* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Endoplasmic Reticulum-Localized Amyloid β
Survey
Document related concepts
Biochemical switches in the cell cycle wikipedia , lookup
Cell membrane wikipedia , lookup
G protein–coupled receptor wikipedia , lookup
Cytokinesis wikipedia , lookup
Hedgehog signaling pathway wikipedia , lookup
Magnesium transporter wikipedia , lookup
Protein phosphorylation wikipedia , lookup
Protein moonlighting wikipedia , lookup
Signal transduction wikipedia , lookup
Endomembrane system wikipedia , lookup
Transcript
Copyright Traffic 2004; 5: 89–101 Blackwell Munksgaard # Blackwell Munksgaard 2004 doi: 10.1046/j.1600-0854.2003.00159.x Endoplasmic Reticulum-Localized Amyloid b-Peptide is Degraded in the Cytosol by Two Distinct Degradation Pathways Anton Schmitz1,*, Andrea Schneider1, Markus P. Kummer1,2 and Volker Herzog1 1 Institut für Zellbiologie, Rheinische Friedrich-WilhelmsUniversität, Ulrich-Haberland-Str. 61a, 53121 Bonn, Germany, 2 Present address: UCSD, Department of Neurosciences, La Jolla, CA 92093–0691 *Corresponding author: Anton Schmitz, anton.schmitz@ uni-bonn.de The paradigm of endoplasmic reticulum (ER)-associated degradation (ERAD) holds that misfolded secretory and membrane proteins are translocated back to the cytosol and degraded by the proteasome in a coupled process. Analyzing the degradation of ER-localized amyloid b-peptide (Ab), we found a divergence from this general model. Cell-free reconstitution of the export in biosynthetically loaded ER-derived brain microsomes showed that the export was mediated by the Sec61p complex and required a cytosolic factor but was independent of ATP. In contrast to the ERAD substrates known so far, the exported Ab was degraded by both, a proteasome-dependent and a proteasome-independent pathway. RNA interference experiments in Ab-transfected cells identified the protease of the proteasome-independent pathway as insulin-degrading enzyme (IDE). The IDE-mediated clearance mechanism for ER-localized Ab represents an as yet unknown type of ERAD which is not entirely dependent on the proteasome. Key words: amyloid b-peptide, endoplasmic reticulum, insulin-degrading enzyme, protein degradation, protein translocation Received 10 October 2003, revised and accepted for publication 15 November 2003 Misfolded secretory and membrane proteins are recognized by quality control mechanisms in the endoplasmic reticulum (ER). They are retained in the ER and finally removed by a series of events named ER-associated degradation (ERAD). According to the current model of ERAD, the misfolded proteins are translocated from the ER back to the cytosol, where they are degraded by the 26S proteasome [for recent reviews, see (1–3)]. Under normal conditions the misfolded proteins are not found in the cytosol, suggesting that their export and their proteasomal degradation are coupled processes. Evidence supporting this view has been obtained for several proteins. For example, a misfolded form of yeast carboxypeptidase Y did not accumulate in the cytosol but was retained in the ER when polyubiquitination, which serves as the main signal for degradation by the 26S proteasome, was inhibited (4,5). The membrane extraction of an artificial degradation substrate (6) and of CD4 (7), as well as the export of an unassembled immunoglobulin light chain (8), were reduced when the catalytic activity of the proteasome was inhibited. However, there is also evidence for a less tight coupling of translocation and proteasomal activity. The MHC class I heavy chain (9) and the cystic fibrosis transmembrane conductance regulator (10) were found in the cytosol after inhibition of proteasomal activity. During the intoxication of their target cells, several bacterial and plant toxins such as cholera toxin or ricin are transported in a retrograde manner through the secretory pathway to the ER, from where they translocate into the cytosol. Their translocation appears to depend on the Sec61p complex (11–13), which also mediates the translocation of misfolded proteins (14–18). In contrast to ERAD substrates, however, these toxins largely escape the proteasomal degradation and act as functional enzymes in the cytosol. One characteristic feature of Alzheimer’s disease (AD) is the accumulation of the amyloid b-peptide (Ab) in extracellular plaques. Ab is generated by proteolytic processing of the amyloid precursor protein (APP) in the secretory and endocytic pathways but not completely released into the extracellular space. Consequently, Ab is found not only in the extracellular space but also in intracellular compartments [for review, see (19)]. Ab, in particular its 42-amino-acidslong form (Ab42), is also generated in the ER but only very inefficiently transported further along the secretory pathway (20–24). Nevertheless, only small amounts of Ab42 are found in the ER, even upon over-expression of APP. Therefore, we asked whether the accumulation of Ab42 in the ER is counteracted by its removal through the ERAD machinery. We show that the export of Ab is distinct in several aspects from that of other ERAD substrates and that the exported Ab is degraded by proteasome-dependent and proteasome-independent pathways. RNA interference experiments identified the protease of the latter pathway as insulindegrading enzyme (IDE). Results Ab 42 is exported from microsomes The hypothesis that ER-localized Ab is degraded by the proteasome implies that it is exported from the ER to the 89 Schmitz et al. cytosol. We obtained first evidence for this hypothesis by morphological analysis of CHO cells transfected with ER-targeted Ab. In these cells Ab was found not only in the ER but also in the cytosol, where it formed aggregates (25). To study the export biochemically, a cell-free export system was applied. In principle, ER-derived brain microsomes were loaded with Ab42 by in vitro translation and the release of the imported Ab42 was monitored. To allow the import of Ab42 into the microsomes, the signal peptide of APP was fused to the N-terminus of Ab42 (Figure 1A). To verify the import, Ab42-loaded microsomes were re-isolated by centrifugation and subjected to protease protection analysis (Figure 1B). Trypsin treatment of the microsomes resulted in the loss of the non-imported, signal peptide containing precursor (sAb42). In contrast, Ab42 was protected from proteolysis in the absence of detergent (lane 2) but was degraded when detergent was added (lane 3), demonstrating complete import of Ab42 into the microsomal lumen. To examine the export of Ab42, the loaded microsomes were re-isolated, washed with high-salt to remove proteins adsorbed to the outside and resuspended in export buffer. After incubation at 37 °C the reaction was separated into pellet and supernatant. During the chase an increasing amount of Ab42 became detectable in the supernatant (Figure 2A, upper panel). This increase was prevented by incubation on ice (data not shown). The total amount of Ab42 present after 6 h chase accounted for 114% (12% SE) compared to the amount present at 0 h chase, indicating that the total amount of Ab42 did not change during the course of the experiment. Kinetic Figure 1: Import of Ab into microsomes. (A) For the construction of sAb42, the signal peptide (broken line) of APP was directly fused to the N-terminus of Ab42 (unbroken line). The arrow indicates the predicted cleavage site of signal peptidase. (B) sAb42 was translated in a reticulocyte lysate in the presence of brain microsomes. After translation, the microsomes were isolated by centrifugation and treated with trypsin in the absence or presence of Triton X-100 (TX-100). The reaction products were analyzed by SDS-PAGE and autoradiography. 90 analysis showed that the export of Ab42 was a fast process occurring largely during the first 30 min (Figure 2B). To demonstrate that the export did not result from rupture of the microsomal membranes, the release of the ER-resident chaperone BiP was determined. BiP was not released into the supernatant (Figure 2A, lower panel). In contrast, more than 90% of BiP was released into the supernatant when the microsomal membrane was disrupted by the addition of detergent (data not shown). The band running below the BiP band is cytosolic Hsc70, which is also recognized by the anti BiP-antiserum. Budding of transport vesicles from the microsomes was unlikely under the experimental conditions used here. Nevertheless, the possibility that the Ab42 in the supernatant was contained within transport vesicles and not translocated across the microsomal membrane was analyzed by protease protection analysis at the end of an export reaction (Figure 2C). Whereas the non-exported Ab42 still residing inside the microsomes was protected from trypsin in the absence of detergent (lane 2), the Ab42 in the supernatant was readily degraded (lane 5). The susceptibility to trypsin showed that the Ab42 present in the supernatant was not contained in vesicles but exported across the microsomal membrane and freely accessible in the buffer. The retention of BiP in the microsomes did not exclude the possibility that Ab42 was released due to leakiness of the microsomes for small molecules. Therefore, the permeability of the microsomes for substances which were much smaller than Ab42 was determined. For this purpose, Ab42-loaded microsomes were incubated with sulfo-NHSbiotin (molecular weight 443.4 Da) and the biotinylation of the lumenal proteins BiP and Glucose-regulated protein 94 (Grp94) was determined by their retention on a streptavidin column followed by immunoblotting. The biotinylation was performed in either the absence or the presence of 13 mM n-octylglucoside, which served as a positive control mimicking leaky microsomes. Biotinylation of a membranefree extract of lumenal proteins was equally efficient in the absence or presence of the detergent, indicating that the biotinylation itself was not affected by the detergent (Figure 3A). In contrast, in the Ab42-loaded microsomes biotinylation of BiP and Grp94 was almost undetectable in the absence of the detergent, but was readily observed in its presence (Figure 3B; compare lane 3 with 4 and lane 5 with 6, respectively). Quantitative analysis showed that 1.1% (0.6% SE, n ¼ 3) of the total Grp94 was accessible for the sulfo-NHS-biotin in the absence of detergent as compared to 20.1% (9.0% SE) in the presence of detergent. Complete biotinylation in the presence of detergent was not achievable because this would have required complete permeabilization of the microsomes and would have resulted in the total loss of Grp94. The finding that the low molecular weight sulfo-NHS-biotin did not penetrate the microsomes as long as the membrane Traffic 2004; 5: 89–101 Cytosolic Degradation of Ab42 Figure 2: Ab42 is exported from microsomes. (A) After loading with Ab42 by in vitro translation, the microsomes were re-isolated, highsalt-washed, resuspended in export buffer containing 5 mg/ml cytosol and divided into aliquots. One aliquot was directly separated into a pellet and a supernatant fraction, the other aliquots were first incubated for the indicated time at 37 °C. The reaction products were analyzed by SDS-PAGE. Ab42 was visualized by autoradiography. BiP was immunodetected using an anti BiP-antiserum which cross-reacts with cytosolic Hsc70. (B) Graphical representation of 5 independent export reactions performed as described above. The error bars represent the standard deviation. (C) To exclude the possible localization of the exported Ab42 in vesicles, a proteolysis protection analysis was performed. At the end of a 1-h export reaction performed as in (A), the microsomal pellet containing the non-exported Ab42 (p) and the supernatant containing the exported Ab42 (s) were treated with trypsin in the absence or presence of Triton X-100 (TX-100). was not permeabilized by the detergent showed that the microsomes were sealed by an intact membrane, and strongly argued against the possibility that the export of Ab42 was a result of microsomal leakiness. To analyze whether the export of Ab42 required ATP, export reactions were performed in the presence of an ATP depleting or regenerating system, respectively. Unexpectedly, the export of Ab42 was independent of ATP (Figure 4A). Ab42 is a relatively small molecule of 4.5 kDa, which may fit into the presumed translocation channel without the need of ATP-dependent unfolding. Therefore, we Traffic 2004; 5: 89–101 reasoned that elongation of the peptide might render its export ATP dependent. To test this assumption, the export of Ab42 which had been N-terminally elongated by 42 (42Ab42) or 81(81Ab42) amino acids was analyzed (Figure 4B). Increasing the length of the peptides resulted in an increasing ATP dependence of their export (Figure 4C). This correlation suggested that it was indeed the small size of Ab42 which allowed its ATP-independent export. As the export of ERAD substrates appears to be mediated by the Sec61p complex, its involvement in the export of Ab42 was investigated. This channel also mediates the 91 Schmitz et al. demonstrated that the export of Ab42 was mediated by the Sec61p complex. As known so far, the export of ERAD substrates depends on cytosolic proteins. To test whether cytosolic proteins were also required for the export of Ab42 high-salt-washed, Ab42-loaded microsomes were incubated in the presence of 5 mg/ml brain cytosol which had been inactivated by heating (95 °C, 10 min). Heat inactivation of the cytosol almost completely inhibited the export (Figure 6A). The stimulatory effect of the untreated, active cytosol was concentration dependent (Figure 6B). Figure 3: Ab42-loaded microsomes are sealed by a tight membrane. (A) Membrane-free lumenal ER-proteins were incubated with sulfo-NHS-biotin in the absence or presence of n-octylglucoside (OG). Biotinylated proteins were detected by streptavidin-peroxidase. (B) Ab42-loaded microsomes were incubated with sulfo-NHS-biotin in the absence or presence of n-octylglucoside (OG). The microsomes were lysed and BiP (lanes 1–4) and Grp94 (lanes 5–8) were identified by immunoblotting either directly in the lysate (total; lanes 1, 2, 7, 8) or after the biotinylated proteins had been separated on a streptavidin column (biotinylated, lanes 3–6). The eluate from the column was normalized to the total amount of BiP or Grp94, respectively, present in the lysate. import of proteins into the ER. Therefore, the export of proteins can be inhibited by blocking the channel with ribosome nascent chain complexes (RNCs), which are short polypeptides still bound to the ribosome (13). For this purpose, the Sec61p complexes of Ab42-loaded microsomes were saturated with RNCs formed of a 75-aminoacids-long fragment of cholera toxin subunit A (CTAD137; Figure 5A). As the signal peptide-containing CTAD137 is stably inserted into the Sec61p channel, it should inhibit the export of Ab42 if the export is mediated by the Sec61p complex. As a control, a 75-amino-acids-long fragment of dihydrofolate reductase (DHFRD112) was used. DHFRD112 which lacks a signal peptide for import into the ER does not stably interact with the channel and thus should be ineffective. Export reactions performed in the presence of the RNCs showed that this was indeed the case (Figure 5B). The export of Ab42 was blocked by CTAD137 (upper panel), but proceeded undisturbed in the presence of DHFRD112 (lower panel). The ongoing recruitment of CTAD137 during the first 30 min of the export reaction is explained by the fact that the microsomes had to be preincubated with the RNCs on ice, which is less efficient than binding at 37 °C. Preincubation at 37 °C was not possible because then Ab42 would have been exported before the blocking of the Sec61p complex was effective. The selective inhibition by signal peptide-containing RNCs 92 The finding of about 40% of total Ab42 in the cytosol showed that the export of Ab42 and its degradation were not as tightly coupled as reported for other ERAD substrates. In some cases, the tight coupling of export and degradation is thought to result from the proteasome being involved not only in the actual degradation but also in the retro-translocation itself. Therefore, the export of Ab42 was performed in the absence of proteasomes. For this purpose, the cytosol was depleted of proteasomes by ultracentrifugation (Figure 6C). However, the export of Ab42 was not affected by the depletion (Figure 6D), indicating that it did not require the proteasome. Exported Ab 42 is degraded by the proteasome and insulin-degrading enzyme To analyze whether Ab42 was, after its export, degraded by the proteasome, HeLa cells were transfected with ER-targeted Ab42 and chased in the presence or absence of the proteasome inhibitor MG-132 (Figure 7). Indeed, the degradation of Ab42 was attenuated by the proteasome inhibitor. However, the inhibition was not complete and about 70% of Ab42 were still degraded in the presence of MG-132, suggesting that an alternative pathway might contribute to the degradation of ER-localized Ab42. Reports exist on the IDE-mediated degradation of Ab in cell culture (26–28) and in a mouse model (29), and the degradation of Ab was reconstituted with purified IDE (26,30). These studies concentrated on extracellular Ab and Ab found in the secretory and endocytic pathways. But as IDE is predominantly localized in the cytosol (31), the degradation of cytosolic Ab42 could also be mediated by IDE. Therefore, RNA interference was used to investigate the involvement of IDE in the degradation of ER-localized Ab42. To deplete cells of IDE, HeLa cells were transfected with IDE-specific small interfering doublestranded RNA (siRNA). Immunoblot analysis verified that transfection with siRNA, but not with single-stranded sense RNA used as a control, induced the depletion of the majority of IDE (Figure 8A). In cell culture Ab42 is found in a soluble and an insoluble form (32). To analyze both fractions, the RIPA-soluble Ab42 was immunoprecipitated, whereas the RIPA-insoluble material was extracted by sonification in 2 concentrated sample buffer containing Traffic 2004; 5: 89–101 Cytosolic Degradation of Ab42 Figure 4: Export of Ab42 is independent of ATP. (A) Ab42-loaded microsomes were used for an export reaction as described in Figure 2(A) in the presence of an ATP-depleting (– ATP) or ATP-regenerating (þ ATP) system, respectively. (B) Ab42 was extended at its N-terminus by either 42 amino acids (42Ab42; upper panel) or 81 amino acids (81Ab42; lower panel). Microsomes were loaded by in vitro translation of the respective construct and used for an export reaction as in (A). (C) Graphical representation of export reactions performed as in (A) and (B). The columns show the export after 1 h incubation. For the calculation of the export rate, the material present in the supernatant at 0 min chase was subtracted. The error bars give the standard error (n ¼ 3). 5% SDS and directly loaded on the gel (see Materials and Methods). By this procedure, Ab42 is the only peptide found in the RIPA-insoluble fraction of sAb42-transfected Traffic 2004; 5: 89–101 HeLa cells without any background of other peptides of the same molecular weight (Figure 8B). The result was verified by formic acid extraction. Immunoprecipitation 93 Schmitz et al. Figure 5: Export of Ab42 is mediated by the Sec61p complex. (A) Microsomes were incubated with CTAD137 isolated from the indicated volumes of a translation reaction. After washing with high-salt buffer the microsomes were isolated, and the CTAD137 bound in a salt-resistant manner was visualized. Saturation of binding was achieved with a 40-ml translation reaction. (B) Ab42-loaded microsomes were reisolated, resuspended in export buffer containing 5 mg/ml cytosol and split into two aliquots. One aliquot was incubated for 10 min on ice with saturating amounts (corresponding to a 40-ml translation reaction) of the signal peptidecontaining CTA1D137 (upper panel), the other with DHFRD112 which did not contain a signal peptide and was used as a control (lower panel). After binding of the RNCs, the microsomes were used for an export reaction as described in Figure 2(A). Ab42 , CTAD137 and DHFRD112 were detected by autoradiography. was omitted because immunoprecipitation of RIPA-insoluble Ab42 is not quantitative [own observation and (32)]. To determine the degradation rate of Ab42, HeLa cells were cotransfected with sAb42 and either IDE-specific siRNA or sense RNA as a control. The cells were labeled for 15 min with [35S]-methionine and chased for up to 2 h. Where indicated, 20 mM MG-132 was present during the entire pulse and chase periods to inhibit the proteasome (Figure 8C). When used alone, either inhibitor, MG-132 (lanes 1–3) or the siRNA treatment (lanes 10–12), inhibited the degradation of Ab42 only weakly. Efficient inhibition required the presence of both inhibitors (lanes 7–9). The inhibition by either inhibitor alone was not statistically significant (Figure 8D). However, when both proteases, IDE and the proteasome, were inhibited the degradation of soluble and total Ab42 was statistically significantly reduced, indicating that both proteases contributed to its degradation. If the two proteases operated sequentially in a common proteolytic pathway, each inhibitor should have led to the same degree of inhibition as their combined use. However, the inhibitory effects were additive, indicating that the proteasome and IDE were part of two independent parallel pathways. 94 For the experiments described above, sAb42 was expressed from the strong CMV promoter. Therefore, the inhibitory effect of MG-132 on the degradation of soluble Ab42, although not found for total Ab42, could have indicated that at lower expression levels Ab would be degraded, like other ERAD substrates, only by the proteasome-dependent pathway, and that it became a substrate for the normally unused IDE-dependent pathway only because the proteasome-dependent pathway was saturated. To exclude this possibility a minimal HSV thymidine kinase promoter, a promoter much weaker than the CMV promoter, was used to express sAb42. The expression level of sAb42 under control of this promoter was reduced to 15% (5% SE) as compared to the expression level under control of the CMV promoter. Under this condition, insoluble Ab42 was not detectable (not shown). The soluble Ab42 was almost completely degraded in the first hour of the chase as long as one protease, the proteasome or IDE, was active (Figure 9A). Neither MG-132 (lanes 1–3, upper panel) nor the siRNA treatment (lanes 4–6, lower panel) resulted in significant inhibition. Only the simultaneous inhibition of both proteases reduced the degradation of Ab42 (lanes 1–3, lower panel). When both proteases Traffic 2004; 5: 89–101 Cytosolic Degradation of Ab42 Figure 6: Export of Ab42 requires cytosol but is independent of the proteasome. (A) After loading with Ab42, the microsomes were washed with 0.5 M potassium acetate to remove adsorbed cytosolic proteins. The microsomes were resuspended in export buffer containing 5 mg/ml untreated or heat-inactivated brain cytosol, respectively. The export reaction was performed as described in Figure 2(A). p, pellet; s, supernatant. (B) Graphical representation of 3 independent export reactions performed as in (A) in the presence of the indicated concentrations of untreated cytosol. The error bars represent the standard error. (C) For the depletion of proteasomes the cytosol was centrifuged at 500 000 g for 1 h. The supernatant was recovered and the absence of proteasomes verified by immunodetection using an antibody against subunit a7 of the 20S core of the proteasome. (D) An export reaction was performed as in (A), the only difference being that the cytosol was depleted of proteasomes by ultracentrifugation. Traffic 2004; 5: 89–101 95 Schmitz et al. were active, the amount of Ab42 detectable directly after the pulse (lane 4, upper panel) was less than in the presence of inhibitors because a major fraction was degraded already during the pulse. This rapid degradation rate implies fast export of Ab42 from the ER, which is in full agreement with the export kinetics determined in the cellfree assay (see Figure 2B). The degradation of Ab42 was statistically significantly reduced only when both proteases, the proteasome and IDE, had been inhibited (Figure 9B). This finding indicated that each of the two pathways was functional independently of the other and could compensate the inhibition of the other. The redundancy of these two pathways explains why inhibition of the proteasome did not alter the turnover of Ab42 in an APP-transfected neuronal cell line (33), a model in which low levels of Ab were generated. In summary, our results show that ER-localized Ab42 is degraded by two distinct cytosolic degradation pathways, one being dependent on the proteasome and the other on IDE. Therefore, one possible explanation for the inefficient degradation is that the exported Ab42 is not ubiquitinated. To our knowledge, with the exception of pro-a factor (35), all ERAD substrates known so far are polyubiquitinated during their export, and the polyubiquitination is required for their efficient degradation. The AAA-ATPase p97/ Cdc48 which binds to mono- and polyubiquitinated proteins when associated with Ufd1/Npl4 (36) has been shown to be required for the export and degradation of several ERAD substrates (37–41). Thus this factor might possibly provide the link between export and efficient targeting to the proteasome. The function of p97 is, however, ATP dependent (42), arguing against its involvement in the export of Ab42. Indeed, recent experiments showed that the export of Ab42 was independent of p97 (data not shown). It should be mentioned that the proteasome is able to degrade at least some proteins without their prior ubiquitination, e.g. ornithine decarboxylase (43) or pro-a factor (35). Thus, the lack of ubiquitination may only partially explain why the proteasome degrades the exported Ab42 less efficiently than other ERAD substrates. Discussion An unexpected feature of the export of Ab42 is its independence of ATP. Although it is generally accepted that protein export from the ER to the cytosol is an ATP-dependent process, it should be noted that the experimental data on the membrane translocation itself are rather limited. To our knowledge, unglycosylated pro-a factor (44) and MHC class I heavy chain (14,45) are the only ERAD substrates for which the ATP dependence of the actual export has been shown. Our results show that increasing the length of the Ab42 peptide resulted in a corresponding increase in the ATP dependence of the export. They are therefore in line with the assumption that it is the small size of Ab42 which allows its ATP-independent export. It is assumed that ATP is required for the unfolding of the export substrate to fit into the translocation channel formed by the Sec61p complex. Alternatively, an ATP-dependent conformational rearrangement of the translocation channel may permit the transit of a folded protein, as has been suggested for MHC class I heavy chain (46). As N-glycans are rather bulky and rigid structures extending from the protein backbone, such a conformational rearrangement may also explain why the export of glycopeptides from the ER is ATP dependent (47,48). However, all these constraints may not pertain to the small and unglycosylated Ab42 peptide, resulting in an ATP-independent export process. The ATP independence of the basic export mechanism does not, however, exclude the existence in vivo of superimposed ATP-dependent regulatory mechanisms which may have not been active in the export assay. In this study, we have identified the export of Ab42 to the cytosol followed by its IDE-mediated degradation as an as yet unknown type of ERAD. The independence of proteasomes distinguishes this pathway from the general paradigm of ERAD, which is characterized by the tight coupling of export and proteasomal degradation. The uncoupling of export from proteasomal degradation has recently been described for ricin (11) and cholera toxin (13,34). Ab42 is, however, the first endogenous protein for which this uncoupling has been demonstrated. The reason why Ab42 is not as efficiently degraded by the proteasome as other ERAD substrates is as yet unknown. In the export experiments the molecular weight of the exported Ab42 was unchanged, and Ab42 could not be precipitated with anti-ubiquitin antibody (data not shown). Figure 7: The degradation of ER-localized Ab42 is only partially sensitive to inhibition of the proteasome. HeLa cells were transfected with ER-targeted Ab42 and chased in the presence or absence of the proteasome inhibitor MG-132. Ab42 was immunoprecipitated and visualized by autoradiography. The numbers give the relative amount of Ab42 (0 h: 100%). 96 The cytosolic degradation of ER-localized Ab42 may be relevant for the pathogenesis of AD. On one hand, it constitutes a clearance mechanism for ER-localized Ab42, which explains why Ab is found in the ER only upon overexpression of APP. Thus it may prevent Ab-induced disturbance of ER function and represents a potentially Traffic 2004; 5: 89–101 Cytosolic Degradation of Ab42 Figure 8: ER-localized Ab42 is degraded by both the proteasome and IDE. (A) HeLa cells were cotransfected with pCIsAb42 and single-stranded sense RNA or IDE-specific siRNA, respectively. Two days after transfection, IDE was detected by immunoblotting. HeLa, untransfected HeLa cells. (B) The RIPA-insoluble fractions of untransfected (HeLa) or sAb42-transfected (Ab) HeLa cells were extracted with either 5% SDS as described in Materials and Methods or with formic acid and separated by gel electrophoresis. (C) HeLa cells were transfected as in (A). Two days after transfection, the cells were labeled for 15 min with [35S]-methionine and chased for the indicated time in the presence or absence of 20 mM MG-132 (MG). Ab42 in the RIPA-soluble (sol) and insoluble (insol) cell fractions was detected as described in Materials and Methods. 75% of the soluble fraction and 25% of the insoluble fraction were loaded on the gel. (D) Statistical evaluation of 4 independent experiments performed as in (C). For each fraction the amount of Ab42 remaining after 1 h chase is shown (100%: Ab42 present in the respective fraction at 0 h chase). Total Ab42 corresponds to the sum of soluble and insoluble Ab42 corrected for the different amounts loaded onto the gel. The bars represent the standard error. The asterisks indicate a level of significance of p < 0.05. neuroprotective mechanism. On the other hand, the export of Ab42 may under certain conditions result in its accumulation in the cytosol. Indeed, Ab has been found in the cytosol upon overexpression of ER-targeted Ab42 (25) or APP (49). Cytosolic Ab is able to induce apoptosis even in the absence of visible aggregates, as was demonstrated by the cytosolic expression of Ab in transgenic mice (50) and cell culture (51) model systems, as well as by microinjection of Ab42 into the cytosol of cultured neurons (52). Traffic 2004; 5: 89–101 However, the relevance of these reports for the pathogenesis of AD was uncertain because Ab is not generated in the cytosol. Our results provide the missing link between the generation of Ab in the secretory pathway and its possible pro-apoptotic effect in the cytosol. They allow for the hypothesis that insufficient degradation of Ab42 exported from the ER to the cytosol could result in a cytosolic pool of Ab42 triggering apoptosis. Indirect support for this hypothesis comes from reports demonstrating that 97 Schmitz et al. amino acid 42 of Ab. Using the Dovetail PCR cloning kit (MBI Fermentas, Vilnius, Lithuania) the amplified sequences were ligated in such a way that no additional amino acids were introduced into the protein (see Figure 1A) and cloned into the vector pCI-neo (Promega, Madison, WI, USA) (pCIsAb42). For 42Ab42 and 81Ab42, the 42 or 81 amino acids directly N-terminal to the Ab region were included. For the construction of pTKsAb42, the sequence of the GeneSwitch protein of pSwitch (Invitrogen, Groningen, the Netherlands) was replaced by the sequence coding for sAb42. The sequences of all constructs were verified by DNA sequencing (GATC Biotech, Konstanz, Germany). For in vitro transcription, the linearized construct was transcribed with the AmpliScribe T7 Kit (Epicentre Technologies, Madison, WI, USA). The siRNA sequence (AAGUCUCCUGAAGACAAGCGA) targeting IDE corresponds to nucleotides 223–243 of human IDE (EMBL/GenBank Acc. No. M21188). Sense and antisense RNAs (Dharmacon Research, Lafayette, CO, USA) were deprotected and annealed as described (57). Figure 9: IDE-mediated degradation occurs at low-level expression of Ab42. (A) For low-level expression of Ab42, HeLa cells were cotransfected with pTKsAb42 and single-stranded sense RNA or IDE-specific siRNA, respectively. The degradation of the soluble Ab42 was analyzed as described in Figure 8(C). Insoluble Ab42 was not detectable. (B) Statistical evaluation of 3 independent experiments performed as in (A). The amount of Ab42 remaining after 1 h chase is shown (0 h: 100%). The bars represent the standard error and the asterisks a level of significance of p <0.05. the activities of the proteasome (53) and of IDE (54) are reduced in the brains from AD patients. In addition, possible genetic linkage of IDE with late-onset AD has been reported (55,56). Materials and Methods Molecular biology methods For the construction of sAb42, the sequences coding exactly for the signal peptide or the Ab42 region of APP, respectively, were amplified by PCR using the cDNA of human APP695 as a template. The 30 -primer for the amplification of the Ab42 region introduced a stop codon after 98 Export assay Microsomes were prepared from the gray matter of porcine brain or from pancreas obtained from the local slaughter house (58). Both types of microsomes gave identical results in the export assay. The export assay was performed as described (13) with minor modifications. All microsomes used for an experiment were loaded in a single in vitro translation reaction containing 50% reticulocyte lysate (Promega), 0.3 mCi/ml [35S]-methionine, 20 ng/ml mRNA, 0.4 U/ml RNase inhibitor (MBI Fermentas) and 0.1 eq/ml microsomes [for definition of eq, see (58)]. After 10 min at 30 °C, the translation was stopped with puromycin (2 mM final concentration). The microsomes were isolated by centrifugation (9000 g, 10 min, 4 °C) and washed in 0.5 M potassium acetate, pH 7.4 for 20 min on ice. After centrifugation the microsomes were resuspended in cold PBS containing the indicated concentration of cytosol and 10 mM EDTA, divided into aliquots (corresponding to a 10-ml translation reaction) and chased at 37 °C. After the chase the microsomes were re-isolated by centrifugation. The proteins in the pellet and the supernatant were separated on a 10% NuPAGE/MES gel system (Invitrogen), blotted onto nitrocellulose membrane, exposed on a Cyclone Storage Phosphor System (Packard, Meriden, CT, USA) and quantified using the OptiQuant 3.0 software (Packard). The ATP-regenerating system contained 2 mM MgCl2, 1 mM ATP, 10 mM creatine phosphate, 200 mg/ml creatine kinase, the ATP-depleting system 2 mM MgCl2, 5 mM glucose, 60 U/ml hexokinase. In these experiments EDTA was omitted. The blocking of the Sec61p complex was performed as described (13), the only modifications being that RNCs isolated from a 40-ml translation reaction were sufficient to saturate the brain microsomes, and that the binding of the RNCs was performed on ice. Traffic 2004; 5: 89–101 Cytosolic Degradation of Ab42 Biotinylation of microsomes Ab42-loaded microsomes were incubated for 30 min at room temperature in 1 mM puromycin, 500 mM KCl, 50 mM Hepes-KOH, pH 7.5. The re-isolated microsomes were incubated for 30 min at 37 °C in PBS containing 5 mM sulfo-NHS-biotin and, where indicated, 13 mM n-octylglucoside. After quenching the unreacted sulfo-NHS-biotin with 50 mM Tris-Cl, pH 8.0 and washing in TBS, the microsomes were lysed in 50 mM Tris-Cl, pH 8.0, 300 mM NaCl, 0.5% NP-40. The biotinylated proteins were captured using streptavidin mMACS paramagnetic beads (Miltenyi Biotech, Bergisch Gladbach, Germany). BiP and Grp94 were visualized in the eluate from the beads and in the total lysate, using anti-BiP antiserum (SantaCruz Biotechnology, Santa Cruz, CA, USA) or anti-Grp94 monoclonal antibody (StressGen, Canada). Protease protection analysis Ab42-loaded microsomes were resuspended in PBS containing 200 mg/ml trypsin and, where indicated, 1% Triton X-100. After incubation for 30 min on ice, the products were analyzed as above. To test whether the exported Ab42 was contained in vesicles, a 45-min export reaction was separated into the microsomal pellet and the supernatant. Both fractions were treated with trypsin or trypsin and detergent as above. Preparation of cytosol The gray matter of porcine brain cortices was homogenized using a Potter-Elvehjem homogenizer in 4 vol. (v/w) homogenization buffer (20 mM Tris-Cl, pH 7.4, 250 mM sucrose, 25 mM KCl, 5 mM MgCl2). The homogenate was centrifuged successively at 10 000 g for 20 min and at 100 000 g for 2 h. The supernatant was precipitated with ammonium sulfate (90% saturation). The resulting pellet was solubilized in homogenization buffer, followed by chromatography on a desalting column equilibrated in homogenization buffer. To remove proteasomes, brain cytosol was centrifuged at 500 000 g for 1 h and the supernatant was used for further experiments. Proteasomes were immunodetected using an antibody against the a7 subunit (Affiniti Research Products, Mamhead, UK). Degradation of Ab 42 For high-level expression, HeLa cells in 6-well plates were transfected with 1 mg pCIsAb42 and 60 pmol siRNA or single-stranded sense RNA, respectively, using 5.5 ml Metafectene (Biontex Laboratories, München, Germany). For low-level expression, pCIsAb42 was substituted by 1 mg pTKsAb42 and 1 mg pSwitch coding for the GeneSwitch protein. Expression of the GeneSwitch protein which enhances the transcription from pTKsAb42 was induced with 10 nM mifepristone (Invitrogen) for 1 day. Traffic 2004; 5: 89–101 Two days after transfection, the cells were labeled for 15 min with 200 mCi/ml [35S]-methionine. After the chase in medium containing unlabeled methionine, the cells were lysed in RIPA buffer (25 mM Tris-Cl, pH 8.0, 125 mM NaCl, 0.5% sodium deoxycholate, 0.1% SDS, 1% NP-40, 20 mM EDTA) with protease inhibitors (10 mM E-64, 1 mM pepstatin A, 2 mM PMSF). The lysate was separated by centrifugation (18 000 g, 15 min) into a RIPA-soluble and a RIPA-insoluble fraction. Ab42 was precipitated from the soluble fraction with antibody 6E10. To detect Ab in the RIPAinsoluble fraction, the pellet was resuspended in 2 sample buffer containing 5% SDS, sonicated with a Branson Sonifier B-12 (Branson, Danbury, CT, USA) for 5 s at position 2 and boiled for 5 min. An amount normalized to the soluble fraction was directly loaded on the gel. Formic acid extraction was performed as described (32). Statistical significance was determined by the One-way ANOVA Bonferroni test using the software Origin 7.0 (OriginLab Corporation, Northampton, MA, USA). Acknowledgments The authors wish to thank Dr E.H. Koo, University of California, San Diego, for the APP-cDNA, K. Bois and C. Mirschkorsch for technical assistance, Dr A. Winkeler for help with the low-level expression of Ab, and Dr J. Höhfeld and Dr A. Haas for valuable discussions. This work was supported by the Deutsche Forschungsgemeinschaft, Priority Programme ‘Cellular Mechanisms of Alzheimer’s Disease’, and the Bonner Forum Biomedizin. References 1. Tsai B, Ye Y, Rapoport TA. Retro-translocation of proteins from the endoplasmic reticulum into the cytosol. Nat Rev Mol Cell Biol 2002;3:246–255. 2. Kostova Z, Wolf DH. For whom the bell tolls: protein quality control of the endoplasmic reticulum and the ubiquitin-proteasome connection. EMBO J 2003;22:2309–2317. 3. McCracken AA, Brodsky JL. Evolving questions and paradigm shifts in endoplasmic-reticulum-associated degradation (ERAD). Bioessays 2003;25:868–877. 4. Biederer T, Volkwein C, Sommer T. Role of Cue1p in ubiquitination and degradation at the ER surface. Science 1997;278:1806–1809. 5. Bordallo J, Plemper RK, Finger A, Wolf DH. Der3p/Hrd1p is required for endoplasmic reticulum-associated degradation of misfolded lumenal and integral membrane proteins. Mol Biol Cell 1998;9:209–222. 6. Mayer TU, Braun T, Jentsch S. Role of the proteasome in membrane extraction of a short-lived ER-transmembrane protein. EMBO J 1998;17:3251–3257. 7. Schubert U, Anton LC, Bacik I, Cox JH, Bour S, Bennink JR, Orlowski M, Strebel K, Yewdell JW. CD4 glycoprotein degradation induced by human immunodeficiency virus type 1 Vpu protein requires the function of proteasomes and the ubiquitin-conjugating pathway. J Virol 1998;72:2280–2288. 8. Chillaron J, Haas IG. Dissociation from BiP and retrotranslocation of unassembled immunoglobulin light chains are tightly coupled to proteasome activity. Mol Biol Cell 2000;11:217–226. 9. Wiertz EJ, Jones TR, Sun L, Bogyo M, Geuze HJ, Ploegh HL. The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 1996;84:769–779. 99 Schmitz et al. 10. Johnston JA, Ward CL, Kopito RR. Aggresomes: a cellular response to misfolded proteins. J Cell Biol 1998;143:1883–1898. 11. Simpson JC, Roberts LM, Romisch K, Davey J, Wolf DH, Lord JM. Ricin A chain utilises the endoplasmic reticulum-associated protein degradation pathway to enter the cytosol of yeast. FEBS Lett 1999;459:80–84. 12. Wesche J, Rapak A, Olsnes S. Dependence of ricin toxicity on translocation of the toxin A-chain from the endoplasmic reticulum to the cytosol. J Biol Chem 1999;274:34443–34449. 13. Schmitz A, Herrgen H, Winkeler A, Herzog V. Cholera toxin is exported from microsomes by the Sec61p complex. J Cell Biol 2000;148: 1203–1212. 14. Wiertz EJ, Tortorella D, Bogyo MYuJ, Mothes W, Jones TR, Rapoport TA, Ploegh HL. Sec61-mediated transfer of a membrane protein from the endoplasmic reticulum to the proteasome for destruction. Nature 1996;384:432–438. 15. Pilon M, Schekman R, Romisch K. Sec61p mediates export of a misfolded secretory protein from the endoplasmic reticulum to the cytosol for degradation. EMBO J 1997;16:4540–4548. 16. Plemper RK, Bohmler S, Bordallo J, Sommer T, Wolf DH. Mutant analysis links the translocon and BiP to retrograde protein transport for ER degradation. Nature 1997;388:891–895. 17. Zhou M, Schekman R. The engagement of Sec61p in the ER dislocation process. Mol Cell 1999;4:925–934. 18. Wilkinson BM, Tyson JR, Reid PJ, Stirling CJ. Distinct domains within yeast Sec61p involved in post-translational translocation and protein dislocation. J Biol Chem 2000;275:521–529. 19. Selkoe DJ. Alzheimer’s disease: genes, proteins and therapy. Physiol Rev 2001;81:741–766. 20. Cook DG, Forman MS, Sung JC, Leight S, Kolson DL, Iwatsubo T, Lee VM, Doms RW. Alzheimer’s A beta (1–42) is generated in the endoplasmic reticulum/intermediate compartment of NT2N cells. Nat Med 1997;3:1021–1023. 21. Hartmann T, Bieger SC, Bruhl B, Tienari PJ, Ida N, Allsop D, Roberts GW, Masters CL, Dotti CG, Unsicker K, Beyreuther K. Distinct sites of intracellular production for Alzheimer’s disease A beta40/42 amyloid peptides. Nat Med 1997;3:1016–1020. 22. Wild-Bode C, Yamazaki T, Capell A, Leimer U, Steiner H, Ihara Y, Haass C. Intracellular generation and accumulation of amyloid beta-peptide terminating at amino acid 42. J Biol Chem 1997;272:16085–16088. 23. Greenfield JP, Tsai J, Gouras GK, Hai B, Thinakaran G, Checler F, Sisodia SS, Greengard P, Xu H. Endoplasmic reticulum and transGolgi network generate distinct populations of Alzheimer beta-amyloid peptides. Proc Natl Acad Sci USA 1999;96:742–747. 24. Wilson CA, Doms RW, Zheng H, Lee VM. Presenilins are not required for A beta 42 production in the early secretory pathway. Nat Neurosci 2002;5:849–855. 25. Bückig A, Tikkanen R, Herzog V, Schmitz A. Cytosolic and nuclear aggregation of the amyloid beta-peptide following its expression in the endoplasmic reticulum. Histochem Cell Biol 2002;118:353–360. 26. Chesneau V, Vekrellis K, Rosner MR, Selkoe DJ. Purified recombinant insulin-degrading enzyme degrades amyloid beta-protein but does not promote its oligomerization. Biochem J 2000;351:509–516. 27. Vekrellis KYeZ, Qiu WQ, Walsh D, Hartley D, Chesneau V, Rosner MR, Selkoe DJ. Neurons regulate extracellular levels of amyloid betaprotein via proteolysis by insulin-degrading enzyme. J Neurosci 2000;20:1657–1665. 28. Sudoh S, Frosch MP, Wolf BA. Differential effects of proteases involved in intracellular degradation of amyloid beta-protein between detergent-soluble and -insoluble pools in CHO-695 cells. Biochemistry 2002;41:1091–1099. 29. Farris W, Mansourian S, Chang Y, Lindsley L, Eckman EA, Frosch MP, Eckman CB, Tanzi RE, Selkoe DJ, Guenette S. Insulin-degrading enzyme regulates the levels of insulin, amyloid beta-protein, and the 100 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. beta-amyloid precursor protein intracellular domain in vivo. Proc Natl Acad Sci USA 2003;100:4162–4167. Kurochkin IV, Goto S. Alzheimer’s beta-amyloid peptide specifically interacts with and is degraded by insulin degrading enzyme. FEBS Lett 1994;345:33–37. Duckworth WC, Bennett RG, Hamel FG. Insulin degradation. progress and potential. Endocr Rev 1998;19:608–624. Skovronsky DM, Doms RW, Lee VM. Detection of a novel intraneuronal pool of insoluble amyloid beta protein that accumulates with time in culture. J Cell Biol 1998;141:1031–1039. Skovronsky DM, Pijak DS, Doms RW, Lee VM. A distinct ER/IC gamma-secretase competes with the proteasome for cleavage of APP. Biochemistry 2000;39:810–817. Rodighiero C, Tsai B, Rapoport TA, Lencer WI. Role of ubiquitination in retro-translocation of cholera toxin and escape of cytosolic degradation. EMBO Rep 2002;21:21. Werner ED, Brodsky JL, McCracken AA. Proteasome-dependent endoplasmic reticulum-associated protein degradation: an unconventional route to a familiar fate. Proc Natl Acad Sci USA 1996;93: 13797–13801. Meyer HH, Wang Y, Warren G. Direct binding of ubiquitin conjugates by the mammalian p97 adaptor complexes, p47 and Ufd1-Npl4. EMBO J 2002;21:5645–5652. Bays NW, Wilhovsky SK, Goradia A, Hodgkiss-Harlow K, Hampton RY. HRD4/NPL4 is required for the proteasomal processing of ubiquitinated ER proteins. Mol Biol Cell 2001;12:4114–4128. Ye Y, Meyer HH, Rapoport TA. The AAA ATPase Cdc48/p97 and its partners transport proteins from the ER into the cytosol. Nature 2001;414:652–656. Braun S, Matuschewski K, Rape M, Thoms S, Jentsch S. Role of the ubiquitin-selective CDC48 (UFD1/NPL4) chaperone (segregase) in ERAD of OLE1 and other substrates. EMBO J 2002;21:615–621. Jarosch E, Taxis C, Volkwein C, Bordallo J, Finley D, Wolf DH, Sommer T. Protein dislocation from the ER requires polyubiquitination and the AAA-ATPase Cdc48. Nat Cell Biol 2002;28:28. Rabinovich E, Kerem A, Frohlich KU, Diamant N, Bar-Nun S. AAA-ATPase p97/Cdc48p, a cytosolic chaperone required for endoplasmic reticulumassociated protein degradation. Mol Cell Biol 2002;22:626–634. Ye Y, Meyer HH, Rapoport TA. Function of the p97-Ufd1-Npl4 complex in retrotranslocation from the ER to the cytosol: dual recognition of nonubiquitinated polypeptide segments and polyubiquitin chains. J Cell Biol 2003;162:71–84. Rosenberg-Hasson Y, Bercovich Z, Ciechanover A, Kahana C. Degradation of ornithine decarboxylase in mammalian cells is ATP dependent but ubiquitin independent. Eur J Biochem 1989;185:469–474. McCracken AA, Brodsky JL. Assembly of ER-associated protein degradation in vitro: dependence on cytosol, calnexin, and ATP. J Cell Biol 1996;132:291–298. Shamu CE, Story CM, Rapoport TA, Ploegh HL. The pathway of US11dependent degradation of MHC class I heavy chains involves a ubiquitinconjugated intermediate. J Cell Biol 1999;147:45–58. Tirosh B, Furman MH, Tortorella D, Ploegh HL. Protein unfolding is not a prerequisite for endoplasmic reticulum-to-cytosol dislocation. J Biol Chem 2003;278:6664–6672. Romisch K, Schekman R. Distinct processes mediate glycoprotein and glycopeptide export from the endoplasmic reticulum in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1992;89:7227–7231. Gillece P, Pilon M, Romisch K. The protein translocation channel mediates glycopeptide export across the endoplasmic reticulum membrane. Proc Natl Acad Sci USA 2000;97:4609–4614. Grant SM, Ducatenzeiler A, Szyf M, Cuello AC. Abeta immunoreactive material is present in several intracellular compartments in transfected, neuronally differentiated, P19 cells expressing the human amyloid beta-protein precursor. J Alzheimer Dis 2000;2:207–222. Traffic 2004; 5: 89–101 Cytosolic Degradation of Ab42 50. LaFerla FM, Tinkle BT, Bieberich CJ, Haudenschild CC, Jay G. The Alzheimer’s A beta peptide induces neurodegeneration and apoptotic cell death in transgenic mice. Nat Genet 1995;9:21–30. 51. Yan SD, Fu J, Soto C, Chen X, Zhu H, Al-Mohanna F, Collison K, Zhu A, Stern E, Saido T, Tohyama M, Ogawa S, Roher A, Stern D. An intracellular protein that binds amyloid-beta peptide and mediates neurotoxicity in Alzheimer’s disease. Nature 1997;389:689–695. 52. Zhang Y, McLaughlin R, Goodyer C, LeBlanc A. Selective cytotoxicity of intracellular amyloid beta peptide1-42 through p53 and Bax in cultured primary human neurons. J Cell Biol 2002;156:519–529. 53. Keller JN, Hanni KB, Markesbery WR. Impaired proteasome function in Alzheimer’s disease. J Neurochem 2000;75:436–439. 54. Perez A, Morelli L, Cresto JC, Castano EM. Degradation of soluble amyloid beta-peptides 1–40, 1–42. and the Dutch variant 1–40Q by Traffic 2004; 5: 89–101 55. 56. 57. 58. insulin degrading enzyme from Alzheimer disease and control brains. Neurochem Res 2000;25:247–255. Bertram L, Blacker D, Mullin K, Keeney D, Jones J, Basu S, Yhu S, McInnis MG, Go RC, Vekrellis K, Selkoe DJ, Saunders AJ, Tanzi RE. Evidence for genetic linkage of Alzheimer’s disease to chromosome 10q. Science 2000;290:2302–2303. Prince JA, Feuk L, Gu HF, Johansson B, Gatz M, Blennow K, Brookes AJ. Genetic variation in a haplotype block spanning IDE influences Alzheimer disease. Hum Mutat 2003;22:363–371. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, Tuschl T. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 2001;411:494–498. Walter P, Blobel G. Preparation of microsomal membranes for cotranslational protein translocation. Meth Enzymol 1983;96:84–93. 101