Dependence Molecules Determines Spectrum of Tapasin Peptide

A Single Polymorphic Residue Within the Peptide-Binding Cleft of MHC Class I Molecules Determines Spectrum of Tapasin Dependence This information is current as of October 2, 2021. Boyoun Park, Sungwook Lee, Euijae Kim and Kwangseog Ahn J Immunol 2003; 170:961-968; ; doi: 10.4049/jimmunol.170.2.961

http://www.jimmunol.org/content/170/2/961 Downloaded from

References This article cites 30 articles, 11 of which you can access for free at: http://www.jimmunol.org/content/170/2/961.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on October 2, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts Errata An erratum has been published regarding this article. Please see next page or: /content/170/9/4869.2.full.pdf

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2003 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

A Single Polymorphic Residue Within the Peptide-Binding Cleft of MHC Class I Molecules Determines Spectrum of Tapasin Dependence1

Boyoun Park, Sungwook Lee, Euijae Kim, and Kwangseog Ahn2

Different HLA class I alleles display a distinctive dependence on tapasin for surface expression and Ag presentation. In this study, we show that the tapasin dependence of HLA class I alleles correlates to the nature of the amino acid residues present at the naturally polymorphic position 114. The tapasin dependence of HLA class I alleles bearing different residues at position 114 decreases in the order of acidity, with high tapasin dependence for acidic amino acids (aspartic acid and glutamic acid), moderate dependence for neutral amino acids (asparagine and glutamine), and low dependence for basic amino acids (histidine and arginine). A glutamic acid to histidine substitution at position 114 allows the otherwise tapasin-dependent HLA-B4402 alleles to load high-affinity peptides independently of tapasin and to have surface expression levels comparable to the levels seen in the presence Downloaded from of tapasin. The opposite substitution, histidine to glutamic acid at position 114, is sufficient to change the HLA-B2705 allele from the tapasin-independent to the tapasin-dependent phenotype. Furthermore, analysis of point mutants at position 114 reveals that tapasin plays a principal role in transforming the peptide-binding groove into a high-affinity, peptide-receptive conformation. The natural polymorphisms in HLA class I H chains that selectively affect tapasin-dependent peptide loading provide insights into the functional interaction of tapasin with MHC class I molecules. The Journal of Immunology, 2003, 170: 961Ð968. http://www.jimmunol.org/ fter targeting to the endoplasmic reticulum (ER),3 nas- tapasin dependence and either cellular abundance of the ligands or cent MHC class I H chains associate with a multiprotein their binding affinities for HLA-B2705 molecule (9). Taking these complex that assists their assembly with peptide and ␤ findings and the relative lack of tapasin polymorphism into ac- A 2 ␤ microglobulin ( 2m). This complex includes calnexin, calreticulin, count (10), the MHC class I allele-dependent requirement for ta- ERp57, transporter associated with Ag processing (TAP), and tapasin is likely a property of the individual MHC class I alleles. pasin (1). Tapasin is indispensable for proper function of the class Recent studies indicate that the presence of serine rather than I Ag presentation pathway. In tapasin-mutant mice, the expression phenylalanine or tyrosine at position 116 in the HLA-B H chain and stability of surface MHC class I molecules are strongly re- correlates with inefficient association with the assembly complex duced (2). Even though tapasin clearly facilitates Ag presentation and correspondingly high surface expression (11, 12). Similarly, by guest on October 2, 2021 by MHC class I molecules, its precise function in this process has substitution of glutamine to alanine at position 115 of HLA-A2 not been fully elucidated. In addition to the uncertainty regarding causes a lack of association with the loading complex but the mu- the precise function of tapasin, various MHC class I molecules tant proteins are expressed on the cell surface at a level similar to differ in their dependency on tapasin for both efficient surface ex- that of wild-type HLA-A2 H chain (13). On the basis of these pression and presentation of antigenic determinants to CTL (3Ð8). findings, positions 115 and 116 are suggested to determine the In tapasin-negative 721.220 cells, tapasin is not required for high tapasin dependence of class I surface expression. However, among HLA-B2705 allele (B2705) surface expression or for presentation HLA alleles, glutamine at position 115 is highly conserved, and of viral determinants to CTL (3). In contrast, for the HLA-B4402 position 116 alone cannot explain the phenotypic differences in allele (B4402), functional Ag presentation and surface expression tapasin dependence between B4402 and B2705 or A0301 alleles are highly dependent on tapasin (3). The factors that determine the because they are identical at position 116 with aspartic acid. More- differences in the relative tapasin dependence of various MHC over, the presence of aspartic acid at 116 resulted in either in- class I molecules remain elusive. Analysis of the amino acid se- creased (in A68) or decreased (in B7) binding of tapasin to MHC quence of some ligands that are loaded into B2705 in the presence class I molecules (12, 14). Thus, other positions in the HLA H and the absence of tapasin shows no clear correlation between chain should act as a major determinant of the tapasin dependence of MHC class I molecules. In this study, we have determined the requirements of MHC class I alleles for tapasin dependence. Graduate School of Biotechnology, Korea University, Seoul, Korea Received for publication September 9, 2002. Accepted for publication November 8, 2002. Materials and Methods The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance DNA constructs with 18 U.S.C. Section 1734 solely to indicate this fact. The cDNA encoding human tapasin was subcloned into the pcDNA3.1/ 1 This work was supported by a grant from the 21C Frontier for Functional Genome hygromycin vector (Invitrogen, Carlsbad, CA). Site-directed mutants of Analysis of Human Genome. B2705 with H to D at position 114 (B27H114D), H to N at position 114 2 Address correspondence and reprint requests to Dr. Kwangseog Ahn, Graduate (B27H114N), H to R at position 114 (B27H114R), D to Y at position 116 School of Biotechnology, Korea University, 1, 5-Ga, Anam-Dong, Sungbuk-Gu, (B27D116Y), and the other substitution mutants of B4402 were made by Seoul 136-701, Korea. E-mail address: [email protected] changing the codons by PCR with Pfu DNA polymerase (Stratagene, La 3 ␤ ␤ Abbreviations used in this paper: ER, endoplasmic reticulum; 2m, 2 microglobu- Jolla, CA). All HLA cDNAs and their mutagenized derivatives were in- lin; TAP, transporter associated with Ag processing; endo H, endoglycosidase H. serted into the mammalian expression vector pcDNA3.1 (Invitrogen).

Stable cell lines and Abs cross-linked proteins were immunoprecipitated with mAb HLA-B Ab-1. Precipitates were separated by 12% SDS-PAGE and transferred to an im- Tapasin expression was restored in 721.220 cells by transfection with the mobilon-P membrane (Millipore, Bedford, MA). The membrane was in- cDNA-encoding tapasin, and transfectants were selected with 0.2 mg/ml cubated with HRP-conjugated streptavidin for1h4¡C, and biotinylated hygromycin, giving rise to the 721.220.Tpn cell line. The HLA-G-specific proteins were visualized by using ECL Western blotting reagent (Pierce). mAb G233 was a gift from Dr. Y. Loke (University of Cambridge, Cam- Peptide translocation was determined after incubating microsomes with bridge, U.K.). Polyclonal rabbit K455 Ab reacts with MHC class I H chains biotin-conjugated reporter peptides in the absence of competitor for 30 min ␤ and 2m in both assembled and nonassembled forms (15). The mouse mAb at 26¡C with or without 1 mM ATP. Microsomal membranes were recov- HLA-B Ab-1 (NeoMarkers, Fremont, CA) recognizes only HLA-B alleles ered by centrifugation at 75,000 ϫ g for 10 min through a 0.5 M sucrose ␤ and mAb W6/32 recognizes only the complex of H chain and 2m. The cushion in cold RM buffer. After washing with cold RM buffer twice, the anti-tapasin Ab (R.gp48N) was a gift from Dr. P. Cresswell (Yale Univer- membrane pellet was directly dissolved in sample buffer. The samples were sity, New Haven, CT). FITC-conjugated goat anti-mouse IgG Abs and analyzed on tricine/SDS-PAGE, appropriate for resolution of low mass HRP-conjugated streptavidin were purchased from Jackson ImmunoRe- polypeptides as described (18), and probed with HRP-conjugated strepta- search Laboratories (West Grove, PA) and Pierce (Rockford, IL), vidin. The relative densities of the peptide bands were determined by use respectively. of an imaging densitometer (GS-700; Bio-Rad, Richmond, CA) and Multi- Pulse-chase and immunoprecipitation Analyst densitometer software (Bio-Rad). Cells (5 ϫ 106) were transfected by electroporation (Gene Therapy Sys- tems, San Diego, CA), starved for 40 min in medium lacking methionine, Results labeled with 0.1 mCi/ml [35S]methionine (TranS-label; NEN, Boston, MA) Variable dependence of HLA class I molecules on tapasin for for 20 min, and chased in normal medium for the indicated times. Cells their cell surface expression were lysed by using 1% Nonidet P-40 (Sigma-Aldrich, St. Louis, MO) in PBS with a protease inhibitor mixture for 30 min at 4¡C. After preclearing Among different HLA class I alleles, dependence on tapasin for cell lysates with protein G-Sepharose (Amersham Pharmacia Biotech, Pis- class I expression is variable. B4402, and to a lesser extent B0801, Downloaded from cataway, NJ), primary Abs and protein G-Sepharose were added to the are tapasin-dependent for cell surface expression, whereas B2705 supernatant and incubated at 4¡C with rotation for 2 h. The beads were is tapasin-independent (3). To evaluate whether these allele-spe- washed three times with 0.1% Nonidet P-40 in PBS. Proteins were eluted from the beads by boiling in SDS sample buffer and separated by 12% cific differences in tapasin dependence could be further extended to SDS-PAGE. The gels were dried, exposed to BAS film for 14 h, and then other MHC class I molecules, we examined the cell surface ex- analyzed with Phosphor Imaging System BAS-2500 (Fuji Film, Tokyo, pression of 10 HLA class I alleles in 721.220 and 721.220.Tpn

Japan). For endoglycosidase H (endo-H) treatment, immunoprecipitates cells. In the absence of tapasin, a spectrum of surface expression of http://www.jimmunol.org/ ␮ were digested with 3 of endo H (Roche, Indianapolis, IN) at 37¡C over- MHC class I alleles was consistently evident in pools of trans- night in 50 mM sodium acetate (pH 5.6), 0.3% SDS, and 150 mM 2-ME (Sigma-Aldrich). fected cells (Fig. 1a). High levels of surface expression were observed for B2705, B2702, A0210, A2401, B0801, and B5401 al- Flow cytometry leles. In contrast to these alleles, Ͼ5-fold lower surface expression The surface expression of MHC class I molecules was determined by ﬂow was observed for B4402, B3501, A3001, and, interestingly, cytometry (FACSCalibur; BD Biosciences, Mountain View, CA). Cells HLA-G Ag, a nonclassical MHC class I molecule (Fig. 1a, middle ϫ 6 (1 10 ) were washed twice with cold PBS containing 1% BSA and column). Signiﬁcantly, expression of tapasin fully restored the sur- incubated for1hat4¡C with a saturating concentration of primary Ab. Normal mouse IgG Ab was used as a negative control for each test. Cells face expression of B4402, B3501, A3001, and HLA-G Ag (Fig. by guest on October 2, 2021 were washed twice with cold PBS containing 1% BSA and then stained 1a, middle column). Surface levels of B0801 and B5401 were also with FITC-conjugated goat anti-mouse IgG Abs for 30 min. A total of increased, albeit to a lesser extent, upon expression of tapasin (Fig. 1a, 10,000 gated events were collected by the FACSCalibur cytometer and right column). However, expression of tapasin did not affect the sur- analyzed with CellQuest software (BD Biosciences). face levels of B2705, B2702, A0210, and A2401 (Fig. 1a, left col- Coimmunoprecipitation and Western blot analysis umn). These results indicate that the degree to which class I surface Cells were lysed in 1% digitonin in buffer containing 25 mM HEPES, 100 expression is affected by tapasin is indeed allele-dependent.

mM NaCl, 10 mM CaCl2, and 5 mM MgCl2 (pH 7.6) supplemented with protease inhibitors. Lysates were precleared with protein G-Sepharose Residue 114 determines spectrum of tapasin dependence of HLA (Amersham Pharmacia Biotech) for1hat4¡C. For immunoprecipitation, class I molecules samples were incubated with the appropriate Abs for2hat4¡C before protein G-Sepharose beads were added. Beads were washed four times As tapasin shows little polymorphism (10), variable dependence with 0.1% digitonin, and bound proteins were eluted by boiling in SDS on tapasin should be explained by amino acid differences among sample buffer. Proteins were separated by 12% SDS-PAGE, transferred the respective class I alleles. These differences are mainly located onto a nitrocellulose membrane, blocked with 5% skim milk in PBS with ␣ ␣ 0.1% Tween 20 for 2 h, and probed with the appropriate Abs for 4 h. in the 1 2 domain of MHC class I molecules. To define the Membranes were washed three times with PBS with 0.1% Tween 20 and amino acids potentially involved in determining tapasin depen- incubated with HRP-conjugated streptavidin (Pierce) for1hat4¡C. Im- dence, we, therefore, aligned the sequences of the ␣1␣2 domain of munoblots were visualized with ECL detection reagent (Pierce). HLA class I molecules (Fig. 1b). Interestingly, tapasin-indepen- Microsomes and peptide-loading assay dent HLA class I alleles contain histidine or arginine at position 114, whereas tapasin-dependent alleles contain acidic amino acid Microsomes from 721.220 and 721.220.Tpn cells expressing B2705, B4402, or their mutant HLA class I H chains, with or without tapasin, were residues, such as aspartic acid or glutamic acid. Alleles that are prepared and purified as previously described (16). Biotinylated peptides weakly dependent on tapasin contain neutrally charged asparagine GRIDKPILK (ligand for B2705) and AEIDKVTGY (ligand for B4402) at this position. Database analysis reveals that position 114 of all were conjugated to the photoreactive cross-linker N-5-azido-2-nitroben- HLA class I molecules is represented by one of the following six zoyloxysuccinimide (ANB-NOS; Pierce) as described (17). For the pep- amino acids: histidine, arginine (basic), aspartic acid, glutamic tide-loading assay, reporter peptides, with or without various concentrations of the unlabeled peptides, were mixed with 15 ␮l of microsomes acid (acidic), and asparagine, glutamine (neutral). ␮ (concentration of 60 A280/ml) in a total volume of 50 l of RM buffer (250 To test whether the side chain of the amino acid at position 114 mM sucrose, 50 mM triethanolamine-HCl, 50 mM potassium acetate, 2 governs the extent of tapasin dependence of HLA class I alleles, mM magnesium acetate, 1 mM DTT, and 10 mM ATP). The mixture was we constructed substitution mutants and examined their depen- incubated for 30 min at 26¡Cinaflat-bottom 96-well tissue culture plate. Samples were maintained on ice during a 3-min exposure to shortwave dence on tapasin for surface expression. Surprisingly, B44D114H (365 nm) ultraviolet irradiation. After centrifugation, membranes were was no longer tapasin-dependent and instead displayed a tapasin- washed once with cold RM buffer and lysed with 1% digitonin, and the independent phenotype for their surface expression (Fig. 2a). The The Journal of Immunology 963

FIGURE 1. Determinant for tapasin dependence of HLA class I molecules. 721.220 and 721.220.Tpn cells were transfected with DNA encoding the indicated HLA class I H chains. a, Surface expression level of MHC class I molecules was measured by FACS analysis. FACS histograms are shown for cells without tapasin (thin lines) and with tapasin (thick Downloaded from lines). Staining of mock-transfected cells is shown in the ﬁlled histograms (control). b, Amino acid comparison of tapasin-independent and -dependent HLA class I alleles. http://www.jimmunol.org/ by guest on October 2, 2021

B44D114N and B27H114N, in which residue 114 was replaced by of MHC class I molecules by examining their acquisition of the neutral-charged asparagine, fell between the B4402 and B2705 endo-H resistance. In the absence of tapasin, the majority of B4402 in the spectrum of tapasin dependence. The reciprocal mutation, H chains remained sensitive to endo-H digestion, even after 90 min B27H114D, rendered the otherwise tapasin-independent B2705 al- (Fig. 3a, upper panel), reflecting complete retention of these mol- lele dependent on tapasin for surface expression. As arginine, a ecules in the ER over this period. This finding suggests that the basic amino acid, is also found in some MHC class I alleles, such impaired intracellular transport of B4402 molecules in the absence as A0101, A0301, A1101, or A2902, we substituted arginine for of tapasin accounts for their lower levels of surface expression. histidine in B2705 (B27H114R) and examined its dependence on Conversely, in the presence of tapasin, B4402 molecules had be- tapasin for surface expression. In the absence of tapasin, the sur- come endo H-resistant by this time, consistent with their efficient face expression of B27H114R was comparable to the level ob- expression at the cell surface (Fig. 3a, lower panel). Pulse-chase served for wild-type B2705 and was not affected by expression of experiments from tapasin-negative and -positive cells revealed tapasin (Fig. 2b), indicating that arginine can functionally replace comparable acquisition of endo-H resistance by B44D114H (Fig. histidine at position 114 in respect to conferring tapasin indepen- 3b), indicating its tapasin independence for normal intracellular dence on MHC class I alleles. These results indicate that the ta- transport. Without tapasin, the maturation kinetics of B2705 and pasin dependency of HLA alleles is determined by the nature of B27H114D were essentially the opposite of the maturation kinetics the amino acid at position 114. of B4402 and B44D114H, respectively. After a 90-min chase, B2705 had become resistant to endo H irrespective of tapasin, Differential effect of tapasin on maturation of HLA class I indicating its transport out of the ER (Fig. 3c). At the 90-min alleles is exerted by residue 114 chase, B27H114D remained sensitive to endo H in the absence of To investigate the mechanism by which the amino acid at position human tapasin but had become resistant to endo H in its presence 114 influences dependence of MHC class I molecules on tapasin (Fig. 3d). In the absence of tapasin, B44D114N, in which aspartic for surface expression, we compared the intracellular maturation acid is replaced with the neutrally charged asparagine at position and transport of wild-type HLA class I alleles with their respective 114 of B4402, was efficiently transported out of the ER (Fig. 3e, point mutants in the absence or presence of tapasin. We used upper panel) at a rate that was similar, albeit lower, to the rate pulse-chase analysis to estimate the extent of intracellular transport observed for B44D114H. Tapasin slightly promoted the transport 964 DETERMINANT FOR TAPASIN DEPENDENCE OF MHC CLASS I MOLECULES

FIGURE 2. Residue 114 governs a spectrum of tapasin dependence for efficient surface expression. 721.220 and 721.220.Tpn cells were transfected with DNA encoding either B4402, B44D114H, or B44D114N (a), and B2705, B27H114D, B27H114N, or B27H114R H chains (b). Cells were stained 48 h posttransfection with the mAb HLA-B Ab-1 followed by FACS analysis. Tapasin-positive (thick lines), tapasin-negative (thin lines), and mock-transfected 721.220 cells (dotted lines) are shown. Downloaded from http://www.jimmunol.org/ of B44D114N (Fig. 3e, lower panel). The intracellular maturation reports (12, 14), these data confirm the importance of position 116 and transport results correlate with the cell surface expression of for chaperone association and indicate that position 116 indeed MHC class I molecules. From these experiments, we have estab- affects the association with tapasin. Taken together, our findings lished that the amino acid residue at position 114 governs a spec- indicate that position 114 influences tapasin association and that trum of dependence on tapasin for biochemical maturation and the residue 114-mediated tapasin dependence of HLA class I al- surface expression. leles correlates with the strength of tapasin interaction.

Residue 114 influences the association of HLA class I alleles Position 114 determines the peptide-loading characteristics of with tapasin HLA class I molecules by guest on October 2, 2021 To further understand the molecular basis for the critical role of In the absence of tapasin, the impaired intracellular transport and, residue 114 in defining the tapasin dependence of HLA class I consequently, the low surface expression of B4402 and molecules, we examined by coimmunoprecipitation the interaction B27H114D might be due to their inability to load peptides. To of wild-type HLA and its mutant derivatives with tapasin in the load- determine the influence of residue 114 on peptide loading as a ing complex. Remarkably, B44D114H, in comparison with wild-type function of tapasin, we examined peptide loading into MHC class B4402, showed an 8-fold lower affinity for tapasin (Fig. 4a, lanes 1 I molecules by using the reporter peptide. We incubated intact and 2), suggesting that introduction of a basic amino acid residue at microsomes derived from B4402 and B44D114H transfectants position 114 disrupts its interaction with tapasin. B44D114N exhib- with or without tapasin and AEIDKVTGY peptide, a high-affinity ited weaker association with tapasin than did B4402 but exhibited ligand for B4402 (19), in the presence of the competitors at dif- stronger association with tapasin than did B44D114H (Fig. 4a, ferent concentrations. The amount of tapasin-dependent peptide lane 3). In contrast, B27H114D had an affinity for tapasin that was loading was determined by measuring the level of incorporation of even higher than the tapasin affinity of the wild-type B2705 in this labeled peptide into W6/32-reactive class I complexes. We were and repeating experiments (Fig. 4b, lanes 1 and 2); B27H114N fell surprised at the marked differences in the ability of B4402 and between B27H114D and B2705 in tapasin affinity. To confirm that B44D114H to load peptide in the absence of tapasin. When no the observed differential binding of HLA class I molecules to ta- competitor was added, the level of binding of the reporter peptide pasin was not due to differences in sample loading, we reacted the by B4402 was only 25% of the binding observed in B44D114H same blots with the K455 Ab and verified that each lane contained (Fig. 5a, lane 0). Furthermore, the reporter peptides loaded into a comparable amount of MHC class I molecules (Fig. 4, a and b, B4402 were completely out-competed by a lower concentration lower panels). Several studies have shown that positions 115 and (1.6 ␮M) of unlabeled peptide, whereas as much as 12.8 ␮Mof 116 affect association of HLA alleles with chaperones and that the unlabeled peptide was not even sufficient for completely out-com- interaction of chaperones with HLA class I H chains is cooperative peting reporter peptide binding to B44D114H. However, in the (11Ð14). We wanted to confirm the effect of residue 116 on the presence of tapasin, no discernible difference in the peptide-bind- association of HLA class I H chains with chaperones and thereby ing ability of B4402 and B44D114H was seen (Fig. 5b). Analysis validate our coprecipitation system. Substitution of aspartic acid to of B2705 and B27H114D as to tapasin dependence for peptide tyrosine at position 116 caused disparate binding of tapasin by loading revealed that wild-type B2705 resembles B44D114H, B2705 and B4402. In the case of 2705, the substitution whereas B27H114D behaves like wild-type B4402. In the absence (B27D116Y) significantly decreased tapasin (Fig. 4b, lane 4), of tapasin, B2705 efficiently binds the reporter peptide, GRIDK whereas the B44D116Y mutant had a stronger association with PILK, a high-affinity ligand for B2705 (20). A high concentration tapasin (Fig. 4a, lane 4). In general agreement with the previous (12.8 ␮M) of unlabeled peptide was not sufficient to out-compete The Journal of Immunology 965 Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 3. Tapasin-dependent effect of residue 114 on intracellular transport of MHC class I molecules. 721.220 (tapasin-negative) and 721.220.Tpn (tapasin-positive) cells were transfected with B4402 (a), B44D114H (b), B2705 (c), B27H114D (d), or B44D114N cDNAs (e). Cells were labeled and chased for the indicated times. Cell lysates were immunoprecipitated with mAb HLA-B Ab-1, and then digested with endo H (ϩ) or mock digested (Ϫ) for 16 h. Endo-H-resistant (r) and endo-H-sensitive (s) proteins bands are indicated.

the reporter peptide binding (Fig. 5c, upper panel). In contrast, and B27H114D in the absence of tapasin are likely the result of substitution of aspartic acid for histidine at position 114 of B2705 their inability for peptide loading. dramatically abolished its ability to load peptides (Fig. 5c, lower To further investigate whether the differential peptide loading panel). However, expression of tapasin greatly enhanced the abil- observed in the absence of tapasin might be due to differences in ity of B27H114D to load peptides to the same level as observed for the luminal availability of the peptide, we quantitated the amount wild-type B2705 (Fig. 5d, lower and upper panels, respectively). of reporter peptides that were translocated into the ER lumen by These results indicate that B4402 molecules are highly dependent UV irradiation time point. Comparable amounts of peptides were on tapasin for efﬁcient peptide loading, but the aspartic acid-to- recovered between B2705 and B27H114D (Fig. 5e, lanes 1 and 2) histidine substitution at position 114 renders the molecules inde- and between B4402 and B44D114H (Fig. 5e, lanes 3 and 4). pendent of tapasin for peptide loading. Accordingly, the impaired Therefore, we exclude the notion that luminal availability of the intracellular transport and reduced surface expression of B4402 peptide plays an important role in dictating class I loading in 966 DETERMINANT FOR TAPASIN DEPENDENCE OF MHC CLASS I MOLECULES

FIGURE 4. Residue 114 and the association of HLA class I molecules with tapasin. 721.220.Tpn cells were transfected with DNA encoding the indicated HLA class I H chains. Cells were lysed by 1% digitonin, and lysates were immunoprecipitated with mAb HLA-B Ab-1. The immunoprecipitated proteins were separated by SDS-PAGE and blotted with anti- tapasin Ab. The intensities of precipitated bands were quantiﬁed. The results of several indepen-

dent experiments are shown as mean values and Downloaded from SD. The relative intensities of coprecipitated tapasin with mutants B4402 and B2705 were compared with those with wild-type B4402 and B2705 (100% control), respectively. http://www.jimmunol.org/ by guest on October 2, 2021 tapasin-deficient cells. In control experiments in which ATP was seemingly forms the open state to fit the appropriate peptides, re- omitted, little peptide was recovered (Fig. 5e, lanes 5 and 6). Be- gardless of tapasin. Accordingly, these HLA class I alleles are cause peptide translocation via TAP requires ATP (21), we con- capable of binding a broad spectrum of the peptide repertoire with- clude that the modified reporter peptides entered the ER lumen in out the assistance of tapasin and are subsequently transported to a TAP-dependent manner. Overall, our data clearly show that the cell surface. In corroboration with this concept, HLA class I functional polymorphism in tapasin dependence for efficient pep- alleles with a neutral-charged amino acid display an intermediate tide loading of MHC class I molecules is defined by the presence level of tapasin dependence for surface expression (3). These find- of different amino acids at position 114. ings intimate that the dependence upon tapasin for loading a repertoire of peptides of sufficient stability to allow egress from the Discussion ER might reside in the ability of the D pocket to facilitate the Based on the available class I crystal structures, some consider- loading of peptides or perhaps to initiate a peptide-induced con- ations on the peculiar behavior of position 114, an integral part of formational change in the class I molecule that signals release from the D pocket of the peptide-binding groove, may be discussed. The the loading complex. We predict that the tapasin rule might also be general structure of different MHC class I molecules is similar, imposed upon other nonclassical MHC class I molecules, such as diverging only at the polymorphic residues, which are relatively HLA-E, HLA-F, as well as the unexamined other classical MHC few and located mostly in the peptide-binding groove (22). A class I molecules. slight structural variation affecting this region can dramatically Recently, Williams et al. (23) observed that unlike tapasin-de- change the requirements for binding peptides and might result in a pendent B4402, B4405 (tyrosine at position 116), which was as- different set of bound peptides. Our findings reveal that tapasin dependence and the nature of the side chain of amino acid residue sumed to be identical to B4402 (aspartic acid at position 116) apart 114 are clearly correlated (the “tapasin-rule”). HLA class I alleles from the polymorphism at position 116, is tapasin-independent for with acidic amino acids at position 114 associate strongly with the cell surface expression. This observation led them to conclude tapasin and are dependent on tapasin for high-affinity peptide load- that amino acid residue 116 determines the tapasin dependence of ing and surface expression. The negatively charged side chain of cell surface expression. This result contrasts with our observations aspartic acid or glutamic acid might pose steric hindrances in the with B44D116Y. We found that the single amino acid change of peptide-binding groove that would interfere with the strong bind- tyrosine for aspartic acid at position 116 does not change the phe- ing of peptides. Upon interaction with tapasin, the peptide-binding notype of B4402 for tapasin dependence (data not shown) and that cleft transits to the open, peptide-receptive form. In contrast to B4402 and B2705, which share the same amino acid residue at tapasin-dependent HLA class I alleles, the peptide-binding groove position 116, exhibit different phenotypes for tapasin dependence of MHC class I molecules containing basic residues at position 114 in their surface expression (Fig. 1). Likewise, HLA-G and A2401 The Journal of Immunology 967

association of certain MHC class I molecules with the components of the loading complex, albeit, in unpredictable manner. Several studies have shown that TAP association of MHC class I molecules requires tapasin (2, 25Ð27), as expected if tapasin bridges MHC class I molecules to TAP. Therefore, a spectrum of residue 114-determined dependence of HLA class I alleles on tapasin might be an indirect effect of the residue on TAP-related function. However, our data argue against this possibility. First, residue 114 appears to have no effect on the function of TAP in translocating peptides into the ER lumen (Fig. 5e). Second, as loading of peptides onto B44D114H and B2705 proceeds normally in the absence of tapasin (Fig. 5, a and c), bridging of MHC class I molecules to TAP, which is one of the proposed functions of tapasin, appears to be unnecessary for enhanced peptide loading, which is consistent with a previous report (3). What is the clinical relevance of our findings on tapasin dependence of MHC class I alleles? HLA alleles that encode tapasin- independent molecules might have evolved in response to evolu- tionary pressures exerted by microbial pathogens. Adenovirus protein E19 targets tapasin to prevent Ag presentation and in turn Downloaded from to evade CTL recognition (28). The flexibility offered by tapasin independence might permit certain MHC class I molecules to con- tinue presenting viral Ags even when tapasin function is disrupted. Peptide loading that is independent of tapasin, although beneficial during infection with such microorganisms, might also result in the up-regulated presentation of poorly tolerized self-peptides in an http://www.jimmunol.org/ inflammatory context. Such up-regulation might be a novel mechanism for triggering autoimmune responses. This hypothesis might be particularly relevant to the development of inflammatory spon- dyloarthritis, known to be strongly associated with the “tapasin- independent” HLA-B27 alleles (29). Differences in the residues at positions 114 and 116 are likely to contribute to allele-specific recognition by different CTL clonotypes, changing the requirements for binding peptides and modifying the set of bound pep- by guest on October 2, 2021 tides. This might provide a possible molecular mechanism by which a molecular mismatch involving positions 114 and 116 of MHC class I H chain influences the outcome of unrelated bone marrow transplantation in donor-recipient pairs, otherwise matched for HLA-DRB and -DQ loci; patients with substitution at positions 114 and 116 are at increased risk for transplant-related mortality (30). The description of the natural polymorphisms in FIGURE 5. Residue 114 and the tapasin dependence of peptide loading. HLA class I H chains that selectively affect tapasin-dependent pep- Microsomal membranes were derived from 721.220 (a and c) and tide loading provides insights into the functional interaction of 721.220.Tpn (b and d) that had been transfected with B4402 (a and b), tapasin with MHC class I molecules and indicates that manipula- B44D114H (a and b), B2705 (c and d), or B27H114D cDNAs (c and d). Peptide loading (aÐd) and peptide transport (e) were assayed as described tion of class I sequences can affect MHC class I maturation and in Materials and Methods. The peptide transport assay was performed us- cell surface expression. ing the microsomal membrane derived from 721.220 cells. The relative amount of translocated peptides was expressed as a percentage of the stron- Acknowledgments gest band (B4402) to yield the graph shown at the bottom. A representative We thank Drs. P. Cresswell and Y. Loke for providing materials. immunoblot and the results of the densitometry are displayed. References 1. Cresswell, P., N. Bangia, T. Dick, and G. Diedrich. 1999. The nature of the MHC class I peptide loading complex. Immunol. Rev. 172:21. 2. Grandea, A. G., III, T. N. Golovina, S. E. Hamilton, V. Sriram, T. Spies, have tyrosine at position 116 but they display the opposite phe- R. R. Brutkiewicz, J. T. Harty, L. C. Eisenlohr, and L. Van Kaer. 2000. Impaired assembly yet normal trafficking of MHC class I molecules in Tapasin mutant notype for tapasin dependence (Fig. 1). In fact, the structures of mice. Immunity 13:213. B4405 and B4402 are quite different. In comparison with B4402, 3. Peh, C. A., S. R. Burrows, M. Barnden, R. Khanna, P. Cresswell, D. J. Moss, and B4405 is devoid of the N-terminal 48 amino acids and the entire J. McCluskey. 1998. HLA-B27-restricted antigen presentation in the absence of ␣ tapasin reveals polymorphism in mechanisms of HLA class I peptide loading. 3 region, including the transmembrane and cytoplasmic domains Immunity 8:531. (24). It is intriguing how Williams et al. (23) could detect the 4. Grandea, A. G., 3rd, P. J. Lehner, P. Cresswell, and T. Spies. 1997. Regulation putative “soluble” B4405 proteins at the cell surface by FACS of MHC class I heterodimer stability and interaction with TAP by tapasin. Im- munogenetics 46:477. analysis. On the basis of our findings and the findings of others, we 5. Lewis, J. W., and T. Elliott. 1998. Evidence for successive peptide binding and propose that residue 114 acts as a major determinant of tapasin quality control stages during MHC class I assembly. Curr. Biol. 8:717. 6. Greenwood, R., Y. Shimizu, G. S. Sekhon, and R. DeMars. 1994. Novel allele- dependence for efficient peptide loading, maturation, and surface specific, post-translational reduction in HLA class I surface expression in a mu- expression of HLA class I alleles while residue 116 can also affect tant human B cell line. J. Immunol. 153:5525. 968 DETERMINANT FOR TAPASIN DEPENDENCE OF MHC CLASS I MOLECULES

7. Lehner, P. J., M. J. Surman, and P. Cresswell. 1998. Soluble tapasin restores motif for HLA-B44, one of the most common HLA-B alleles in the Caucasian MHC class I expression and function in the tapasin-negative cell line .220. Im- population. Biochemistry 34:10130. munity 8:221. 20. Urvater, J. A., H. Hickman, J. L. Dzuris, K. Prilliman, T. M. Allen, K. J. Schwartz, 8. Kenneally, A., B. Liang, and L. D. Barber. 2000. The peptide-binding motif of D. Lorentzen, C. Shufflebotham, E. J. Collins, D. L. Neiffer, et al. 2001. Gorillas with HLA-B*3505. Immunogenetics 51:866. spondyloarthropathies express an MHC class I molecule with only limited sequence 9. Purcell, A. W., J. J. Gorman, M. Garcia-Peydro, A. Paradela, S. R. Burrows, similarity to HLA-B27 that binds peptides with arginine at P2. J. Immunol. 166:3334. G. H. Talbo, N. Laham, C. A. Peh, E. C. Reynolds, J. A. Lopez De Castro, and 21. Jun, Y., E. Kim, M. Jin, H. C. Sung, H. Han, D. E. Geraghty, and K. Ahn. 2000. J. McCluskey. 2001. Quantitative and qualitative influences of tapasin on the Human cytomegalovirus gene products US3 and US6 down-regulate trophoblast class I peptide repertoire. J. Immunol. 166:1016. class I MHC molecules. J. Immunol. 164:805. 10. Williams, A. P., S. Bevan, M. Bunce, R. Houlston, K. I. Welsh, and T. Elliott. 22. Madden, D. R., D. N. Garboczi, and D. C. Wiley. 1993. The antigenic identity of 2000. Identification of novel tapasin polymorphisms and linkage disequilibrium peptide-MHC complexes: a comparison of the conformations of five viral pep- to MHC class I alleles. Immunogenetics 52:9. tides presented by HLA-A2. Cell 75:693. 11. Neisig, A., R. Wubbolts, X. Zang, C. Melief, and J. Neefjes. 1996. Allele-specific 23. Williams, A. P., C. A. Peh, A. W. Purcell, J. McCluskey, and T. Elliott. 2002. differences in the interaction of MHC class I molecules with transporters asso- Optimization of the MHC class I peptide cargo is dependent on tapasin. Immunity ciated with antigen processing. J. Immunol. 156:3196. 16:509. 12. Turnquist, H. R., H. J. Thomas, K. R. Prilliman, C. T. Lutz, W. H. Hildebrand, 24. Petersdorf, E. W., T. Setoda, A. G. Smith, and J. A. Hansen. 1994. Analysis of and J. C. Solheim. 2000. HLA-B polymorphism affects interactions with multiple HLA-B*44 alleles encoded on extended HLA haplotypes by direct automated endoplasmic reticulum proteins. Eur. J. Immunol. 30:3021. sequencing. Tissue Antigens 44:211. 13. Beissbarth, T., J. Sun, P. B. Kavathas, and B. Ortmann. 2000. Increased efficiency 25. Ortmann, B., J. Copeman, P. J. Lehner, B. Sadasivan, J. A. Herberg, of folding and peptide loading of mutant MHC class I molecules. Eur. J. Immu- A. G. Grandea, S. R. Riddell, R. Tampe, T. Spies, J. Trowsdale, and P. Cresswell. nol. 30:1203. 1997. A critical role for tapasin in the assembly and function of multimeric MHC 14. Turnquist, H. R., E. L. Schenk, M. M. McIlhaney, H. D. Hickman, W. H. Hildebrand, class I-TAP complexes. Science 277:1306. and J. C. Solheim. 2002. Disparate binding of chaperone proteins by HLA-A sub- types. Immunogenetics 53:830. 26. Garbi, N., P. Tan, A. D. Diehl, B. J. Chambers, H. G. Ljunggren, F. Momburg, 15. Andersson, M., S. Paabo, T. Nilsson, and P. A. Peterson. 1985. Impaired intra- and G. J. Hammerling. 2000. Impaired immune responses and altered peptide cellular transport of class I MHC antigens as a possible means for adenoviruses repertoire in tapasin-deficient mice. Nat. Immunol. 1:234. to evade immune surveillance. Cell 43:215. 27. Grandea, A. G., 3rd, M. J. Androlewicz, R. S. Athwal, D. E. Geraghty, and Downloaded from 16. Saraste, J., G. E. Palade, and M. G. Farquhar. 1986. Temperature-sensitive steps T. Spies. 1995. Dependence of peptide binding by MHC class I molecules on in the transport of secretory proteins through the Golgi complex in exocrine their interaction with TAP. Science 270:105. pancreatic cells. Proc. Natl. Acad. Sci. USA 83:6425. 28. Bennett, E. M., J. R. Bennink, J. W. Yewdell, and F. M. Brodsky. 1999. Cutting 17. Li, S., K. M. Paulsson, S. Chen, H. O. Sjogren, and P. Wang. 2000. Tapasin is edge: adenovirus E19 has two mechanisms for affecting class I MHC expression. required for efficient peptide binding to transporter associated with antigen pro- J. Immunol. 162:5049. cessing. J. Biol. Chem. 275:1581. 29. Lopez de Castro, J. A. 1998. The pathogenetic role of HLA-B27 in chronic 18. Schagger, H., and G. von Jagow. 1987. Tricine-sodium dodecyl sulfate-polyac- arthritis. Curr. Opin. Immunol. 10:59.

rylamide gel electrophoresis for the separation of proteins in the range from 1 to 30. Ferrara, G. B., A. Bacigalupo, T. Lamparelli, E. Lanino, L. Delﬁno, A. Morabito, http://www.jimmunol.org/ 100 kDa. Anal. Biochem. 166:368. A. M. Parodi, C. Pera, S. Pozzi, M. P. Sormani, et al. 2001. Bone marrow trans- 19. DiBrino, M., K. C. Parker, D. H. Margulies, J. Shiloach, R. V. Turner, plantation from unrelated donors: the impact of mismatches with substitutions at W. E. Biddison, and J. E. Coligan. 1995. Identiﬁcation of the peptide binding position 116 of the human leukocyte antigen class I heavy chain. Blood 98:3150. by guest on October 2, 2021 The Journal of Immunology

CORRECTIONS

Boyoun Park, Sungwook Lee, Euijae Kim, and Kwangseog Ahn. A Single Polymorphic Residue Within the Peptide- Binding Cleft of MHC Class I Molecules Determines Spectrum of Tapasin Dependence. The Journal of Immunology 2003;170:961–968.

In the original article, both the concepts and language in the last paragraph of the Discussion were taken without attribution from the previous publications (References 3 and 9). We deeply regret and apologize for this lack of appropriate attribution.

Sang-Heon Lee, Dong Kyung Chang, Ajay Goel, C. Richard Boland, William Bugbee, David L. Boyle, and Gary S. Firestein. Microsatallite Instability and Suppressed DNA Repair Enzyme Expression in Rheumatoid Arthritis. The Jour- nal of Immunology 2003;170:2214–2220.

In Figure 2B, the mismatch repair enzyme expression for osteoarthritis (OA) and rheumatiod arthritis (RA) was compared in a bar graph. The designation for RA and OA were reversed in the ﬁgure. The corrected ﬁgure is shown below.