© 2016. Published by The Company of Biologists Ltd | Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

RESEARCH ARTICLE The murine cytomegalovirus immunoevasin gp40 binds MHC class I molecules to retain them in the early secretory pathway Linda Janßen1, Venkat Raman Ramnarayan1, Mohamed Aboelmagd1, Maro Iliopoulou1, Zeynep Hein1, Irina Majoul2, Susanne Fritzsche1, Anne Halenius3 and Sebastian Springer1,*

ABSTRACT ERGIC or cis-Golgi, and we demonstrate that a sequence in the In the presence of the murine cytomegalovirus (mCMV) gp40 (m152) linker between the folded lumenal domain of gp40 and the , murine major histocompatibility complex (MHC) class I transmembrane sequence is required for this retention. molecules do not reach the cell surface but are retained in an early compartment of the secretory pathway. We find that gp40 does not RESULTS impair the folding or high-affinity peptide binding of the class I Gp40 retains MHC class I in the early secretory pathway molecules but binds to them, leading to their retention in the To assess the effect of gp40 on murine class I molecules, we m152 (ER), the ER-Golgi intermediate compartment expressed in K41 cells (murine fibroblasts) by lentiviral (ERGIC) and the cis-Golgi, most likely by retrieval from the cis-Golgi to transduction. The surface levels of the endogenous class I allotypes b b b b the ER. We identify a sequence in gp40 that is required for both its own H-2D (D ) and H-2K (K ) were reduced to background levels in retention in the early secretory pathway and for that of class I most cells, as observed by flow cytometry with the allotype-specific β b molecules. beta-2 microglobulin ( 2m)-dependent antibodies B22.249 (for D ) and Y3 (for Kb) (Fig. 1A). Kb, but not Db, was fully resistant to gp40 KEY WORDS: ERGIC and cis-Golgi retention, , in some cells (arrow), especially in confluent cultures, and there was Early secretory pathway, Immune evasion, Murine cytomegalovirus no evidence of intermediate retention phenotypes. We conclude that whereas gp40-mediated retention of class I molecules can be highly INTRODUCTION effective, it varies between class I allotypes and growth conditions. To escape from destruction by the cellular adaptive , To find the intracellular steady-state location of the gp40-retained inhibit almost every step of major histocompatibility class I, we performed fluorescence microscopy with the same complex (MHC)-class-I-mediated antigen presentation (Ambagala antibodies, and we also stained for gp40 with an antibody against a et al., 2005). Herpesviruses encode multiple interfering C-terminal hemagglutinin (HA) tag (Fig. 1B,C; Fig. S1). Gp40 was (immunoevasins) (Basta and Bennink, 2003). The glycoprotein observed in a compact juxtanuclear location that colocalized well 40 kDa (gp40) of the murine cytomegalovirus (mCMV), encoded with the ERGIC-53 (also known as LMAN1) marker (Fig. 1C, top by the m152 , was the first inhibitor of MHC-class-I-mediated row) and with the Golgi marker GM130 (also known as GOLGA2) antigen presentation and function to be described in (Fig. S2). Kb was found in the same location (Fig. 1B, top row, CMV (Krmpotićet al., 2002; Lu et al., 2006). Gp40 inhibits the Fig. 1C, center row; Fig. S1). In some cells (∼30%), Db localized transport of peptide-loaded class I molecules (proteins) to the cell exclusively to the same juxtanuclear region (Fig. 1B, bottom row; surface and retains them in the early secretory pathway, but its Fig. S1), but in the majority of cells, Db exhibited, in addition, a molecular mechanism of action is still unknown (Pinto and spread-out reticular pattern reminiscent of the ER (Fig. 1B, center Hill, 2005). Earlier investigations have shown that gp40-retained row, Fig. 1C, bottom row, Fig. 1D; Fig. S1). In agreement with class I molecules are bound to high-affinity peptide but fail earlier data for H-2Kd (Ziegler et al., 1997), we conclude that Db and to proceed beyond the ER-Golgi intermediate compartment Kb are indeed retained in an intracellular compartment where they (ERGIC). In contrast, gp40 itself has been shown to progress to colocalize with gp40, and we also find that the exact steady-state the for degradation. The authors found an interaction of subcellular location of a gp40-retained class I molecule depends on gp40 with calnexin but not with class I molecules (Ziegler et al., the allotype (Kavanagh et al., 2001). 2000, 1997). To understand the influence of gp40 on the kinetics of class I Gp40 also retains the class-I-related stress marker RAE-1. transport from the ER to the Golgi, we next performed radiolabeling Recently, a crystal structure of gp40 in complex with RAE-1 has and pulse-chase experiments with endoglycosidase F1 digests been published (Zhi et al., 2010). This has prompted us to re- (EndoF1; see Materials and Methods). In wild-type cells, Db and Kb examine the question of gp40–class-I interaction. We show here glycans become resistant to digestion with EndoF1 in the medial that, indeed, gp40 binds to class I molecules to retain them in Golgi due to the action of mannosidase II. When they reach the the early secretory pathway, most likely by retrieval from the trans-Golgi, further carbohydrates including sialic acids are added; this sialylation results in an upward shift of the EndoF1-resistant

1 band on SDS-PAGE gels (Fritzsche and Springer, 2013). In gp40- Department of Life Sciences and Chemistry, Jacobs University Bremen, b 28759 Bremen, Germany. 2Institute of Biology, Center for Structural and Cell Biology expressing cells, D showed no sialylation after 2 h of chase in Medicine, University of Lübeck, 23562 Lübeck, Germany. 3Institute of Virology, (Fig. 1E, bottom row); thus, it had not passed the trans-Golgi. University Medical Center, University of Freiburg, 79104 Freiburg, Germany. In contrast, Kb showed a small amount of EndoF1 resistance *Author for correspondence ([email protected]) and sialylation (Fig. 1E, second row, sia). This agrees with the above observation that in the presence of gp40, Kb can still reach the

Received 10 June 2015; Accepted 26 October 2015 surface in some cells (Fig. 1A, arrow). Journal of Cell Science

219 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

Fig. 1. See next page for legend.

Remarkably, at the 2-h time point, both Db and Kb acquired partial gp40-free control samples, where Db and Kb progressed directly (incomplete) EndoF1 resistance, visible as two (Kb) or three (Db) from the EndoF1-sensitive form to the fully EndoF1-resistant and intermediate bands between the EndoF1-resistant and -sensitive sialylated form. At longer chase times of up to 16 h, the partially forms (Fig. 1E, arrowheads), corresponding to the number of EndoF1-resistant forms of Db and Kb persisted (Fig. 1F, b glycosylation sites for each allotype. This was never seen in the arrowheads), and sialylation still did not occur (D ), or only for a Journal of Cell Science

220 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

Fig. 1. Gp40 retains MHC class I molecules. (A) Gp40 abolishes class I was also visible for a variant of gp40 with a C-terminal cytosolic surface expression. Gp40 was expressed in K41 cells, and surface expression KKSQ sequence for ER retention (Fig. S3C). Thus, in our system, at of endogenous class I was determined by cell surface staining with B22.249 b b least some gp40 is retained in a pre-medial-Golgi compartment, most (D ) and Y3 (K ) and flow cytometry. Gray shading, a second antibody control; – solid line, empty vector control; dashed line, gp40-expressing cells. One likely in an ER Golgi cycle, just like the class I molecules that it b representative experiment out of five is shown. (B) In the presence of gp40, Kb retains. Interestingly, the gp40-KKSQ variant retained both D and b and Db localize to juxtanuclear compartments, and Db additionally localizes to K very efficiently, which suggests that gp40 does not need to the ER. Gp40-expressing K41 cells were fixed, permeabilized and stained with progress beyond the cis-Golgi to be effective (Fig. S3D). FITC-conjugated anti-HA antibody (gp40), and with B22 (Db)orY3(Kb) and Cy3-conjugated secondary antibody. (C) To show ERGIC localization, cells Gp40 does not impair class I maturation or peptide binding were stained with Cy3-conjugated ERGIC-53 antiserum and either with FITC- conjugated anti-HA antibody or Y3 followed by Alexa-Fluor-488-conjugated Many viral proteins interfere with class I synthesis and folding or secondary antibody. More images are shown in Fig. S1 and are available from with peptide loading onto class I molecules, for example by the authors. Scale bars: 20 µm. (D) For quantification of the colocalization inhibiting or removing TAP (Loch and Tampe, 2005; Momburg between gp40 and Db and Kb, individual cells were examined by eye and and Hengel, 2002). Given that our results suggested that gp40 acts assigned to one of three categories, according to the colocalization of the in the early secretory pathway, we first asked whether it inhibits class I molecule with gp40: +, total colocalization with gp40, no ER localization; class I folding and peptide binding. We applied a 5-min radioactive − o, mostly colocalization with gp40, little to some ER localization; , little pulse and followed the folding of the class I heavy chains by colocalization with gp40, mostly ER localization. For the quantification, β additional microscopic images were used, which are not shown. Light gray immunoprecipitating with 2m-dependent antibodies at different b b bars, Db; dark gray bars, Kb.(E–G) Pulse-chase analysis of class I molecules chase times and quantifying total D or K . gp40 did not cause a and gp40. Cells were pulse labeled for 5 (E,G) or 10 min (F), chased for the significant change in in vivo folding kinetics for either allotype indicated times, and lysed. Proteins of interest were immunoprecipitated with (Fig. 2A). To assess the binding affinity of the overall peptide b b anti-HA antibody (gp40), B22 (D )orY3(K), digested with EndoF1 as load of Db and Kb, we then performed thermal denaturation indicated, and separated on SDS-PAGE. Sia, sialylated band; black experiments by heating cell lysates to different temperatures and arrowheads, partial EndoF1 resistance; s, EndoF1-sensitive band; asterisks, unspecific bands. One representative experiment out of three is shown. immunoprecipitating with Y3 and B22 (Gao et al., 2002; Garstka et al., 2011). There was no evidence of any impairment of peptide loading in the presence of gp40 for either allotype such as small fraction of the molecules (Kb). To determine whether the it is seen for or TAP deficiency (Fig. 2B; compare with partially EndoF1-resistant forms were still intracellular or had Fig. 3B and with figures S2D and S3H in Fritzsche et al., 2015). progressed to the cell surface, we pulse-labeled gp40-expressing Thus, in agreement with earlier reports (del Val et al., 1992; Ziegler cells and chased for 120 min. We then incubated the intact cells with et al., 2000), we find that gp40 does not impair class I folding, biotinylated Kb-specific peptide for 120 min, washed, and lysed the maturation, or peptide binding. cells, and we found that only the sialylated, but not the EndoF1- sensitive or the partially EndoF1-resistant forms of Kb, precipitated Gp40-mediated retention of class I does not use the with streptavidin-conjugated agarose (Fig. S3A). Thus, the partially peptide-loading complex EndoF1-resistant forms of class I in gp40-containing cells Class I molecules loaded with suboptimal peptide ligands are (arrowheads in Fig. 1E,F) are not cell surface forms but are retained in the early secretory pathway of wild-type cells by two trapped in the cell interior, just like the partially EndoF1-resistant different mechanisms (Springer, 2015): first, by the class-I-specific class I proteins in transporter associated with antigen processing chaperone tapasin (Paulsson et al., 2002), and second, by the lectin (TAP)-deficient cells, which circulate between the ER and the cis- calreticulin (Howe et al., 2009), probably in concert with the UDP- Golgi (Fritzsche et al., 2015). Given that at least Db does not become glucose:glycoprotein glucosyltransferase (UGT1) (Zhang et al., EndoF1-resistant at all in gp40-containing cells, we conclude that 2011). Both tapasin and calreticulin are members of the class I the compact juxtanuclear steady-state location observed for Db and peptide-loading complex (PLC). To investigate whether gp40 Kb in Fig. 1B is a pre-medial-Golgi compartment, most likely the somehow appropriates these or similar class I retention ERGIC and/or the cis-Golgi, as suggested previously (del Val et al., mechanisms that already exist in the cell, we studied the effect of 1992), and in agreement with our microscopic analysis. gp40 in cell lines that lack functional calreticulin, tapasin, TAP, As shown in earlier reports, Db and Kb were not rapidly degraded calnexin, or UGT1 (Fig. 2C). In every case, gp40 caused retention during the chase but instead were slightly more metabolically stable of both Db and Kb to the same extent as in wild-type cells. We than in wild-type cells (Fig. S3B and Ziegler et al., 2000), conclude that no single known class-I-associated protein of the early suggesting that they are not rapidly routed to lysosomes in the secretory pathway is required for gp40-mediated retention. presence of gp40. Class I molecules are released from the PLC after binding high- Taken together with the microscopy results, the results of the affinity peptide (Ortmann et al., 1994; Springer, 2015). We next pulse-chase experiments suggest that in the presence of gp40, Db asked whether this interaction is prolonged by gp40 to achieve and Kb do not become transported to the medial Golgi but are class I retention, and so we performed a pulse-chase experiment, present in the ERGIC and cis-Golgi (Db and Kb) and the ER (Db)at precipitated tapasin from the cell lysate, and analyzed the class I steady state, most likely circulating between these compartments. molecules bound to it in a re-precipitation experiment (Fig. 2D; We next wondered whether gp40 travels from the ER to the Golgi schematic in Fig. S3H). Some Db was indeed bound to tapasin in synchrony with the class I molecules it retains, and we followed it throughout the chase, but this was the case both in the presence and in a pulse-chase experiment under the same conditions as above. absence of gp40. Binding of Kb to tapasin appeared stronger in the Intriguingly, a form of gp40 that was fully EndoF1 resistant and presence of gp40, but the Kb band intensity clearly decreased over sialylated (Fig. 1G, sia) and several partially EndoF1-resistant forms, time, which suggests eventual dissociation of Kb from tapasin and not resembling those of class I molecules (arrowheads), appeared after an irreversible association. We conclude that gp40 does not use the 30 min, and both fully and partially resistant forms persisted PLC, or any of its proteins individually, to retain class I in the early throughout the 2-h chase. Partial EndoF1 resistance during the chase secretory pathway. Journal of Cell Science

221 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

Fig. 2. The peptide-loading complex is not involved in gp40-mediated retention. (A) Gp40 does not impair class I maturation. K41 cells with or without gp40 were labeled for 5 min, chased for the indicated times and lysed. Proteins were immunoprecipitated with conformation-specific antibodies B22.249 (Db)orY3(Kb), digested with PNGase, separated by SDS-PAGE and quantified. The signal intensity was normalized to the start of the chase. The mean±s.d. of three independent experiments is shown. (B) Gp40 does not impair class I peptide binding. K41 cells with or without gp40 were labeled for 5 min, chased for the indicated times and lysed. Lysates were heated to the indicated temperatures for 10 min; then, proteins were immunoprecipitated with conformation-specific antibodies B22.249 (Db)orY3(Kb), digested with PNGase, separated by SDS-PAGE and quantified. The signal intensity was normalized to signal intensity of 4°C. The mean±s.d. of two independent experiments is shown. (C) Gp40 does not use any PLC components to retain class I. Gp40-expressing murine fibroblasts with a functional deficiency in the indicated protein were surface-stained for flow cytometry with B22.249 (Db) and Y3 (Kb). Cells were analyzed by flow cytometry and results are presented as in Fig. 1A. One representative experiment out of at least two is shown. (D) Gp40 does not force permanent class I association with the PLC. K41 cells with or without gp40 were labeled for 10 min, chased for the indicated times, and lysed in 1% digitonin buffer. Tapasin was immunoprecipitated from the lysate (1st IP), and co-precipitated Db and Kb were re-precipitated (reIP) as described in the Materials and Methods and Fig. S3H, treated with PNGase, and separated by SDS-PAGE. One representative experiment out of two is shown.

b b Gp40 interacts with D and K epitopes of these antibodies, which lie in the α1/α2 superdomain of Direct binding to class I is a hallmark of many viral Db and Kb. In contrast, the monoclonal antibody 28-14-8S, which b immunoevasins (Bennett et al., 1999; Furman et al., 2002; Jones binds to the α3 domain of D , efficiently co-precipitated gp40 et al., 1996). To test whether gp40 interacts with class I molecules, (Fig. 3C). Our data thus show that gp40 interacts with class I; the we immunoprecipitated gp40–HA from lysates of radiolabeled simplest interpretation is that it binds directly to the α1/α2 cells and re-precipitated with an antiserum against a common superdomain of Db and Kb, as proposed previously for Ld (Wang sequence in the cytosolic tails of Db and Kb (Db/Kb serum). Both et al., 2012). Db and Kb co-precipitated with gp40; the interaction was also seen when we first immunoprecipitated with Db/Kb serum and then re- The class-I–gp40 interaction persists in the early secretory precipitated with anti-HA antibodies (Fig. 3A). Given that Db and pathway gp40 have an almost identical molecular mass on SDS gels, we We next hypothesized that gp40 and the class I molecules bound to it demonstrated in HeLa cells that the Db/Kb serum does not cross- might be retained as a complex in the early secretory pathway, and so react with gp40 (Fig. S3E). Gp40 also co-precipitated with we decided to investigate the EndoF1 resistance pattern of class-I- tapasin, calnexin, and calreticulin (Fig. S3F). Interaction with associated gp40 in pulse-chase experiments with re-precipitations tapasin was genuine and not a post-lysis artifact (shown in a (Fig. 4). Association of gp40 with Db and Kb was observed at the start mixed lysate experiment, Fig. S3G), but it was not required for of the chase, suggesting that it occurs shortly after synthesis of the gp40-mediated class I retention because in tapasin-deficient cells, three proteins. Those gp40 molecules that were bound to class I class I molecules still bound to gp40 (Fig. 3B) and were still (Fig. 4D) became partially EndoF1-resistant over time, but they retained (Fig. 2C). showed little sialylation (as compared to the entire cohort of gp40 at Intriguingly, first-round immunoprecipitations with the the same time point, Fig. 4C), suggesting that most did not progress monoclonal antibodies B22.249 and Y3 co-precipitated much less beyond the ER–Golgi cycle. Likewise, in the class I molecules bound b b b gp40 than the D /K serum. This suggests that gp40 might mask the to gp40, the sialylated band of K was very weak (Fig. 4B, arrow; Journal of Cell Science

222 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

Fig. 3. Gp40 interacts with Kb and Db. (A) Gp40 coprecipitates with Db and Kb. Gp40-expressing K41 cells were radiolabeled for 30 min and lysed. Proteins of interest were immunoprecipitated (1st IP) with serum against the cytosolic tail of Db and Kb or anti-HA monoclonal antibody (for HA–gp40). Immunoprecipitates were dissociated with SDS, and proteins were re-precipitated (reIP) as indicated, PNGase digested and separated by SDS-PAGE. (B) Gp40–class-I α interaction does not require tapasin. Experiment as in Fig. 3A, but with gp40-expressing tapasin-deficient mouse fibroblasts. (C) Gp40 probably binds to the 1 α b b b b b and 2 domain. As in Fig. 3A, but for the first immunoprecipitation, conformation-specific antibodies Y3 (K ), B22 (D ), 28-14-8s (D ), and D /K antiserum were used; re-precipitation was done with anti-HA monoclonal antibody or Db/Kb antiserum as indicated. Proteins were digested with PNGase and separated by SDS-PAGE. compare Fig. 4F, arrow), suggesting that gp40 associates only with the (Fig. 5B). The rapid export of gp40-(G4S)9 indicates that it passes intracellular forms of class I. The simplest explanation of these data is the cellular quality control and that the introduced does not that gp40 and class I form a complex that is retained in the ER, ERGIC lead to misfolding. Gp40-(G4S)9 no longer decreased the steady- and cis-Golgi, probably by cycling through these compartments. state surface levels of Db or Kb (Fig. 5C,D), and it bound to class I molecules much more weakly than wild-type gp40 (Fig. 5E). A sequence in the linker of gp40 is essential for gp40–class-I To find out whether this weak binding of gp40-(G4S)9 to complex retention class I molecules was the cause or the consequence of its rapid We next decided to investigate the mechanism of retention of gp40 export from the early secretory pathway, we forced it to remain in in the early secretory pathway. Given that the sequence of gp40 the ER together with class I molecules by treating the cells with contains no known retention signal, we tested a panel of mutants brefeldin A (BFA). Immunoprecipitation showed that under these (data not shown). In one such mutant, we replaced the 43-amino- conditions, it bound to class I molecules with the same strength as acid linker between the folded luminal domain and the the wild type (Fig. 5F); this suggests that its rapid export deprives transmembrane domain (Wang et al., 2012) by a (glycine4- gp40-(G4S)9 of the opportunity to bind tightly to class I molecules, serine)9 sequence to yield the gp40-(G4S)9 mutant (Fig. 5A). In a resulting in lack of class I retention. pulse-chase, gp40-(G4S)9 was much more strongly sialylated than Taken together, these experiments suggest that retention of the wild-type gp40, suggesting fast progress to the trans-Golgi, and it gp40–class-I complex in the early secretory pathway is achieved was much shorter-lived, suggesting rapid degradation in lysosomes through a sequence in the linker of gp40, and that the linker does not

Fig. 4. The gp40–class-I complex circulates through the early secretory pathway. Gp40-expressing K41 cells were labeled for 10 min, chased for the indicated times and lysed in 1% digitonin lysis buffer. Proteins of interest were immunoprecipitated (1st IP) with conformation- specific antibodies Y3 (Kb, F), B22 (Db, E), Db/Kb antiserum (A,D) or anti-HA (gp40) (B,C) antibody; after denaturation in 1% SDS, either gp40 (C,D) or Db/Kb (A,B,E,F) were re-precipitated (reIP), EndoF1 digested as indicated, and separated by SDS-PAGE. Journal of Cell Science

223 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

Fig. 5. Gp40-(G4S)9 rapidly exits the early secretory pathway and does not retain class I. (A) Schematic representation of gp40. Indicated are the lumenal domain, the linker sequence, the transmembrane domain and cytoplasmic tail (TMD/CT), the HA tag at the C terminus, and the residue numbers of the mature protein. (B) Gp40 (G4S)9 rapidly exits the early secretory pathway. K41 cells expressing gp40-(G4S)9 were pulse labeled for 5 min and chased for the indicated times, and lysed with 1% Triton X-100 buffer. Gp40 was immunoprecipitated with anti-HA antibody, digested with EndoF1 as indicated and separated by SDS-PAGE. Sia, sialylated band; black arrowhead, EndoF1-resistant band; s, EndoF1-sensitive band. One representative experiment out of three is shown. b b (C) Gp40 (G4S)9 does not retain class I. K41 cells expressing empty vector, wild-type gp40 (wt) or gp40-(G4S)9 were surface stained with B22.249 (D ) and Y3 (K ). Cells were analyzed by flow cytometry. Gray shading, second antibody control; solid line, empty vector control; dashed line, cells expressing wt gp40; long-dashed line, cells expressing gp40-(G4S)9. A representative experiment of eight is shown. (D) Gp40 (G4S)9 does not retain class I. The mean±s.e.m. of eight independent experiments performed as in Fig. 5C showing the mean fluorescence intensity (MFI) normalized to empty vector cells. (E) Gp40-(G4S)9 binds weakly to class I molecules. K41 cells expressing gp40 wt or gp40-(G4S)9 were labeled for 15 min and lysed in 1% digitonin buffer, and proteins were immunoprecipitated with Db/Kb serum or anti-HA monoclonal antibody (for gp40–HA); after denaturation in 1% SDS, gp40 or Db/Kb were reprecipitated as indicated, PNGase digested, and separated by SDS-PAGE. (F) BFA treatment increases the interaction between gp40-(G4S)9 and MHC class I molecules. Cells were treated with BFA during radiolabeling (10 min), then lysed, immunoprecipitated (1st IP) with Db/Kb antiserum, and re-precipitated (2nd IP) with anti-HA monoclonal antibody. mediate binding of gp40 to class I molecules, but it is required for cells, which might explain why gp40 does not retain human class I the retention of gp40 itself. molecules (Ziegler et al., 1997; data not shown). We think that in the cell, the class-I–gp40 complex is restricted to DISCUSSION a compartment prior to the medial Golgi because class-I-bound We demonstrate here for the first time a physical interaction of gp40 and gp40-bound class I both acquire little sialylation over time murine class I molecules with mCMV gp40 that causes their (which would signify arrival in the trans-Golgi) and do not even retention in the cell interior. From a crystal structure of gp40 with become completely EndoF1 resistant (Fig. 4). As assessed by the class-I-like protein RAE-1, Margulies and collaborators have microscopy, in the presence of gp40, Db is mostly found in the ER, b predicted that gp40 binds directly to the top of the α1/α2 whereas K is mostly in the ERGIC and cis-Golgi (Fig. 1B,C). superdomain of class I molecules (Wang et al., 2012). Our data in Given that Kb is exported from the ER with greater efficiency than Fig. 3C support this hypothesis: both the monoclonal antibodies Y3 Db (Fritzsche et al., 2015), this suggests a dynamic scenario of and B22.249, which bind to the α1/α2 superdomain of class I class I retention by retrieval of gp40–class-I complexes from the cis- molecules, preclude gp40 co-precipitation, whereas the antibody Golgi to the ER; in such an export-and-retrieval cycle, the higher ER b 28-14-8S, which binds to the α3 domain, co-precipitates gp40 exit rate of K might shift its steady-state location further towards (Allen et al., 1984; Nathenson et al., 1989). In the structural model the ERGIC and cis-Golgi. The faster anterograde transport of Kb of Wang et al. (2012), there is space for class-I-bound peptide, might also cause the escape of Kb from gp40-mediated retention that which agrees with the observation that gp40 does not inhibit peptide we observed in some cells (Fig. 1A, arrow). When Kb is forced by binding (Fig. 2A,B; del Val et al., 1992). In previous co- gp40 to circulate between ER and Golgi, its increased concentration precipitation experiments, a gp40–class-I interaction was not in the early secretory pathway might cause the tighter association found (Ziegler et al., 2000). Given that the authors of that study with the PLC that we observe (Fig. 2D). did not re-precipitate, they were unable to use the heavy chain signal Our finding of localization of gp40 in the ERGIC and cis-Golgi to detect co-precipitated class I (it appears at the same molecular at steady state agrees with previous microscopy results that show mass as gp40); thus, they depended on the β2m signal, which might that it had an excellent colocalization with p58 (also known as have been too weak to observe in their system. We have also found LMAN-1) (Ziegler et al., 1997). In that work, the steady-state that gp40 does not bind to the human class I molecules in HeLa location of gp40 shifted to the lysosomes if protein degradation Journal of Cell Science

224 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

secretory pathway, in analogy to the ‘dynamic matrix’ model (Nehls et al., 2000). So how does gp40 retain class I molecules? Our data suggest that the major fraction of gp40 is retained in the ER and associates with class I very soon after its synthesis (Fig. 4). The complex then circulates in the early secretory pathway for several hours (Figs 4 and 1F). Retention of gp40 itself in this cycle is necessary for class I retention (Figs 5 and 6). Eventually, the complex of class I and gp40 travels to the lysosomes for rapid degradation, perhaps by direct transfer from the trans-Golgi and so avoiding the cell surface. Some gp40 molecules, in contrast, are not retained in the early secretory pathway, and due to their rapid exit, they do not associate with class I molecules. In earlier published work, the large amount of overexpressed gp40 that was rapidly transported to the lysosomes might have obscured those gp40 molecules that remained in the early secretory pathway to retain class I molecules. In this work, we have demonstrated a new mechanism used by cytomegaloviruses in their multi-pronged attack on class I antigen presentation. Futureworkmightfocusonfurther exploring the intriguing trafficking path of gp40 and the factors that govern it. Gp40-mediated retention or retardation of class I might be used to study the connection between the rate of Fig. 6. Proposed mechanism for gp40-mediated MHC class I retention. In cell surface transport of class I and its binding of peptide ligands β the absence of gp40 (left), MHC class I heavy chains (dark red) bind to 2m (Praveen et al., 2010), and perhaps as a tool in a therapeutic (pink) and high-affinity peptide (blue) before being exported to the cell surface. context to manipulate class I peptide selection in autoimmunity, Once they pass the medial Golgi, they become fully EndoF1 resistant. In cancer and . gp40-expressing cells (right), fully assembled and peptide-loaded MHC class I molecules are bound in the ER by gp40 (orange), which itself is bound by an unknown retrieval factor (green) that returns the complex to the ER. Cycling in MATERIALS AND METHODS the early secretory pathway leads to a slow maturation of the glycans of both Antibodies, peptides and reagents gp40 and MHC class I and results in their partial EndoF1 resistance. Chemicals were purchased from AppliChem (Darmstadt, Germany) or Carl Roth (Karlsruhe, Germany). Mouse monoclonal hybridoma supernatants Y3 was inhibited with leupeptin (Ziegler et al., 2000); this suggests (Hammerling et al., 1982), B22.249 (Lemke et al., 1979) and HA 12CA5 that, eventually, most gp40 leaves the ER–Golgi cycle to become (Niman et al., 1983) were as described previously. Tapasin serum 2668 was kindly donated by Xiaoli Wang (Dept. of Pathology and Immunology, degraded in the lysosomes. In our experiments, when expressed Washington University School of Medicine, USA). Rabbit anti-calnexin from a retroviral vector, gp40 is rather long-lived (Fig. 1G), serum was kindly provided by David Williams (Dept. of Biochemistry, whereas in Ziegler et al. (2000), where cells are transfected with University of Toronto, Toronto, Canada). PA3-900 antiserum (Pierce an expression plasmid, the bulk of gp40 is degraded after 2 h Antibodies), monoclonal antibody against GM130 (BD Transduction (without cycloheximide) or is EndoH resistant, suggesting transit Laboratories, No. 610822), and Cy3-conjugated ERGIC-53 antiserum through the medial Golgi. We think that overexpression of gp40 (Sigma-Aldrich) were purchased. Rabbit antiserum against H-2Db and might saturate the retention mechanism that holds gp40 in the H-2Kb was generated by Charles Rivers Laboratories (Kisslegg, Germany) early secretory pathway and thus lead to the transport of excess against the peptide C-RRRNTGGKGGDYALAPGSQ corresponding to the b b gp40 to the lysosomes for degradation. Even in our expression membrane proximal C-terminal cytosolic tail of both H-2D and H-2K – system, retention of gp40 is not complete because we observe (residues 331 349, Swiss-Prot P01901); biotinylated SIINFEKL peptide was synthesized by Genecust (Luxembourg). some sialylated gp40 in the pulse-chase experiments (Fig. 1G). In class-I-deficient murine fibroblasts, this sialylated gp40 fraction was the same (data not shown), which suggests that class I Cells K41 and K42 cells (Gao et al., 2002) were kindly provided by Tim Elliott molecules do not determine the rate of exit of gp40 from the early (Faculty of Medicine and Institute for Life Sciences, University of secretory pathway. Southhampton, UK), MEF TPNd (Grandea et al., 2000) and MEF TAPd Taken together, our data and those from the literature suggest that were kindly provided by Luc van Kaer (Dept. of Pathology, Microbiology gp40 is temporarily held in the early secretory pathway by a and Immunology, Vanderbilt University School of Medicine, USA), MEF saturable retention mechanism. This retention clearly depends on an CNXd (Kraus et al., 2010) were a gift from Jody Groenendyk (Dept. of amino acid sequence in the gp40 linker (Fig. 5), but its molecular Biochemistry, University of Alberta, USA). MEF UGT1d (Solda et al., mechanism is not obvious because gp40 lacks any known retention 2007) were a gift from Tatiana Soldà (Institute for Research in Biomedicine, or retrieval motif. The linker itself might contain such a motif, Protein Folding and Quality Control, Switzerland Università della perhaps in a three-dimensional structural element; alternatively, the Svizzera Italiana, Switzerland). Cells were grown at 37°C and 5% CO2 in ’ ’ linker sequence might structurally support, or contribute to, a motif high-glucose (4.5 g/l) Dulbecco s modified Eagle s medium (DMEM) medium (GE Healthcare) supplemented with 10% FCS (Biochrom, that is elsewhere in the gp40 protein. Berlin, Germany), 2 mM glutamine, 100 U/ml penicillin and 100 µg/ml As we and others have shown, gp40 interacts with calnexin and streptomycin. For flow cytometry experiments with MEF TAP- and calreticulin, which have ER retention signals (Ziegler et al., 2000; MEF TPN-deficient cells, cells were incubated before staining at 25°C data not shown and Fig. S3F), but neither protein is required for in CO2-independent medium (Life Technologies, Darmstadt, Germany) gp40-mediated class I retention (Fig. 2C). Perhaps gp40 can bind supplemented as above to accumulate MHC class I at the cell surface alternatively to several different proteins to remain in the early (Ljunggren et al., 1990). Journal of Cell Science

225 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

Retroviral expression, microscopy and flow cytometry delVal,M.,Hengel,H.,Häcker,H.,Hartlaub,U.,Ruppert,T.,Lucin,P.and Retroviral expression, microscopy and flow cytometry were performed as in Koszinowski, U. H. (1992). Cytomegalovirus prevents antigen presentation Hein et al., 2014. by blocking the transport of peptide-loaded major histocompatibility complex class I molecules into the medial-Golgi compartment. J. Exp. Med. 176, 729-738. Pulse-chase experiments Fritzsche, S. and Springer, S. (2013). Investigating MHC class I folding and Pulse-chase experiments were performed as in Fritzsche and Springer, 2014. trafficking with pulse-chase experiments. Mol. Immunol. 55, 126-130. For Fig. S3A, 5 µM Kb-specific biotinylated peptide (ovalbumin 357-64 Fritzsche, S. and Springer, S. (2014). Pulse-chase analysis for studying protein synthesis and maturation. Curr. Protoc. Protein Sci. 78, 30.33.31-30.33.23. peptide, sequence SIINFEKbioL, with biotin attached to the lysine side chain Fritzsche, S., Abualrous, E. T., Borchert, B., Momburg, F. and Springer, S. by a 6-aminohexanoic acid linker) was added to the cells, followed by lysis (2015). Release from endoplasmic reticulum matrix proteins controls cell surface in the presence of 10 µM non-biotinylated peptide (to prevent post-lysis transport of MHC class I molecules. Traffic 16, 591-603. binding), precipitation with streptavidin–agarose and SDS-PAGE (lane 3). Furman, M. H., Dey, N., Tortorella, D. and Ploegh, H. L. (2002). The human For the other lanes, biotinylated peptide was added only after lysis (without cytomegalovirus US10 gene product delays trafficking of major histocompatibility non-biotinylated peptide) in order to precipitate all forms of Kb. complex class I molecules. J. Virol. 76, 11753-11756. – Gao, B., Adhikari, R., Howarth, M., Nakamura, K., Gold, M. C., Hill, A. B., Knee, Precipitation was with streptavidin agarose or monoclonal antibody Y3 R., Michalak, M. and Elliott, T. (2002). Assembly and antigen-presenting function and protein-A–agarose, as indicated. For Fig. 5F, cells were radiolabeled for of MHC class I molecules in cells lacking the ER chaperone calreticulin. Immunity 15 min in medium containing 10 µg/ml BFA. Afterwards, cells were lysed, 16, 99-109. and gp40–class-I complexes were precipitated as described below. Garstka, M. A., Fritzsche, S., Lenart, I., Hein, Z., Jankevicius, G., Boyle, L. H., Elliott, T., Trowsdale, J., Antoniou, A. N., Zacharias, M. et al. (2011). Tapasin dependence of major histocompatibility complex class I molecules correlates with Co-immunoprecipitation and re-immunoprecipitation their conformational flexibility. FASEB J. 25, 3989-3998. Labeling, pulse chase and co-immunoprecipitation were performed as in Grandea, A. G., Golovina, T. N., Hamilton, S. E., Sriram, V., Spies, T., Halenius et al., 2011 except that protein A agarose (Merck Millipore) was Brutkiewicz, R. R., Harty, J. T., Eisenlohr, L. C. and Van Kaer, L. (2000). used instead of protein-A–sepharose. Precipitated proteins were then eluted Impaired assembly yet normal trafficking of MHC class I molecules in Tapasin from the agarose beads by boiling in 50 µl denaturation buffer (1% SDS, mutant mice. Immunity 13, 213-222. 2 mM DTT) at 95°C for 10 min. Samples were cooled on ice, and SDS was Halenius, A., Hauka, S., Dolken, L., Stindt, J., Reinhard, H., Wiek, C., Hanenberg, H., Koszinowski, U. H., Momburg, F. and Hengel, H. (2011). neutralized with a 20-fold volume (1 ml) of 0.1% Triton X-100 in PBS. Human cytomegalovirus disrupts the major histocompatibility complex class I Samples were centrifuged at 1000 g for 10 min, and 900 µl were transferred peptide-loading complex and inhibits tapasin gene transcription. J. Virol. 85, to protein-A–agarose beads prebound to the respective antibody for the re- 3473-3485. immunoprecipitation and incubated for 1 h at 4°C rotating. The beads were Hammerling, G. J., Rusch, E., Tada, N., Kimura, S. and Hammerling, U. (1982). washed twice in PBS with 0.1% Triton X-100, and precipitated proteins Localization of allodeterminants on H-2Kb antigens determined with monoclonal antibodies and H-2 mutant mice. Proc. Natl. Acad. Sci. USA 79, 4737-4741. were eluted by boiling in 20 µl denaturation buffer at 95°C for 5 min for Hein, Z., Uchtenhagen, H., Abualrous, E. T., Saini, S. K., Janssen, L., Van SDS-PAGE. Hateren, A., Wiek, C., Hanenberg, H., Momburg, F., Achour, A. et al. (2014). Peptide-independent stabilization of MHC class I molecules breaches cellular Acknowledgements quality control. J. Cell Sci. 127, 2885-2897. We would like to thank Hartmut Hengel for advice on the project and the Howe, C., Garstka, M., Al-Balushi, M., Ghanem, E., Antoniou, A. N., Fritzsche, manuscript; Constanze Wiek and Helmut Hanenberg for the retroviral S., Jankevicius, G., Kontouli, N., Schneeweiss, C., Williams, A. et al. (2009). expression system; Peter Reinink for computational biology support; those Calreticulin-dependent recycling in the early secretory pathway mediates optimal mentioned in the Materials and Methods and Hesso Farhan for donating peptide loading of MHC class I molecules. EMBO J. 28, 3730-3744. reagents; Uschi Wellbrock for excellent technical assistance; and Ina Huppertz, Jones, T. R., Wiertz, E. J., Sun, L., Fish, K. N., Nelson, J. A. and Ploegh, H. L. Maria Bottermann, Andrei Iosif S¸mid, and Florin Tudor Ilca for preparatory and (1996). Human cytomegalovirus US3 impairs transport and maturation of major histocompatibility complex class I heavy chains. Proc. Natl. Acad. Sci. USA 93, additional laboratory work on the project. 11327-11333. Kavanagh, D. G., Gold, M. C., Wagner, M., Koszinowski, U. H. and Hill, A. B. Competing interests (2001). The multiple immune-evasion of murine cytomegalovirus are not The authors declare no competing or financial interests. redundant: m4 and m152 inhibit antigen presentation in a complementary and cooperative fashion. J. Exp. Med. 194, 967-978. Author contributions Kraus, A., Groenendyk, J., Bedard, K., Baldwin, T. A., Krause, K.-H., Dubois- L.J., V.R. and S.S. designed the experiments. L.J. and V.R. performed the Dauphin, M., Dyck, J., Rosenbaum, E. E., Korngut, L., Colley, N. J. et al. biochemical experiments. L.J., M.I., M.A. and I.M. performed the microscopy (2010). Calnexin deficiency leads to dysmyelination. J. Biol. Chem. 285, experiments. L.J., V.R., A.H., Z.H. and S.F. analysed the data. L.J., V.R. and S.S. 18928-18938. ́ ́ wrote the manuscript. Krmpotic, A., Busch, D. H., Bubic, I., Gebhardt, F., Hengel, H., Hasan, M., Scalzo, A. A., Koszinowski, U. H. and Jonjić,S.(2002). MCMV glycoprotein gp40 confers resistance to CD8+ T cells and NK cells in vivo. Nat. Immunol. Funding 3, 529-535. ̈ This work was supported by the Tonjes Vagt Foundation of Bremen [grant number Lemke, H., Hammerling, G. J. and Hammerling, U. (1979). Fine specificity XXIX to S.S.]. analysis with monoclonal antibodies of antigens controlled by the major histocompatibility complex and by the Qa/TL region in mice. Immunol. Rev. 47, Supplementary information 175-206. ̈ ́ ̈ Supplementary information available online at Ljunggren, H.-G., Stam, N. J., Ohlen, C., Neefjes, J. J., Hoglund, P., Heemels, ̈ http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.175620/-/DC1 M.-T., Bastin, J., Schumacher, T. N. M., Townsend, A., Karre, K. et al. (1990). Empty MHC class I molecules come out in the cold. Nature 346, 476-480. Loch, S. and Tampe, R. (2005). Viral evasion of the MHC class I antigen- References processing machinery. Pflugers Arch. 451, 409-417. Allen, H., Wraith, D., Pala, P., Askonas, B. and Flavell, R. A. (1984). Domain Lu, X., Pinto, A. K., Kelly, A. M., Cho, K. S. and Hill, A. B. (2006). Murine interactions of H-2 class I antigens alter cytotoxic T-cell recognition sites. Nature cytomegalovirus interference with antigen presentation contributes to the 309, 279-281. inability of CD8T cells to control virus in the salivary gland. J. Virol. 80, Ambagala, A. P. N., Solheim, J. C. and Srikumaran, S. (2005). Viral interference 4200-4202. with MHC class I antigen presentation pathway: the battle continues. Vet. Momburg, F. and Hengel, H. (2002). Corking the bottleneck: the transporter Immunol. Immunopathol. 107, 1-15. associated with antigen processing as a target for immune subversion by viruses. Basta, S. and Bennink, J. R. (2003). A survival game of hide and seek: Curr. Top. Microbiol. Immunol. 269, 57-74. cytomegaloviruses and MHC class I antigen presentation pathways. Viral Nathenson, S. G., Kesari, K., Sheil, J. M. and Ajitkumar, P. (1989). Use of Immunol. 16, 231-242. mutants to analyze regions on the H-2Kb molecule for interaction with immune Bennett, E. M., Bennink, J. R., Yewdell, J. W. and Brodsky, F. M. (1999). Cutting receptors. Cold Spring Harb. Symp. Quant. Biol. 54, 521-528. edge: adenovirus E19 has two mechanisms for affecting class I MHC expression. Nehls, S., Snapp, E. L., Cole, N. B., Zaal, K. J. M., Kenworthy, A. K., Roberts,

J. Immunol. 162, 5049-5052. T. H., Ellenberg, J., Presley, J. F., Siggia, E. and Lippincott-Schwartz, J. Journal of Cell Science

226 RESEARCH ARTICLE Journal of Cell Science (2016) 129, 219-227 doi:10.1242/jcs.175620

(2000). Dynamics and retention of misfolded proteins in native ER membranes. Springer, S. (2015). Transport and quality control of MHC class I molecules in the Nat. Cell Biol. 2, 288-295. early secretory pathway. Curr. Opin. Immunol. 34, 83-90. Niman, H. L., Houghten, R. A., Walker, L. E., Reisfeld, R. A., Wilson, I. A., Hogle, Wang, R., Natarajan, K., Revilleza, M. J. R., Boyd, L. F., Zhi, L., Zhao, H., J. M. and Lerner, R. A. (1983). Generation of protein-reactive antibodies by short Robinson, H. and Margulies, D. H. (2012). Structural basis of mouse peptides is an event of high frequency: implications for the structural basis of cytomegalovirus m152/gp40 interaction with RAE1gamma reveals a paradigm immune recognition. Proc. Natl. Acad. Sci. USA 80, 4949-4953. for MHC/MHC interaction in immune evasion. Proc. Natl. Acad. Sci. USA 109, Ortmann, B., Androlewicz, M. J. and Cresswell, P. (1994). MHC class I/beta 2- E3578-E3587. microglobulin complexes associate with TAP transporters before peptide binding. Zhang, W., Wearsch, P. A., Zhu, Y., Leonhardt, R. M. and Cresswell, P. (2011). Nature 368, 864-867. A role for UDP-glucose glycoprotein glucosyltransferase in expression and Paulsson, K. M., Kleijmeer, M. J., Griffith, J., Jevon, M., Chen, S., Anderson, quality control of MHC class I molecules. Proc. Natl. Acad. Sci. USA 108, P. O., Sjogren, H.-O., Li, S. and Wang, P. (2002). Association of tapasin and 4956-4961. COPI provides a mechanism for the retrograde transport of major Zhi, L., Mans, J., Paskow, M. J., Brown, P. H., Schuck, P., Jonjić, S., Natarajan, histocompatibility complex (MHC) class I molecules from the Golgi complex to K. and Margulies, D. H. (2010). Direct interaction of the mouse the endoplasmic reticulum. J. Biol. Chem. 277, 18266-18271. cytomegalovirus m152/gp40 immunoevasin with RAE-1 isoforms. Biochemistry Pinto, A. K. and Hill, A. B. (2005). Viral interference with antigen presentation to 49, 2443-2453. CD8+ T cells: lessons from cytomegalovirus. Viral Immunol. 18, 434-444. Ziegler, H., Thäle, R., Lucin, P., Muranyi, W., Flohr, T., Hengel, H., Farrell, H., Praveen, P. V. K., Yaneva, R., Kalbacher, H. and Springer, S. (2010). Tapasin Rawlinson, W. and Koszinowski, U. H. (1997). A mouse cytomegalovirus edits peptides on MHC class I molecules by accelerating peptide exchange. glycoprotein retains MHC class I complexes in the ERGIC/cis-Golgi Eur. J. Immunol. 40, 214-224. compartments. Immunity 6, 57-66. Solda, T., Galli, C., Kaufman, R. J. and Molinari, M. (2007). Substrate- Ziegler, H., Muranyi, W., Burgert, H.-G., Kremmer, E. and Koszinowski, U. H. specific requirements for UGT1-dependent release from calnexin. Mol. Cell 27, (2000). The luminal part of the murine cytomegalovirus glycoprotein gp40 238-249. catalyzes the retention of MHC class I molecules. EMBO J. 19, 870-881. Journal of Cell Science

227