- Email: [email protected]
Remnant epitopes, autoimmunity and glycosylation
Biochimica et Biophysica Acta 1760 (2006) 610 – 615 http://www.elsevier.com/locate/bba
Review
Remnant epitopes, autoimmunity and glycosylation Ghislain Opdenakker a,⁎, Chris Dillen a , Pierre Fiten a , Erik Martens a , Ilse Van Aelst a , Philippe E. Van den Steen a , Inge Nelissen a , Sofie Starckx a , Francis J. Descamps a , Jialiang Hu a , Helene Piccard a , Jo Van Damme b , Mark R. Wormald c , Pauline M. Rudd c , Raymond A. Dwek c b
a Rega Institute for Medical Research, Laboratory of Immunobiology, University of Leuven, Minderbroedersstraat 10, B-3000 Leuven, Belgium Rega Institute for Medical Research, Laboratory of Molecular Immunology, University of Leuven, Minderbroedersstraat 10, B-3000 Leuven, Belgium c Glycobiology Institute, University of Oxford, South Parks Road, Oxford OX1 3QU, UK
Received 21 October 2005; received in revised form 12 December 2005; accepted 12 December 2005 Available online 6 January 2006
Abstract The role of extracellular proteolysis in innate and adaptive immunity and the interplay between cytokines, chemokines and proteinases are gradually becoming recognized as critical factors in autoimmune processes. Many of the involved proteinases, including those of the plasminogen activator and matrix metalloproteinase cascades, and also several cytokines and chemokines, are glycoproteins. The stability, interactions with inhibitors or receptors, and activities of these molecules are fine-controlled by glycosylation. We studied gelatinase B or matrix metalloproteinase9 (MMP-9) as a glycosylated enzyme involved in autoimmunity. In the joints of rheumatoid arthritis patients, CXC chemokines, such as interleukin-8/CXCL8, recruit and activate neutrophils to secrete prestored neutrophil collagenase/MMP-8 and gelatinase B/MMP-9. Gelatinase B potentiates interleukin-8 at least tenfold and thus enhances neutrophil and lymphocyte influxes to the joints. When cartilage collagen type II is cleaved at a unique site by one of several collagenases (MMP-1, MMP-8 or MMP-13), it becomes a substrate of gelatinase B. Human gelatinase B cleaves the resulting two large collagen fragments into at least 33 peptides of which two have been shown to be immunodominant, i.e., to elicit activation and proliferation of autoimmune T cells. One of these two remnant epitopes contains a glycan which is important for its immunoreactivity. In addition to the role of gelatinase B as a regulator in adaptive immune processes, we have also demonstrated that it destroys interferon-β, a typical innate immunity effector molecule and therapeutic cytokine in multiple sclerosis. Furthermore, glycosylated interferon-β, expressed in Chinese hamster ovary cells, was more resistant to this proteolysis than recombinant interferon-β from bacteria. These data not only prove that glycosylation of proteins is mechanistically important in the pathogenesis of autoimmune diseases, but also show that targeting of glycosylated proteinases or the use of glycosylated cytokines seems also critical for the treatment of autoimmune diseases. © 2005 Elsevier B.V. All rights reserved. Keywords: Gelatinase B/MMP-9; Remnant epitope; TIMP-1; Collagenase; Neutrophil
1. Introduction To understand autoimmune processes, it is critical to know that autoreactivity is normal. The generation of antibodies and T cell receptors is based on genetic principles in which multiple gene segments are mutated and recombined, thus encoding a multifold of (receptor) molecules, called the immunologic repertoire [1,2]. Some of these gene products possess intrinsic reactivity to (parts of) host molecules. Many of the ⁎ Corresponding author. E-mail address: [email protected] (G. Opdenakker). 0304-4165/$ - see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bbagen.2005.12.014
lymphocytes, but certainly not all, that encode self-reactive receptors are eliminated by control processes of tolerance induction [3]. It has been estimated that, after passage through the thymus, about 10% of the T lymphocytes are self-reactive [4]. The mechanisms of peripheral tolerance [5,6], e.g., ignorance and anergy, further reduce the chance of developing autoimmune diseases. Ignorance is best defined as the presence of anatomical barriers [7], such as the blood–brain barrier, which prevent the interaction of the self-reactive lymphocyte with the presented antigen, whereas anergy [8] is the absence of immune reaction in the absence of a second stimulus (infection or inflammation).
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
Autoimmune diseases are multifactorial [9], i.e., they are based on the interplay between both genetic constitution and environment of the host. T cell reactive autoimmune antigens are presented within the grooves of the highly polymorphic major histocompatibility complex (MHC) molecules [10], and association with the MHC locus (predisposing and protective) has been demonstrated for autoimmune diseases, including multiple sclerosis (MS) [11], rheumatoid arthritis [12], diabetes [13] and inflammatory bowel disease [14]. The environmental factors in autoimmune diseases are much less clear so far. The study of host cells by (immuno) histopathology and host molecules by biological and biochemical analysis forms the best basis to understand the pathophysiology of autoimmunity. 2. The REGA model We have studied host cells and molecules in autoimmune diseases to define the important role of cytokines, chemokines and extracellular proteases in autoimmunity [15]. Inflammatory processes lead to the production of cytokines and chemokines, which recruit and stimulate cells to release proteinases. These include serine proteinases (e.g., plasminogen activators, granzymes, neutrophil elastase and plasmin) that act together in a cascade with the matrix metalloproteinases (MMPs). This cascade forms an efficient way of amplifying the activity of the enzymes in the chain, as each enzyme continues to activate its substrate. In common with other cascades, such as the complement and blood clotting systems, the extracellular matrix degrading cascade is kept under control by specific inhibitors [15]. The action of the extracellular proteinases results in the proteolysis of matrix molecules or other substrates, and leads to quantitative and qualitative changes of the “antigen repertoire”. These Remnant Epitopes Generate Autoimmunity, hence, the process was named the REGA model [15]. When a substrate is cleaved at multiple sites, a molar excess of peptides becomes available for processing and presentation by classical or nonclassical pathways. Furthermore, proteolysis might liberate hidden (cryptic) antigens from a folded protein, thus adding to qualitative changes in the “antigen repertoire” (Fig. 1). From the scheme in Fig. 1, it can be deduced that inhibition of target proteinases may be beneficial in autoimmune diseases. The REGA model has been tested in vivo with the use of animal models. In comparison with wildtype animals, young gelatinase B/MMP-9-deficient mice were resistant to the development of experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis [16], and to autoimmune arthritis induced by antibodies against collagen [17]. These experiments form a good basis for the development of gelatinase B inhibitors for the treatment of multiple sclerosis and rheumatoid arthritis [18]. For example, MMP inhibitors have been used with success in the treatment of EAE [19,20].
611
Fig. 1. The role of cytokines and extracellular proteinases in the REGA model. The trigger of an autoimmune process is not known, but studies of molecular and cellular changes in patients and animal models have pointed to the involvement of cytokines, chemokines and proteinases in autoimmune processes. Many cytokines enhance inflammation and chemokines recruit and activate leukocytes. This leads to the secretion of proteinases that cleave (glyco)proteins into remnant epitopes. Examples of substrates yielding such remnant epitopes generating autoimmunity (REGA) include myelin basic protein, αB-crystallin and, possibly also, interferon-β in multiple sclerosis and type II collagen in rheumatoid arthritis. As a consequence, T cell activation maintains the process by generating proinflammatory cytokines. The REGA model has therapeutic consequences, since inhibition of proteinases, such as gelatinase B, results in beneficial effects [18,30]. Figure modified from reference 15.
3. The glycoprotein gelatinase B: structural and functional aspect of the sugars 3.1. Gelatinase B as a complex glycosylated matrix metalloproteinase MMP-9 contains a prodomain with an occupied N-linked sequon (Asn38–Leu–Thr) and an active domain with a second N-linked oligosaccharide (located at Asn120–Ile–Thr) [21,22]. The Zn2+-binding domain contains three conserved histidines for the coordination of the catalytic Zn2+-ion. Together with parts of the active site domain, this constitutes the catalytic site, of which the three-dimensional structure is highly similar in the different MMPs [23] (Fig. 2a). The active site has a cleft in which the (Zn2+-binding) substrate can bind before it is cleaved. In the proenzyme, this cleft is occupied by the propeptide with a Zn2+-binding cysteine residue, which prevents catalytic activity. After the proteolytic removal of the propeptide, a water molecule binds to the Zn2+-ion allowing the catalysis to proceed [24,25]. In gelatinases, three fibronectin repeats, positioned between the active site and the Zn2+-binding domain, increase the affinity for gelatin. Gelatinase B contains an extended threonine–proline–serine-rich region between the aminoterminal active site and the carboxyterminal hemopexin domain. This region is extensively O-glycosylated in neutrophil gelatinase B [21,26,27]. Gelatinases are, in terms of domain structure, already rather complex MMPs [28]. By comparison of the structure of unglycosylated gelatinase A/MMP-2 [29] with a model of glycosylated gelatinase B [30], it also became clear that both molecules are structurally quite different from each other. Gelatinase B is thus not only the most complex MMP in
612
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
Fig. 2. Molecular models of gelatinase B from human neutrophils. Neutrophilic granulocytes secrete three forms of gelatinase B: monomers, covalent homodimers and covalent complexes with NGAL [32,33]. In panel a, a view of the glycosylated monomer is presented. This model is based on all available protein and carbohydrate structures of gelatinases A and B. The peptide is shown in green, the N-linked glycans in dark blue and the O-linked glycans in brown. The N-terminal and C-terminal domains are based on the crystal structure of the N-terminal part of MMP-9 [23] and the crystal structure of the C-terminal domain of MMP-2 [29]. Analytical ultracentrifugation (AUC) studies (unpublished data) show that the O-link domain connecting the two is not extended and the whole molecule is relatively compact. Panel b. AUC studies show that the dimer is compact and overall more spherical than the monomers, and that dimerisation does not lead to an extension of the O-link domain. The AUC results can be reproduced by the side-to-side interaction of two compact monomers, although there are other models of the dimerisation that would also fit the experimental data.
terms of domain structure, it is also – as far as is known – the most extensively glycosylated MMP. In view of the heterogeneity and complexity of the attached sugars [27], so far, only mixtures of glycoforms/glycotypes of gelatinase B have been obtained. This might constitute one of the possible explanations why attempts to crystallize intact gelatinase B for detailed structural analysis have failed so far.
About 93% of the N-linked oligosaccharides contain core fucose. Human neutrophil gelatinase B is significantly less sialylated (9%) than MMP-1 from fibroblasts and HT-1080 cells (50% and 34% respectively). Human neutrophil gelatinase B and MMP-1 from both sources contain glycans with outer arm fucose-linked α1–3 to GlcNAc [26,27,34]. 3.3. Functional analysis of the glycans of gelatinase B
3.2. Structures of the glycans of neutrophil gelatinase B Natural gelatinase B can be purified from human neutrophils without any degradation or activation and in reasonable quantities [31]. Upon stimulation, e.g., with interleukin-8, neutrophils release large amounts of gelatinase B and do not secrete gelatinase A or TIMP-1. This is in contrast with many other cell types that constitutively produce gelatinase A and coexpress TIMP-1 and gelatinase B upon stimulation. This makes the purification procedure straightforward for neutrophil gelatinase B. Three different forms of neutrophil gelatinase B can be distinguished: monomers, disulfide-bound dimers and a covalent complex of gelatinase B with neutrophil gelatinase Bassociated lipocalin (NGAL). The latter complex can be removed using monoclonal antibodies against NGAL [32,33]. A model of MMP-9 as a dimer is presented in Fig. 2b. Approximately 85% of the sugars of human neutrophil gelatinase B are O-linked structures attached to 14 potential Oglycosylation sites in a stretch of about 50 amino acids. Most of these sugars are type 2 structures with Galβ1-4GlcNAc (Nacetyl-lactosamine) extensions, with or without sialic acid and fucose. The two N-glycans are about 80% complex biantennary structures and 20% are tri-antennary structures.
The glycosylation of the propeptide of gelatinase B raises the possibility of a function for such a glycan in protein folding and regulating enzyme activation and activity [21]. Since the propeptide mimics a peptide inhibitor [35] and glycosylation protects glycoproteins against degradation, future experiments are directed towards testing the inhibitory activity of a glycosylated propeptide against the activities of gelatinase B and other MMPs. Gelatinase B can be activated by various proteases, including gelatinase A (MMP-2) [36] and stromelysin-1 (MMP-3) [37]. Therefore, we studied the influence of N-deglycosylation and desialylation on the activation of progelatinase B and on the activity of the active form. No significant differences were detected between the N-deglycosylated gelatinase B forms and native gelatinase B. Also, no differences were found for the activation by gelatinase A or stromelysin-1 of N-deglycosylated or desialylated progelatinase B, compared with native gelatinase B, in zymography and fluorogenic substrate conversion assay [21]. Therefore, neither the N-linked glycan on the prodomain nor the sialic acids that terminate many of the glycans attached to gelatinase B have any direct influence on the activity and activation of gelatinase B.
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
Recent data using recombinant variants of gelatinase B, with or without the O-glycosylated domain, indicate that this domain is not critical for the hydrolysis of more than 10 different known substrates (P. Van den Steen et al., unpublished). The glycans of gelatinase B contain terminal sialic acids, α2–3 or α2–6 linked both on N-linked and on O-linked glycans [26,27]. These sialic acids were removed with Streptococcus sp. sialidase. Sialic acids often influence the interaction/recognition between glycoproteins. Therefore, the sensitivity of activated desialylated gelatinase B to inhibition by TIMP-1 was compared with the activated intact glycosylated MMP-9 in an inhibition assay. Desialylated gelatinase B was significantly less inhibited by TIMP-1 and this enzyme form thus possessed higher activity than the intact glycoprotein, in the presence of TIMP-1. Neither deglycosylation of Asn38 in the propeptide nor desialylation of gelatinase B glycans influence the activation rate of natural gelatinase B by gelatinase A and stromelysin-1 [21]. 4. Gelatinase B cleaves molecules of the innate and adaptive immune system 4.1. Gelatinase B destroys interferon-β and potentiates specific cytokines and chemokines One of the molecules currently used in the treatment of multiple sclerosis is interferon-β. Unfortunately, its mechanism of action is not understood. Patients with multiple sclerosis are treated with rather high doses of interferon-β, either produced in recombinant form in bacteria (Betaferon) or Chinese hamster ovary cells (Avonex, Rebif). Of note, the unglycosylated Betaferon induces neutralizing antibodies (which prevent further use as a therapeutic) in about 30% of treated patients, whereas the glycosylated interferon-β from CHO cells is almost devoid of such induction [38]. Although we do not know whether this is related to differences in glycosylation, aggregation, stability or other reasons, we discovered that gelatinase B readily cleaves some aminoterminal residues from the unglycosylated interferon-β, whereas the glycosylated interferon-β is resistant to this proteolysis. If we add the well established information that gelatinase B levels and activity are increased and TIMP-1 levels are decreased in patients with multiple sclerosis [18,39,40], it is possible that remnant epitopes of interferon-β are much more easily formed in patients treated with Betaferon versus Avonex or Rebif. On the basis of a three-dimensional model of glycosylated interferon-β and taking into account that N-linked sugars can protect relatively large protein surfaces, we proposed a mechanism for the aminoterminal inactivation of unglycosylated recombinant interferon-β and the role of the attached sugars in resistance against proteolysis [41]. In addition to the degradation of interferon-β, gelatinase B cleaves other molecules with functions in the immune system, such as α1-proteinase inhibitor, thereby contributing to a more positive balance of proteases versus inhibitors [42], and cytokines, such as pro-IL-1β [43] and pro-TNF-α [44,45].
613
Further, gelatinase B potentiates IL-8 by aminoterminal clipping, thereby providing a positive feedback loop as IL8 induces the degranulation of gelatinase B from neutrophils [33]. 4.2. Gelatinase B generates remnant epitopes Prototypic immunodominant proteins in multiple sclerosis and rheumatoid arthritis are αB-crystallin and collagen type II, respectively. Recent microarray analysis demonstrated that αB-crystallin mRNA was the most upregulated mRNA in MS lesions compared with control brain tissue [46]. A decade ago, the αB-crystallin protein was described as an autoantigen in multiple sclerosis [47]. We found proof of concept of the REGA model with αB-crystallin [48]. In fact, the cleavage of αB-crystallin by gelatinase B resulted in remnant peptides, which corresponded to the immunodominant T cell activating peptides in Biozzi ABH, SJL and 129 mouse strains [48] and also to another immunodominant epitope in humans [49]. In rheumatoid arthritis, cartilage collagen type II is degraded by neutrophil collagenase/MMP-8 into an aminoterminal 1/4th and a carboxyterminal 3/4th fragment [50]. This single cleavage leads to a local unwinding and generates an efficient substrate for gelatinase B. The gelatin is degraded into more than 30 fragments, at least two of which are immunodominant. Earlier [51] and recent experiments [52,53] showed that at least one immunodominant remnant epitope is glycosylated. This example extends the important earlier findings [54–56] that not only peptides, but also glycopeptides may be presented in MHC molecules by antigen presenting cells and thus activate autoreactive T lymphocytes, thereby maintaining or enhancing the autoimmune process. 5. Conclusions Glycosylation and recognition of oligosaccharides by selectins have been recognized as a control mechanism of autoimmunity [57]. Here, we extend this thesis with novel mechanisms and additional examples. Glyco-enzymes and glycosylated cytokines are critical regulator and effector molecules in autoimmune diseases, such as multiple sclerosis and rheumatoid arthritis. The activity and stability of these molecules are fine-controlled by glycosylation and this might have therapeutic implications. Finally, glycopeptides are natural remnant epitopes in human autoimmune processes. Thus, the importance of glycobiology in autoimmune diseases is clear at all levels: the innate and adaptive immune system, T cell receptors, antibodies and autoantigens. Acknowledgements Supported by Fortis AB, the National Fund for Scientific Research (F.W.O.-Vlaanderen), the Geconcerteerde OnderzoeksActies (2002–2006), and the Centre of Excellence (EF/ 05/015) Belgium, and the Glycobiology Institute, Oxford, UK
614
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
endownment. PVdS is a postdoctoral research fellow of the F. W.O.-Vlaanderen.
[24]
References [1] A.C. Hayday, D.J. Diamond, G. Tanigawa, J.S. Heilig, V. Folsom, H. Saito, S. Tonegawa, Unusual organization and diversity of T-cell receptor alpha-chain genes, Nature 316 (1985) 828–832. [2] M.H. Bakkus, Ig gene sequences in the study of clonality, Pathol. Biol. (Paris) 47 (1999) 128–147. [3] S.P. Cobbold, E. Adams, S.E. Marshall, J.D. Davies, H. Waldmann, Mechanisms of peripheral tolerance and suppression induced by monoclonal antibodies to CD4 and CD8, Immunol. Rev. 149 (1996) 5–33. [4] K.A. Hogquist, T.A. Baldwin, S.C. Jameson, Central tolerance: learning self-control in the thymus, Nat. Rev., Immunol. 5 (2005) 772–782. [5] B. Arnold, Levels of peripheral T cell tolerance, Transpl. Immunol. 10 (2002) 109–114. [6] B. Arnold, G. Schonrich, G.J. Hämmerling, Multiple levels of peripheral tolerance, Immunol. Today 14 (1993) 12–14. [7] K.W. Wucherpfennig, Autoimmunity in the central nervous system: mechanisms of antigen presentation and recognition, Clin. Immunol. Immunopathol. 72 (1994) 293–306. [8] J. Lohr, B. Knoechel, V. Nagabhushanam, A.K. Abbas, T-cell tolerance and autoimmunity to systemic and tissue-restricted self-antigens, Immunol. Rev. 204 (2005) 116–127. [9] J.M. van Noort, S. Amor, Cell biology of autoimmune diseases, Int. Rev. Cytol. 178 (1998) 127–206. [10] H.O. McDevitt, Discovering the role of the major histocompatibility complex in the immune response, Annu. Rev. Immunol. 18 (2000) 1–17. [11] L. Steinman, Multiple sclerosis: a coordinated immunological attack against myelin in the central nervous system, Cell 85 (1996) 299–302. [12] R. Holmdahl, Dissection of the genetic complexity of arthritis using animal models, J. Autoimmun. 21 (2003) 99–103. [13] H.J. Cordell, J.A. Todd, Multifactorial inheritance in type 1 diabetes, Trends Genet. 11 (1995) 499–504. [14] D.A. Watts, J. Satsangi, The genetic jigsaw of inflammatory bowel disease, Gut (Suppl. 3) (2002) 31–36. [15] G. Opdenakker, J. Van Damme, Cytokine-regulated proteases in autoimmune diseases, Immunol. Today 15 (1994) 103–107. [16] B. Dubois, S. Masure, U. Hurtenbach, L. Paemen, H. Heremans, J. van den Oord, R. Sciot, T. Meinhardt, G. Hämmerling, G. Opdenakker, B. Arnold, Resistance of young gelatinase B-deficient mice to experimental autoimmune encephalomyelitis and necrotizing tail lesions, J. Clin. Invest. 104 (1999) 1507–1515. [17] T. Itoh, H. Matsuda, M. Tanioka, K. Kuwabara, S. Itohara, R. Suzuki, The role of matrix metalloproteinase-2 and matrix metalloproteinase-9 in antibody-induced arthritis, J. Immunol. 169 (2002) 2643–2647. [18] G. Opdenakker, I. Nelissen, J. Van Damme, Functional roles and therapeutic targeting of gelatinase B and chemokines in multiple sclerosis, Lancet Neurol. 2 (2003) 747–756. [19] K. Gijbels, R.E. Galardy, L. Steinman, Reversal of experimental autoimmune encephalomyelitis with a hydroxamate inhibitor of matrix metalloproteases, J. Clin. Invest. 94 (1994) 2177–2182. [20] A.K. Hewson, T. Smith, J.P. Leonard, M.L. Cuzner, Suppression of experimental allergic encephalomyelitis in the Lewis rat by the matrix metalloproteinase inhibitor Ro31-9790, Inflamm. Res. 44 (1995) 345–349. [21] P.E. Van den Steen, G. Opdenakker, M.R. Wormald, R.A. Dwek, P.M. Rudd, Matrix remodelling enzymes, the protease cascade and glycosylation, Biochim. Biophys. Acta 1528 (2001) 61–73. [22] L.P. Kotra, L. Zhang, R. Fridman, R. Orlando, S. Mobashery, NGlycosylation pattern of the zymogenic form of human matrix metalloproteinase-9, Bioorg. Chem. 30 (2002) 356–370. [23] P.A. Elkins, Y.S. Ho, W.W. Smith, C.A. Janson, K.J. D'Alessio, M.S. McQueney, M.D. Cummings, A.M. Romanic, Structure of the C-
[25]
[26]
[27]
[28]
[29]
[30] [31]
[32]
[33]
[34]
[35]
[36] [37]
[38]
[39]
[40]
[41]
terminally truncated human ProMMP9, a gelatin-binding matrix metalloproteinase, Acta Crystallogr., D Biol. Crystallogr. 58 (2002) 1182–1192. H.E. Van Wart, H. Birkedal-Hansen, The cysteine switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family, Proc. Natl. Acad. Sci. U. S. A. 87 (1990) 5578–5582. O. Kleifeld, P.E. Van den Steen, A. Frenkel, F. Cheng, H.L. Jiang, G. Opdenakker, I. Sagi, Structural characterization of the catalytic active site in the latent and active natural gelatinase B from human neutrophils, J. Biol. Chem. 275 (2000) 34335–34343. T.S. Mattu, L. Royle, J. Langridge, M.R. Wormald, P.E. Van den Steen, J. Van Damme, G. Opdenakker, D.J. Harvey, R.A. Dwek, P.M. Rudd, OGlycan analysis of natural human neutrophil gelatinase B using a combination of normal phase- HPLC and online tandem mass spectrometry: implications for the domain organization of the enzyme, Biochemistry 39 (2000) 15695–15704. P.M. Rudd, T.S. Mattu, S. Masure, T. Bratt, P.E. Van den Steen, M.R. Wormald, B. Küster, D.J. Harvey, N. Borregaard, J. Van Damme, R.A. Dwek, G. Opdenakker, Glycosylation of natural human neutrophil gelatinase B and neutrophil gelatinase B-associated lipocalin, Biochemistry 38 (1999) 13937–13950. P.E. Van den Steen, B. Dubois, I. Nelissen, P.M. Rudd, R.A. Dwek, G. Opdenakker, Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9), Crit. Rev. Biochem. Mol. Biol. 37 (2002) 375–536. E. Morgunova, A. Tuuttila, U. Bergmann, M. Isupov, Y. Lindqvist, G. Schneider, K. Tryggvason, Structure of human pro-matrix metalloproteinase-2: activation mechanism revealed, Science 284 (1999) 1667–1670. G. Opdenakker, P.E. Van den Steen, J. Van Damme, Gelatinase B: a tuner and amplifier of immune functions, Trends Immunol. 22 (2001) 571–579. S. Masure, P. Proost, J. Van Damme, G. Opdenakker, Purification and identification of 91-kDa neutrophil gelatinase. Release by the activating peptide interleukin-8, Eur. J. Biochem. 198 (1991) 391–398. L. Kjeldsen, A.H. Johnsen, H. Sengelov, N. Borregaard, Isolation and primary structure of NGAL, a novel protein associated with human neutrophil gelatinase, J. Biol. Chem. 268 (1993) 10425–10432. P.E. Van den Steen, P. Proost, A. Wuyts, J. Van Damme, G. Opdenakker, Neutrophil gelatinase B potentiates interleukin-8 tenfold by aminoterminal processing, whereas it degrades CTAP-III, PF-4, and GRO-alpha and leaves RANTES and MCP-2 intact, Blood 96 (2000) 2673–2681. J. Saarinen, H.G. Welgus, C.A. Flizar, N. Kalkkinen, J. Helin, N-glycan structures of matrix metalloproteinase-1 derived from human fibroblasts and from HT-1080 fibrosarcoma cells, Eur. J. Biochem. 259 (1999) 829–840. J. Hu, P. Fiten, P.E. Van den Steen, P. Chaltin, G. Opdenakker, Simulation of evolution-selected propeptide by high-throughput selection of a peptidomimetic inhibitor on a capillary DNA sequencer platform, Anal. Chem. 77 (2005) 2116–2124. R. Fridman, M. Toth, D. Pena, S. Mobashery, Activation of progelatinase B (MMP-9) by gelatinase A (MMP-2), Cancer Res. 55 (1995) 2548–2555. Y. Ogata, J.J. Enghild, H. Nagase, Matrix metalloproteinase 3 (stromelysin) activates the precursor for the human matrix metalloproteinase 9, J. Biol. Chem. 267 (1992) 3581–3584. F. Monzani, G. Meucci, N. Caraccio, M. Saviozzi, A. Casolaro, G. Moscato, F. Lombardo, S. Mosti, C. Scagnolari, F. Bruschi, G. Antonelli, E. Ferrannini, L. Murri, Discordant effect of IFN-beta1a therapy on antiIFN antibodies and thyroid disease development in patients with multiple sclerosis, J. Interferon Cytokine Res. 22 (2002) 773–781. K. Gijbels, S. Masure, H. Carton, G. Opdenakker, Gelatinase in the cerebrospinal fluid of patients with multiple sclerosis and other inflammatory neurological disorders, J. Neuroimmunol. 41 (1992) 29–34. D. Leppert, J. Ford, G. Stabler, C. Grygar, C. Lienert, S. Huber, K.M. Miller, S.L. Hauser, L. Kappos, Matrix metalloproteinase-9 (gelatinase B) is selectively elevated in CSF during relapses and stable phases of multiple sclerosis, Brain 121 (1998) 2327–2334. I. Nelissen, E. Martens, P.E. Van den Steen, P. Proost, I. Ronsse, G. Opdenakker, Gelatinase B/matrix metalloproteinase-9 cleaves interferon-β and is a target for immunotherapy, Brain 126 (2003) 1371–1381.
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615 [42] P.E. Desrochers, K. Mookhtiar, H.E. Van Wart, K.A. Hasty, S.J. Weiss, Proteolytic inactivation of α1-proteinase inhibitor and α1-antichymotrypsin by oxidatively activated human neutrophil metalloproteinases, J. Biol. Chem. 267 (1992) 5005–5012. [43] U. Schönbeck, F. Mach, P. Libby, Generation of biologically active IL-1b by matrix metalloproteinases: a novel caspase-1-independent pathway of IL-1β processing, J. Immunol. 161 (1998) 3340–3346. [44] A.J. Gearing, P. Beckett, M. Christodoulou, M. Churchill, J. Clements, A.H. Davidson, A.H. Drummond, W.A. Galloway, R. Gilbert, J.L. Gordon, T.M. Leber, M. Mangan, K. Miller, P. Nayee, K. Owen, S. Patel, W. Thomas, G. Wells, L.M. Wood, K. Woolley, Processing of tumour necrosis factor-α precursor by metalloproteinases, Nature 370 (1994) 555–557. [45] A.J.H. Gearing, P. Beckett, M. Christodoulou, M. Churchill, J.M. Clements, M. Crimmin, A.H. Davidson, A.H. Drummond, W.A. Galloway, R. Gilbert, J.L. Gordon, T.M. Leber, M. Mangan, K. Miller, P. Nayee, K. Owen, S. Patel, W. Thomas, G. Wells, L.M. Wood, K. Woolley, Matrix metalloproteinases and processing of proTNF-α, J. Leukoc. Biol. 57 (1995) 774–777. [46] D. Chabas, S.E. Baranzini, D. Mitchell, C.C. Bernard, S.R. Rittling, D.T. Denhardt, R.A. Sobel, C. Lock, M. Karpuj, R. Pedotti, R. Heller, J.R. Oksenberg, L. Steinman, The influence of the proinflammatory cytokine, osteopontin, on autoimmune demyelinating disease, Science 294 (2001) 1731–1735. [47] J.M. van Noort, A.C. van Sechel, J.J. Bajramovic, M. el Ouagmiri, C.H. Polman, H. Lassmann, R. Ravid, The small heat-shock protein αBcrystallin as candidate autoantigen in multiple sclerosis, Nature 375 (1995) 798–801. [48] S. Starckx, P.E. Van den Steen, R. Verbeek, J.M. van Noort, G. Opdenakker, A novel rationale for inhibition of gelatinase B in multiple sclerosis: MMP-9 destroys αB-crystallin and generates a promiscuous T cell epitope, J. Neuroimmunol. 141 (2003) 47–57.
615
[49] Y.K. Chou, G.G. Burrows, D. LaTocha, C. Wang, S. Subramanian, D.N. Bourdette, A.A. Vandenbark, CD4 T-cell epitopes of human αB-crystallin, J. Neurosci. Res. 75 (2004) 516–523. [50] P.E. Van den Steen, P. Proost, D.D. Brand, A.H. Kang, J. Van Damme, G. Opdenakker, Generation of glycosylated remnant epitopes from human collagen type II by gelatinase B, Biochemistry 43 (2004) 10809–10816. [51] E. Michaëlsson, V. Malmstrom, S. Reis, A. Engstrom, H. Burkhardt, R. Holmdahl, T cell recognition of carbohydrates on type II collagen, J. Exp. Med. 180 (1994) 745–749. [52] J. Bäcklund, S. Carlsen, T. Hoger, B. Holm, L. Fugger, J. Kihlberg, H. Burkhardt, R. Holmdahl, Predominant selection of T cells specific for the glycosylated collagen type II epitope (263–270) in humanized transgenic mice and in rheumatoid arthritis, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 9960–9965. [53] L.K. Myers, J. Myllyharju, M. Nokelainen, D.D. Brand, M.A. Cremer, J.M. Stuart, M. Bodo, K.I. Kivirikko, A.H. Kang, Relevance of posttranslational modifications for the arthritogenicity of type II collagen, J. Immunol. 172 (2004) 2970–2975. [54] G.Y. Ishioka, A.G. Lamont, D. Thomson, N. Bulbow, F.C. Gaeta, A. Sette, H.M. Grey, MHC interaction and T cell recognition of carbohydrates and glycopeptides, J. Immunol. 148 (1992) 2446–2451. [55] J.S. Haurum, G. Arsequell, A.C. Lellouch, S.Y. Wong, R.A. Dwek, A.J. McMichael, T. Elliott, Recognition of carbohydrate by major histocompatibility complex class I-restricted, glycopeptide-specific cytotoxic T lymphocytes, J. Exp. Med. 180 (1994) 739–744. [56] A. Glithero, J. Tormo, J.S. Haurum, G. Arsequell, G. Valencia, J. Edwards, S. Springer, A. Townsend, Y.L. Pao, M. Wormald, R.A. Dwek, E.Y. Jones, T. Elliott, Crystal structures of two H-2Db/glycopeptide complexes suggest a molecular basis for CTL cross-reactivity, Immunity 10 (1999) 63–74. [57] J.B. Lowe, Glycosylation, immunity, and autoimmunity, Cell 104 (2001) 809–812.
Review
Remnant epitopes, autoimmunity and glycosylation Ghislain Opdenakker a,⁎, Chris Dillen a , Pierre Fiten a , Erik Martens a , Ilse Van Aelst a , Philippe E. Van den Steen a , Inge Nelissen a , Sofie Starckx a , Francis J. Descamps a , Jialiang Hu a , Helene Piccard a , Jo Van Damme b , Mark R. Wormald c , Pauline M. Rudd c , Raymond A. Dwek c b
a Rega Institute for Medical Research, Laboratory of Immunobiology, University of Leuven, Minderbroedersstraat 10, B-3000 Leuven, Belgium Rega Institute for Medical Research, Laboratory of Molecular Immunology, University of Leuven, Minderbroedersstraat 10, B-3000 Leuven, Belgium c Glycobiology Institute, University of Oxford, South Parks Road, Oxford OX1 3QU, UK
Received 21 October 2005; received in revised form 12 December 2005; accepted 12 December 2005 Available online 6 January 2006
Abstract The role of extracellular proteolysis in innate and adaptive immunity and the interplay between cytokines, chemokines and proteinases are gradually becoming recognized as critical factors in autoimmune processes. Many of the involved proteinases, including those of the plasminogen activator and matrix metalloproteinase cascades, and also several cytokines and chemokines, are glycoproteins. The stability, interactions with inhibitors or receptors, and activities of these molecules are fine-controlled by glycosylation. We studied gelatinase B or matrix metalloproteinase9 (MMP-9) as a glycosylated enzyme involved in autoimmunity. In the joints of rheumatoid arthritis patients, CXC chemokines, such as interleukin-8/CXCL8, recruit and activate neutrophils to secrete prestored neutrophil collagenase/MMP-8 and gelatinase B/MMP-9. Gelatinase B potentiates interleukin-8 at least tenfold and thus enhances neutrophil and lymphocyte influxes to the joints. When cartilage collagen type II is cleaved at a unique site by one of several collagenases (MMP-1, MMP-8 or MMP-13), it becomes a substrate of gelatinase B. Human gelatinase B cleaves the resulting two large collagen fragments into at least 33 peptides of which two have been shown to be immunodominant, i.e., to elicit activation and proliferation of autoimmune T cells. One of these two remnant epitopes contains a glycan which is important for its immunoreactivity. In addition to the role of gelatinase B as a regulator in adaptive immune processes, we have also demonstrated that it destroys interferon-β, a typical innate immunity effector molecule and therapeutic cytokine in multiple sclerosis. Furthermore, glycosylated interferon-β, expressed in Chinese hamster ovary cells, was more resistant to this proteolysis than recombinant interferon-β from bacteria. These data not only prove that glycosylation of proteins is mechanistically important in the pathogenesis of autoimmune diseases, but also show that targeting of glycosylated proteinases or the use of glycosylated cytokines seems also critical for the treatment of autoimmune diseases. © 2005 Elsevier B.V. All rights reserved. Keywords: Gelatinase B/MMP-9; Remnant epitope; TIMP-1; Collagenase; Neutrophil
1. Introduction To understand autoimmune processes, it is critical to know that autoreactivity is normal. The generation of antibodies and T cell receptors is based on genetic principles in which multiple gene segments are mutated and recombined, thus encoding a multifold of (receptor) molecules, called the immunologic repertoire [1,2]. Some of these gene products possess intrinsic reactivity to (parts of) host molecules. Many of the ⁎ Corresponding author. E-mail address: [email protected] (G. Opdenakker). 0304-4165/$ - see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bbagen.2005.12.014
lymphocytes, but certainly not all, that encode self-reactive receptors are eliminated by control processes of tolerance induction [3]. It has been estimated that, after passage through the thymus, about 10% of the T lymphocytes are self-reactive [4]. The mechanisms of peripheral tolerance [5,6], e.g., ignorance and anergy, further reduce the chance of developing autoimmune diseases. Ignorance is best defined as the presence of anatomical barriers [7], such as the blood–brain barrier, which prevent the interaction of the self-reactive lymphocyte with the presented antigen, whereas anergy [8] is the absence of immune reaction in the absence of a second stimulus (infection or inflammation).
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
Autoimmune diseases are multifactorial [9], i.e., they are based on the interplay between both genetic constitution and environment of the host. T cell reactive autoimmune antigens are presented within the grooves of the highly polymorphic major histocompatibility complex (MHC) molecules [10], and association with the MHC locus (predisposing and protective) has been demonstrated for autoimmune diseases, including multiple sclerosis (MS) [11], rheumatoid arthritis [12], diabetes [13] and inflammatory bowel disease [14]. The environmental factors in autoimmune diseases are much less clear so far. The study of host cells by (immuno) histopathology and host molecules by biological and biochemical analysis forms the best basis to understand the pathophysiology of autoimmunity. 2. The REGA model We have studied host cells and molecules in autoimmune diseases to define the important role of cytokines, chemokines and extracellular proteases in autoimmunity [15]. Inflammatory processes lead to the production of cytokines and chemokines, which recruit and stimulate cells to release proteinases. These include serine proteinases (e.g., plasminogen activators, granzymes, neutrophil elastase and plasmin) that act together in a cascade with the matrix metalloproteinases (MMPs). This cascade forms an efficient way of amplifying the activity of the enzymes in the chain, as each enzyme continues to activate its substrate. In common with other cascades, such as the complement and blood clotting systems, the extracellular matrix degrading cascade is kept under control by specific inhibitors [15]. The action of the extracellular proteinases results in the proteolysis of matrix molecules or other substrates, and leads to quantitative and qualitative changes of the “antigen repertoire”. These Remnant Epitopes Generate Autoimmunity, hence, the process was named the REGA model [15]. When a substrate is cleaved at multiple sites, a molar excess of peptides becomes available for processing and presentation by classical or nonclassical pathways. Furthermore, proteolysis might liberate hidden (cryptic) antigens from a folded protein, thus adding to qualitative changes in the “antigen repertoire” (Fig. 1). From the scheme in Fig. 1, it can be deduced that inhibition of target proteinases may be beneficial in autoimmune diseases. The REGA model has been tested in vivo with the use of animal models. In comparison with wildtype animals, young gelatinase B/MMP-9-deficient mice were resistant to the development of experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis [16], and to autoimmune arthritis induced by antibodies against collagen [17]. These experiments form a good basis for the development of gelatinase B inhibitors for the treatment of multiple sclerosis and rheumatoid arthritis [18]. For example, MMP inhibitors have been used with success in the treatment of EAE [19,20].
611
Fig. 1. The role of cytokines and extracellular proteinases in the REGA model. The trigger of an autoimmune process is not known, but studies of molecular and cellular changes in patients and animal models have pointed to the involvement of cytokines, chemokines and proteinases in autoimmune processes. Many cytokines enhance inflammation and chemokines recruit and activate leukocytes. This leads to the secretion of proteinases that cleave (glyco)proteins into remnant epitopes. Examples of substrates yielding such remnant epitopes generating autoimmunity (REGA) include myelin basic protein, αB-crystallin and, possibly also, interferon-β in multiple sclerosis and type II collagen in rheumatoid arthritis. As a consequence, T cell activation maintains the process by generating proinflammatory cytokines. The REGA model has therapeutic consequences, since inhibition of proteinases, such as gelatinase B, results in beneficial effects [18,30]. Figure modified from reference 15.
3. The glycoprotein gelatinase B: structural and functional aspect of the sugars 3.1. Gelatinase B as a complex glycosylated matrix metalloproteinase MMP-9 contains a prodomain with an occupied N-linked sequon (Asn38–Leu–Thr) and an active domain with a second N-linked oligosaccharide (located at Asn120–Ile–Thr) [21,22]. The Zn2+-binding domain contains three conserved histidines for the coordination of the catalytic Zn2+-ion. Together with parts of the active site domain, this constitutes the catalytic site, of which the three-dimensional structure is highly similar in the different MMPs [23] (Fig. 2a). The active site has a cleft in which the (Zn2+-binding) substrate can bind before it is cleaved. In the proenzyme, this cleft is occupied by the propeptide with a Zn2+-binding cysteine residue, which prevents catalytic activity. After the proteolytic removal of the propeptide, a water molecule binds to the Zn2+-ion allowing the catalysis to proceed [24,25]. In gelatinases, three fibronectin repeats, positioned between the active site and the Zn2+-binding domain, increase the affinity for gelatin. Gelatinase B contains an extended threonine–proline–serine-rich region between the aminoterminal active site and the carboxyterminal hemopexin domain. This region is extensively O-glycosylated in neutrophil gelatinase B [21,26,27]. Gelatinases are, in terms of domain structure, already rather complex MMPs [28]. By comparison of the structure of unglycosylated gelatinase A/MMP-2 [29] with a model of glycosylated gelatinase B [30], it also became clear that both molecules are structurally quite different from each other. Gelatinase B is thus not only the most complex MMP in
612
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
Fig. 2. Molecular models of gelatinase B from human neutrophils. Neutrophilic granulocytes secrete three forms of gelatinase B: monomers, covalent homodimers and covalent complexes with NGAL [32,33]. In panel a, a view of the glycosylated monomer is presented. This model is based on all available protein and carbohydrate structures of gelatinases A and B. The peptide is shown in green, the N-linked glycans in dark blue and the O-linked glycans in brown. The N-terminal and C-terminal domains are based on the crystal structure of the N-terminal part of MMP-9 [23] and the crystal structure of the C-terminal domain of MMP-2 [29]. Analytical ultracentrifugation (AUC) studies (unpublished data) show that the O-link domain connecting the two is not extended and the whole molecule is relatively compact. Panel b. AUC studies show that the dimer is compact and overall more spherical than the monomers, and that dimerisation does not lead to an extension of the O-link domain. The AUC results can be reproduced by the side-to-side interaction of two compact monomers, although there are other models of the dimerisation that would also fit the experimental data.
terms of domain structure, it is also – as far as is known – the most extensively glycosylated MMP. In view of the heterogeneity and complexity of the attached sugars [27], so far, only mixtures of glycoforms/glycotypes of gelatinase B have been obtained. This might constitute one of the possible explanations why attempts to crystallize intact gelatinase B for detailed structural analysis have failed so far.
About 93% of the N-linked oligosaccharides contain core fucose. Human neutrophil gelatinase B is significantly less sialylated (9%) than MMP-1 from fibroblasts and HT-1080 cells (50% and 34% respectively). Human neutrophil gelatinase B and MMP-1 from both sources contain glycans with outer arm fucose-linked α1–3 to GlcNAc [26,27,34]. 3.3. Functional analysis of the glycans of gelatinase B
3.2. Structures of the glycans of neutrophil gelatinase B Natural gelatinase B can be purified from human neutrophils without any degradation or activation and in reasonable quantities [31]. Upon stimulation, e.g., with interleukin-8, neutrophils release large amounts of gelatinase B and do not secrete gelatinase A or TIMP-1. This is in contrast with many other cell types that constitutively produce gelatinase A and coexpress TIMP-1 and gelatinase B upon stimulation. This makes the purification procedure straightforward for neutrophil gelatinase B. Three different forms of neutrophil gelatinase B can be distinguished: monomers, disulfide-bound dimers and a covalent complex of gelatinase B with neutrophil gelatinase Bassociated lipocalin (NGAL). The latter complex can be removed using monoclonal antibodies against NGAL [32,33]. A model of MMP-9 as a dimer is presented in Fig. 2b. Approximately 85% of the sugars of human neutrophil gelatinase B are O-linked structures attached to 14 potential Oglycosylation sites in a stretch of about 50 amino acids. Most of these sugars are type 2 structures with Galβ1-4GlcNAc (Nacetyl-lactosamine) extensions, with or without sialic acid and fucose. The two N-glycans are about 80% complex biantennary structures and 20% are tri-antennary structures.
The glycosylation of the propeptide of gelatinase B raises the possibility of a function for such a glycan in protein folding and regulating enzyme activation and activity [21]. Since the propeptide mimics a peptide inhibitor [35] and glycosylation protects glycoproteins against degradation, future experiments are directed towards testing the inhibitory activity of a glycosylated propeptide against the activities of gelatinase B and other MMPs. Gelatinase B can be activated by various proteases, including gelatinase A (MMP-2) [36] and stromelysin-1 (MMP-3) [37]. Therefore, we studied the influence of N-deglycosylation and desialylation on the activation of progelatinase B and on the activity of the active form. No significant differences were detected between the N-deglycosylated gelatinase B forms and native gelatinase B. Also, no differences were found for the activation by gelatinase A or stromelysin-1 of N-deglycosylated or desialylated progelatinase B, compared with native gelatinase B, in zymography and fluorogenic substrate conversion assay [21]. Therefore, neither the N-linked glycan on the prodomain nor the sialic acids that terminate many of the glycans attached to gelatinase B have any direct influence on the activity and activation of gelatinase B.
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
Recent data using recombinant variants of gelatinase B, with or without the O-glycosylated domain, indicate that this domain is not critical for the hydrolysis of more than 10 different known substrates (P. Van den Steen et al., unpublished). The glycans of gelatinase B contain terminal sialic acids, α2–3 or α2–6 linked both on N-linked and on O-linked glycans [26,27]. These sialic acids were removed with Streptococcus sp. sialidase. Sialic acids often influence the interaction/recognition between glycoproteins. Therefore, the sensitivity of activated desialylated gelatinase B to inhibition by TIMP-1 was compared with the activated intact glycosylated MMP-9 in an inhibition assay. Desialylated gelatinase B was significantly less inhibited by TIMP-1 and this enzyme form thus possessed higher activity than the intact glycoprotein, in the presence of TIMP-1. Neither deglycosylation of Asn38 in the propeptide nor desialylation of gelatinase B glycans influence the activation rate of natural gelatinase B by gelatinase A and stromelysin-1 [21]. 4. Gelatinase B cleaves molecules of the innate and adaptive immune system 4.1. Gelatinase B destroys interferon-β and potentiates specific cytokines and chemokines One of the molecules currently used in the treatment of multiple sclerosis is interferon-β. Unfortunately, its mechanism of action is not understood. Patients with multiple sclerosis are treated with rather high doses of interferon-β, either produced in recombinant form in bacteria (Betaferon) or Chinese hamster ovary cells (Avonex, Rebif). Of note, the unglycosylated Betaferon induces neutralizing antibodies (which prevent further use as a therapeutic) in about 30% of treated patients, whereas the glycosylated interferon-β from CHO cells is almost devoid of such induction [38]. Although we do not know whether this is related to differences in glycosylation, aggregation, stability or other reasons, we discovered that gelatinase B readily cleaves some aminoterminal residues from the unglycosylated interferon-β, whereas the glycosylated interferon-β is resistant to this proteolysis. If we add the well established information that gelatinase B levels and activity are increased and TIMP-1 levels are decreased in patients with multiple sclerosis [18,39,40], it is possible that remnant epitopes of interferon-β are much more easily formed in patients treated with Betaferon versus Avonex or Rebif. On the basis of a three-dimensional model of glycosylated interferon-β and taking into account that N-linked sugars can protect relatively large protein surfaces, we proposed a mechanism for the aminoterminal inactivation of unglycosylated recombinant interferon-β and the role of the attached sugars in resistance against proteolysis [41]. In addition to the degradation of interferon-β, gelatinase B cleaves other molecules with functions in the immune system, such as α1-proteinase inhibitor, thereby contributing to a more positive balance of proteases versus inhibitors [42], and cytokines, such as pro-IL-1β [43] and pro-TNF-α [44,45].
613
Further, gelatinase B potentiates IL-8 by aminoterminal clipping, thereby providing a positive feedback loop as IL8 induces the degranulation of gelatinase B from neutrophils [33]. 4.2. Gelatinase B generates remnant epitopes Prototypic immunodominant proteins in multiple sclerosis and rheumatoid arthritis are αB-crystallin and collagen type II, respectively. Recent microarray analysis demonstrated that αB-crystallin mRNA was the most upregulated mRNA in MS lesions compared with control brain tissue [46]. A decade ago, the αB-crystallin protein was described as an autoantigen in multiple sclerosis [47]. We found proof of concept of the REGA model with αB-crystallin [48]. In fact, the cleavage of αB-crystallin by gelatinase B resulted in remnant peptides, which corresponded to the immunodominant T cell activating peptides in Biozzi ABH, SJL and 129 mouse strains [48] and also to another immunodominant epitope in humans [49]. In rheumatoid arthritis, cartilage collagen type II is degraded by neutrophil collagenase/MMP-8 into an aminoterminal 1/4th and a carboxyterminal 3/4th fragment [50]. This single cleavage leads to a local unwinding and generates an efficient substrate for gelatinase B. The gelatin is degraded into more than 30 fragments, at least two of which are immunodominant. Earlier [51] and recent experiments [52,53] showed that at least one immunodominant remnant epitope is glycosylated. This example extends the important earlier findings [54–56] that not only peptides, but also glycopeptides may be presented in MHC molecules by antigen presenting cells and thus activate autoreactive T lymphocytes, thereby maintaining or enhancing the autoimmune process. 5. Conclusions Glycosylation and recognition of oligosaccharides by selectins have been recognized as a control mechanism of autoimmunity [57]. Here, we extend this thesis with novel mechanisms and additional examples. Glyco-enzymes and glycosylated cytokines are critical regulator and effector molecules in autoimmune diseases, such as multiple sclerosis and rheumatoid arthritis. The activity and stability of these molecules are fine-controlled by glycosylation and this might have therapeutic implications. Finally, glycopeptides are natural remnant epitopes in human autoimmune processes. Thus, the importance of glycobiology in autoimmune diseases is clear at all levels: the innate and adaptive immune system, T cell receptors, antibodies and autoantigens. Acknowledgements Supported by Fortis AB, the National Fund for Scientific Research (F.W.O.-Vlaanderen), the Geconcerteerde OnderzoeksActies (2002–2006), and the Centre of Excellence (EF/ 05/015) Belgium, and the Glycobiology Institute, Oxford, UK
614
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615
endownment. PVdS is a postdoctoral research fellow of the F. W.O.-Vlaanderen.
[24]
References [1] A.C. Hayday, D.J. Diamond, G. Tanigawa, J.S. Heilig, V. Folsom, H. Saito, S. Tonegawa, Unusual organization and diversity of T-cell receptor alpha-chain genes, Nature 316 (1985) 828–832. [2] M.H. Bakkus, Ig gene sequences in the study of clonality, Pathol. Biol. (Paris) 47 (1999) 128–147. [3] S.P. Cobbold, E. Adams, S.E. Marshall, J.D. Davies, H. Waldmann, Mechanisms of peripheral tolerance and suppression induced by monoclonal antibodies to CD4 and CD8, Immunol. Rev. 149 (1996) 5–33. [4] K.A. Hogquist, T.A. Baldwin, S.C. Jameson, Central tolerance: learning self-control in the thymus, Nat. Rev., Immunol. 5 (2005) 772–782. [5] B. Arnold, Levels of peripheral T cell tolerance, Transpl. Immunol. 10 (2002) 109–114. [6] B. Arnold, G. Schonrich, G.J. Hämmerling, Multiple levels of peripheral tolerance, Immunol. Today 14 (1993) 12–14. [7] K.W. Wucherpfennig, Autoimmunity in the central nervous system: mechanisms of antigen presentation and recognition, Clin. Immunol. Immunopathol. 72 (1994) 293–306. [8] J. Lohr, B. Knoechel, V. Nagabhushanam, A.K. Abbas, T-cell tolerance and autoimmunity to systemic and tissue-restricted self-antigens, Immunol. Rev. 204 (2005) 116–127. [9] J.M. van Noort, S. Amor, Cell biology of autoimmune diseases, Int. Rev. Cytol. 178 (1998) 127–206. [10] H.O. McDevitt, Discovering the role of the major histocompatibility complex in the immune response, Annu. Rev. Immunol. 18 (2000) 1–17. [11] L. Steinman, Multiple sclerosis: a coordinated immunological attack against myelin in the central nervous system, Cell 85 (1996) 299–302. [12] R. Holmdahl, Dissection of the genetic complexity of arthritis using animal models, J. Autoimmun. 21 (2003) 99–103. [13] H.J. Cordell, J.A. Todd, Multifactorial inheritance in type 1 diabetes, Trends Genet. 11 (1995) 499–504. [14] D.A. Watts, J. Satsangi, The genetic jigsaw of inflammatory bowel disease, Gut (Suppl. 3) (2002) 31–36. [15] G. Opdenakker, J. Van Damme, Cytokine-regulated proteases in autoimmune diseases, Immunol. Today 15 (1994) 103–107. [16] B. Dubois, S. Masure, U. Hurtenbach, L. Paemen, H. Heremans, J. van den Oord, R. Sciot, T. Meinhardt, G. Hämmerling, G. Opdenakker, B. Arnold, Resistance of young gelatinase B-deficient mice to experimental autoimmune encephalomyelitis and necrotizing tail lesions, J. Clin. Invest. 104 (1999) 1507–1515. [17] T. Itoh, H. Matsuda, M. Tanioka, K. Kuwabara, S. Itohara, R. Suzuki, The role of matrix metalloproteinase-2 and matrix metalloproteinase-9 in antibody-induced arthritis, J. Immunol. 169 (2002) 2643–2647. [18] G. Opdenakker, I. Nelissen, J. Van Damme, Functional roles and therapeutic targeting of gelatinase B and chemokines in multiple sclerosis, Lancet Neurol. 2 (2003) 747–756. [19] K. Gijbels, R.E. Galardy, L. Steinman, Reversal of experimental autoimmune encephalomyelitis with a hydroxamate inhibitor of matrix metalloproteases, J. Clin. Invest. 94 (1994) 2177–2182. [20] A.K. Hewson, T. Smith, J.P. Leonard, M.L. Cuzner, Suppression of experimental allergic encephalomyelitis in the Lewis rat by the matrix metalloproteinase inhibitor Ro31-9790, Inflamm. Res. 44 (1995) 345–349. [21] P.E. Van den Steen, G. Opdenakker, M.R. Wormald, R.A. Dwek, P.M. Rudd, Matrix remodelling enzymes, the protease cascade and glycosylation, Biochim. Biophys. Acta 1528 (2001) 61–73. [22] L.P. Kotra, L. Zhang, R. Fridman, R. Orlando, S. Mobashery, NGlycosylation pattern of the zymogenic form of human matrix metalloproteinase-9, Bioorg. Chem. 30 (2002) 356–370. [23] P.A. Elkins, Y.S. Ho, W.W. Smith, C.A. Janson, K.J. D'Alessio, M.S. McQueney, M.D. Cummings, A.M. Romanic, Structure of the C-
[25]
[26]
[27]
[28]
[29]
[30] [31]
[32]
[33]
[34]
[35]
[36] [37]
[38]
[39]
[40]
[41]
terminally truncated human ProMMP9, a gelatin-binding matrix metalloproteinase, Acta Crystallogr., D Biol. Crystallogr. 58 (2002) 1182–1192. H.E. Van Wart, H. Birkedal-Hansen, The cysteine switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene family, Proc. Natl. Acad. Sci. U. S. A. 87 (1990) 5578–5582. O. Kleifeld, P.E. Van den Steen, A. Frenkel, F. Cheng, H.L. Jiang, G. Opdenakker, I. Sagi, Structural characterization of the catalytic active site in the latent and active natural gelatinase B from human neutrophils, J. Biol. Chem. 275 (2000) 34335–34343. T.S. Mattu, L. Royle, J. Langridge, M.R. Wormald, P.E. Van den Steen, J. Van Damme, G. Opdenakker, D.J. Harvey, R.A. Dwek, P.M. Rudd, OGlycan analysis of natural human neutrophil gelatinase B using a combination of normal phase- HPLC and online tandem mass spectrometry: implications for the domain organization of the enzyme, Biochemistry 39 (2000) 15695–15704. P.M. Rudd, T.S. Mattu, S. Masure, T. Bratt, P.E. Van den Steen, M.R. Wormald, B. Küster, D.J. Harvey, N. Borregaard, J. Van Damme, R.A. Dwek, G. Opdenakker, Glycosylation of natural human neutrophil gelatinase B and neutrophil gelatinase B-associated lipocalin, Biochemistry 38 (1999) 13937–13950. P.E. Van den Steen, B. Dubois, I. Nelissen, P.M. Rudd, R.A. Dwek, G. Opdenakker, Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9), Crit. Rev. Biochem. Mol. Biol. 37 (2002) 375–536. E. Morgunova, A. Tuuttila, U. Bergmann, M. Isupov, Y. Lindqvist, G. Schneider, K. Tryggvason, Structure of human pro-matrix metalloproteinase-2: activation mechanism revealed, Science 284 (1999) 1667–1670. G. Opdenakker, P.E. Van den Steen, J. Van Damme, Gelatinase B: a tuner and amplifier of immune functions, Trends Immunol. 22 (2001) 571–579. S. Masure, P. Proost, J. Van Damme, G. Opdenakker, Purification and identification of 91-kDa neutrophil gelatinase. Release by the activating peptide interleukin-8, Eur. J. Biochem. 198 (1991) 391–398. L. Kjeldsen, A.H. Johnsen, H. Sengelov, N. Borregaard, Isolation and primary structure of NGAL, a novel protein associated with human neutrophil gelatinase, J. Biol. Chem. 268 (1993) 10425–10432. P.E. Van den Steen, P. Proost, A. Wuyts, J. Van Damme, G. Opdenakker, Neutrophil gelatinase B potentiates interleukin-8 tenfold by aminoterminal processing, whereas it degrades CTAP-III, PF-4, and GRO-alpha and leaves RANTES and MCP-2 intact, Blood 96 (2000) 2673–2681. J. Saarinen, H.G. Welgus, C.A. Flizar, N. Kalkkinen, J. Helin, N-glycan structures of matrix metalloproteinase-1 derived from human fibroblasts and from HT-1080 fibrosarcoma cells, Eur. J. Biochem. 259 (1999) 829–840. J. Hu, P. Fiten, P.E. Van den Steen, P. Chaltin, G. Opdenakker, Simulation of evolution-selected propeptide by high-throughput selection of a peptidomimetic inhibitor on a capillary DNA sequencer platform, Anal. Chem. 77 (2005) 2116–2124. R. Fridman, M. Toth, D. Pena, S. Mobashery, Activation of progelatinase B (MMP-9) by gelatinase A (MMP-2), Cancer Res. 55 (1995) 2548–2555. Y. Ogata, J.J. Enghild, H. Nagase, Matrix metalloproteinase 3 (stromelysin) activates the precursor for the human matrix metalloproteinase 9, J. Biol. Chem. 267 (1992) 3581–3584. F. Monzani, G. Meucci, N. Caraccio, M. Saviozzi, A. Casolaro, G. Moscato, F. Lombardo, S. Mosti, C. Scagnolari, F. Bruschi, G. Antonelli, E. Ferrannini, L. Murri, Discordant effect of IFN-beta1a therapy on antiIFN antibodies and thyroid disease development in patients with multiple sclerosis, J. Interferon Cytokine Res. 22 (2002) 773–781. K. Gijbels, S. Masure, H. Carton, G. Opdenakker, Gelatinase in the cerebrospinal fluid of patients with multiple sclerosis and other inflammatory neurological disorders, J. Neuroimmunol. 41 (1992) 29–34. D. Leppert, J. Ford, G. Stabler, C. Grygar, C. Lienert, S. Huber, K.M. Miller, S.L. Hauser, L. Kappos, Matrix metalloproteinase-9 (gelatinase B) is selectively elevated in CSF during relapses and stable phases of multiple sclerosis, Brain 121 (1998) 2327–2334. I. Nelissen, E. Martens, P.E. Van den Steen, P. Proost, I. Ronsse, G. Opdenakker, Gelatinase B/matrix metalloproteinase-9 cleaves interferon-β and is a target for immunotherapy, Brain 126 (2003) 1371–1381.
G. Opdenakker et al. / Biochimica et Biophysica Acta 1760 (2006) 610–615 [42] P.E. Desrochers, K. Mookhtiar, H.E. Van Wart, K.A. Hasty, S.J. Weiss, Proteolytic inactivation of α1-proteinase inhibitor and α1-antichymotrypsin by oxidatively activated human neutrophil metalloproteinases, J. Biol. Chem. 267 (1992) 5005–5012. [43] U. Schönbeck, F. Mach, P. Libby, Generation of biologically active IL-1b by matrix metalloproteinases: a novel caspase-1-independent pathway of IL-1β processing, J. Immunol. 161 (1998) 3340–3346. [44] A.J. Gearing, P. Beckett, M. Christodoulou, M. Churchill, J. Clements, A.H. Davidson, A.H. Drummond, W.A. Galloway, R. Gilbert, J.L. Gordon, T.M. Leber, M. Mangan, K. Miller, P. Nayee, K. Owen, S. Patel, W. Thomas, G. Wells, L.M. Wood, K. Woolley, Processing of tumour necrosis factor-α precursor by metalloproteinases, Nature 370 (1994) 555–557. [45] A.J.H. Gearing, P. Beckett, M. Christodoulou, M. Churchill, J.M. Clements, M. Crimmin, A.H. Davidson, A.H. Drummond, W.A. Galloway, R. Gilbert, J.L. Gordon, T.M. Leber, M. Mangan, K. Miller, P. Nayee, K. Owen, S. Patel, W. Thomas, G. Wells, L.M. Wood, K. Woolley, Matrix metalloproteinases and processing of proTNF-α, J. Leukoc. Biol. 57 (1995) 774–777. [46] D. Chabas, S.E. Baranzini, D. Mitchell, C.C. Bernard, S.R. Rittling, D.T. Denhardt, R.A. Sobel, C. Lock, M. Karpuj, R. Pedotti, R. Heller, J.R. Oksenberg, L. Steinman, The influence of the proinflammatory cytokine, osteopontin, on autoimmune demyelinating disease, Science 294 (2001) 1731–1735. [47] J.M. van Noort, A.C. van Sechel, J.J. Bajramovic, M. el Ouagmiri, C.H. Polman, H. Lassmann, R. Ravid, The small heat-shock protein αBcrystallin as candidate autoantigen in multiple sclerosis, Nature 375 (1995) 798–801. [48] S. Starckx, P.E. Van den Steen, R. Verbeek, J.M. van Noort, G. Opdenakker, A novel rationale for inhibition of gelatinase B in multiple sclerosis: MMP-9 destroys αB-crystallin and generates a promiscuous T cell epitope, J. Neuroimmunol. 141 (2003) 47–57.
615
[49] Y.K. Chou, G.G. Burrows, D. LaTocha, C. Wang, S. Subramanian, D.N. Bourdette, A.A. Vandenbark, CD4 T-cell epitopes of human αB-crystallin, J. Neurosci. Res. 75 (2004) 516–523. [50] P.E. Van den Steen, P. Proost, D.D. Brand, A.H. Kang, J. Van Damme, G. Opdenakker, Generation of glycosylated remnant epitopes from human collagen type II by gelatinase B, Biochemistry 43 (2004) 10809–10816. [51] E. Michaëlsson, V. Malmstrom, S. Reis, A. Engstrom, H. Burkhardt, R. Holmdahl, T cell recognition of carbohydrates on type II collagen, J. Exp. Med. 180 (1994) 745–749. [52] J. Bäcklund, S. Carlsen, T. Hoger, B. Holm, L. Fugger, J. Kihlberg, H. Burkhardt, R. Holmdahl, Predominant selection of T cells specific for the glycosylated collagen type II epitope (263–270) in humanized transgenic mice and in rheumatoid arthritis, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 9960–9965. [53] L.K. Myers, J. Myllyharju, M. Nokelainen, D.D. Brand, M.A. Cremer, J.M. Stuart, M. Bodo, K.I. Kivirikko, A.H. Kang, Relevance of posttranslational modifications for the arthritogenicity of type II collagen, J. Immunol. 172 (2004) 2970–2975. [54] G.Y. Ishioka, A.G. Lamont, D. Thomson, N. Bulbow, F.C. Gaeta, A. Sette, H.M. Grey, MHC interaction and T cell recognition of carbohydrates and glycopeptides, J. Immunol. 148 (1992) 2446–2451. [55] J.S. Haurum, G. Arsequell, A.C. Lellouch, S.Y. Wong, R.A. Dwek, A.J. McMichael, T. Elliott, Recognition of carbohydrate by major histocompatibility complex class I-restricted, glycopeptide-specific cytotoxic T lymphocytes, J. Exp. Med. 180 (1994) 739–744. [56] A. Glithero, J. Tormo, J.S. Haurum, G. Arsequell, G. Valencia, J. Edwards, S. Springer, A. Townsend, Y.L. Pao, M. Wormald, R.A. Dwek, E.Y. Jones, T. Elliott, Crystal structures of two H-2Db/glycopeptide complexes suggest a molecular basis for CTL cross-reactivity, Immunity 10 (1999) 63–74. [57] J.B. Lowe, Glycosylation, immunity, and autoimmunity, Cell 104 (2001) 809–812.