sign in sign up
<<
Home , Glycopeptide, Glycoprotein

Commentary

Processing of on and in - presenting cells

Ole Werdelin*†, Morten Meldal‡, and Teis Jensen§

*Institute for Medical Microbiology and , University of Copenhagen, DK-2200 Copenhagen, Denmark; ‡Department of Chemistry, The Carlsberg Laboratory, DK-2500 Copenhagen, Denmark; and §Pharmexa A͞S, DK-2970 Hørsholm, Denmark

ntil about 12 years ago, almost all an appropriate MHC class II molecule fine specificity of glycopeptide-specific T Uexperimental work with antigens ca- and obtain presentation to T helper cells. cells. Most common monosaccharides dif- pable of stimulating T cells had been The MHC class II molecule functions as a fer from one another by the orientation of performed with and , as capable of binding 10–25 residue hydroxy groups. Progress in technology well as haptenated proteins and peptides. long fragments of antigens with a has made it relatively easy to synthesize By far most of the work had been per- broad specificity and transporting them to with the same formed with mice that had been immu- the surface of the antigen-presenting cell sequence, but with closely related sugar nized and examined for responses in as- (APC) for presentation to T cells. In an groups differing only in the orientation of says for proliferation. Pure immunogenic glycopeptide antigen, the the hydrophilic hydroxy groups attached were found to be incapable peptide provides the binding motif that to the same site. Experi- of major histocompatibility complex enables the glycopeptide to bind to the ments with such sugar-variant glycopep- (MHC) binding and T cell stimulation (1, MHC molecule, and the group tides have revealed the glycan group as an 2). However, within the last 10 years it has provides an important part of the struc- integrated part of the T cell epitope that become evident that both CD4ϩ and ture that constitutes the epitope, i.e., the ϩ may be recognized with high specificity. CD8 T cells can recognize glycopeptides structure that is recognized by the T cell One example is a large collection of gly- carrying mono- and disaccharides in a through its T cell receptor (TCR). Eu- copeptide-specific T cell hybridomas that MHC-restricted manner provided the gly- karyote glycosylation may be N- were unable of recognizing a glycopeptide

can group is attached linked to or identical with the cognate glycopeptide COMMENTARY to the peptide at suit- O-linked to , except for the orientation of a single hy- able positions. In , or hydroxy- T cell recognition of glycans droxy group (9). The glycan specificity is such glycopeptides, lysines. Studies of gly- reflected in the overall amino acid com- the primed T cells may be crucial for T cell copeptide recognition position of the central parts of the TCR’s recognize the glycan responses to autoantigens. by T cells have shown complementarity determining region structure with high fi- that a glycan group lo- (CDR3) of glycopeptide-specific T cell delity (see below). cated outside the pep- hybridomas and clones. Conserved amino The question of T cell tide binding core of the acid motifs and of small polar recognition of glycopeptides may be im- MHC class II molecule will not be specif- amino acids, which are frequently found in portant in the immune defense against ically recognized, though it may change and other glycan-recognizing microorganisms, because many microbial the overall conformation of the MHC proteins, have been identified within antigens are in fact glycosylated. T cell bound peptide and in this way indirectly ␣␤TCR CDR3 regions of glycopeptide- recognition of glycans may also play an influence the structure of the MHC- specific T cells (12–14). In addition, amino bound T cell epitope (5, 6). In contrast, a important role in the immune defence acids that are flanking the glycan group glycan group positioned within the MHC against tumors, because one of the most and are oriented away from the binding binding core may influence the T cell consistent traits of a cancer cell is an cleft of the MHC molecule are recog- abnormal glycosylation of the proteins of recognition in several ways. If the glycan is linked to an amino acid functioning as an nized. The crystal structure of MHC class the malignant cell (3). In this issue of ͞ MHC anchor residue the glycopeptide I glycopeptide complexes also show that PNAS, Ba¨cklund et al. (4) provide evi- may be incapable of MHC binding and glycans can be accommodated by the TCR dence that T cell recognition of protein thus become nonimmunogenic (2, 7). (15, 16). However, not all peptide- glycans may be crucial also for T cell attached glycans can elicit a T cell re- However, if the glycan is linked to an ␣␤ responses to autoantigens in the course of amino acid pointing away from the bind- sponse (17). It appears that T cells are autoimmune diseases. Below we will de- ing site of the MHC molecule, the binding unable to recognize large and highly com- scribe and discuss the general rules for is maintained, and in antigens with the plex glycan structures. A possible molec- MHC class II restricted T cell recognition glycan attached to central residue within ular explanation for this may be that the of glycans, the fate of glycoprotein glycans the MHC core, the glycan becomes the during antigen processing, and the role of dominant structure in the epitope, which antigen glycosylation in tolerance to au- is recognized with very high fidelity by T See companion article on page 9960. toantigens and tumor antigens. cells specific for the particular glycopep- †To whom reprint requests should be sent at the present address: Institute for Medical Microbiology and Immunol- A sine qua non for a compound to be tide (7–11). ogy, The Panum Institute, Room 22.5.24, Blegdamsvej 3, able to stimulate an antigen-specific T Glycopeptides with simple sugars have DK-2200 N Copenhagen, Denmark. E-mail: o.werdelin@ helper cell response is that it can bind to been suitable for studies of the antigen immi.ku.dk.

www.pnas.org͞cgi͞doi͞10.1073͞pnas.152345899 PNAS ͉ July 23, 2002 ͉ vol. 99 ͉ no. 15 ͉ 9611–9613 Downloaded by guest on September 29, 2021 clones only recognizing the unglycosy- lated form of the glycopeptide used for the immunization strongly indicates that deg- lycosylation of some of the antigen mole- cules had occurred during the priming phase of the immune response. The sig- nificance of deglycosylation may be quite different for various antigens, but should be considered. In the present study by Ba¨cklund et al. (4), some of the important glycopeptides obviously survive antigen processing with the glycan intact. Studies by Chicz et al. (20, 21) have shown that in natural glycoprotein pro- cessing some glycan groups remain at- tached to the antigen fragments bound onto MHC class II molecules. These work- ers eluted material bound onto the human HLA-DR and HLA-DQ molecules and identified some of them as glycopeptides containing N-linked GlcNAc residues. So far, however, N-linked carbohydrates have not been identified been identified on material eluted from MHC class I mole- cules. The majority of the MHC class I binding peptides are derived from cytoso- lic proteins that have been targeted by Fig. 1. Processing of in APCs. (A) The intact glycoprotein has been taken up by the APC and is transported through the endocytic pathway. (B–D) The peptide fragments after alternate glycan ubiquitinylation and degraded by the pro- processing. In all three, the peptide is in extended conformation. (B) The complex glycan group has teasome. The reason why these protein survived the processing and is left intact on the peptide fragment. (C) Only some of the glycan has survived fragments are devoid of sugars may be that bound to the glycosylated segment. (D) The glycan has been removed entirely. all sugars are removed by a cytosolic N- glycanase before the cytosolic protein in- teracts with the proteasome. By contrast, central CDR3 region of the TCR cannot with low pH optimum are added to the elution of peptides from MHC class I accommodate very large glycans, while endosome and the pH decreases progres- molecules have revealed that around other parts of the receptor at the same sively, leading to activation of the proteo- 0.1% of all class I bound peptides carry time interact with the ␣-helices of the lytic . The enzymes, which include small O-linked N-acetylglucosamine (O- presenting MHC molecule. Studies of the endoproteases and exoproteases of many ␤GlcNAc) residues (22). This finding is in immunogenicity of linear and branched different substrate specificities, attack and agreement with the fact that glycosylation sugars of varying length demonstrated fragment the antigen into peptides. Suit- of serine and threonine residues on that the ␣␤ TCR may recognize exten- able peptides bind to empty MHC class II nuclear and cytosolic proteins by O- sions of glycan groups consisting of up to molecules, which are accumulating within ␤GlcNAc are abundant in all multicellular three or four sugars, whereas even larger the acidic compartments and the peptides eukaryotes. In Ba¨cklund et al.’s paper in glycan structures lead to loss of MHC class are from that point on protected against this issue of PNAS (4) the authors dem- II restricted T cell recognition (16–18). further (19). Finally, the onstrate that glycans not only remain at- Such large glycans may activate MHC MHC-peptide complexes are transported tached to the peptide backbone of the unrestricted ␥␦T cells, which recognize to the cell surface and presented to the T large glycoprotein antigen, II the glycan antigen regardless of whether it cell system. (CII), during the processing in the APC, is associated with the APC through an Some observations pointing to the fate but such glycans are apparently also pre- MHC binding peptide or through a of glycans during antigen processing have sented to the T cell system and stimulate tail. been made with a mouse hemoglobin- specific CD4ϩ T helper cells. T cells with Antigen processing and presentation is derived decaglycopeptide O-glycosylated specificity for glycosylated epitopes of CII a thoroughly studied area in immunology with ␣-D-GalNAc on a central threonine. can be detected both in mice immunized and our knowledge of the mechanisms Mice immunized with this glycopeptide with human CII and in patients suffering whereby peptides are generated for MHC gave a proliferative T cell response to the from rheumatoid arthritis, a disease in class II and class I molecules is amazing. cognant glycopeptide, but they also re- which CII is implicated as an important However, a less investigated area is the sponded to the unglycosylated decapep- autoantigen. One evident implication of effect of protein glycosylation on the pro- tide (7). This was not a simple cross- the alternate ways of glycan processing is cessing of antigen in APCs, where a glycan response from some glycopeptide specific that a glycan-containing T cell epitope can interfere with the proteolytic frag- clones. That became clear when T cell may be converted and presented as several mentation of the large glycoprotein anti- hybridomas were raised against the glyco- variants. This in turn diversifies the im- gen and influence the pattern of T cell peptide and analyzed for antigen-fine mune response, and it has practical con- epitopes that are created. Glycoprotein specificity (9). The majority of the raised sequences for the design of glycopeptide- antigens are ingested by APCs by endo- T cell hybridomas responded to the gly- based vaccines. Fig. 1 illustrates the cytosis and transported in the endocytic copeptide but not to the peptide, and a few alternative possibilities for the processing pathway from the cell surface toward the hybridomas responded to the peptide but of glycoproteins. lysosomal compartments of the cell. Dur- not to the glycopeptide. No hybridoma The question of antigen glycosylation is ing the transport, proteolytic enzymes responded to both. The presence of T cell central to the work presented by Ba¨cklund

9612 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.152345899 Werdelin et al. Downloaded by guest on September 29, 2021 et al. (4) and seems to be relevant to the sequence 261–278. In this epitope there to have been purged of clones with spec- breaking of immunological tolerance to are two hydroxylysines that are glycosy- ificity toward the unglycosylated variants CII. CII is an important autoantigen in lated on some CII molecules but not on of the arthritogenic CII epitope, some patients with rheumatoid arthritis, and others. Backlund et al. discovered that the clones with specificity toward the glycosy- immunization of some mouse strains with CII-immunized mice were almost com- lated variant have escaped thymic deletion CII triggers an inflammatory arthritis pletely tolerant to the unglycosylated and are fully functional. This leaves one widely regarded as a model for rheuma- epitope variant but not to the glycosylated wondering whether tolerance mechanisms toid arthritis. By genetic manipulation, epitope variant. They also showed that may be less efficient for T cell clones with Ba¨cklund et al. created mice equipped patients with severe rheumatoid arthritis specificity for glycosylated epitopes. Al- though such clones may be damaging for with human HLA-DR molecules and ca- predominantly recognized the glycosy- the patients with rheumatoid arthritis they pable of giving HLA-DR4-restricted T cell lated CII epitope. Here, it is of great may play a beneficial role in patients with responses. Into these mice they then in- importance that the epitope actually is cancer. Many tumors express glycosylated troduced the gene for human CII. The shown to be present in patients with rheu- tumor antigens. One such tumor-associ- expected outcome is that such mice will matoid arthritis. However, the comple- ated antigen is the MUC1 molecule, which develop immunological tolerance to CII. mentary absence or reduced presence in is heavily O-glycosylated with Tn (␣-D- The mice indeed failed to give T cell mature individuals without rheumatoid GalNAc) and T (␤-D-Gal(1–3)␣-D-Gal- responses after appropriate immuniza- arthritis could have been demonstrated to NAc). Might some of the antitumor T cell tion, but it turned out that the tolerance give this observation more significance. responses observed in patients with breast was incomplete. The crucial arthritogenic Thus, although the T cell repertoire in cancer be directed against MUC1 epitopes T cell epitope on CII is located on the patients with rheumatoid arthritis seems containing the Tn or T glycans?

1. Harding, C. V., Roof, R. W., Allen, P. M. & 9. Jensen, T., Hansen, P., Galli-Stampino, L., Pao, Y. L., Wormald, M., Dwek, R. A., Jones, E. Y. Unanue, E. R. (1991) Proc. Natl. Acad. Sci. USA Mouritsen, S., Frische, K., Meinjohanns, E., & Elliott, T. (1999) Immunity 10, 63–74. 88, 2740–2744. Meldal, M. & Werdelin, O. (1997) J. Immunol. 16. Speir, J. A., Abdel-Motal, U. M., Jondal, M. & 2. Ishioka, G. Y., Lamont, A. G., Thomson, D., 158, 3769–3778. Wilson, I. A. (1999) Immunity 10, 51–61. Bulbow, N., Gaeta, F. C., Sette, A. & Grey, H. M. 10. Deck, M. B., Sjolin, P., Unanue, E. R. & Kihlberg, 17. Galli-Stampino, L., Frische, K., Meinjohanns, E., (1992) J. Immunol. 148, 2446–2451. J. (1999) J. Immunol. 162, 4740–4744. Meldal, M., Jensen, T., Werdelin, O. & Mouritsen, 3. Hakomori, S. (1989) Adv. Cancer Res. 52, 257–331. 11. Haurum, J. S., Arsequell, G., Lellouch, A., Wong, S. (1997) Cancer Res. 57, 3214–3222. 4. Ba¨cklund, J., Carlsen, S., Ho¨ger, T., Holm, B., S. Y., Dwek, R. A., McMichael, A. J. & Elliott, T. 18. Abdel-Motal, U. M., Berg, L, Rosen, A., Bengts- Fugger, L., Kihlberg, J., Burkhardt, H. & Holm- (1994) J. Exp. Med. 180, 739–744. son, M., Thorpe, C. J., Kihlberg, J., Dahmen, J dahl, R. (2002) Proc. Natl. Acad. Sci. USA 99, 12. Corthay, A., Ba¨cklund, L., Broddefalk, J., Michae¨ls- Magnusson, G., Karlsson, K. A. & Jondal, M. 9960–9965. son, E., Goldschmidt, T. J., Kihlberg, J. & Holmdahl, (1996) Eur. J. Immunol. 26, 544–551. 5. Harding, C. V., Kihlberg, J., Elofsson, M., Mag- R. (1998) Eur. J. Immunol. 28, 2580–2590. 19. Mouritsen, S., Meldal, M., Werdelin, O., Hansen, nusson, G. & Unanue, E. R. (1992) J. Immunol. 13. Jensen, T., Hansen, P., Faurskov Nielsen, A., A. S. & Buus, S. (1992) J. Immunol. 149, 1987–1993. 151, 2419–2425. Meldal, M., Komba, S. & Werdelin, O. (1999) Eur. 20. Chicz, R. M., Urban, R. G., Gorga, J. C., Vignali,

6. Mouritsen, S., Meldal, M., Christiansen-Brams, I., J. Immunol. 29, 2759–2768. D. A., Lane, W. S. & Strominger, J. L. (1993) J. COMMENTARY Elsner, H. & Werdelin, O. (1994) Eur. J. Immunol. 14. Haurum, J. S., Tan, L., Arsequell, G., Frodsham, Exp. Med. 178, 27–47. 24, 1066–1072. P., Lellouch, A. C., Moss, P. A., Dwek, R. A., 21. Chicz, R. M., Lane, W. S., Robinson, R. A., 7. Jensen, T., Galli-Stampino, L., Mouritsen, S., McMichael, A. J. & Elliott, T. (1995) Eur. J. Im- Trucco, M., Strominger, J. L. & Gorga, J. C. Frische, K., Peters, S., Meldal, M. & Werdelin, O. munol. 25, 3270–3276. (1994) Int. Immunol. 6, 1639–1649. (1996) Eur. J. Immunol. 26, 1342–1349. 15. Glithero, A., Tormo, J., Haurum, J. S., Arsequell, G., 22. Haurum, J. S., Bjerring Højer, I., Arsequell, G., 8. Deck, B., Elofsson, M., Kihlberg, J. & Unanue, Valencia, G., Edwards, J., Springer, S. Townsend, A., Neisig, A., Valencia, G., Zeuthen, J., Neefjes, J. & E. R. (1995) J. Immunol. 155, 1074–1078. Elliott, T. (1999) J. Exp. Med. 190, 145–150.

Werdelin et al. PNAS ͉ July 23, 2002 ͉ vol. 99 ͉ no. 15 ͉ 9613 Downloaded by guest on September 29, 2021