The C-terminal Globular Domain of the Prion Protein Is Necessary and Sufficient for Import into the Endoplasmic Reticulum*

The mammalian prion protein (PrP) is composed of an unstructured flexible N-terminal region and a C-terminal globular domain. We examined the import of PrP into the endoplasmic reticulum (ER) of neuronal cells and show that information present in the C-terminal globular domain is required for ER import of the N terminus. N-terminal fragments of PrP, devoid of structural domains located in the C terminus, remained in the cytosol with an uncleaved signal peptide and were rapidly degraded by the proteasome. Conversely, the separate C-terminal domain of PrP, comprising the highly ordered helix 2-loop-helix 3 motif, was entirely imported into the ER. As a consequence, two PrP mutants linked to inherited prion disease in humans, PrP-W145Stop and PrP-Q160Stop, were partially retained in the cytosol. The cytosolic fraction was characterized by an uncleaved N-terminal signal peptide and was degraded by the proteasome. Our study identified a new regulatory element in the C-terminal globular domain of PrP necessary and sufficient to promote import of PrP into the ER.

The mammalian prion protein (PrP) is composed of an unstructured flexible N-terminal region and a C-terminal globular domain. We examined the import of PrP into the endoplasmic reticulum (ER) of neuronal cells and show that information present in the C-terminal globular domain is required for ER import of the N terminus. N-terminal fragments of PrP, devoid of structural domains located in the C terminus, remained in the cytosol with an uncleaved signal peptide and were rapidly degraded by the proteasome. Conversely, the separate C-terminal domain of PrP, comprising the highly ordered helix 2-loop-helix 3 motif, was entirely imported into the ER. As a consequence, two PrP mutants linked to inherited prion disease in humans, PrP-W145Stop and PrP-Q160Stop, were partially retained in the cytosol. The cytosolic fraction was characterized by an uncleaved N-terminal signal peptide and was degraded by the proteasome. Our study identified a new regulatory element in the C-terminal globular domain of PrP necessary and sufficient to promote import of PrP into the ER.
The mammalian prion protein (PrP C ) 1 is a glycosylphosphatidylinositol (GPI)-anchored protein, mainly present at the plasma membrane of neuronal and lymphatic cells. Structural studies with recombinantly expressed PrP revealed a large flexibly disordered N-terminal region and a structured C-terminal domain. This autonomously folding domain contains three ␣-helical regions and a short, two-stranded ␤-sheet (1-3). Biogenesis of PrP C is characterized by a series of post-translational modifications. Two signal sequences, located at the Nand the C terminus, respectively, direct the import of PrP into the endoplasmic reticulum (ER) and mediate the attachment of a GPI anchor at Ser 230 (4). In addition, the C-terminal domain of PrP C is modified with two N-linked complex carbohydrate moieties (Asn 180 and Asn 196 ) (5-8) and a disulfide bond between Cys 178 and Cys 213 . Folding of PrP C and the degree of post-translational modifications are intimately connected. PrP⌬GPI, a mutant devoid of the C-terminal signal sequence, not only lacks the membrane anchor but in addition remains predominantly unglycosylated (9 -12). Similarly, modifying folding of PrP C by destabilizing or deleting helix 1 interferes with the GPI anchor attachment and the conversion of core glycans into complex structures (9). In the same study we found that several PrP mutants adopted a misfolded conformation in the ER, yet were transported through the secretory pathway and secreted. Of note, the N-terminal domain is dispensable for the maturation of PrP C ; PrP⌬N is complex glycosylated and localized at the plasma membrane via the GPI anchor.
Prion diseases in humans and animals are characterized by the appearance of abnormal isoforms of PrP C . A diagnostic marker is the accumulation of PrP Sc , a detergent-insoluble and partially proteinase K-resistant PrP isoform, which is the major component of infectious prions (13)(14)(15)(16). Two other PrP species have been described to induce neurodegeneration in transgenic mice in the absence of PrP Sc : an unusual transmembrane topology designated Ctm PrP as well as PrP aggregates in the cytosol (17,18). Increased synthesis of Ctm PrP is also linked to an inherited prion disease in humans (19). A possible pathogenic role of misfolded PrP conformers at or in the ER membrane was emphasized by a recent study using a yeast model. We could show that PrP misfolding occurred during post-translational targeting to the ER membrane. As a consequence, PrP bearing an uncleaved N-terminal signal peptide associated with ER membranes and interfered with cell growth (20).
Targeting of secretory proteins to the ER is mediated by a N-terminal signal peptide. In yeast, targeting may occur co-or post-translationally, dependent on the hydrophobicity of the signal peptide (21) (reviewed in [22][23][24][25][26]. Co-translational import, the main pathway in mammalian cells, involves binding of the signal recognition particle (SRP) to the emerging Nterminal signal peptide, pausing of elongation and targeting of the ribosome-polypeptide complex to the Sec61 translocon. In addition to its targeting role the signal sequence can modulate the translocation process and influence the topology of the polypeptide chain (27,28).
PrP C has the unusual potential to adopt three distinct topological forms during import into the ER. The GPI-anchored version Sec PrP, which is fully translocated, predominates under physiological conditions. In addition to the pathogenic Ctm PrP (C terminus facing the ER lumen), a second transmembrane topology, designated Ntm PrP (N terminus facing the ER lumen), can be formed. The relative ratio between these three topological isoforms is under the control of the N-terminal signal peptide and an internal putative transmembrane (TM) domain (19,28). Interestingly, increased synthesis of Ctm PrP has also been described in scrapie-infected mice (17).
In this study we have analyzed the ER import of PrP in vivo and identified a new regulatory element within the C-terminal domain of PrP. Complete or partial deletions of this domain abolished or reduced import of PrP into the ER and gave rise to a N-terminally unprocessed PrP species, which was degraded by the proteasome. As a consequence, ER import of two pathogenic PrP mutants linked to inherited prion diseases in humans, W145Stop and Q160Stop, was compromised; only a fraction was fully translocated into the ER, the other fraction containing an uncleaved signal peptide was present in the cytosol. For the generation of A115 containing the growth hormone signal sequence, the signal sequence of PrP (aa 2-22) was replaced by the rat growth hormone signal sequence (MAADSQTPWLLTFSLLCLLW-PQEAGA). N115-GFP was constructed by fusing the coding sequence of green fluorescent protein (GFP) to the C terminus of A115Stop. All amino acid numbers refer to the mouse PrP sequence (GenBank TM accession number M18070).
Cell Culture and Transfections-N2a cells were cultivated as described earlier (32). Cells were transfected by a liposome-mediated method using LipofectAMINE Plus reagent according to the manufac-turerЈs instructions (Invitrogen).
Detergent Solubility Assay and Secretion Analysis-As previously described (9), N2a cells were washed twice with cold phosphate-buffered saline (PBS), scraped off the plate, pelleted by centrifugation, and lysed in cold buffer A (0.5% Triton X-100 and 0.5% sodium deoxycholate in phosphate-buffered saline). The lysate was centrifuged at 15,000 ϫ g for 20 min at 4°C; after boiling in Laemmli sample buffer supernatants and pellets were analyzed by immunoblotting. To examine the secretion of PrP into the cell culture supernatant, cells were cultivated in cell culture medium without supplements for 3 h at 37°C. The medium was collected, and PrP was precipitated with trichloroacetic acid and then analyzed by Western blotting.
Western Blot Analysis-Following SDS-PAGE, proteins were transferred onto a nitrocellulose membrane and analyzed as previously described (33).
Metabolic Labeling and Immunoprecipitation and Proteasomal Inhibitor Treatment-For metabolic labeling cells were starved for 30 min in methionine-free modified Eagle's medium (MEM, Invitrogen) and subsequently labeled for 30 min with 0.3 Ci of Promix L-[ 35 S]cell labeling mix (Amersham Biosciences; Ͼ 37 TBq/mmol) in methioninefree MEM. Immunoprecipitation was carried out as previously described (32). When indicated, the proteasomal inhibitor MG 132 (Calbiochem) was present during the starving and the labeling periods (50 M, dissolved in Me 2 SO). For the chase, the labeling medium was removed, the cells were washed twice and then incubated in complete medium for the time indicated. To detect secreted proteins, the labeling medium was collected and further analyzed by immunoprecipitation. For endoH digestion immunoprecipitation products were adjusted to 0.5% SDS, boiled for 10 min, and digested with endoglycosidase H (New England Biolabs) for 1 h at 37°C as specified by the manufacturer. Immunoprecipitation products were analyzed by SDS-PAGE. Gels were impregnated with Amplify (PerkinElmer Life Sciences) and exposed to film. For quantification, signals were detected by autoradiography and quantified using a phosphorimager (FLA-2000, Fuji; software AIDA, Raytest). The amount of each protein in the presence of MG 132 was set as 100%. The amount of protein in the absence of MG132 was calculated as a percentage of the initial value and subtracted from 100% to give the value for proteasomal degradation. For the ER constructs the two forms (signal sequence cleaved and uncleaved) were analyzed separately. All quantifications were based on at least three independent experiments.
In Vitro Translation and Proteinase K Digestion-The in vitro translation was carried out by using the TNT T7 Quick-coupled Transcrip-tion/Translation System (Promega) by following the manufacturer's instructions. When indicated, Canine Pancreatic Microsomal Membranes (Promega) were added. After translation for 70 min at 30°C the samples were split; one half was left untreated and the other half was incubated with 0.5 mg/ml proteinase K (Roche Applied Science) for 1 h on ice. To stop the reaction, 2 mmol of phenylmethylsulfonyl fluoride (Serva) was added. After immunoprecipitation of PrP using the monoclonal antibody 3F4, the samples were separated by SDS-PAGE and analyzed as described in the previous section.
Membrane Fractionation and Proteinase K Digestion of Cell Lysates-The cellular localization of the PrP mutants was analyzed by Renografin density gradient centrifugation (20,34,35). Cells were washed twice with cold phosphate-buffered saline, scraped off the plate, and pelleted by centrifugation. The cells were resuspended in TE and lysed by three cycles of freezing and thawing. A postnuclear supernatant (cell lysate) was prepared by centrifugation at 500 ϫ g for 1 min at 4°C. The cell lysate was mixed with 76% Renografin (0.5 ml each) and successively overlaid with 1 ml of 34, 30, 26, and 22% Renografin, respectively. Gradients were then centrifuged at 150,000 ϫ g for 20 h at 4°C. 13 fractions were removed from the top of the gradient. Membrane fractions were diluted 10-fold with TE, pelleted at 100,000 ϫ g for 1 h at 4°C, and resuspended in Laemmli sample buffer. For the proteolysis experiment, cells were lysed as described above, mixed with 76% Renografin (final concentration 22%) and incubated with proteinase K (Roche Applied Science) (1:250) for 1 h on ice. The reaction was terminated by the addition of phenylmethylsulfonyl fluoride (Serva) and boiling in Laemmli sample buffer. Residual PrP was detected by Western blotting.

C-terminal Truncations
Interfere with N-terminal Processing of PrP-In a previous study we analyzed the biogenesis of several PrP mutants with deletions in the N-terminal domain including the putative transmembrane (TM) domain and helix 1. It turned out that all these mutants were efficiently imported into the ER, even though maturation of PrP C was impaired by deletion or destabilization of helix 1 (9). To analyze the impact of the highly ordered helix 2-loop-helix 3 motif on the biogenesis of PrP C we constructed several successively larger C-terminal deletion mutants. Two versions were generated for each construct, one version with the authentic Nterminal signal sequence to direct PrP into the ER (ER form) and another version for cytosolic expression of PrP by deleting the ER signal sequence (cyto form) (Fig. 1A). For the inclusion of the two pathogenic human PrP mutants W145Stop and Q160Stop in our analysis, we cloned the mouse equivalents W144Stop and Q159Stop. As a control, we decided to use S230Stop, a mutant lacking the C-terminal GPI signal sequence (previously designated PrP⌬GPI) instead of full-length PrP for the following reasons. All of our C-terminal truncation mutants are devoid of the two consensus sites for N-linked glycosylation as well as the GPI anchor attachment signal. Similarly, S230Stop lacks the GPI anchor and is mainly unglycosylated, however, it is efficiently imported into the ER and secreted (9 -11). Therefore, by using S230Stop as a control instead of complex glycosylated and GPI-anchored wild-type (wt) PrP we can exclude possible effects of the post-translational modifications.
For the initial biochemical analysis these PrP mutants were transiently expressed in N2a cells and screened by Western blotting using the monoclonal anti-PrP antibody 3F4. Of note, 3F4 does not recognize endogenous mouse PrP C (30). Corroborating earlier results (9), endoH and PNGaseF digestion revealed that S230Stop was mainly unglycosylated (data not shown); however, the N-terminal signal sequence of S230Stop was cleaved, as indicated by the size of the ER form in comparison to that of the respective cytosolic form on SDS-PAGE (Fig. 1B, S230Stop). For both Q159Stop and W144Stop we observed two distinct bands of the ER forms, with the faster migrating species showing the same relative size as the cytosolic form (Fig. 1B, Q159Stop, W144Stop). These findings indi-cate that only a fraction of Q159Stop and W144Stop was N-terminally processed. The shorter PrP mutants A115Stop and P101Stop were not detectable by Western blotting. The analysis of the cell culture supernatant of PrP-expressing cells revealed that the ER forms of S230Stop, Q159Stop, and W144Stop were secreted (Fig. 1B).
PrP Species with an Uncleaved N-terminal Signal Peptide Are Rapidly Degraded by the Proteasome-For a more detailed analysis transfected N2a cells were metabolically labeled with [ 35 S]methionine for 30 min and then PrP was analyzed by immunoprecipitation. To address a possible role of proteasomal degradation, this analysis was also performed in the presence of the proteasomal inhibitor MG132 (36). Several findings emerged from this approach. Firstly, it became evident that the ER form and the cyto form of A115Stop were synthesized, however, they were rapidly degraded by the proteasome (Fig. 2,  A, B, C, A115Stop). Secondly, the ER form of A115Stop appeared larger by SDS-PAGE compared with the cyto form, indicating that the N-terminal ER targeting signal peptide of A115Stop has not been processed. Thirdly, for both Q159Stop and W144Stop the proteasomal inhibitor specifically increased the relative amount of the N-terminally unprocessed ER form (Fig. 2, A, B, Q159Stop, and W144Stop). Stabilization by MG132 was also observed for all the cyto forms of the PrP mutants. It appeared that the size of the C-terminal truncations inversely correlates with the stability of cytosolic PrP; S230Stop was the most stable and A115Stop the least stable mutant (Fig. 2C).
In conjunction with the Western blot analysis these data indicate that an intact C-terminal globular domain is necessary and sufficient to allow efficient ER import of PrP to occur. In contrast to N-terminally processed PrP, PrP species with an uncleaved signal peptide are subjected to proteasomal degradation, similarly to PrP expressed in the cytosol.
N-terminally Unprocessed PrP Is Not Imported into the ER-Our experiments revealed that the A115Stop mutant as well as a fraction of the Q159Stop and the W144Stop mutant contained an uncleaved ER signal peptide, raising the possibility that these PrP mutants were not imported into the ER.
To address this question in more detail, we first performed in vitro import experiments in rabbit reticulocyte lysate supplemented with dog pancreas microsomes. In the presence of microsomes the ER forms of S230Stop, Q159Stop, and W144Stop were N-terminally processed, partially glycosylated (only S230Stop) and protected against proteolytic digestion, revealing import into the microsomes (Fig. 3A). In contrast, no import was seen for the ER form of the A115Stop mutant. Moreover, the N-terminally unprocessed forms of Q159Stop and W144Stop To quantify the degradation of PrP by the proteasome the relative amount of PrP present in MG132-treated cells was set as 100%. For the ER constructs the two species (signal sequence cleaved and uncleaved) were analyzed separately. The quantification of the species indicated by white and black arrowheads in panel A is representated by white (signal sequence cleaved) and black (signal sequence uncleaved) bars in panel B. Degradation was calculated from three independent experiments. Please note that the entire fraction of S230Stop is N-terminally processed, while A115Stop is only present with an uncleaved signal peptide.
were not protected against proteolytic digestion, i.e. were not present in the lumen of the microsomes.
Next we performed a Renografin density gradient centrifugation with cell extracts prepared from transiently transfected N2a cells in order to analyze the cellular localization of different PrP mutants. As expected, the secretory form of S230Stop colocalized with the chaperone Grp94, a marker for the ER fractions, while the cytosolic form remained in the bottom fractions (Fig. 3B). The secretory form of W144Stop was present in the ER fraction as well, however, only the N-terminally processed fraction. Unprocessed W144Stop remained in the bottom fractions, colocalizing with the cytosolically expressed W144Stop. Due to its instability, A115Stop could not be analyzed by this method.
The C-terminal Globular Domain Is Sufficient for Import into ER-Based on these experiments two different mecha- FIG. 3. N-terminally unprocessed PrP is not imported into the ER. A, the indicated constructs (ER form) were translated in vitro using the TNT Quick System. Microsomal membranes (ϩ microsomes) were included to analyze cotranslational import. Imported proteins were identified by a protease protection assay; the samples were incubated in the presence of 0.1 mg/ml proteinase K for 1 h on ice (PK ϩ). As a control, proteins were translated in the absence of microsomal membranes (Ϫ microsomes). Radioactively labeled proteins were separated on SDS-PAGE. CHO, glycosylated form of S230Stop and ⌬27-156/S230Stop. B, N2 cells were transiently transfected with the PrP constructs indicated on the left. Subcellular localization of the proteins was analyzed by a Renografin density gradient centrifugation of cell extracts. Fractions were removed from the top of the gradient and analyzed by Western blotting using the antibodies indicated on the right side of the figure. C, protease protection assay of membrane vesicles prepared from transfected N2a cells. Cell lysates were mixed with 76% Renographin (final concentration 22%) and incubated with proteinase K (1:250) for 1 h on ice. Residual PrP was detected by Western blotting. nisms could explain the defective import of PrP mutants with C-terminal truncations. Either the decreasing length of the polypeptide chain impairs the ER import, or information present in the C-terminal globular domain is required for efficient import.
To address a possible role of the polypeptide chain length, we generated two N-terminal deletion mutants, ⌬N/S230Stop, which lacks the unstructured N-terminal domain, and ⌬27-156/S230Stop with a larger N-terminal deletion including the putative TM domain and helix 1 (Fig. 4A). Transiently transfected N2a cells were metabolically labeled and PrP was analyzed by immunoprecipitation. ⌬N/S230Stop was efficiently transported in the ER and secreted, similarly to S230Stop (Fig.  4B). The mainly unglycosylated phenotype of ⌬N/S230Stop adds to the notion that helix 1 interferes with the core glycosylation of PrP devoid of a membrane anchor (9). ⌬27-156/ S230Stop was glycosylated (Fig. 4C), indicating efficient ER import, and was not subjected to proteasomal degradation (data not shown). We have previously shown that deletion or destabilization of helix 1 restores core glycosylation of S230Stop (9), explaining why ⌬27-156/S230Stop was glycosylated.
We next analyzed if a heterologous C terminus might promote ER import of the N terminus of PrP. We therefore fused the coding sequence of green fluorescent protein (GFP) to the C terminus of A115Stop to generate 115N-GFP (Fig. 5A). The Western blot analysis of transiently transfected N2a cells indicated that the heterologous C terminus was effective in mediating translocation and stabilization of the N terminus of PrP (Fig. 5B).
To exclude the possibility that the N-terminal signal sequence of PrP was responsible for the defective import of A115Stop, we replaced the PrP signal sequence by that of rat growth hormone (GH) (Fig. 5A), which has previously been shown to enhance ER import (37). As a control, S230Stop bearing the GH signal peptide instead of the PrP signal peptide was efficiently imported into the ER, similarly to S230Stop with the authentic signal peptide (Fig. 5C). However, the GH signal sequence was not able to promote the translocation of A115Stop; GH-SS-A115Stop contained an uncleaved signal peptide and was rapidly degraded by the proteasome as was A115Stop with the authentic signal sequence (Fig. 5C). DISCUSSION A crucial step in the biogenesis of PrP C is its import into the ER. Aberrant regulation at this step can lead to the formation of neurotoxic transmembrane topologies like Ctm PrP, or, in the case of unproductive import, to the generation of cytosolic PrP. Previous studies revealed that the N-terminal signal sequence and the putative TM domain contain information for the formation of the topological forms of PrP (19,28). We now show that a third control element, located in the distal C-terminal domain, determines the efficiency of ER import and thereby regulates the amount of cytosolic PrP.
Information in the C-terminal Domain of PrP Is Required for ER Import of the N Terminus-In this study we analyzed the biogenesis of N-terminal fragments of PrP and found that the C-terminal globular domain contains information necessary and sufficient for import into the ER. With successively larger FIG. 4. The C-terminal domain of PrP is suffcient for import into the ER. A, schematic representation of the PrP mutants analyzed. B and C, N2a cells were transiently transfected with the indicated PrP constructs. Cells were metabolically labeled with [ 35 S]methionine for 30 min and PrP was analyzed by immunoprecipitation using the monoclonal 3F4 antibody or the polyclonal anti-PrP antiserum A7 (⌬27-156/S230Stop). B, the unstructured N-terminal domain is dispensable for efficient import of PrP into the ER. After metabolic labeling the cells were incubated in fresh medium for the indicated time (chase). PrP present in the cell culture medium (M) or cell lysate (L) was immunoprecipitated and analyzed on SDS-PAGE. CHO indicates the core glycosylated fraction. C, after metabolic labeling and immunoprecipitation the samples were treated with endoglycosidase H (EndoH ϩ) for 1 h at 37°C and analyzed on SDS-PAGE.
C-terminal deletions the ER import of the remaining N-terminal domain became less efficient. S230Stop was completely imported; Q159Stop and W144Stop displayed an intermediate phenotype with a fraction of unprocessed polypeptide in the cytosol. Finally, ER import of A115Stop was completely abolished. In contrast to the N-terminal fragments, a separate C-terminal fragment of PrP (⌬27-156/S230Stop) was efficiently imported into the ER, implying that the length of the polypeptide chain is not limiting for efficient ER import.
It was previously shown that a portion of W144Stop contains an uncleaved signal peptide (38). Similar to our study, Zanusso et al. (38) described a rapid degradation of this mutant by the proteasome, but also secretion of W144Stop molecules with an uncleaved signal peptide. We did not observe considerable secretion of this mutant in our experimental approach. This discrepancy might be explained by the fact that Zanusso et al. used a different cell line as well as stable transfectants; moreover, significant secretion of W144Stop was only observed in the presence of proteasomal inhibitors. We now show that the Q159Stop mutant displays a similar phenotype as the W144Stop mutant and provide a mechanistic explanation for the impaired ER import. The effect of the C-terminal domain of PrP on ER import is clearly different from the previously identified role of the signal sequence and the internal putative TM domain. These two elements modulate the formation of the three topological isoforms of PrP, Sec PrP, Ctm PrP, and Ntm PrP (19,28). We have no evidence that deletions in the C-terminal domain change the topology of PrP. Instead, the C-terminal domain seems to ensure that PrP is effectively imported into the ER and does not remain in the cytosol. The C-terminal GPI anchor signal sequence does not significantly influence the translocation process, because all of our constructs were devoid of this element. In addition to the GPI anchor, the C-terminal domain contains two consensus sites for N-linked glycosylation. So far, there is no indication that glycosylation has an impact on ER import, rather events during import influence the degree of glycosylation. ⌬27-156/S230Stop was entirely glycosylated, while S230Stop and ⌬N/S230Stop, mutants with the same signal sequence and with glycan acceptor sites at the same distance from the C terminus, remained mainly unglycosylated. However, all mutants were efficiently imported into the ER, similarly to PrP mutants with a deleted or destabilized helix 1 (9).
What kind of information could be encoded in the C-terminal domain? When comparing the two PrP mutants A115Stop and ⌬27-156/S230Stop, a striking difference is the lack of struc-tural elements in the N-terminal fragment A115Stop, whereas the highly ordered helix 2-loop-helix 3 motif is present in the C-terminal fragment ⌬27-156/S230Stop. In this context it is important to note that folding of the C-terminal globular domain of PrP is one of the most rapid folding reactions measured to date (39). Based on the dimension of the ribosomal exit tunnel and the pore of the translocon (40 -43), it is conceivable that the emerging polypeptide chain adopts a secondary structure before it is fully translocated into the ER lumen. Indeed, the analysis of the chimeric fusion protein 115N-GFP indicated that a heterologous domain can stabilize the N terminus of PrP and supports its translocation into the ER. How folding of the nascent polypeptide chain could influence the efficiency of ER import remains to be determined. In this context it is important to remember that ER import of PrP seems to be more complex than that of other secretory proteins. In in vitro studies using reconstituted lipid vesicles, ER import of PrP required the translocon-associated protein (TRAP) complex (44) in addition to the essential translocon components Sec61p complex, the signal recognition particle receptor and the translocating chain-associated protein (TRAM) (45)(46)(47)(48).
At present we do not know at which stage ER import of the PrP mutants is defective. Do these mutants never engage in ER import, is the import aborted at an early step during translocation, or is there a retrograde transport from the ER? The in vitro import studies might argue against the last possibility, because PrP molecules with an uncleaved signal peptide were not protected against proteolytic digestion, i.e. were not entirely translocated into the microsomes.
FIG. 6. Determinants of PrP import in vivo. This scheme represents a summary of the PrP mutants analyzed in this study. Based on the efficiency of ER import three different classes of PrP mutants can be distinguished. The first group of PrP mutants comprises the highly stuctured helix 2-loop-helix 3 motif and was entirely imported into the ER. The second group, represented by the two pathogenic PrP mutants Q159Stop and W144Stop, is partially devoid of the C-terminal globular domain. These mutants are present as two distinct forms, one form contains the signal peptide, while the other is N-terminally processed. The third group is devoid of any structural elements of the C-terminal domain; PrP mutants that belong to this group are entirely present in the cytosol and contain an uncleaved signal peptide.
Proteasomal Degradation of PrP-Another focus of this study was the proteasomal degradation of PrP. We found that the cytosolic PrP mutants are degraded by the proteasome and that the stability of PrP correlates with the length of the C-terminal domain; cyto-S230Stop was the most stable and cyto-A115Stop the least stable mutant. As a result, the steady state levels of S230Stop were high enough to be detectable by Western blotting, while A115Stop could only be analyzed in the presence of MG132 after immunoprecipitation (Fig. 6).
When analyzing the PrP mutants targeted to the ER we observed that cleavage of the signal peptide correlates with efficient ER import; PrP polypeptides with a cleaved signal peptide were not significantly stabilized by the proteasomal inhibitor MG132. The analysis of the pathogenic mutants W144Stop and Q159Stop corroborated the notion that unproductive ER import and proteasomal degradation was linked to an uncleaved signal peptide and not to the differences in the primary sequence. Both mutants were present both with and without the signal peptide, however, mainly the polypeptides bearing an uncleaved signal peptide could be stabilized by MG132. In this context it is important to note that for these experiments cells were exposed to MG132 no longer than 60 min.
ER Import of PrP and Prion Diseases-Our study adds to the notion that some PrP mutations linked to inherited prion diseases in humans alter the ER import of PrP. While mutations in the putative TM domain, like A117V, increase the formation of the pathogenic transmembrane topology Ctm PrP, in W144Stop and Q159Stop information necessary for efficient ER import is obviously missing. As a consequence, expression of these mutants generates significant amounts of cytosolic PrP with an uncleaved signal peptide. Could the cytosolic localization of PrP contribute to the pathogenesis of prion diseases? Previous studies indicated that in the presence of proteasomal inhibitors even wild-type PrP is present in the cytosol (49 -51). Furthermore, cytosolic PrP has been shown to be toxic to cultured cells (18) and to induce neurodegeneration in transgenic animals (18). It has also been suggested that cytosolic PrP gains a toxic potential through transport into the nucleus (52,53). Our recent study in yeast might indicate an alternative pathway. We could show that cytosolic PrP with an uncleaved signal peptide decreased cell viability. Even though we did not observe formation of Ctm PrP, the cytotoxic effects were linked to a prolonged association of cytosolic PrP with ER membranes (20). Interestingly, the pathogenic Ctm PrP seems to contain the signal peptide as well (54). It will now be important to analyze if retrograde transport of Ctm PrP contributes to the generation of cytosolic PrP and to delineate the mechanism by which cytosolic PrP impairs cell viability.