 and -ß proteins toward the smooth microsomal fraction, their increases in the lighter fractions were variable, and none achieved the high levels seen in adrenal microsomes (Fig. 7).
Adrenal microsomal subfractions from other species
To eliminate the possibility that the presence of translocon-associated proteins in adrenal smooth microsomes was confined to the guinea pig, we examined similarly prepared microsomes from adrenals of several species: rat, dog, cattle, rabbit, sheep, and pig. In most, as in the guinea pig adrenal, OST and Sec61 complex subunits as well as the molecular chaperones BiP, GRP94, and calnexin, were found to be in equal or greater concentration in smooth microsomes compared with rough microsomes (Fig. 8). The sole exceptions to the equal or greater concentration of these proteins in the smooth microsomes were in the rat adrenal. In the rat adrenal, the OST components as well as BiP and GRP94 were in greater concentration in the rough microsomes. In all of the species examined, however, the SER markers CYP17 and 3ßHSD were in greater concentration in the smooth microsomes than in rough microsomes. In all cases, the ribosomal protein S3 was localized to the ribosome bearing fractions (data not shown). This gives increased confidence that the broader distribution of translocon-associated proteins in the ER is a property of most adrenocortical cells and perhaps of steroid-secreting cells in general.
|
). It was greater than the capacity for N-glycosylation by the adrenal intermediate smooth/rough or rough microsomes. Omission of activated sugars did not significantly change these results, indicating that there are dolichol-linked oligosaccharides already present in the adrenal microsomes (data not shown). Given the levels of OST components detectable by immunoblotting in the adrenal rough microsomes, the activity measured in these microsomes was disproportionately low. This result was very reproducible and may accurately reflect the composition of these microsomes; e.g. some lack of other element(s), as yet undefined, necessary for full activity. However, we cannot exclude the possibility that the low activity resulted from some artifact of preparation.
|
). There were, however, some differences in the TLC patterns of products formed by pancreatic and adrenal microsomes. Of the two major spots resolved, the upper one was preferentially visible in dog pancreatic microsomes, whereas the lower was more prominent in the adrenal smooth microsomes [Fig. 9, B (lanes DP and S) and C (pancreas and adrenal, lane C)].
We then examined the nature of the glycosyl chain more closely by enzymatic digestion. Endoglycosidase H (EndoH) cleaves all N-linked carbohydrate chains normally occurring in the ER between the two innermost GlcNAc residues, leading to very fast migration of the digested product. EndoH digestion was used to confirm that the isolated glycotripeptides contain typical N-linked carbohydrate chains (Fig. 9C, pancreas and adrenal, lane E). In contrast, Jack bean -mannosidase is an exomannosidase that cleaves only terminal -linked mannoses. In dog pancreas-derived glycotripeptides, incubation with this enzyme led to a complete digestion down to the GlcNAc2ßMan core (Fig. 9C, pancreas, lane M). In contrast, when glycotripeptides derived from adrenal smooth microsomes were subjected to this digest (Fig. 9C, adrenal, lane M), an additional spot with lower Rf value was seen. This partially resistant material presumably retains one or more of the terminal glucose residues that are initially present on the transferred N-glycosyl chain (Glc3Man9GlcNAc2). This assumption was confirmed by comparison with glycotripeptides generated in the presence of castanospermine, a glucosidase inhibitor (data not shown). Thus, the slower migration of adrenal-derived material can be explained by a low activity of at least one glucosidase (I and/or II) in adrenal microsomes. If glucosidase II were preferentially affected, retention of monoglucosyl residues could result in more extended association with chaperones such as calnexin, abundant in the smooth microsomes, facilitating retention of glycosylated proteins in the SER (56, 57).
Ribosome targeting and binding
Having found that the OST components formed a functional unit in adrenal smooth microsomes, we then sought to determine whether the SR and Sec61 complexes function in this setting. To determine SR activity, we took advantage of a previously developed photocross-linking assay in which one can assess SR activity by looking for a decrease in the amount of signal peptide-associated SRP54 (23). Truncated RNCs for 86pPL containing photocross-linker-modified lysines, were prepared in a reticulocyte lysate system. In absence of microsomal membranes a 62-kDa photoadduct formed representing the 86pPL signal peptide cross-linked to SRP 54 kDa (24, 25). Addition of SR-containing dog rough microsomes reduces the intensity of the SRP cross-link due to the release of SRP from the nascent chain by SR present in the microsomal membranes (23). In this assay, adrenal smooth microsomes were capable of decreasing the intensity of the SRP cross-link, although not to the same extent as dog pancreatic microsomes (Fig. 10A). Because the levels of SR are lower in adrenal smooth microsomes than in liver or pancreas rough microsomes, it is possible that some ribosome-nascent chain-SRP complexes interacted with SR-deficient Sec61 complexes and were therefore unable to release SRP.
|
).
Cotranslational translocation, signal peptide cleavage, and N-glycosylation
To confirm that the activities of translocon, SPC and OST components could be coordinated in adrenal smooth microsomes, we assayed the ability of adrenal microsomal subfractions to translocate peptides synthesized on ribosomes bound to these membranes as well as to cleave signal peptides and N-glycosylate incoming, newly synthesized polypeptides. The adrenal smooth microsomal subfractions were at least as capable of these functions as adrenal rough microsomal subfractions (Fig 11, A and B, respectively). However, the levels of both activities were considerably lower than in dog pancreatic rough microsomes. Comparison of the levels of N-glycosylation of the opsin fragment with the levels of oligosaccharyltransferase activity measured using the OTP acceptor (Fig. 9A), suggest that the amount of processing observed reflects the rate of targeting and translocation of the newly synthesized heterologous polypeptides rather than the full capacity of these enzyme complexes.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and -ß, key components of the translocation channel, as well as subunits of the OST complex, which is involved in N-glycosylation of newly synthesized polypeptides, are found in adrenal smooth microsomes, derived largely from the abundant SER in these cells. These proteins have been considered to be localized primarily in ribosome-bearing rough microsomes and to be markers of the RER (60, 61, 62, 63), yet in adrenocortical cells they are distributed throughout the tubular ER network, which is predominantly smooth surfaced. This discrepancy forced reevaluation of the nature of the SER and its distinction from RER in steroid-secreting cells as well as the functions of OST and Sec61 in membranes thought to be devoid of de novo protein synthesis.
Despite the broader distribution of Sec61 and OST complex proteins in adrenal microsomes, two features distinguished isolated adrenal rough from smooth microsomes: bound ribosomes and TRAP. These features are shared with rough microsomes prepared from pancreas and liver and are consistent with active cotranslational translocation (30).
However, it seems that there are at least two forms of SER. One, found in steroid-secreting cells, possesses abundant machinery for the translocation and processing of proteins targeted for the ER. The other, characteristic of protein-secreting cells, does not. The high levels of these proteins in adrenal smooth microsomes cannot be attributed to contamination by rough microsomes. The small amount of RER present in adrenocortical cells and the clear localization of ribosomal protein to the rough microsomal fraction preclude this as does their distribution throughout the ER visualized by immunocytochemistry. Furthermore, the broader distribution of these proteins in adrenal ER is not related to the amount of SER per se. Smooth microsomes from livers of animals treated with xenobiotics that expand the SER in hepatocytes did not contain high levels of these proteins.
Although not as dramatic as the data presented here, there have been a few previous reports of smooth endomembrane systems possessing RER-specific proteins in other cell types. Small amounts of translocon-associated proteins have been seen in SER found in mammalian dendrites (Sec61) (82) and Caenorhabditis elegans neurites [ribosome-associated membrane protein 4 (RAMP4), TRAPß, translocating-chain associating membrane (TRAM)] (83). However, as shown in C. elegans, the bulk of the ribosomes and translocon-associated proteins colocalize in the neuron cell bodies that contain abundant RER (83). In studies using less specific reagents, an antibody made against RER membrane proteins reacted with the sarcoplasmic reticulum in skeletal muscle and smooth-surfaced cisternal stacks in Purkinje neurons (84, 85). These compartments both contain high levels of specific multispanning transmembrane (TM) proteins (Ca2+ATPase and inositol 1,4,5-trisphosphate receptor, respectively) (for discussion of these and other ER domains, see Ref.86).
We hypothesize that the components of the translocation apparatus found in adrenal smooth microsomes participate in synthesis of membrane proteins targeted for particular domains in the ER and/or quality control of these proteins. Morphologically distinct domains clearly exist in adrenocortical cells (6, 7) and smaller domains of enzyme complexes have been postulated (87). Two lines of evidence suggest some correlation with cholesterol production. First, translocation components were more prominent in adrenal smooth microsomes of species in which the adrenals synthesize a large proportion of their own cholesterol (88, 89, 90, 91). They were less prominent in SER of the rat adrenal that depends on plasma lipoproteins, rather than de novo synthesis, for cholesterol used in steroid synthesis (89, 90). Second, Sec61 and OST were not as prominent in the equally abundant SER induced in hepatocytes by xenobiotics. These SER membranes showed relative increases in CYPs involved in metabolism of foreign compounds but did not show a relative increase in the concentration of HMGR.
We have demonstrated that Sec61, OST, and other components of the translocation apparatus in the adrenal SER form functional complexes capable of ribosome binding, translocation, signal peptide cleavage, and N-glycosylation. These data confirm and extend an earlier report from Boime and colleagues (92) and our own preliminary report (93). Taken together, they strongly suggest that the adrenal SER is potentially capable of ER-targeted protein synthesis.
Steroid-secreting cells are rich in membrane proteins involved in sterol and steroid synthesis that are potential substrates for such activity, e.g. cytochromes P450 and HMGR as well as sterol regulatory element-binding protein (SREBP), SREBP cleavage-activating protein (SCAP), and Insig-1, proteins involved in the sterol sensitive regulation of the levels of HMGR and other enzymes involved in the cholesterol biosynthetic pathway (94, 95, 96, 97). When adrenals are stimulated with ACTH, HMGR activity and protein content in the SER increase dramatically (11). The increase in HMGR activity parallels that of cytochromes P45017 and -21, proteins involved in steroidogenesis. Coincident with these increases in enzyme activity, the volume of the SER, as assessed by stereological analysis, increases 2-fold, occupying up to 70% of the cytoplasmic volume (6). The SER also undergoes dramatic changes in morphology, often forming large complex arrays (6).
The RER does not undergo such dramatic changes. RER volume increases but remains sparse, occupying less than 0.8% of the cytoplasmic volume, and no significant changes occur in its morphology (6). It remains predominantly as ribosomes bound on short isolated cisternae or in patches scattered along predominantly smooth-surfaced tubules. Longer RER cisternae, although more easily identified, are relatively rare in control and ACTH-treated cells. These morphological data are consistent with data extracted from our previous biochemical studies (65, 66, 98). Although there is an increase in the percentage of the ribosome-bearing subfractions in the total microsomal fraction with ACTH treatment (from 32 to 40%), the majority of the increase is in the intermediate subfraction (from 23 to 30.5%), which would be derived from the dispersed elements, less densely covered with ribosomes.
One model that reconciles the relatively sparse, widely distributed RER elements seen in adrenocortical cells and their nonconcordant increase in comparison with the SER with the distribution of proteins involved in translocation and processing throughout the ER is that protein synthesis on membrane-bound ribosomes can potentially occur at multiple points on the tubular ER membranes of these cells but occurs only in a small fraction at any one time. The rough microsomal fraction then represents a snapshot of where cotranslational translocation was occurring at the time of fractionation. Although some rough microsomes could be derived from RER devoted to protein synthesis, the majority would be from microsomes that are rough protemp. The Sec61 and associated complexes present in the smooth microsomes would represent translocon components awaiting arrival of targeted ribosome/nascent chain complexes. Stimulation of synthesis would increase the number of protemp ribosome bearing complexes, but their sum over the period of induced membrane synthesis would be greater that that captured at any one moment. Viewed from this perspective, the elements involved in translocation and processing found throughout the SER in steroid-secreting cells would be functionally similar to those seen in the RER of protein-secreting cells but dynamically distinct.
However, the broad distribution of translocation apparatus and oligosaccaryltransferase proteins in the SER may suggest that they have additional functions in this site. The high levels of BiP, GRP94, and calnexin in the SER are consistent with the functioning of this compartment not only in protein folding and assembly but also in regulated forms of endoplasmic reticulum-associated degradation (50, 51, 52). The highly regulated degradative pathways of HMGR and the degradation of some cytochrome P450s occur via the ubiquitin proteasome pathway (99, 100, 101). Because proteosomes are in the cytosol, this requires retrograde transport of the protein out of the ER. The Sec61 channel complex has been implicated in retrograde translocation (14, 15). Recent studies show that the membrane protein Derlin-1 is required for retrograde translocation of membrane proteins (102, 103). It is linked via another recently discovered protein VIMP [VCP (valosin-containing protein)/p97-interacting membrane protein] to the cytosolic AAA ATPase, p97, which is essential for movement of polypeptides from the membrane to the cytosol (102). The involvement of Sec61 vs. Derlin-1 may depend on the endoplasmic reticulum-associated degradation substrate (104). More speculative is a role for OST components in quality control. Recently, RI and other OST complex proteins have been found tightly associated with specific membrane proteins after their integration into the membrane (105). In one case, US11, the protein also associates with p97 and the ER chaperones BiP and calnexin (106), suggesting a link between the OST complex and quality control processes (105). Yeast OST subunits, OST3 and OST6, and their mammalian homologs, N33 and implantation-associated protein, contain thiroedoxin domains. It has been suggested that they may be involved in disulfide oxidoreductase regulatory mechanisms (105). This could be important if retention and/or release of ER proteins is dependent on disulfide bonds with associated proteins.
In conclusion, the abundant SER in steroid-secreting cells contains high levels of proteins involved in translocation and processing of ER-targeted proteins. In these cells, ribosomes are scattered in patches along the predominantly smooth-surfaced tubular network. This form of RER increases when the cells are treated with ACTH. We suggest that these smaller RER elements represent transitory complexes between ribosomes and the translocation apparatus participating in local synthesis of TM proteins involved in sterol and steroid synthesis within specific domains of the ER (see Refs.13 , 107 , and 108 for review of TM protein synthesis) as well as proteins involved in membrane protein folding and/or identification of misfolded proteins and their retrograde transport (104). Targeting of the translocation substrates to specific ER domains could involve not only SRP and SR but also mRNA trafficking (109) and signal sequence information (110, 111) as well as cytoplasmic factors and the membrane protein and lipid composition of the ER domains (111). The equilibrium of bound ribosomes in any one domain could be determined not only by the rate of synthesis and competition for synthesis of cytoplasmic proteins, as suggested by Potter and Nicchitta (112), but also by competition for retrograde transport of proteins out of the ER. Alternative roles in synthesis vs. degradation could be regulated by levels of cholesterol (113) or other lipid moieties (114) in the ER membranes, determined, in part, by levels of trophic hormones.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Present address for A.S.: Department of Anatomy and Cell Biology, Temple University, Philadelphia, Pennsylvania.
Present address for K.v.L.: Department of Radiology, Neuroprotection Research Laboratory, Massachusetts General Hospital, Charlestown, Massachusetts.
Results from this work were presented in part at the Annual Meeting of the Society for Cell Biology, San Francisco, CA, December 2000 and 2002, and the IX Conference on the Adrenal Cortex, San Francisco, CA, June 2002 (93 ).
First Published Online June 9, 2005
Abbreviations: BiP, Ig heavy chain-binding protein; bw, body weight; CTF, C-terminal fragment; CYP, cytochrome P450; DAB, 3,3'-diamino-benzidine tetrahydrochloride; DAD, defender against apoptotic death; ECL, enhanced chemiluminescence; EndoH, endoglycosidase H; ER, endoplasmic reticulum; ERAD, ER-associated degradation; GRP, glucose regulated protein; HMGR, 3-hydroxy-3-methylglutaryl-coenzyme A reductase; 3ßHSD, 3ß-hydroxysteroid dehydrogenase; Hsp, heat shock protein; 3MC, 3-methylcholanthrene; op-156, the first 156 codons of bovine opsin; OST, oligosaccharyl-transferase; OTP, octanoyl tripeptide; PB, phenobarbital; PKRM, puromycin/KOAc washed microsomes; 86pPL, 86-amino-acid preprolactin; RAMP4, ribosome-associated membrane protein 4; RER, rough endoplasmic reticulum; RI and RII, ribophorins I and II; RNC, ribosome-nascent chain complex; SDS, sodium dodecyl sulfate; SER, smooth endoplasmic reticulum; SP, signal peptide; SPC, signal peptide cleavage protein, signal peptidase; SR, SRP receptor; SRP, signal recognition particle; TCA, trichloroacetic acid; TLC, thin-layer chromatography; TM, transmembrane; TRAP, translocon-associated protein.
1 The recovery of rough microsomes isolated by subcellular fractionation of the guinea pig adrenal is greater than what would be predicted based on previous stereological analysis of RER and SER volumes in the inner cortex (6 ). This reflects, in part, the fact that the zona glomerulosa cells, which possess more RER than the inner cortical cells, were not included in the morphological analysis. It also reflects an underestimate of RER volume by the morphological analysis. Small cisternae and small patches, or even single ribosomes, bound to tubular elements would not have been adequately accounted for by the grid technique employed in this stereological study. 
2 The proliferation of SER coincident with induction of CYP1A observed here in guinea pig liver differs from previous reports for rodent liver and 293T cells (see Ref.8 ). In those experiments, increases in CYP1A were not accompanied by increases in the SER. This suggests that there may be differences in the effects of aromatic hydrocarbons on CYP1A induction and consequent SER proliferation, depending on the species or cell type and specific substance used.
Received March 30, 2005.
Accepted for publication May 27, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
