Collagen-induced arthritis (CIA) is an experimental model of autoimmune-mediated polyarthritis that can be induced in susceptible strains of mice by immunization with type II collagen (CII), the major constituent protein of articular cartilage (1). CIA shares important clinical, histologic, and immunologic features with rheumatoid arthritis (RA) (2). Both CIA and RA are strongly associated with class II immune response genes. Susceptibility to CIA is associated with H-2q and H-2r, whereas susceptibility to RA is associated with the presence of HLA–DRB*0101 (DR1) and HLA–DRB*0401 (DR4). CIA is caused by autoimmunity to CII. Although autoimmune reactions to CII are also present in RA, their significance is not clear. To determine the role of autoimmunity to CII in RA, it may be necessary to suppress the specific immune response in patients and determine if disease activity is affected.
The development of transgenic mice expressing HLA class II molecules has made it easier to address the question of how to suppress DR-mediated specific immune reactions experimentally. We have previously shown that CIA in H-2q mice can be prevented by injection of a peptide analog of the immunodominant epitope of CII recognized by I-Aq. This peptide apparently acts by causing immune deviation. It is important because it can both prevent the development of CIA when administered prior to immunization and alter the course of disease when given after immunization. Development of a similar peptide that can alter DR-mediated immune responses might provide a means of preventing CII autoimmunity in RA patients.
Our studies with HLA–DR transgenic mice show that the mice are susceptible to CIA, and DR1 and DR4 can bind and present peptides derived from human CII (3, 4). Using proliferation and cytokine assays, we identified CII (263–270) (FKGEQGPK) as the core of the immunodominant T cell determinant presented by HLA–DR1 and DR4. Based on these data, we developed synthetic analog peptides containing substitutions at selected residues designed to down-regulate the immune response to CII and CIA in the context of DR1. These peptides were tested for their ability to prevent arthritis in DR1 transgenic mice.
- Top of page
- MATERIALS AND METHODS
The recent development of transgenic mice expressing HLA class II molecules has made it possible to study the DR1-restricted immune responses to collagen in a more systematic manner. Some investigators have been concerned that T cell responses generated by the mouse may not be representative of those generated in humans. Using an HLA–A2.1 transgenic mouse model and T cell responses to a hepatitis C virus, clones from humans (bearing HLA–A2) infected with hepatitis were compared with T cell responses of clones from HLA–A2 transgenic mice infected with the same virus (11). The T cell repertoire was flexible enough to allow a similar, almost identical, response when the same major histocompatibility complex (MHC) molecule was presenting the peptide (11). These data suggest that the HLA molecule plays the primary role in determining which peptides are recognized by T cells and that the transgenic mouse model is a valid model for the study of human HLA-restricted T cell determinants (11).
We have identified a synthetic analog peptide, CII (256–276, N263, D266), that is capable of reducing the incidence and severity of arthritis as well as the antibody response to CII when administered to CIA-susceptible DR1 transgenic mice. This peptide was effective in down-regulating the immune responses to CII and arthritis, even when administered 2 weeks following CII immunization. CII (256–276, N263, D266) represents the first peptide analog of collagen that has been shown to suppress arthritis in the context of a human HLA molecule that is known to be associated with RA. This peptide may have therapeutic potential.
Although the precise mechanism by which CII (256–276, N263, D266) exerts its effect is not clear, it may act as an altered peptide ligand (APL). Investigators using other model antigens have indicated that APLs that differ from their respective wild-type peptides can influence T cell activation and cytokine production by cloned T cells (12, 13). For example, an APL containing a single amino acid substitution at the T cell receptor (TCR) contact residue of a cytochrome c peptide was shown to induce immune deviation to a Th2 response, compared with the wild-type agonist peptide, which induced a Th1 response (14, 15). APLs containing a substitution of TCR contact residues in other peptides resulted in marked variability in the TCR recognition of these peptides (15–17). A hemoglobin peptide containing specific amino acid substitutions caused T cells to secrete IL-4 without proliferating (12). Alterations in the structure of hemoglobin 64–76 induced T cells to become cytolytic without proliferating or secreting cytokines (18). These data suggest that minor variations in the peptide binding affinity or in the physicochemical properties of amino acid residues involved in MHC binding can lead to disparate immunologic responses (12, 14, 16, 18, 19).
These observations have stimulated considerable interest in the use of APLs as immunotherapeutic agents in experimental models of autoimmune diseases (20, 21). An APL of myelin proteolipid protein, generated by substitution at a principal TCR contact residue of the encephalitogenic peptide, has been shown to prevent autoimmune encephalitis (22) in mice. Those investigators demonstrated that T cell clones reactive with this APL secreted Th2 cytokines (IL-4 and IL-10) (22). Coimmunization of susceptible mice with myelin basic protein and an APL containing a single substitution at an anchor residue for I-Au similarly blocked the development of experimental autoimmune encephalomyelitis (23). An APL of an acetylcholine receptor peptide effectively down-regulated experimental autoimmune myasthenia gravis in C57BL/6 mice (24). The suppression was accompanied by up-regulated secretion of transforming growth factor β (TGFβ).
The use of APL in the treatment of human disease is limited. Two trials of APL used to treat multiple sclerosis were terminated early because of concerns about disease activity increasing (20, 25). One of these showed that patients treated with the APL had an increase in immunity both to the APL and to myelin basic protein. For this reason, we directly tested the immune potential of CII (256–276, N263, D266) by using it to immunize DR1 transgenic mice. Even when given with CFA, this peptide failed to induce significant Th1 immunity to itself or to intact CII. These data suggest that CII (256–276, N263, D266) may behave somewhat differently than the APL used to treat multiple sclerosis.
Unlike APL used in studies of other models of autoimmune disease, CII (256–276, N263, D266) contains 2 amino acid substitutions, both MHC-binding residues. Jardetzky and colleagues, using a panel of endogenous peptides, described 4 pockets of the class II molecule that can interact with residues of peptides (10). Our data suggest that the phenylanine residue at position 263 and the aspartic acid residue at position 266 most likely bind to the P1 and the P4 binding pockets predicted by the crystal structure data. We find that an aspartic acid substitution at position 266 exhibits decreased binding to DR1 relative to the native peptide, while an alanine substitution has increased binding. Although the binding differences are small, the difference in the effectiveness of the peptides in preventing arthritis is substantial. It is clear that the differences in effectiveness are related entirely to relative binding efficiencies. Although the predominant effect of the CII (256–276, N263, D266) substitution at residue 266 appears to be mediated by a decrease in binding to DR1, we cannot rule out the possibility that an aspartic acid at 266 might affect the neighboring 267 TCR contact residue. Jardetzky and colleagues also predicted that residues P5 (corresponding to CII 267) and P8 (corresponding to CII 270) are the residues most likely to interact with the TCR, since they are positioned to point away from the DR1 binding sites. Based on a comparison of the effect of CII (256–276, N263, D266) with that of peptides that can suppress CIA in mice with other MHC genotypes, all of the effective peptides have decreased MHC binding.
It is evident that there is not a total absence of binding since CII (256–276, N263, D266) does mediate a T cell response. In fact, it seems to stimulate increased production of IL-4 relative to that induced by other analogs. IL-4 is a pleiotropic T cell–derived cytokine. Although it was originally identified as a B cell growth factor, it is now known to be a potent antiinflammatory cytokine (26). It is produced by activated CD4+ T cells and stimulates proliferation, differentiation, and activation of several cell types (27). Major functions include suppression of metalloproteinase production (28), protection against extracellular matrix degradation, and inhibition of osteoclast activity, which blocks bone resorption in vitro (29). Treatment of autoimmune arthritis with the gene for IL-4 has successfully ameliorated CIA in several different studies (28, 30, 31). One would expect that the increased IL-4 detected by culture of CII-autoimmune T cells with CII (256–276, N263, D266) is indicative of in vivo production of this cytokine and protection against inflammation and arthritis. In addition to the in vitro data that show CII (256–276, N263, D266) stimulates production of IL-4, the in vivo data using IL-4−/− mice confirm that its action is mediated through an IL-4–dependent mechanism. These data are consistent with those from a previous report by Yoshino and Yoshino, who used the anti–IL-4 (11B11) antibody to disrupt suppression in another animal model of arthritis, specifically antigen-induced arthritis, using oral administration of the antigen (32). However, these data do not rule out the possibility that other Th2 cytokines, such as IL-10 or TGFβ, may play a role.
Susceptibility to RA is strongly associated with the expression of specific HLA class II alleles, especially HLA–DR1 and DR4 (33–35). Polyarticular juvenile RA is similarly associated with DR1 (36). Evidence has recently been presented showing a correlation of T cell responses to CII (262–270) and the presence of DRB1*0101 and *0401 in RA patients (37, 38). While this does not ensure that a similar parallelism exists between these DR transgenic mice and RA patients, it certainly provides a reasonable basis for development of reagents that might have therapeutic importance. There are strong theoretical and practical reasons to suggest that immunologically specific therapy for autoimmune diseases is preferable to the use of immunologically nonspecific drugs and antibodies. The analog peptide we have identified represents a promising specific immunotherapy for patients with autoimmunity to CII mediated by DR1.