Research ArticleStress responses

The endoplasmic reticulum–residing chaperone BiP is short-lived and metabolized through N-terminal arginylation

See allHide authors and affiliations

Science Signaling  02 Jan 2018:
Vol. 11, Issue 511, eaan0630
DOI: 10.1126/scisignal.aan0630

Instability of the ER-residing proteins

Some ER proteins are subjected to a posttranslational modification known as N-terminal arginylation. Shim et al. found that the ER chaperone BiP was unexpectedly short-lived and that N-terminal arginylation promoted its relocalization to the cytosol, where it was degraded. ER stress, particularly when combined with proteasomal inhibition, increased the N-terminal arginylation of BiP. This pathway was inhibited by HERP, a component of the ER degradation pathway. These results suggest that ER proteins are more unstable than was previously appreciated, which may enable cells to quickly return the abundance of ER chaperones to basal amounts after ER stress has been resolved.

Abstract

BiP and other endoplasmic reticulum (ER)–resident proteins are thought to be metabolically stable and to function primarily in the ER lumen. We sought to assess how the abundance of these proteins dynamically fluctuates in response to various stresses and how their subpopulations are relocated to non-ER compartments such as the cytosol. We showed that the molecular chaperone BiP (also known as GRP78) was short-lived under basal conditions and ER stress. The turnover of BiP was in part driven by its amino-terminal arginylation (Nt-arginylation) by the arginyltransferase ATE1, which generated an autophagic N-degron of the N-end rule pathway. ER stress elicited the formation of R-BiP, an effect that was increased when the proteasome was also inhibited. Nt-arginylation correlated with the cytosolic relocalization of BiP under the types of stress tested. The cytosolic relocalization of BiP did not require the functionality of the unfolded protein response or the Sec61- or Derlin1-containing translocon. A key inhibitor of the turnover and Nt-arginylation of BiP was HERP (homocysteine-responsive ER protein), a 43-kDa ER membrane–integrated protein that is an essential component of ER-associated protein degradation. Pharmacological inhibition of the ER-Golgi secretory pathway also suppressed R-BiP formation. Finally, we showed that cytosolic R-BiP induced by ER stress and proteasomal inhibition was routed to autophagic vacuoles and possibly additional metabolic fates. These results suggest that Nt-arginylation is a posttranslational modification that modulates the function, localization, and metabolic fate of ER-resident proteins.

INTRODUCTION

About one-third of the human proteome is destined for the endoplasmic reticulum (ER)–Golgi secretory pathway (1). Upon translation, the nascent polypeptides of these secretory proteins are translocated into the ER lumen, where their signal peptides are cleaved off by the signal peptide peptidase, exposing new N-terminal residues on mature peptides (fig. S1A). Translocated polypeptides undergo folding and other posttranslational modifications with the assistance by ER-residing folding factors, such as BiP, GRP94, calreticulin, protein disulfide isomerase (PDI), and ERdj5 (fig. S1B) (2). Although the functions of these ER-residing proteins are thought to be largely confined to the ER lumen, these proteins are present and exert specific biological functions in non-ER cellular components. For example, the major Ca2+-binding molecular chaperone BiP redistributes to other subcellular compartments under ER stress and viral infection, such as the cytoplasm, mitochondria, plasma membrane, nucleus, and extracellular space (313). PDI, an essential redox chaperone in the ER lumen, has also been found in various subcellular compartments such as cytosol, mitochondria, and nucleus (14). Non-ER PDIs can function as a reductase and chaperone with binding partners and clients other than the ER luminal PDI (14, 15). A subpopulation of calreticulin relocates to the cytoplasm and plasma membrane in a manner independent of ubiquitination or the proteasome (1618). To date, little is known about the mechanisms by which these ER-residing proteins relocate to non-ER compartments.

Various cellular stresses cause the accumulation of excessive misfolded proteins in the ER lumen. Cells operate complicated protein quality control systems that recognize misfolded proteins and either refold or degrade them (fig. S1C). One such quality control system is the unfolded protein response (UPR) (19). In unstressed cells, the UPR is normally suppressed because BiP binds and inactivates the three UPR effectors, protein kinase RNA-like endoplasmic reticulum kinase (PERK), inositol-required enzyme-1 (IRE-1), and activating transcription factor 6 (ATF6) (20). Upon ER stress, BiP is redirected from the luminal domains of PERK, ATF6, and IRE-1 to accumulating misfolded proteins, leading to the transcriptional induction of folding factors such as BiP (21). Once the UPR achieves ER homeostasis, those transcriptionally induced molecular chaperones and other ER-residing proteins must return to their basal amounts. For example, the abundance of BiP, which normally represents about 0.5% of the human proteome, increases up to 100 folds upon ER stress (22, 23). However, the dynamic fluctuation of ER proteins during the UPR is contradictory to the absence of proteolytic machinery in the ER lumen and to a general notion that ER-resident proteins permanently stay in the ER lumen without being degraded unless they are damaged. BiP has been claimed to be a long-lived protein with a half-life of longer than 24 hours (24, 25). One outstanding question that remains to be addressed is how the abundance of ER-residing proteins fluctuates during ER stress.

Misfolded clients in the ER lumen are sorted by molecular chaperones such as BiP and degraded by ER-associated protein degradation (ERAD) (21). During ERAD, the clients, which typically associate with molecular chaperones such as BiP and ERdj5, are recognized by the ERAD lectins OS-9 or XTP3-B (26). The unfolded substrates are retrotranslocated to the cytosol through the translocon whose major components include Sec61 and Derlin1 (fig. S1D) (27, 28). While passing through the translocon, the polypeptides are ubiquitinated by the ER membrane–associated ubiquitin ligases, such as HRD1 and GP78 (29). Whereas terminally misfolded clients are unfolded and threaded into the translocon for ubiquitination and proteasomal degradation, those chaperones in complex with the clients are thought to maintain their folding and, thus, stay in the lumen without being codegraded by ERAD. The retrotranslocon has an inner diameter as narrow as 15 Å and, thus, does not allow the passing of BiP and other normally folded proteins with average diameters of 60 Å, at least under normal conditions (fig. S1D) (3032). The metabolic fates of chaperones in complex with ERAD substrates remain largely unclear.

Homocysteine-inducible ER stress protein (HERP), which has an estimated molecular weight of 43 kDa, but which migrates as a 54-kDa protein during SDS–polyacrylamide gel electrophoresis (SDS-PAGE), is an ER membrane–integrated protein that acts as a component of ERAD (33, 34). HERP is induced by ER stress and plays an essential role in ERAD through the interaction with HRD1, the translocon component Derlin1, p97, and the proteasome (34, 35). The deficiency of HERP accelerates ER stress–induced apoptosis and suppresses ER stress–induced inflammatory reactions (36, 37). Although primarily associated with the ER membrane, HERP has also been found in the trans-Golgi network (TGN) (38). Upon ER stress, the ubiquitin-like (UBL) domain in HERP works as a ubiquitin moiety and is recognized by the CUE domain of E3 ligase GP78 (39), and Lys48-linked polyubiquitin chains on Lys61 in HERP by E3 ligase GP78 promotes its degradation (40). Lys63-linked polyubiquitination of HERP by the E3 ligase POSH regulates the translocation of HERP from the TGN to the ER (38).

The N-end rule pathway is a proteolytic system in which a single N-terminal amino acid acts as a degron, called an N-degron, which determines the half-life of a short-lived protein (4143). In mammals, N-terminal degradation determinants include Arg, Lys, His (type-1; positively charged), Phe, Tyr, Trp, Leu, and Ile (type-2; bulky hydrophobic) (4447). The human genome encodes a set of recognition components (UBR1, UBR2, UBR4, and UBR5) called N-recognins, which can recognize type-1 and/or type-2 N-degrons to facilitate ubiquitination and proteasomal degradation of short-lived proteins carrying N-degrons (4853). These degrons can be directly exposed at the N termini when otherwise stable proteins are cleaved by endopeptidases (42). Alternatively, the degrons can be generated through posttranslational modifications of specific N-terminal residues, including Asp, Glu, Cys, Asn, and Gln (54). One such modification is N-terminal arginylation (Nt-arginylation), in which the amino acid l-Arg is conjugated to the Nt-Asp or Nt-Glu residues by ATE1-encoded R-transferases (55, 56). The substrates of Nt-arginylation include various cytosolic proteins such as a set of RGS proteins (57), the Drosophila apoptosis inhibitor 1 (DIAP1) (42), and β-actin (58). We have found that a set of ER-residing molecular chaperones, such as BiP, calreticulin, and PDI, are Nt-arginylated by ATE1 (4). Nt-arginylation of BiP is induced in response to cytosolic double-stranded DNA (dsDNA), leading to the cytosolic accumulation of Nt-arginylated BiP, R-BiP. The Nt-Arg residue of R-BiP binds p62 (also known as SQSTM1 and Sequestosome-1) and subsequently is delivered to the autophagosomes for lysosomal degradation (fig. S1, F to I) (4). The physiological functions of R-BiP in non-ER subcellular compartments remain to be further characterized.

Here, we show that BiP was a short-lived protein. BiP is Nt-arginylated in response to ER stress and proteasomal inhibition as part of protein quality control, leading to the accumulation of the Nt-arginylated form in the cytosol. The turnover of BiP is attributed to, at least in part, Nt-arginylation by ATE1. Our results suggested that the cytosolic relocalization of BiP (or as R-BiP) did not require the function of translocon and was counteracted by HERP, a component of ERAD. Finally, we showed that BiP (or R-BiP) relocated to the cytosol in part through the Golgi body. Our results together provide insight into the mechanisms underlying the constitutive and stress-induced turnover of BiP and possibly many other ER-residing proteins.

RESULTS

BiP is short-lived

Although there is a general notion that BiP and other ER-resident proteins are metabolically stable and their primary functions are confined to the ER lumen, it remains unknown how their abundance dynamically fluctuates during cellular stress responses. We performed a pulse chase analysis of the degradation rates of BiP in PC3 cells treated with the translation inhibitor cycloheximide. Time course immunoblotting showed that BiP abundance gradually decreased with an estimated half-life (t0–8, hours) of about 2 to 3 hours [Fig. 1, A (long exposure, lanes 1 to 4) and B (blue line)]. Although BiP abundance was relatively low, a similar decay pattern was observed in human embryonic kidney–293 (HEK293) [Fig. 1, C (long exposure, lane 2 compared to 1) and D (blue line)] and HeLa cells [Fig. 1, E (long exposure, lanes 2 compared to 1) and F (blue line)]. We therefore asked whether BiP was also short-lived when the UPR was induced using the ER stressor thapsigargin, a noncompetitive inhibitor of the sarco/ER Ca2+ adenosine triphosphatase (ATPase) (SERCA) (59), which perturbs Ca2+ flux in the ER. Cycloheximide degradation assays showed that transcriptionally induced BiP was substantially degraded during the UPR in HEK293 [Fig. 1, C (lanes 5–8) and D (pink line)] and HeLa cells [Fig. 1, E (lanes 5–8) and F (pink line)], but not in PC3 cells [Fig. 1, A (lanes 5–8) and B (pink line)]. In contrast to BiP, calreticulin and calnexin were apparently stable under all conditions tested (Fig. 1, A, C, and E). These results suggest that BiP is degraded in an inducible manner under ER stress, with a rate that varies on the cell line and physiological state.

Fig. 1 BiP is short-lived.

(A) Cycloheximide degradation assay. PC3 cells were cultured in the absence or presence of 1 μM thapsigargin for 16 hours. Cells were treated with cycloheximide (30 ng/ml; CHX) for the indicated periods of time and immunoblotted. Representative of three biological replicates. (B) Quantification of (A). Shown are the percentages of BiP amounts relative to GAPDH (glyceraldehyde-3-phosphate dehydrogenase) or actin amounts. (C and D) The same assays as in (A) and (B) using human embryonic kidney–293 (HEK293) cells. Representative of three biological replicates. (E and F) The same assays as in (A) and (B) using HeLa cells. Representative of three biological replicates. (G) Pulse chase degradation assay. HeLa cells were incubated with 0.2 μM thapsigargin in the absence or presence of 10 μM MG132. Newly synthesized proteins were pulsed/labeled with 35S-Met/Cys for 12 min. Cells were incubated with cycloheximide for the indicated periods of time, then subjected to immunoprecipitation and autoradiography of soluble 35S-Met/Cys–labeled BiP. Representative of three biological replicates. (H) Quantitation of (G). Shown are the percentages of BiP signals relative to that at time zero in control cells.

As cycloheximide can nonspecifically affect the ER’s functions, we measured the degradation rates of newly synthesized BiP polypeptide chains labeled with 35S-Met/Cys. Time course immunoprecipitation and autoradiography of soluble 35S-Met/Cys–labeled BiP from thapsigargin-treated HeLa cells revealed a half-life (t0–60, min) of ~45 min (Fig. 1G). The decay rate was similar even in the presence of MG132 (Fig. 1, G and H), suggesting that a portion of newly synthesized BiP molecules may be degraded independently of the ubiquitin-proteasome system. The half-life of newly synthesized BiP was relatively shorter as compared with that of the total BiP population (t0–8, ~2 to 3 hours), which can be in part attributed to cotranslational quality control before translocation into the ER lumen. Nonetheless, given that more than half of BiP molecules labeled for 12 min were degraded within an hour, we suggest that newly synthesized BiP proteins in the lumen may be degraded more rapidly as compared with the “old” or “residential” species.

A subpopulation of BiP is conjugated with the amino acid l-Arg after ER stress, independently of the UPR

We have previously shown that BiP can be Nt-arginylated in response to cytosolic dsDNA (4). To determine the role of Nt-arginylation in the turnover of BiP, we further characterized the regulation and metabolic fates of R-BiP. First, we reproduced our previous finding that R-BiP was induced upon overexpressing the native form of ATE11A7A, an R-transferase isoform that contains a pair of alternative exons 1A and 7A (Fig. 2A, lane 7 compared to lane 1). Our earlier work has shown that this isoform mediates the Nt-arginylation of BiP more efficiently than other isoforms (4). On the basis of these results, we searched for the physiological signals that induced the Nt-arginylation of BiP and other ER-residing chaperones. The treatment of various stressors on cells expressing ATE11A7A revealed ER stress and proteasomal inhibition as two major stress types that induced R-BiP (Fig. 2A, lanes 8, 9, and 12).

Fig. 2 The Nt-arginylation of BiP is induced by the accumulation of misfolded proteins.

(A) HeLa cells transfected with (plasmid) enhanced green fluorescent protein (pEGFP) or ATE11A7A were treated with 1 μM thapsigargin, tunicamycin (1 μg/μl), 2 mM dithiothreitol (DTT), 5 mM 2-deoxyglucose (2-DG), or 2 μM MG132 and were analyzed by immunoblotting. Representative of three biological replicates. (B) HeLa cells were treated with thapsigargin and/or MG132 for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates. (C) HeLa cells were transfected with pEGFP or Null Hong Kong–green fluorescent protein (NHK-GFP) plasmid and cultured for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates. (D) HeLa cells transfected with pEGFP or NHK-GFP plasmid were treated with thapsigargin and/or MG132 and subjected to immunoblotting analysis. Representative of three biological replicates. (E) Cells transfected with scrambled or small interfering RNA (siRNA) targeting BiP were treated with thapsigargin and MG132 and analyzed by immunoblotting. Representative of three biological replicates. (F) HeLa cells were transfected with a plasmid encoding GFP or ATE11A7A or treated with thapsigargin and MG132, and then immunoblotted. Representative of three biological replicates. (G) HeLa cells transiently expressing GFP or ATE11A7A were analyzed by quantitative reverse transcription polymerase chain reaction (RT-PCR) analysis. Representative of three biological replicates. (H) Quantitative RT-PCR was performed on total RNAs from HeLa cells transfected with siRNAs targeting ATE1. Representative of three biological replicates.

To test whether R-BiP was induced by ER stress, we analyzed the abundance of R-BiP in thapsigargin-treated HeLa cells over time. The induction of R-BiP was detected often, but not always, 6 hours after treatment (Fig. 2B, lane 2 compared to lane 1) to varying extents depending on cell types and physiological states. To determine whether R-BiP could be induced by the accumulation of misfolded proteins in the ER lumen, we transiently expressed Null Hong Kong (NHK), a truncated mutant form of human α1-antitrypsin that is degraded by ERAD (60). The expression of NHK increased the abundance of both BiP and R-BiP in a time-dependent manner (Fig. 2C). These results suggest that the physiological signals that induce R-BiP include the accumulation of misfolded proteins in the ER lumen.

To further characterize the possible role of R-BiP in response to proteotoxicity, we treated HeLa cells with both thapsigargin and the proteasome inhibitor MG132. This combination is expected to cause the accumulation of misfolded proteins in both the ER lumen and cytosol. Proteasome inhibition in combination with ER stress synergistically increased R-BiP (Fig. 2B, lane 12 compared to lanes 4 and 8). A similar synergistic induction of R-BiP was observed when proteasomal inhibition was combined with the overexpression of NHK in the ER lumen as compared with NHK overexpression alone (Fig. 2D, lane 4 compared to lane 3) and proteasomal inhibition alone (Fig. 2D, lane 4 compared to lane 2). Knockdown of BiP confirmed that R-BiP was derived from BiP (Fig. 2E). These results suggest that the induction of R-BiP correlates to the severity of proteotoxicity.

To determine whether Nt-arginylation occurred as part of the UPR, we examined the abundance of CHOP, PDI, and calnexin in HeLa cells transiently expressing ATE11A7A. Immunoblotting analyses showed that ATE11A7A did not alter the abundance of these proteins (Fig. 2F). A similar result was obtained at the mRNA level using reverse transcription polymerase chain reaction (RT-PCR) analyses of BiP, ATF4, XBP1, and CHOP in HeLa cells overexpressing ATE11A7A (Fig. 2G and fig. S2A). Knockdown of ATE1 also did not significantly influence the mRNA expression of BiP, CHOP, and ATF4 (Fig. 2H). As a control, we confirmed that thapsigargin strongly induced the mRNA expression of these UPR genes (fig. S2, A and C). Overall, these results suggest that Nt-arginylation occurs independently of the UPR.

Nt-arginylation mediates the cytosolic relocalization of BiP independently of the functionality of the ERAD core machinery

We have previously found that R-BiP induced by cytosolic dsDNA is mainly retrieved from the cytosol (4). To generalize this finding with other stress types such as ER stress and/or proteasomal inhibition, we performed coimmunostaining analysis of HeLa cells treated with MG132 and/or thapsigargin. Proteasomal inhibition or ER stress alone did not elicit detectible R-BiP signals. The combinational treatment strongly induced the formation of R-BiP puncta with sizes of 0.2 to 1.0 μm (Fig. 3A). In contrast to the inducible staining pattern of R-BiP, the BiP staining pattern remained largely unchanged (Fig. 3B). The use of an antibody specific to the KDEL sequence unique to ER-residing proteins (61) confirmed that the BiP staining was limited to the ER (Fig. 3B). Notably, R-BiP puncta and the KDEL-positive bulk ER signals were almost completely exclusive (Fig. 3A, enlarged views). Our results together suggest that R-BiP might not be detectable in the ER lumen because its Nt-arginylation was linked to cytosolic relocalization.

Fig. 3 ER stress and proteasomal inhibition–induced R-BiP is located in the cytosol.

(A and B) Immunostaining analysis of HeLa cells treated with 1 μM thapsigargin and 2 μM MG132. Scale bars, 10 μm. Representative of three biological replicates. (C) HeLa cells were treated with 1 μM thapsigargin and/or 2 μM MG132 and fractionated using differential centrifugation. Total (T) and cytosolic (C) fractions were analyzed using immunoblotting analysis. Representative of three biological replicates. (D) HeLa cells were treated with 1 μM thapsigargin or 2 μM MG132 and separated into cytosolic and microsomal fractions using differential centrifugation. Representative of three biological replicates.

To further characterize the subcellular localization of R-BiP, we fractionated HeLa cells treated with thapsigargin and/or MG132, which showed that a subpopulation of BiP relocated to the cytosol when cells were treated with thapsigargin and MG132 (Fig. 3C). Consistently, the majority of R-BiP was retrieved from the cytosol under these stresses (Fig. 3D). Our earlier (4) and current results suggest that Nt-arginylation might drive the cytosolic relocalization of BiP.

It remains unknown whether ER-residing chaperones in complex with terminally misfolded clients are codegraded by ERAD (fig. S1D) or stay in the ER lumen to return to the folding cycle. Given the cytosolic localization of R-BiP, we monitored BiP abundance after silencing each of the ERAD-related components (Fig. 4A). The knockdown of these ERAD components did not significantly affect the mRNA expression of total BiP, CHOP, and ATF4, although small Xbp1 mRNA species became more prominent (Fig. 4B and fig. S3A). Immunoblotting analysis showed that the knockdown of the E3 ubiquitin ligase HRD1, OS-9, XTP3-B, the AAA-ATPase p97, the E3 ubiquitin ligase adaptor Sel1L, the transcription factor CHOP, or the kinase PERK did not affect R-BiP abundance (Fig. 4C and fig. S4, A to D). This pattern remained largely unchanged even after cells were challenged with thapsigargin and/or MG132 (Fig. 4C). These results suggest that the retrotranslocation of BiP (or as R-BiP) does not require the key ubiquitin ligase HRD1 and other ERAD components essential for ubiquitination and degradation of terminally misfolded clients.

Fig. 4 The formation of R-BiP is independent from the ERAD core machinery.

(A) A diagram showing ERAD components, for which HERP acts as a scaffolding protein. (B) Quantitative RT-PCR was performed on total RNA from HeLa cells transfected with the indicated siRNAs. Representative of three biological replicates. (C) HeLa cells transfected with the indicated siRNAs were treated with 1 μM thapsigargin and analyzed by immunoblotting. Representative of three biological replicates. (D) HeLa cells transfected with the indicated siRNAs were treated with 1 μM thapsigargin and analyzed by immunoblotting. Representative of three biological replicates. (E) HeLa cells treated with thapsigargin and MG132 were incubated with brefeldin A (BFA), followed by immunoblotting. Quantification of the intensities of R-BiP bands relative to those of actin is shown. Representative of three biological replicates. (F) The same assays as in (E) using Exo 1 instead of brefeldin A. Quantification of the intensities of R-BiP bands relative to those of actin is shown. Representative of three biological replicates.

The substrates of ERAD are unfolded and subsequently retrotranslocated to the cytosol through the translocon channel that contains Sec61 or Derlin1 (Fig. 4A) (62). In cells treated with thapsigargin, R-BiP abundance was not reduced by knockdown of Sec61 alone, Derlin1 alone, or both (Fig. 4D). Under normal conditions, R-BiP abundance even increased by the knockdown of translocon components. These results suggest that the cytosolic relocalization of BiP (or R-BiP) may not require the functionality of the Sec61- or Derlin1-containing translocon. However, we do not rule out the possibility that BiP can pass through its relaxed form or a different type of translocon, which was not tested in this study (fig. S1E).

Cytosolic relocalization of R-BiP may involve the ER-Golgi secretory pathway

Brefeldin A inhibits the formation of coat protein I (COPI)–mediated transport vesicles and, thus, the transport of secretory proteins between the ER and the Golgi (63). Treatment with brefeldin A moderately reduced R-BiP abundance without markedly affecting that of BiP (Fig. 4E, lane 4 compared to lane 3). A similar decrease of R-BiP was observed with Exo 1, a chemical inhibitor of exocytosis from the ER toward the Golgi body (Fig. 4F, lane 4 compared to lane 3) (64). Although we do not exclude the possibility of off-target or secondary effects of brefeldin A and Exo 1, these results suggest that a subpopulation of BiP (or R-BiP) may be relocated to the cytosol in part through the Golgi secretory pathway (fig. S1J).

The autophagic degradation of R-BiP contributes to the turnover of BiP

We have previously shown that R-BiP induced by cytosolic dsDNA is targeted to autophagy (4). To generalize this finding with different stress types, such as proteasome inhibition in combination with ER stress, we compared the localization of R-BiP with those of autophagy markers. The immunostaining analysis showed that a portion of the R-BiP staining colocalized with cytosolic puncta positive for p62 (Fig. 5, A and B) as well as LC3 (Fig. 5, C and D). Consistent with these results, the immunoblotting analysis showed that R-BiP was co-induced with the LC3-II in HeLa cells treated with both thapsigargin and MG132 and that their induction correlated with each other in a time-dependent manner, from 6 to 9 hours after treatment (Figs. 2B and 5E). Thus, a subpopulation of cytosolic R-BiP species was targeted to and degraded by p62-mediated selective autophagy, possibly as part of a protein quality control response to proteotoxicity. The pharmaceutical inhibition of autophagy using bafilomycin A1 resulted in BiP stabilization (Fig. 5F). Moreover, R-BiP abundance was markedly increased in ATG5−/− mouse embryonic fibroblasts (MEFs) (Fig. 5G). These results suggest that BiP entered autophagic flux after being Nt-arginylated in response to ER stress associated with proteasomal inhibition. Although our results suggest that a subpopulation of cytosolic R-BiP was targeted to and degraded by autophagy, we also observed that the majority (higher than 70%) of R-BiP–positive puncta-like signals were mutually exclusive with p62 or LC3 puncta (Fig. 5, A to D), suggesting that R-BiP may have diverse metabolic fates in the cytosol.

Fig. 5 Nt-arginylation is targeted to autophagy and mediates the turnover of BiP.

(A) Immunostaining analysis of R-BiP and p62 in HeLa cells treated with thapsigargin and MG132. Scale bars, 10 μm. Representative of three biological replicates. (B) Quantitation of R-BiP punctate signals that colocalize with p62 punctate signals in (A). (C) Immunostaining analysis of R-BiP and LC3 in HeLa cells treated with thapsigargin and MG132. Scale bars, 10 μm. Representative of three biological replicates. (D) Quantitation of R-BiP punctate signals that colocalize with LC3 punctate signals in (C). (E) HeLa cells were incubated with thapsigargin and MG132 for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates. (F) HeLa cells were incubated with MG132 or bafilomycin A1, treated with cycloheximide for the indicated periods of time, and analyzed by immunoblotting. Representative of three biological replicates. (G) Wild-type (WT) and ATG5−/− MEFs were incubated with thapsigargin and MG132, treated with cycloheximide for the indicated periods of time, and analyzed by immunoblotting. Representative of three biological replicates. (H) HeLa cells transfected with pcDNA or ATE11A7A were incubated with MG132, treated for the indicated periods of time with cycloheximide, and analyzed by immunoblotting analysis. Representative of three biological replicates. (I) HeLa cells were incubated with thapsigargin and MG132, treated with cycloheximide for the indicated periods of time, and analyzed by immunoblotting. Representative of three biological replicates.

To determine the role of Nt-arginylation in the turnover of lumenal BiP, we compared the dynamic changes in BiP and R-BiP in cells treated with thapsigargin and MG132. The immunoblotting analysis showed that R-BiP was detectably induced 6 to 9 hours after treatment (Fig. 5E). (R-BiP may have important physiological functions at nondetectable amounts.) We observed that the induction of R-BiP often, but not always, occurred at the expense of bulk BiP (Fig. 2B, lane 12 compared to lane 11). Specifically, BiP abundance, perhaps as part of the UPR, peaked at around 12 hours (Fig. 5E). By contrast, R-BiP peaked as BiP returned to basal amounts (Fig. 5E). These results further support our model that Nt-arginylation mediates the turnover of BiP under cellular stresses.

To compare the metabolic stability of BiP and R-BiP, we performed cycloheximide degradation assay using HeLa cells overexpressing ATE11A7A. BiP and R-BiP abundance was reduced by 50% after ~4 and 8 hours, respectively (Fig. 5H). However, in theory, R-BiP could still be generated through the “posttranslational” conjugation of l-Arg to BiP even under translational inhibition (fig. S1F). R-BiP markedly accumulated even under translational inhibition in cells treated with thapsigargin and MG132 (Fig. 5I). To determine whether R-BiP was subjected to autophagic flux, we performed an analogous assay in wild-type and ATG5−/− MEFs. R-BiP abundance in ATG5−/− MEFs lacking autophagic flux gradually increased after translational inhibition (Fig. 5G). These results suggest that the autophagic degradation of R-BiP contributes to the turnover of BiP in the ER lumen.

The generation of R-BiP is counter-regulated by HERP

HERP is a 43-kDa protein that plays an important role in the UPR and ERAD through its interaction with various components, such as the E3 ligase HRD1, the translocon component Derlin1, p97, and the proteasome (34, 35), although the underlying molecular mechanisms are poorly understood. We noted that the knockdown of HERP resulted in an increase in R-BiP abundance under normal conditions (Fig. 4C, lane 2). By contrast, the knockdown of the ubiquitin ligase GP78, which mediates the ubiquitination of HERP (39), caused an opposite effect, namely, decreased R-BiP abundance (Fig. 4C, lane 5). These results suggest that the GP78-HERP circuit may counteract the generation of R-BiP, possibly during the process of cytosolic relocalization.

To further characterize the role of HERP in Nt-arginylation and cytosolic relocalization of BiP, we monitored HERP and R-BiP abundance in HeLa cells treated with various stressors. Notably, R-BiP and HERP were co-induced in a spatiotemporally correlated pattern under ER stress and/or proteasome inhibition (Fig. 6, A to C). Immunoblotting analysis of HeLa cells treated with thapsigargin showed that HERP was induced at 3 to 6 hours after treatment, followed by R-BiP induction at 9 to 15 hours after treatment (Fig. 6B). HERP normally has a molecular mass of 43 kDa, but when induced by thapsigargin, HERP migrated with mobility consistent with a 54-kDa protein (Fig. 6B). The induction of 54-kDa HERP was accompanied by decreased abundance in 43-kDa HERP (Fig. 6B), suggesting that 54-kDa HERP may be generated through the posttranslational conjugation of a small protein, such as ubiquitin. This trend was more obvious in cells treated with both MG132 and thapsigargin (Fig. 6D). R-BiP and 54-kDa HERP showed similar coinduction patterns in cells treated with MG132 (Fig. 6C).

Fig. 6 The Nt-arginylation of BiP and the increases in HERP amounts are co-induced under ER stress synergistically associated with proteasome inhibition.

(A) HeLa cells were treated with 1 μM thapsigargin, tunicamycin (1 μg/μl), 1 μM A23187, 1 μM MG132, 100 μM chloroquine, 200 μM H2O2, 300 μM CoCl2, or 1 μM doxycycline and analyzed by immunoblotting. Representative of three biological replicates. (B) HeLa cells were treated with 1 μM thapsigargin for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates. (C) HeLa cells were treated with 2 μM MG132 for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates. (D) HeLa cells were treated with thapsigargin and/or MG132 for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates.

Knockdown of HERP facilitated the formation of R-BiP (Fig. 7, A to D), the abundance of which remained constant under translational inhibition at the expense of that of lumenal BiP (Fig. 7, A to D, and fig. S5A). By contrast, the transient transfection of cDNA encoding 43-kDa HERP inhibited the formation of R-BiP (Fig. 7, E to H), which was associated with an increase of both 43- and 54-kDa HERP. Cycloheximide degradation assays showed that 54-kDa HERP had a half-life of 1 to 2 hours in PC3, HEK293, and HeLa cells (fig. S6, A to C). By contrast, 43-kDa HERP was apparently stable over a period of 8 hours (fig. S6, A to C). Because HERP has a transmembrane domain, we sought to rule out the possibility that overexpressed HERP exerted nonspecific effects on the ER membrane functions. Overexpression of other ER membrane–associated proteins, such as the ubiquitin ligases GP78 and HRD1 (29) and the cytosolic AAA+ ATPase protein p97 (26), did not affect the formation of R-BiP under normal conditions or under ER stress (Fig. 7I). We also noted that, in contrast to HERP, the overexpression of GP78, HRD1, and p97 even exerted an opposite effect and induced R-BiP especially in normal conditions (Fig. 7I, lanes 3 to 5), possibly as a result of ER stress. These results collectively suggest that the ERAD machinery component HERP is an inhibitor of R-BiP formation, possibly during the cytosolic relocalization of BiP.

Fig. 7 HERP counteracts the Nt-arginylation of BiP.

(A) HeLa cells were transfected with scrambled or HERP siRNA, treated with cycloheximide for the indicated periods of time, and analyzed by immunoblotting. Representative of three biological replicates. (B) Quantitation of (A). (C) HeLa cells were transfected with scrambled or HERP siRNA, treated with thapsigargin for the indicated periods of time, and analyzed by immunoblotting. Representative of three biological replicates. (D) Quantitation of (C). (E) HeLa cells transfected with pcDNA or HERP-HA were left untreated or treated with cycloheximide for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates. (F) Quantitation of (E). (G) HeLa cells transfected with pcDNA or HERP-HA plasmid were treated with thapsigargin for the indicated periods of time and analyzed by immunoblotting. Representative of three biological replicates. (H) Quantitation of (G). (I) HeLa cells transfected with the plasmid encoding HERP-HA, GP78-Flag, HRD1-Flag, or p97-His were treated with thapsigargin and analyzed by immunoblotting. Representative of three biological replicates. (J) HeLa cells were transfected with the plasmid expressing full-length HERP, ΔN114 lacking the N-terminal 114 residues, ΔN200 lacking the N-terminal 200 residues, and ΔC191 lacking the C-terminal 191 residues. Cells were treated with thapsigargin and MG132 and analyzed by immunoblotting. A nonspecific protein band is indicated by an asterisk. Representative of three biological replicates. (K) HeLa cells transfected with scrambled or HERP siRNA were treated with thapsigargin in the absence or presence of brefeldin A and analyzed by immunoblotting. Representative of three biological replicates.

HERP contains a UBL domain and resides in the ER and the TGN (38), where it plays a critical role when secretory proteins, mostly in vesicles, are sorted to the lysosome, endosome, or plasma membrane. The translocation of HERP species in the TGN to the ER depends on the POSH-mediated conjugation of Lys63-linked polyubiquitin chains to HERP (38). To determine whether the UBL domain of HERP plays a role in the cytosolic localization of BiP, we overexpressed two UBL-deficient mutants (ΔN114 and ΔN200) of HERP in which the first 114 or 200 residues were deleted, respectively. These HERP mutants lost the ability to counteract R-BiP (Fig. 7J, lanes 3 and 4 compared to lane 2), including in cells treated with thapsigargin and MG132 (Fig. 7J, lanes 8 and 9 compared to lane 7). By contrast, a C-terminal deletion mutant (ΔC191) of HERP decreased R-BiP abundance (Fig. 7J, lane 5 compared to lane 1; lane 10 compared to lane 6). These results suggest that the UBL domain of HERP may inhibit the cytosolic relocalization of BiP. Finally, we asked whether R-BiP induced by HERP knockdown involves the ER-Golgi secretory pathway. Brefeldin A reduced R-BiP abundance under normal conditions (Fig. 7K, lane 4 compared to lane 3) and under ER stress (Fig. 7K, lane 8 compared to lane 7) in cells with HERP knockdown. These results together implicate HERP as an inhibitor in the cytosolic relocalization of BiP in part through the Golgi secretory pathway.

DISCUSSION

The ER is involved in folding and posttranslational modification of newly synthesized proteins, the storage and release of Ca2+, and the synthesis of steroids and phospholipids (65). These functions are assisted by various ER-residing proteins, such as the major Ca2+-binding chaperone BiP (65). The fundamental questions in the field of proteolysis concern the mechanisms underlying the turnover of ER-residing molecular chaperones involved in protein quality control. To date, it is generally thought that they work in the lumen without being degraded as long as folding is functional. Here, we found that BiP was a short-lived protein with an estimated half-life ranging from 2 hours and to longer than 5 hours in various cell types, including in cell experiencing ER stress (Fig. 1). Our pulse chase analyses showed that newly synthesized BiP polypeptides were even more rapidly degraded with a half-life (t0–60, min) of ~45 min (Fig. 1, G and H). Although this acute degradation is likely to involve cotranslational quality control of misfolded nascent polypeptide chains that emerge from the ribosome, our results suggest that newly synthesized BiP molecules were subjected to a relatively more dynamic degradative flux. We suggest that other ER-residing proteins may also be similarly short-lived.

Various studies have independently observed that normally ER-residing proteins can be retrieved in various subcellular compartments, such as the cytosol, mitochondria, plasma membrane, nucleus, and extracellular space (4, 7, 14, 18). To date, little is known about the mechanisms that support their relocalization and physiological functions outside the ER. Our results suggest that Nt-arginylation mediates at least in part the turnover of BiP, and possibly other ER proteins as well, through cytosolic relocalization and autophagic degradation (Figs. 3 and 5). This conclusion is supported by several lines of evidence. First, the induction of R-BiP often occurred at the expense of lumenal BiP when transcriptionally induced BiP was decreased to basal amounts (Fig. 5). Second, R-BiP accumulated in the cytosol (Fig. 3), which would result in the deletion of BiP from the ER lumen. Third, R-BiP was targeted to autophagic vacuoles for lysosomal degradation (Fig. 5) (4). It has been a paradox that many ER-residing proteins are transcriptionally induced during the UPR and return to their basal amounts in the absence of proteolytic machinery in the ER lumen. Our results suggest that Nt-arginylation may be a post-UPR degradative mechanism that mediates decreases in the abundance of transcriptionally induced BiP after ER homeostasis is achieved. It remains to be determined to what degree Nt-arginylation of R-BiP and other ER proteins contributes to the turnover of lumenal BiP.

During the UPR, terminally misfolded clients are degraded by ERAD, an ER membrane–associated proteolytic system (19), although it remains unclear whether molecular chaperones, such as BiP, are codegraded with the ERAD substrates or dissociated from the substrates and stay in the lumen. Given that the retrotranslocon has an inner diameter of 15 Å (3032), normally folded chaperones are unlikely to pass through such a narrow channel without adjusting its diameter. In case of the ribosome-free translocon that has a normal internal diameter of 9 to 15 Å, its diameter can expand up to 40 to 60 Å upon binding to the ribosome (32, 66). Thus, there is a possibility that at least a subpopulation of the retrotranslocon can be relaxed, enabling the passing of normally folded chaperones for Nt-arginylation and cytosolic relocalization (fig. S1E). At the same time, our results suggest that the cytosolic relocalization and Nt-arginylation of BiP may occur through the Golgi secretory pathway possibly as a membranous vesicle (fig. S1J and Figs. 4 and 7).

Although the ER membrane–integral protein HERP interacts with ERAD components (34, 35), its function has been unclear. Our results suggest that HERP inhibits the Nt-arginylation of BiP (Fig. 7). Because HERP traffics between the TGN and ER, which is modulated by the polyubiquitination of its UBL domain (38), one would speculate that trafficking is critical to the modulation of Nt-arginylation by HERP. Our results suggested that the UBL-deficient mutants (ΔN114 and ΔN200) of HERP lost their ability to counteract R-BiP (Fig. 7J). It still remains to be further characterized whether HERP modulates the cytosolic relocalization either through a “relaxed” retrotranslocon as a free polypeptide or through the Golgi secretory pathway as a vesicular form.

MATERIALS AND METHODS

Plasmids, antibodies, and other reagents

The plasmid encoding ATE11A7A was cloned into pcDNA3.1 as described (55). The NHK–GFP (green fluorescent protein) plasmid is a gift from N. Hosokawa (Kyoto University, Japan). The plasmids expressing HERP-HA or HERP deletion mutants were provided by J. B. Yoon (Yonsei University, South Korea). Rabbit polyclonal antibodies specific to the N-terminally arginylated form of BiP, R-BiP, was raised using the peptide sequence REEEDKKEDVGC corresponding to the N-terminal region of R-BiP through a custom service in AbFrontier Inc. (Seoul, South Korea), as described previously (4). This antibody is now commercially available at AbFrontier (cat# AR05-PA0001) and Millipore (cat# ABS2103). The following antibodies were also used: FK2 (Enzo, BML-PW8810), BiP (Abcam, ab21685), KDEL (Abcam, ab12223), calreticulin (BD Biosciences, 612136), GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (Abcam, ab9485), LC3 (Sigma, L7543), calnexin (Santa Cruz, sc-23954), Sec61 (Bioworld, BS71505), Derlin1 (Bethyl Laboratories, A302-849A), GP78 (Santa Cruz, sc-166358), HRD1 (Bioworld, BS70501), OS-9 (Abcam, ab19853), XTP3B (Abcam, ab102046), p97 (Abcam, ab11433), Sel1L (Bioworld, BS71687), CHOP (Cell Signaling, #2895), PERK (Santa Cruz, sc-377400), tubulin (Santa Cruz, sc-8035), GFP (Santa Cruz, sc-8334), ATG5 (Novus Biologicals, NB110-53818), HERP (Santa Cruz, sc-100721), actin (Sigma, A1978), T7 (Millipore, AB3790), and His (Santa Cruz, sc-8036). Rabbit polyclonal mATE1 antibody was raised through a custom service at AbFrontier by expressing a 22-kDa fragment corresponding to residues 313 to 516. Secondary antibodies used include Alexa 488–conjugated goat anti-rabbit immunoglobulin G (IgG) (Invitrogen, A11008) and Alexa 555–conjugated goat anti-mouse IgG (Invitrogen, A21422). Other reagents were purchased as follows: thapsigargin (Sigma, T9033), bafilomycin A1 (Sigma, B1793), Exo 1 (Sigma, E8280), MG132 (Calbiochem, 474790), hydrogen peroxide (Sigma, H1009), CoCl2 (Sigma, 232696), cycloheximide (Sigma, C7698), brefeldin A (Santa Cruz, sc-200861), RNAiMax (Invitrogen, 13778), and Lipofectamine 2000 (Invitrogen, 11668).

Cell culture and transfection

HeLa, PC3, and HEK293 cells were obtained from the American Type Culture Collection. Wild-type and ATG5−/− MEFs were obtained from RIKEN. All the cell lines used in this study were determined to be negative in the mycoplasma test using a MycoAlert detection kit (Lonza, LT07-118). The cells were grown in Dulbecco’s modified Eagle’s medium or RPMI containing 10% fetal bovine serum (FBS) (HyClone) and penicillin/streptomycin at 37°C under 5% CO2 (v/v). Cells were transfected with the plasmids using Lipofectamine reagent (Invitrogen) according to the manufacturer’s instructions. MG132, thapsigargin, tunicamycin, A23187, chloroquine, hydrogen peroxide, CoCl2, and doxycycline were treated for 16 hours. Brefeldin A (5 μg/ml) and Exo 1 (100 μM) were treated for 30 min.

RNA interference analysis

Predesigned small interfering RNAs (siRNAs) (100 pmol) were transfected to HeLa cells using RNAiMAX reagent (Invitrogen) according to the manufacturer’s instruction. The siRNAs used in this study are as follows: siATE1 (Invitrogen, s21887, Bioneer, South Korea, 1009388), siBiP (Invitrogen, s6980), siCHOP (Invitrogen, s3995, Bioneer, 1039867), siPERK (Invitrogen, s18102, Bioneer, 1046367), siSec61 (Bioneer, South Korea, 1134675, 1134681), siDerlin1 (Bioneer, 1040911, 1040912), siGP78 (Bioneer, 1004952, 1004960), siHRD1 (Bioneer, 1147563, 1147565), siOS-9 (Bioneer, 1110757, 1110758), siXTP3-B (Bioneer, 1020842, 1020843), sip97 (Bioneer, 1161653, 1161655), siSel1L (Bioneer, 1134774, 1134776), and siHERP (Bioneer, 1068058, 1068059).

Western blot analysis

Cell lysates were prepared by sonication or boiling in radioimmunoprecipitation assay buffer (50 mM tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS) (CellNest, South Korea). Cell lysates were subjected to SDS-PAGE and transferred onto polyvinylidene difluoride membrane using a Bio-Rad semidry transfer unit (Bio-Rad). Blots were blocked with 3% (w/v) nonfat dry milk in PBS-T solution [0.1% (w/v) Tween 20 in phosphate-buffered saline]. After washing with PBS-T twice, blots were incubated with primary antibodies, followed by horseradish peroxidase–conjugated secondary antibodies. Immunoreactive bands were detected using enhanced chemiluminesence reagents (Thermo Fisher Scientific).

Cycloheximide chase assay

Cells were incubated in the absence or presence of 1 μM thapsigargin and/or 2 μM MG132 for 18 to 24 hours. Cells were washed with PBS and treated with cycloheximide (30 ng/ml) for 1 to 8 hours. Cell extracts were prepared and analyzed using immunoblotting analysis.

Immunocytochemistry

Cells were cultured on poly-l-lysine–coated slides in six-well plates and treated with 2 μM MG132 and/or 1 μM of thapsigargin for 16 hours to induce cellular stress. Cells were fixed in 4% paraformaldehyde in PBS for 0.5 hours at room temperature. After washing twice with PBS, the cells were treated with blocking solution (5% FBS and 0.3% Triton X-100 in PBS) for 1 hour followed by incubation with primary and, subsequently, secondary antibodies. For the confocal images, the LSM700 upright laser scanning confocal microscope (Zeiss) equipped with C-Apochromat 40×/NA1.2 water immersion lens was used. Images from the confocal microscope were analyzed using Zen Lite 2012 (Black edition, Version 1.1.13064.302, Zeiss).

Subcellular fractionation

Cells were pelleted and resuspended using buffer F (20 mM Hepes, pH 7.5, 10 mM KCl, 1.5 mM MgCl2, 1 mM EDTA, and 1 mM phenylmethylsulphonyl fluoride) followed by incubation on ice for 30 min in the presence of 250 mM sucrose. Cell lysates were centrifuged at 1000g for 5 min at 4°C. After centrifugation, the resulting supernatant was further centrifuged at 8000g for 10 min at 4°C. The supernatant was once again centrifuged at 100,000g for 3 hours at 4°C to separate soluble cytosol fraction from pelleted microsomal fraction.

Pulse chase analysis

HeLa cells were labeled with [35S] methionine/cysteine (35S-EXPRESS; PerkinElmer) for 12 min at 37°C, followed by a chase for the indicated time in the presence of cycloheximide. Prepared extracts were subjected to immunoprecipitation with anti-BiP antibody. Labeled proteins were performed with SDS-PAGE, autoradiography, and quantitation by ImageJ program, as described previously (55).

Quantitative RT-PCR

Total RNAs were prepared from cells by using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. After the preparation of total RNAs from the cells, cDNAs were synthesized using reverse transcriptase (Invitrogen) and amplified using PCR with a pair of primers in table S1. The target cDNA sequences were amplified using AMPIGENE qPCR Green Mix Lo-Rox (Enzo Life Sciences) in the ABI 7300 Real-time PCR System (Applied Biosystems). The amplification data were normalized to the housekeeping gene actin and analyzed using the 2−ΔΔCt method described by Livak and Schmittgen (67). Each experiment was performed in triplicate, which was repeated at least three times.

Statistical analysis

Data are means ± SD of three independent experiments. Statistical analysis was performed with Prism 6 software (Graph Pad) by analysis of variance (ANOVA) and ImageJ program. Differences with P < 0.05 were considered statistically significant (***P < 0.001, **P < 0.01, *P < 0.05).

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/11/511/eaan0630/DC1

Fig. S1. A model for the mechanism underlying the turnover and Nt-arginylation of BiP.

Fig. S2. The overexpression of ATE1 or knockdown of ATE1 does not induce the UPR.

Fig. S3. The knockdown of ERAD components does not induce BiP at the mRNA level.

Fig. S4. Validation of knockdown efficiency for ERAD components.

Fig. S5. HERP knockdown increases the abundance of R-BiP.

Fig. S6. The 54-kDa HERP protein is induced and rapidly degraded under ER stress.

Table S1. Sequences of the primers used in this study.

REFERENCES AND NOTES

Acknowledgments: We thank J. B. Yoon (Yonsei University, South Korea) for providing us with the plasmids expressing HERP-HA or HERP deletion mutants. Funding: This work was supported by the Seoul National University Nobel Laureates Invitation Program, the Basic Science Research Programs of the National Research Foundation (NRF) funded by the Ministry of Science, Information & Communication Technology (ICT), and Future Planning (MSIP) (NRF-2016R1A2B3011389 to Y.T.K. and NRF-2015R1A6A3A01020301 to S.M.S.), Seoul National University Hospital (to Y.T.K.), Research Resettlement Fund for the new faculty of Seoul National University (to Y.T.K.), the R&D Convergence Program (CAP-16-03-KRIBB) of the National Research Council of Science and Technology (NST) of the Republic of Korea and Korea Research Institute of Bioscience & Biotechnology (KRIBB) Research Initiative Program. A.C. was supported by the Dr. Miriam and Sheldon Adelson Medical Research Foundation (AMRF) and the Bio and Medical Technology Development Program (project no. 2012M3A9B6055305) through the Korean Ministry of Education, Science and Technology, Korea. A.C. is an Israel Cancer Research Fund USA professor. Author contributions: Immunoblotting analyses were performed by S.M.S., H.R.C., K.W.S., H.C.-M., and D.K. Immunofluorescence analyses were performed by S.T.K. and S.R.M. Pulse chase analysis was performed by Y.J.L. RT-PCR and quantitative RT-PCR were performed by S.M.S., H.R.C., and K.W.S. Antibody recognizing arginylated BiP was generated by J.H. Statistical analysis was performed by S.T.K., A.C., Y.T.K., and B.Y.K. provided guidance and specialized expertise. S.M.S. and Y.T.K. designed the experiments and wrote the paper. Competing interests: The authors declare that they have no competing interests.
View Abstract

Navigate This Article