|
|
|
|
|
|
|
|
|||||
| Home Help Feedback Subscriptions Archive Search Table of Contents | |||||
First published online 27 November 2007
doi: 10.1242/jcs.011247
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
1 Department of Biochemistry, George S. Wise Faculty of Life Sciences, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
2 Department of Pathology, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Author for correspondence (e-mail: shoshbn{at}tauex.tau.ac.il)
Accepted 3 October 2007
 |
Summary |
|---|
 Top Summary IntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
Key words: ERAD, Retention signals, Degrons, Proteasome
|
Introduction |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
; Ellgaard and Helenius, 2003; Kostova and Wolf, 2003; McCracken and Brodsky, 2003; Sitia and Braakman, 2003; Trombetta and Parodi, 2003; Sayeed and Ng, 2005; Bar-Nun, 2005). Retention of proteins destined for vesicular export is a prerequisite for ERAD; however, retention does not necessarily culminate in degradation. Therefore, it is of interest to determine whether cis-acting motifs that are recognized as `retention signals', preventing protein secretion, overlap with cis-acting motifs that act as `degrons', targeting proteins for degradation.
Degrons are defined as sequences or domains that are necessary and sufficient for directing otherwise stable proteins for degradation (Varshavsky, 1991). A degron that targets soluble lumenal proteins for ERAD might play a role in any step along this complex pathway, steps that are not necessarily coupled and schematically entail substrate selection, dislocation to the cytosol, ubiquitylation and degradation by the proteasome. Thus, a degron might constitute the recognition signal for the substrate, it might lead the dislocation process, it might recruit the ubiquitylation machinery and/or provide the platform and Lys residue(s) for ubiquitin conjugation and/or it might target the substrate to the proteasome.
The heavy chain of the secretory immunoglobulin M (sIgM), µs, is one of the few soluble lumenal ERAD substrates studied in mammalian cells (Amitay et al., 1992; Elkabetz et al., 2003). This protein provides an attractive model to study the interrelations between retention signals and degrons that operate in ERAD. During differentiation of B lymphocytes, the fate of µs switches from retention and degradation in B cells to stability and efficient secretion in plasma cells (Amitay et al., 1991; Amitay et al., 1992; Shachar et al., 1992). Moreover, in the same B cell, where µs is retained and degraded, the membrane isoform, µm, is a stable protein that reaches the cell surface, where it constitutes the B cell receptor. Thus, the intracellular fate of µs appears to be mediated by B-cell-specific components and to rely on µstp, the C-terminal tailpiece of µs that distinguishes it from µm. This unique sequence comprises 20 residues that include a penultimate Cys575 and is designated `µstpCys'. The µstpCys motif is a necessary cis-acting retention signal; either its truncation or Cys575 mutation to Ser or Ala hampers retention of µs in B cells (Sitia et al., 1990). Moreover, µstpCys, unlike µstpSer or µstpAla, is also sufficient to prevent secretion of IgG2b in B cells or in the non-lymphoid COS-7 cells and to retain the lysosomal protein cathepsin D to a pre-Golgi compartment in COS-7 cells (Sitia et al., 1990; Fra et al., 1993; Isidoro et al., 1996). In addition, the µstpCys motif targets the retained cathepsin D, but not the retained IgG2b, for rapid degradation in a pre-Golgi compartment (Fra et al., 1993). In this work, we fuse the µstpCys to two bona fide secretory proteins and study, in various non-lymphoid cells, its role as a retention signal, we localize the retained protein to the ER and we show that the µstpCys sequence acts as an ERAD degron.
That µstpCys acts as a cis-acting motif also allows investigation of the relevance of degron-linked N-glycans to ERAD. It is well established that N-glycans are key players in ER quality control, both as monitors of the folding of glycoproteins in the ER and as signals that target glycoproteins to ERAD (Helenius and Aebi, 2001; Ellgaard and Helenius, 2003; Helenius and Aebi, 2004; Moremen and Molinari, 2006; Olivari et al., 2006). Furthermore, correct positioning of N-glycans appears crucial for retention and degradation, as shown for the yeast lumenal ERAD substrate carboxypeptidase Y* (CPY*), whose most C-terminal of four N-glycans is necessary and sufficient to direct this protein for degradation (Kostova and Wolf, 2005; Spear and Ng, 2005). Nonetheless, it is currently unknown whether actual linking of N-glycans to degrons is of any significance. The µstp element harbors a single, most C-terminal N-glycan linked to Asn563 in the context of µs, in addition to four N-glycans linked to µs upstream of the µstp element (Brenckle and Kornfeld, 1980; Sidman et al., 1981). Both the µstp N-glycan as well as the penultimate Cys residue are highly conserved throughout evolution and are found in all species (de Lalla et al., 1998), although this N-glycan plays no role in maintaining the function of Cys575 in retention (Fra et al., 1993). Nevertheless, the µstp N-glycan was shown to be important for sIgM J-chain association to form pentamers, yet it was dispensable for secretion of sIgM hexamers or a µs double mutant (S565A C575A) (Cals et al., 1996; de Lalla et al., 1998). Taking into account the contribution of N-glycans linked to µs to the secretion and stability of this heavy chain (Sidman et al., 1981; Sidman, 1981), here we examine whether the N-glycan linked to the µstp motif, being the most C-terminal or the sole N-glycan, plays any role in sorting two bona fide secretory proteins to either a secretory or ERAD fate.
Our results establish the role of the µstpCys element as a portable retention signal. We also demonstrate that the µstpCys sequence acts as a portable degron that targets secretory proteins to ERAD through proteasomal degradation. Two interesting features of µstpCys are studied with regard to its role as a degron. First, we exclude the µstpCys degron as an obligatory ubiquitylation site. Second, we show that the N-glycan linked to the µstp element plays an important role in the sorting of µstp-fusion proteins for either secretion or ERAD.
|
Results |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
) from which the transmembrane segment was truncated (designated TPO) and a secreted version of the relatively stable yellow fluorescent protein that is led to the ER lumen by the hen egg lysozyme signal sequence (designated ssYFP). Two versions of µstp, wild-type µstpCys (penultimate Cys) and mutant µstpSer (penultimate Ser), were appended to the C-termini of TPO and ssYFP, downstream of a Myc tag, generating TPO-µstpCys and TPO-µstpSer (Fig. 1A) and ssYFP-µstpCys and ssYFP-µstpSer (Fig. 2A). Importantly, the fluorescence of the ssYFP fusion proteins reflects their proper folding and allows direct visualization of their localization within living cells by laser scanning confocal microscopy.
|
|
70% of TPO-µstpSer was recovered from the medium, fully accounting for its decrease within the cells, whereas less than 10% of TPO-µstpCys was secreted from HEK293T or COS-7 cells (Fig. 1B,C), suggesting that this protein was degraded.
Grafting the µstpCys sequence onto TPO allows us to test directly whether this cis-acting retention signal acts also as a portable ERAD degron. We have previously shown in 38C B cells that µs is a bona fide lumenal ERAD substrate because this retained protein is eliminated rapidly by the ubiquitin-proteasome system (Amitay et al., 1991; Amitay et al., 1992; Shachar et al., 1992; Elkabetz et al., 2003). To determine whether TPO-µstpCys was degraded in a fashion similar to that of µs, we performed pulse-chase experiments in the presence of proteasome inhibitors. The rapidly degraded TPO-µstpCys was stabilized in the presence of proteasome inhibitors, extending its half-life from 4 to 10 hours, and its secretion was improved slightly (Fig. 1D,E).
Similar experiments were performed with Myc-tagged ssYFP-µstpCys and ssYFP-µstpSer, and, for comparison, with ssYFP-Myc that lacked µstp altogether. Radiolabeled cells were chased and ssYFP-µstpCys, ssYFP-µstpSer and ssYFP-Myc were immunoprecipitated from cells and medium. After 7.5 hours of chase, more than 50% of ssYFP-µstpCys was retained within cells, whereas 70% of ssYFP-µstpSer and 63% of ssYFP-Myc were secreted to the medium (Fig. 2B,C). Notably, in agreement with the previously reported inefficient processing of the µstp N-glycan (Brenckle and Kornfeld, 1980; Cals et al., 1996), the N-glycans of the secreted ssYFP-µstpSer were a mixture of high-mannose (Fig. 2B, lower panel, arrow) and complex (Fig. 2B, lower panel, arrowhead) species. Again, ssYFP-µstpCys was not secreted, yet its intracellular levels decreased during the chase (Fig. 2B,C), suggestive of degradation. Indeed, in the presence of proteasome inhibitors, the 9 hour half-life of ssYFP-µstpCys was prolonged tenfold (Fig. 2B,C). Thus, µstpCys acts as a portable ERAD degron that targets both TPO and ssYFP for proteasomal degradation.
The function of µstpCys as an ER retention signal was demonstrated by real-time visualization of ssYFP in COS-7 cells (Fig. 3). Fluorescence microscopy revealed that, while the wild-type ssYFP-µstpCys was restricted to the ER, the mutant ssYFP-µstpSer was distributed both within the ER and Golgi, as was ssYFP-Myc (Fig. 3). These latter two fusion proteins colocalized with the trans-Golgi resident galactosyl transferase fused to cyan fluorescent protein (GalT-CFP), whereas the ssYFP-µstpCys was excluded from the GalT-CFP sites (Fig. 3). Importantly, the comparable fluorescence of the ssYFP fusion proteins indicates that µstpCys does not interfere with the proper folding of this reporter protein. Therefore, the retention of the fusion proteins at the ER is brought about by µstpCys per se. Hence, by analogy to the fate of µs in B cells (Sitia et al., 1990), wild-type µstpCys, unlike mutant µstpSer, is a sufficient cis-acting retention signal when fused to several reporter secretory proteins. The µstpCys sequence retains the reporter secretory proteins at the ER and appears to act as a retention signal and an ERAD degron recognized in various non-lymphoid cell lines.
|
The single lysine in µstp is irrelevant for degradation
Most ERAD substrates, including µs (Amitay et al., 1991; Amitay et al., 1992; Shachar et al., 1992; Elkabetz et al., 2003), are ubiquitylated on lysine residues before their proteasomal degradation. Following our finding that the µstpCys fusion proteins studied here are degraded by the proteasome (Figs 1, 2), we asked whether µstpCys, acting as an ERAD degron, also serves as the ubiquitylation site of its fusion proteins. Conceivably, sequences that function as degrons could recruit the ubiquitylation machinery to the ERAD substrates and/or provide the site for ubiquitin conjugation. As µstpCys contains a single lysine, we examined whether this residue serves as the ubiquitylation site by substituting it with arginine, generating the TPO-µstpCysKR mutant. Pulse-chase experiments showed indistinguishable proteasomal degradation of wild-type TPO-µstpCys (Fig. 4A) and the TPO-µstpCysKR mutant (Fig. 4B). The similar short half-life of both substrates was prolonged equally by 2.5-fold in the presence of proteasome inhibitors (Fig. 4C). Hence, the µstp lysine residue appears to play no role in degradation of TPO-µstpCys, suggesting that degrons, which by definition confer degradation, are not necessarily the sequences that undergo ubiquitylation.
|
; Ellgaard and Helenius, 2003; Helenius and Aebi, 2004), and their positioning within protein substrates might be crucial for degradation (Kostova and Wolf, 2005; Spear and Ng, 2005). Here, we investigated whether the N-glycan linked to the µstp motif plays any role in the retention and degradation of µstp-fusion proteins. This is especially relevant because the single N-glycan of µstp is the most C-terminal also in the context of the µs heavy chain and is highly conserved throughout evolution (Sidman et al., 1981; de Lalla et al., 1998). Substituting the asparagine for glutamine in the tripeptide sequon Asn-Val-Ser in the µstp of the Cys and Ser versions of the two reporter fusion proteins generated four `NQ' mutants. In TPO-µstp, this asparagine residue is most C-terminal of four additional potential N-glycosylation sites (Rawitch et al., 1992). For ssYFP, which has no N-glycosylation sites of it own, µstp provides the only site for N-glycan attachment in ssYFP-µstp. To confirm that the asparagine residue within the µstp motif was glycosylated also in the context of the fusion proteins, the electrophoretic mobility of the NQ mutants was compared with that of their wild-type counterparts. This analysis is based on mobility retardation, which is contributed by each of the N-glycans. The slightly faster migration of the TPO-µstpCysNQ mutant, compared with the mobility of wild-type TPO-µstpCys and the TPO-µstpCysKR mutant (Fig. 5A), suggested that the asparagine residue within µstp was indeed glycosylated. Mutant TPO-µstpCysNQ was still sensitive to endoglycosidase H (endo H) (Fig. 5A), indicating that N-glycans were linked to at least some of the additional four N-glycosylation sites in the TPO protein. Likewise, the ssYFP-µstpCysNQ mutant migrated faster than wild-type ssYFP-µstpCys (Fig. 6A), and its electrophoretic mobility, which was similar to that of de-glycosylated ssYFP-µstpCys, was not altered upon digestion with endo H (Fig. 6A). This demonstrated that the asparagine residue in µstp underwent glycosylation also in the context of ssYFP-µstpCys, establishing it as the sole N-glycosylation site in this fusion protein.
|
|
Ablation of the N-glycosylation site in µstp exerted distinct effects on the intracellular fate of the µstp-fusion proteins. It significantly accelerated the turnover of both TPO-µstpCysNQ (Fig. 5B,C) and ssYFP-µstpCysNQ (Fig. 6B,C), relative to their corresponding wild-type proteins. The half-life of TPO-µstpCysNQ was shortened to 2.5 hours (Fig. 5C) and that of ssYFP-µstpCysNQ was 2 hours (Fig. 6C). Moreover, this mutation also considerably altered the intracellular fate of the SerNQ chimeras. While TPO-µstpSerNQ was still secreted, albeit poorly (data not shown), ssYFP-µstpSerNQ was tightly retained within cells (Fig. 6D). Confocal microscopy revealed that, unlike the ssYFP-µstpSer that was localized both to the ER and the trans-Golgi (Fig. 3), ssYFP-µstpSerNQ was restricted to the ER (Fig. 6E), as was the wild-type ssYFP-µstpCys (Fig. 3). Moreover, ssYFP-µstpSerNQ became a rapidly degraded protein (Fig. 6B,C), exhibiting turnover rates (t1/2=1.7 hours) similar to those measured for the CysNQ construct (t1/2=2.1 hours), and both substrates were stabilized in the presence of proteasome inhibitors (Fig. 6B,C). Thus, in the case of the µstp-fusion reporter proteins, the absence of the most C-terminal N-glycan did not hamper their elimination by ERAD, but in fact, it accelerated the turnover of these NQ chimeras and even converted secreted ssYFP-µstpSer to an ERAD substrate.
Substituting the asparagine for glutamine in the µstp motif abolished the N-glycosylation site, but also introduced a mutation that could have generated a retention signal that is unrelated to the glycosylation status of µstp. To demonstrate that the improved retention and accelerated degradation were indeed related to the absence of the N-glycan, we generated unglycosylated ssYFP-µstpCys and ssYFP-µstpSer by two additional approaches. First, N-glycosylation was abolished altogether by treating cells with tunicamycin. As expected, this treatment generated mostly unglycosylated ssYFP-µstp nascent proteins, but a small fraction (20%) was still glycosylated, probably reflecting the pool of the dolichol pyrophosphate oligosaccharide donor (Fig. 7A,D). The residual glycosylation of ssYFP-µstp chimeras provided an interesting reference to the retention and degradation of the unglycosylated fusion proteins. Accelerated proteasomal degradation (t1/2=2.2 hours) was observed for the unglycosylated ssYFP-µstpCys (Fig. 7A,B, triangles), whereas the 20% glycosylated ssYFP-µstpCys was more stable (Fig. 7A). Importantly, tunicamycin exerted no effect on the turnover of the already unglycosylated ssYFP-µstpCysNQ mutant (Fig. 7C). This indicates that the effect of tunicamycin is specific to the glycosylation status of the ssYFP-µstp chimeras, rather than being a consequence of abolished N-glycosylation of other glycoproteins. The effect of tunicamycin on ssYFP-µstpSer revealed that the secretion of the unglycosylated form was severely hampered (Fig. 7D,E, compare triangles and circles), resembling the tight retention of ssYFP-µstpSerNQ (Fig. 6D). Conversely, a small fraction of the glycosylated ssYFP-µstpSer was secreted (Fig. 7D,E, squares) and was even processed from a high-mannose to a complex (Fig. 7D, arrowhead) N-glycan. However, this processing was inefficient, resembling the processing of the secreted ssYFP-µstpSer in untreated cells (Fig. 7D, compare arrow and arrowhead; see also Fig. 2B, lower panel). Moreover, unlike the residual glycosylated fusion protein that was relatively stable, the unglycosylated ssYFP-µstpSer was degraded rapidly (t1/2=3.5 hours), and this degradation was inhibited by proteasome inhibitors (Fig. 7D,E).
|
To confirm further that the cysteine-independent retention and accelerated degradation were the consequence of the absence of the µstp N-glycan, we employed an additional mutagenesis to abolish specifically the glycosylation of the µstp motif, by substituting the serine in the tripeptide sequon Asn-Val-Ser for alanine, generating ssYFP-µstpCysSA and ssYFP-µstpSerSA. This approach was used previously to generate µs that lacked the µstp N-glycan and to show that this N-glycan was dispensable for secretion of sIgM hexamers or the µs double mutant S565A C575A (Cals et al., 1996; de Lalla et al., 1998). The electrophoretic mobility and sensitivity to endo H confirmed that the µstp N-glycosylation was ablated by this substitution (Fig. 8A). This ablation exerted effects similar to those observed with the NQ mutants or the tunicamycin treatment. The turnover of ssYFP-µstpCysSA was significantly accelerated (t1/2 2.5 hours) and ssYFP-µstpSerSA was retained within the cells (Fig. 8D) and rapidly degraded (t1/2=2.6 hours) (Fig. 8B,C). Degradation of both SA mutants was proteasomal, as indicated by its inhibition by proteasome inhibitors (Fig. 8B,C). Finally, the ER retention of ssYFP-µstpSerSA (Fig. 8E) observed by confocal microscopy resembled the ER localization of ssYFP-µstpSerNQ (Fig. 6E) and wild-type ssYFP-µstpCys (Fig. 3), contrary to the ER and trans-Golgi localization of the wild-type ssYFP-µstpSer (Fig. 3). To conclude, in the absence of the sole N-glycan of the µstp motif, ssYFP-µstpSer was no longer secreted and both ssYFP-µstpSer and ssYFP-µstpCys were rapidly degraded proteins (Figs 6, 7, 8). Thus, the µstp N-glycan facilitates secretion and impedes elimination by ERAD of secretory reporter proteins fused to the µstp sequence.
|
|
Discussion |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
|
The function of µstpCys as an element necessary for retention of µs in B cells (Sitia et al., 1990) and sufficient for retention of lysosomal cathepsin D in COS-7 cells (Isidoro et al., 1996) is extended in the current study. We show that µstpCys confers ER retention upon two bona fide secretory proteins, TPO and ssYFP, when fused to their C-termini. Secretion of the chimeras is prevented with similar efficiency in several non-lymphoid cell lines (COS-7, HEK293T or HeLa). Consistent with previous reports, this µstpCys-mediated retention depends on the penultimate Cys residue, as its replacement by a Ser residue allows secretion of TPO-µstpSer and ssYFP-µstpSer.
A particularly intriguing finding in this study is the role played by the µstp N-glycan in sorting for retention or secretion (Table 1). Our results indicate that this N-glycan is required for secretion of the fusion protein, as its absence results in intracellular retention. However, the latter is observed only when this N-glycan is the sole one that is linked to the protein, as in the case of ssYFP. This observation is independent of the mode in which the N-glycan is abolished. Substitutions of the asparagine or serine in the tripeptide sequon Asn-Val-Ser as well as N-glycosylation inhibition by tunicamycin resulted in retention of ssYFP-µstpSer. Conversely, proteins bearing additional glycosylation sites, albeit lacking the µstp N-glycan, such as TPO-µstpSerNQ or the µs double mutant (S565A C575A) (Fra et al., 1993; Cals et al., 1996; de Lalla et al., 1998), were still secreted. Notably, the contribution of the N-glycan to secretion is revealed only in the context of the µstp, as ssYFP-Myc, lacking µstp altogether, is secreted. Nonetheless, the involvement of the N-glycan in secretion is counteracted by the penultimate Cys residue, and the net result is retention of ssYFP-µstpCys.
The contribution of the µstp N-glycan to secretion is revealed only when the Cys residue is replaced by Ser, as ssYFP-µstpSer is secreted, whereas ssYFP-µstpSer lacking N-glycan is retained (Table 1). This indicates that abolishing N-glycosylation renders the µstp a retention signal, regardless of the penultimate Cys. This finding is especially interesting as both the µstp N-glycan and penultimate Cys are highly conserved throughout evolution and the µstp N-glycan is important for sIgM association with the J-chain for pentamerization (de Lalla et al., 1998). However, in the context of the µs heavy chain, this N-glycan is relevant neither for maintaining the Cys575 function in retention nor for secretion of sIgM hexamers or the µs double mutant (S565A C575A) (Fra et al., 1993; Cals et al., 1996; de Lalla et al., 1998), indicating that the unglycosylated µstp adopts a secretion-competent conformation. Therefore, retention of the unglycosylated ssYFP-µstp suggests that, in the context of µstp, the N-glycan plays a positive role in secretion, rather than a negative role in generating a misfolded µstp that can serve as a retention signal.
Our study demonstrates that µstpCys is a portable ERAD degron, acting in cis and conferring proteasomal degradation upon two secretory reporter proteins in non-lymphoid cells. Moreover, the turnover rates of these µstp fusion proteins are similar to those measured for the endogenous µs heavy chain in B cells (Amitay et al., 1991; Amitay et al., 1992; Elkabetz et al., 2003). Another portable C-terminal ERAD degron was discovered recently in cyclooxygenase 2 (COX2) (Mbonye et al., 2006). A segment of 19 amino acids (19-AA), located just inside the C-terminus, mediated the entry of COX2 to ERAD. This degron was necessary for rapid degradation of COX2 and conferred rapid turnover upon the otherwise stable cyclooxygenase 1 (COX1) when inserted near its C-terminus (ins594-612 COX1) (Mbonye et al., 2006). Similar to the COX2 degron, µstp shares no apparent properties with other degrons. Unlike the artificial C-terminal degron SL17 (Gilon et al., 1998) or the N-terminal Sgk1 degron (Arteaga et al., 2006), neither the COX2 degron nor µstp are hydrophobic. Also, neither the COX2 degron nor µstp forms amphipathic helices, as opposed to the N-terminal Deg1 (Johnson et al., 1998) or Sgk1 degrons. Nevertheless, the µstp element is extremely conserved in mammals, indicating the significance of its entire sequence.
ERAD degrons might recruit the ubiquitylation machinery and/or provide the platform for ubiquitin conjugation. It has been shown that the N-terminal transmembrane segment of 3-hydroxy-3-methylglutaryl-coenzyme A reductase serves as a necessary and sufficient ERAD degron that also provides the lysine residues required for the sterol-dependent ubiquitylation and degradation of this protein (Doolman et al., 2004; Sever et al., 2003). However, in the case of µstpCys, it appears that the polyubiquitin chain(s) are not conjugated to µstpCys itself as degradation of TPO-µstpCys is not affected by substituting the single lysine residue within the µstp sequence. Hence, ubiquitin is either always conjugated to other lysine residues located within the reporter protein or ubiquitin conjugation is promiscuous enough to select alternative lysine(s) as acceptor site(s) if the preferred lysine within the µstp is mutated. The notion that ubiquitin conjugation to site(s) proximal to the degron is not obligatory for degradation was also reported for the yeast ERAD substrate CPY* (Spear and Ng, 2005).
The dual role played by µstp-Cys as a retention signal and a degron can be attributed to distinct elements within the µstp-Cys motif (Table 1). As described above, the penultimate Cys is crucial for retention, and the N-glycan of µstp is required for secretion. However, reporter proteins that are tightly retained, either owing to the presence of the Cys residue or absence of the N-glycan, are degraded by the proteasome, suggesting that elements other than the penultimate Cys or the N-glycan within µstp also contribute to the targeting to ERAD. The penultimate Cys appears to function mostly in retention, possibly by means of its association with ER thiol oxidoreductases (Anelli et al., 2002; Anelli et al., 2003), but does not seem to contribute to degradation. Even when the penultimate Cys is replaced, µstpSer can still target ssYFP for degradation provided retention is conferred by the absence of the N-glycan. On the other hand, the N-glycan appears to play two roles, contributing both to secretion and stability of the reporter proteins. Clearly, in the absence of the N-glycan, degradation of the already tightly retained µstpCys fusion proteins is accelerated, and ssYFP-µstpSer is converted from a secreted protein into one that is retained and rapidly degraded.
Our finding that the µstp N-glycan plays a stabilizing role adds another layer to the role played by N-glycans in ER quality control and in ERAD (Helenius and Aebi, 2001; Ellgaard and Helenius, 2003; Helenius and Aebi, 2004; Molinari et al., 2003; Moremen and Molinari, 2006; Olivari et al., 2006). Results similar to ours were reported for RI332, a lumenal truncated variant of ribophorin I, which was an ERAD substrate in HeLa cells (de Virgilio et al., 1999). Removal of its single N-glycosylation site resulted in degradation that was no longer biphasic, but turned into a rapid monophasic turnover of the mutant RI332-Thr. Interaction with calnexin was implicated in regulating the slower proteolytic phase (de Virgilio et al., 1999). The contribution of the most C-terminal N-glycan to ERAD was shown in yeast as well as in mammalian cells. Truncation of the C-terminal domain of CPY* stabilized this ERAD substrate. As it turned out, the lysine-devoid C-terminal domain still allowed degradation of CPY*, whereas the most C-terminal N-glycan was necessary and sufficient for ERAD (Kostova and Wolf, 2005; Spear and Ng, 2005). Interesting interrelationships between an ERAD degron and its N-glycan are shown here and were recently reported for COX1 and COX2 (Mbonye et al., 2006). Mutation of Asn594, an N-glycosylation site at the beginning of the 19-AA degron, stabilized both COX2 and ins594-612 COX1. Nonetheless, COX mutants that were glycosylated at Asn594 but lacked the remainder of the 19-AA cassette were also stable, suggesting that glycosylation of Asn594 was necessary for COX2 degradation, but that at least some other part of the 19-AA segment was also required (Mbonye et al., 2006). Conversely, in our study, abolishing the single N-glycan in µstp, which is the only site for N-glycosylation in ssYFP-µstpCys and the most C-terminal (out of five potential sites) in TPO-µstpCys, did not hamper degradation but, in fact, accelerated the turnover of both chimeras. Hence, the role of N-glycans in ERAD cannot be generalized and appears to be affected by other elements in the degron and/or reporter protein.
Comparison between the turnover rates of the ssYFP-µstpCys with or without the N-glycan points to N-glycan-related processes as potential rate-limiting steps and might provide mechanistic insights into their degradation. Within the ER lumen, ERAD substrates are subjected to several N-glycan-related quality control processes, including the calnexin-calreticulin cycle, trimming by mannosidase and recognition by EDEM (Helenius and Aebi, 2001; Ellgaard and Helenius, 2003; Helenius and Aebi, 2004; Moremen and Molinari, 2006; Olivari et al., 2006). Passage across the ER membrane, poly-ubiquitylation and de-glycosylation by peptide N-glycanase (Suzuki et al., 2002) appear to occur in a coupled fashion coordinated by the p97/Cdc48 complex (Bar-Nun, 2005). This AAA-ATPase provides the driving force for dislocation (Ye et al., 2001; Elkabetz et al., 2004), associates with ERAD-dedicated E3 ligases (Zhong et al., 2004; Ye et al., 2005; Lilley and Ploegh, 2005; Schuberth and Buchberger, 2005; Neuber et al., 2005), mediates the binding of peptide N-glycanase to proteasomes (Li et al., 2005; Allen et al., 2006) and interacts with ubiquitin (Dai and Li, 2001) and deubiquitylating enzymes such as Otu1 and ataxin-3 (Doss-Pepe et al., 2003; Rumpf and Jentsch, 2006; Zhong and Pittman, 2006; Wang et al., 2006). It remains to be established which of the N-glycan-related processes is the rate-limiting step that is bypassed by the proteins fused to the unglycosylated µstp.
|
Materials and Methods |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
); generously provided by R. Magnusson], the transmembrane domain (residues 849-933) was truncated and residues 1-848 were fused to a Myc tag sequence (EQKLISEEDLN) followed by the µstpCys/Ser sequence. The latter were generated by PCR from plasmid pSV-Vµ1 (Neuberger, 1983), generously provided by R. Sitia. To introduce mutations into µstpCys, a fragment of about 1 kb encoding the last third of TPO-Myc-µstpCys was excised from the TPO-Myc-µstpCys pcDNA3 vector, using ClaI and XbaI (ClaI-XbaI fragment), and cloned into pBCSK(+) (Stratagene) at these sites. PCR with TPO-Myc-µstpCys as a template was used to generate the KR mutation, using forward primer no. 1 and reverse primer no. 2, and the NQ mutation, using forward primer no. 3 and reverse primer no. 4 (for sequences of PCR primers and oligonucleotides, see Table 2). The PCR products were cut with SphI and NdeI and cloned into these sites in the pBCSK(+) vector containing the ClaI-XbaI fragment. The ClaI-XbaI fragment was then cut and cloned back into the TPO-Myc-µstpCys pcDNA3 vector and the resulting plasmids were sequenced.
|
Plasmid pEYFP-C1 (Clontech) encodes yellow fluorescent protein (YFP), and ssYFP was generated by inserting the hen egg lysozyme signal sequence upstream of the YFP sequence. The ssYFP-encoding sequence was excised with NheI and HindIII and cloned into NheI- and HindIII-digested pcDNA3, generating pcDNA3-ssYFP. The Myc-µstp was amplified by PCR using TPO-Myc-µstpCys pcDNA3 as a template, introducing HindIII and BamHI sites at the 5' and 3' ends, respectively. Both Myc-µstpCys and Myc-µstpSer were generated using the same forward primer no. 5 and two different reverse primers, no. 6 for µstpCys and no. 7 for µstpSer. The ssYFP-Myc-µstp expression plasmids (pcDNA3-ssYFP-Myc-µstpCys/Ser) were constructed by cloning the HindIII- and BamHI-digested PCR products into the HindIII- and BamHI-digested pcDNA3-ssYFP. For comparison, pcDNA3-ssYFP-Myc (with no µstp) was constructed by ligating the HindIII- and BamHI-digested Myc sequence (sense strand no. 8 and antisense strand no. 9) into the HindIII- and BamHI-digested pcDNA3-ssYFP.
For constructing the ssYFP-Myc-µstp NQ mutants, an intermediate plasmid, pBC-Myc-µstp, was constructed by cloning a PCR product into pBCSK(+). The PCR reaction used primers no. 5 and no. 6 and pcDNA3-ssYFP-Myc-µstpCys as a template, and the HindIII- and BamHI-digested PCR products were inserted into pBCSK(+) digested with the same enzymes. To introduce the NQ mutation into the µstp motif, both µstpCysNQ and µstpSerNQ were amplified using pBC-Myc-µstp as a template, the same forward primer no. 3 and two different reverse primers, no. 6 for µstpCysNQ and no. 7 for µstpSerNQ. The NdeI- and BamHI-digested PCR products were cloned into the same sites in the pBC-Myc-µstpCys. The Myc-µstpCysNQ and Myc-µstpSerNQ fragments were excised from the respective pBC-Myc-µstpNQ plasmids by HindIII and BamHI and cloned back into the pcDNA3-ssYFP. The ssYFP-Myc-µstp SA mutants were constructed by PCR reaction using primers no. 11 and no. 12 and pcDNA3-ssYFP-Myc-µstpCys and ssYFP-Myc-µstpSer as templates, employing a QuikChange site-directed mutagenesis kit (Stratagene).
Cell lines and transfection
COS-7, HeLa or HEK293(T) cells were grown at 37°C in a humidified 5% CO2 incubator in Dulbecco's modified Eagle's medium (DMEM; Invitrogen Life Technologies), supplemented with 10% (v/v) fetal calf serum, 1 mM L-glutamine, 2 µg/ml penicillin, 20 µg/ml streptomycin and 2.5 µg/ml nystatin (Biological Industries, Beit Ha'Emek). When they reached 40% confluence, the cells were transiently transfected with 4-10 µg of the appropriate DNA construct by the calcium phosphate method (Chen and Okayama, 1988) or using jetPEI (Polyplus-Transfection). Analysis of the cells was performed 24-48 hours post-transfection. Stable transfectants were selected with G418.
Protein secretion and degradation, immunoprecipitation, immunoblotting and endoglycosidase H treatment
Secretion was followed by estimation of radiolabeled or unlabeled proteins accumulated in the medium. Degradation was followed by pulse-chase experiments of radiolabeled cells. Cells were starved for 1 hour in methionine/cysteine-deficient medium, pulse-labeled for 1 hour with 50 µCi/ml of [35S]methionine-[35S]cysteine (Promix), and chased in the presence of excess unlabeled methionine-cysteine. Where indicated, tunicamycin (10 µg/ml) or proteasome inhibitors N-acetyl-leucyl-leucyl-norleucinal (ALLN; 50 µM) and carboxybenzyl-leucil-leucil-leucinal (MG-132; 5 µM) were present throughout the starvation, pulse-labeling and chase. At various time points, medium was collected, cells were washed twice with PBS and lysed in fresh ice-cold lysis buffer, as described previously (Amitay et al., 1991). The µstp fusion proteins were immunoprecipitated from cell lysates and media using an excess of mouse anti-Myc (clone 9E10) or rabbit anti-GFP (Abcam), followed by protein A-sepharose (Repligen). Immunoprecipitated material was collected by centrifugation (500 g, 2 minutes, 4°C) and washed three times with PBS containing 1% (v/v) Nonidet P-40. Immunoprecipitated proteins or cell extracts were resolved by SDS-PAGE and electroblotted onto nitrocellulose. Radiolabeled blots were first exposed to autoradiography and then probed with either horseradish peroxidase (HRP)-conjugated specific antibodies or with specific primary antibodies followed by the respective HRP-conjugated secondary antibodies. The HRP was visualized by the enhanced chemiluminescence (ECL) reaction. The primary antibodies used were: mouse anti-Myc (clone 9E10) and rabbit anti-µstp (Rabinovich et al., 2002). The HRP-conjugated secondary antibodies used were: goat anti-mouse IgG (Jackson) and goat-anti-rabbit IgG (Sigma). The HRP-conjugated specific antibody used was mouse anti-Myc-HRP (Upstate Millipore). For endo H treatment, immunoprecipitated material was incubated for 1 hour at 37°C with 0.25 IU/ml endo H (New England Biolabs) according to the manufacturer's protocol.
Confocal laser-scanning microscopy and live-cell imaging
Transfected cells, grown in Labtek chambers in DMEM (without phenol red) supplemented with 20 mM HEPES pH 7.4, were imaged with a Zeiss LSM PASCAL equipped with an Axiovert 200 inverted microscope. Argon 458 nm and 514 nm laser lines were used for ECFP and EYFP, respectively. Confocal images were captured using a 63x 1.4 NA objective with a pinhole diameter between 1 and 2 Airy units. Images were generated and analyzed using the Zeiss LSM software and modified using Adobe Photoshop.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Present address: Laboratory of Stem Cell and Tumor Biology, Division of Neurosurgery and Developmental Biology Program, Sloan-Kettering Institute, New York, NY, USA
|
References |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
Allen, M. D., Buchberger, A. and Bycroft, M. (2006). The PUB domain functions as a p97 binding module in human peptide N-glycanase. J. Biol. Chem. 281, 25502-25508.
Amitay, R., Bar-Nun, S., Haimovich, J., Rabinovich, E. and Shachar, I. (1991). Post-translational regulation of IgM expression in B lymphocytes. J. Biol. Chem. 266, 12568-12573.
Amitay, R., Shachar, I., Rabinovich, E., Haimovich, J. and Bar-Nun, S. (1992). Degradation of secretory IgM in B lymphocytes occurs in a post-endoplasmic reticulum compartment and is mediated by a cysteine protease. J. Biol. Chem. 267, 20694-20700.
Anelli, T., Alessio, M., Mezghrani, A., Simmen, T., Talamo, F., Bachi, A. and Sitia, R. (2002). ERp44, a novel endoplasmic reticulum folding assistant of the thioredoxin family. EMBO J. 21, 835-844.[CrossRef][Medline]
Anelli, T., Alessio, M., Bachi, A., Bergamelli, L., Bertoli, G., Camerini, S., Mezghrani, A., Ruffato, E., Simmen, T. and Sitia, R. (2003). Thiol-mediated protein retention in the endoplasmic reticulum: the role of ERp44. EMBO J. 22, 5015-5022.[CrossRef][Medline]
Arteaga, M. F., Wang, L., Ravid, T., Hochstrasser, M. and Canessa, C. M. (2006). An amphipathic helix targets serum and glucocorticoid-induced kinase 1 to the endoplasmic reticulum-associated ubiquitin-conjugation machinery. Proc. Natl. Acad. Sci. USA 103, 11178-11183.
Bar-Nun, S. (2005). The role of p97/Cdc48p in endoplasmic reticulum-associated degradation: from the immune system to yeast. Curr. Top. Microbiol. Immunol. 300, 95-125.[Medline]
Bonifacino, J. S. and Weissman, A. M. (1998). Ubiquitin and the control of protein fate in the secretory and endocytic pathways. Annu. Rev. Cell Dev. Biol. 14, 19-57.[CrossRef][Medline]
Brenckle, R. and Kornfeld, R. (1980). Structure of the oligosaccharides of mouse immunoglobulin M secreted by the MOPC 104E Plasmacytoma. Arch. Biochem. Biophys. 201, 160-173.[CrossRef][Medline]
Cals, M.-M., Guenzi, S., Carelli, S., Simmen, T., Sparvoli, A. and Sitia, R. (1996). IgM polymerization inhibits the Golgi-mediated processing of the mu-chain carboxy-terminal glycans. Mol. Immunol. 33, 15-24.[CrossRef][Medline]
Chen, C. A. and Okayama, H. (1988). Calcium phosphate-mediated gene transfer: a highly efficient transfection system for stably transforming cells with plasmid DNA. Biotechniques 6, 632-638.[Medline]
Dai, R. M. and Li, C. C. (2001). Valosin-containing protein is a multi-ubiquitin chain-targeting factor required in ubiquitin-proteasome degradation. Nat. Cell Biol. 3, 740-744.[CrossRef][Medline]
de Lalla, C., Fagioli, C., Cessi, F. S., Smilovich, D. and Sitia, R. (1998). Biogenesis and function of IgM: the role of the conserved mu-chain tailpiece glycans. Mol. Immunol. 35, 837-845.[CrossRef][Medline]
de Virgilio, M., Kitzmuller, C., Schwaiger, E., Klein, M., Kreibich, G. and Ivessa, N. E. (1999). Degradation of a short-lived glycoprotein from the lumen of the endoplasmic reticulum: the role of N-linked glycans and the unfolded protein response. Mol. Biol. Cell 10, 4059-4073.
Doolman, R., Leichner, G. S., Avner, R. and Roitelman, J. (2004). Ubiquitin is conjugated by membrane ubiquitin ligase to three sites, including the N terminus, in transmembrane region of mammalian 3-hydroxy-3-methylglutaryl coenzyme A reductase: implications for sterol-regulated enzyme degradation. J. Biol. Chem. 279, 38184-38193.
Doss-Pepe, E. W., Stenroos, E. S., Johnson, W. G. and Madura, K. (2003). Ataxin-3 interactions with rad23 and valosin-containing protein and its associations with ubiquitin chains and the proteasome are consistent with a role in ubiquitin-mediated proteolysis. Mol. Cell. Biol. 23, 6469-6483.
Elkabetz, Y., Kerem, A., Tencer, L., Winitz, D., Kopito, R. R. and Bar-Nun, S. (2003). Immunoglobulin light chains dictate vesicular transport-dependent and -independent routes for IgM degradation by the ubiquitin-proteasome pathway. J. Biol. Chem. 278, 18922-18929.
Elkabetz, Y., Shapira, I., Rabinovich, E. and Bar-Nun, S. (2004). Distinct steps in dislocation of luminal endoplasmic reticulum-associated degradation substrates: roles of endoplamic reticulum-bound p97/Cdc48p and proteasome. J. Biol. Chem. 279, 3980-3989.
Ellgaard, L. and Helenius, A. (2003). Quality control in the endoplasmic reticulum. Nat. Rev. Mol. Cell Biol. 4, 181-191.[CrossRef][Medline]
Fayadat, L., Siffroi-Fernandez, S., Lanet, J. and Franc, J. L. (2000). Degradation of human thyroperoxidase in the endoplasmic reticulum involves two different pathways depending on the folding state of the protein. J. Biol. Chem. 275, 15948-15954.
Fra, A. M., Fagioli, C., Finazzi, D., Sitia, R. and Alberini, C. M. (1993). Quality control of ER synthesized proteins: an exposed thiol group as a three-way switch mediating assembly, retention and degradation. EMBO J. 12, 4755-4761.[Medline]
Gilon, T., Chomsky, O. and Kulka, R. G. (1998). Degradation signals for ubiquitin system proteolysis in Saccharomyces cerevisiae. EMBO J. 17, 2759-2766.[CrossRef][Medline]
Helenius, A. and Aebi, M. (2001). Intracellular functions of N-linked glycans. Science 291, 2364-2369.
Helenius, A. and Aebi, M. (2004). Roles of N-linked glycans in the endoplasmic reticulum. Annu. Rev. Biochem. 73, 1019-1049.[CrossRef][Medline]
Isidoro, C., Maggioni, C., Demoz, M., Pizzagalli, A., Fra, A. M. and Sitia, R. (1996). Exposed thiols confer localization in the endoplasmic reticulum by retention rather than retrieval. J. Biol. Chem. 271, 26138-26142.
Johnson, P. R., Swanson, R., Rakhilina, L. and Hochstrasser, M. (1998). Degradation signal masking by heterodimerization of MATalpha2 and MATa1 blocks their mutual destruction by the ubiquitin-proteasome pathway. Cell 94, 217-227.[CrossRef][Medline]
Kostova, Z. and Wolf, D. H. (2003). For whom the bell tolls: protein quality control of the endoplasmic reticulum and the ubiquitin-proteasome connection. EMBO J. 22, 2309-2317.[CrossRef][Medline]
Kostova, Z. and Wolf, D. H. (2005). Importance of carbohydrate positioning in the recognition of mutated CPY for ER-associated degradation. J. Cell Sci. 118, 1485-1492.
Li, G., Zhou, X., Zhao, G., Schindelin, H. and Lennarz, W. J. (2005). Multiple modes of interaction of the deglycosylation enzyme, mouse peptide N-glycanase, with the proteasome. Proc. Natl. Acad. Sci. USA 102, 15809-15814.
