Download PDF

DOI: 10.17653/2374-9105.SS0002

Collagen-induced arthritis is an animal model for rheumatoid arthritis with unknown etiology. A strong genetic predisposition is implicated in CIA as certain strains of mice develop CIA which is dependent on the Major Histocompatibility Complex (MHC) class II polymorphism. H2-Aq and H-2Ar are associated with CIA susceptibility while H2-Ab, H2-Ad and H2-Af are associated with resistance to develop arthritis1 . CIA is a T cell dependent disease. MHC molecules are involved in positive and negative selection of T cells in thymus and also play a central role in regulation of immune responses. Antigen-specific T cell response requires 2 signals. The first is via the molecules that bind and present diverse peptides and second is through the co-stimulatory molecules present on the antigen presenting cells. The major function of MHC molecules is to generate response to pathogens and foreign antigens while maintaining self-tolerance in the immune system. Many mechanisms have been reported for generating self-tolerance including clonal deletion and anergy 24 . It has been reported that self MHC-derived peptides can be presented by class II molecules5 . We have shown presentation of Eβ-derived peptides by the H2-Aq molecule6 . Elution studies of naturally processed proteins from MHC class II molecules show presence of self-MHC derived peptides suggest that these peptides may be involved in tolerance or autoimmune responses .7,8 Similarly, allogeneic Aαk 60-90 peptide and autologous Aαd 60-70 peptide can bind to H-2Ad molecules. H2-Ab class II molecules can bind a peptide derived from the conserved Eα-chain with high affinity and compete for antigen presentation9 . Similarly, I-Ag7-derived peptides have been shown to bind to syngeneic and allogeneic peptides and modulate diabetes in NOD mice by eliciting T cell responses10 . The binding of self MHC peptides to autologous molecules is stable as evidenced by the observations that monoclonal antibodies can be raised against self Eα:Aβ complexes and that T cells can be induced to proliferate against self as well as allogeneic MHCclass II complexes. We have reported that allogeneic Eαk peptides are immunogenic and elicit measurable T cell responses in H-2E mice11 . We and others have reported that different class II molecules can present isotypic class II peptides. In humans, HLA-DR peptides can be presented by HLA-DQ molecules and elicit T cell responses in vitro12 . DBA/1 (H-2q) mice are highly susceptible to arthritis following immunization using heterologous type II collagen. In addition, a small percentage of DBA/1 mice develop spontaneous arthritis13 . Using a panel of synthetic Aq peptides, we have determined the immunogenic potential of the peptides and the presence of T cells reactive to self-MHC in DBA/1 mice that will explain spontaneous disease.

Materials and Methods


DBA/1J (H2-Aq), B10.Q (H2-Aq), B10 (H2-Ab), B10.K (H2-Ak), B10.M (H2-Af), B10.D2 (H2-Ad) and B10.S (H2-As) inbred strains of mice were obtained from Jackson laboratories. Six to eight week old mice were bred and maintained in the Immunogenetics Mouse Colony at the Mayo Clinic. All mouse husbandry and research protocols were monitored and approved by an Institutional Animal Care Committee.


All peptides including those spanning residues 1-30, 30-60, and 60-90 of the α- and β- chains of the Aq molecule (Table 1) were synthesized and purified at the Peptide Core Facility of the Mayo Clinic.

Immunizations and in vitro cultures:

Mice were immunized and the draining lymph node cells (LNCs) challenged in vitro according to previously described protocols.12 Briefly, mice were immunized with 100 μg of peptide emulsified using complete Freund’s adjuvant. Seven days post injection, the draining lumbar, caudal, inguinal, and popliteal lymph nodes were harvested and challenged with peptides in vitro for 48 hr. Eighteen hours before termination of experiment 3H-thymidine was added to the cultures. The T cell activation was assessed by determining the amount of 3H-thymidine incorporation. T cell responses following peptide challenge are expressed as Δcpm =(mean cpm of triplicate cultures containing peptide)-(mean cpm of triplicate cultures containing medium).

Mixed Lymphocyte reactions:

DBA/1 lymph node cells (105 ) were co-cultured either with medium or 1 x 106 irradiated (3300 rads) spleen cells as antigen presenting cells (APCs) from self or other strains in a volume of 200 µl. Forty µg of peptides (20 µl volume) were added to the wells. Control wells included just media with cells but without the peptides. 3H-thymidine was added after 48 hr of culture. Cells were harvested and the extent of 3H-thymidine incorporation determined after 72 hr of in vitro culture. The graph depicts Δcpm (mean cpm of the cultures with cells and peptides- mean cpm of culture with cells but no peptide).

Statistical Analysis:

All comparisons were done using Student’s t test. A P value of <.05 was considered significant.


Immunization using peptides elicits allo- but not auto- T cell responses:

We tested CIA susceptible DBA/1 and other inbred mice for response to 30mers peptides spanning residues 1-90 derived from Aαq. At this region the Aαq and Aαb differ for 8 aa, some of which are conservative changes. The sequences for the peptides are given in Table 1.


Table 1: Sequences of synthetic H2Aq peptides

The results of in vitro T cell proliferation in DBA/1 and a series of inbred lines of mice with peptide Aαq 1-30 are summarized in Figure 1. H-2d (B10.D2) and H-2k (B10.K) mice responded (Δcpm 32528±2821 and 30976±1239 respectively) vigorously to a recall peptide challenge while H-2f (B10.M) and H-2s (B10.S) mice generated a moderate response. B10 H-2b (B10) and H-2q (B10.Q and DBA/1) mice were unresponsive following immunization with self Aαq 1-30 peptide.

A comparison of T cell response to peptide Aαq 30-60 among various strains showed strongest response by B10.M mice (Δcpm 46233±4498) with moderate response by B10.S mice (Δcpm 23365±2440). All other strains of mice including H-2q mice were hypo-responsive. Similar to Aαq30-60 peptide, with the exception of B10.M and B10.S mice, all other mice did not respond (Δcpm<9,500) following Aαq 60-90 peptide. Among H-2q mice, DBA/1 mice exhibited a stronger propensity towards autoreactivity (Δcpm of 9,633±786) than did B10.Q mice (Δcpm of 875±123).

Overall, B10.M and B10.S showed much stronger responses to peptides derived from Aα 1-90 while DBA/1 and B10.Q, CIA susceptible strains, showed lower responses with statistically significant differences.

Immunization using Aβq peptides elicits both auto-as well as allo- T cell responses:

In contrast to the results obtained using Aαq peptides, much stronger and broader reactivity were measured following immunizations using Aβq peptides. Strong T cell proliferative responses were measured following Aβq1-30 peptide challenge (Δcpm’s 23,000-33,000) in H-2d,f,k mice (Figure 2). Both H-2b and H-2s mice were unresponsive (cpm’s<5,500). Interestingly, DBA/1 (Δcpm 33,434±3271) generated a much stronger response compared to B10.Q (Δcpm 2336±218) suggesting presence of autoreactive T cells to the self Aβq 1-30 peptide. Nearly all of the tested mouse strains responded to varying degrees following immunization using Aβq 30-60 peptide, although H-2 b,k,s inbred lines of mice were less responsive.

In H-2Aq mice, B10.Q mice did not respond (Δcpm 5,488) to an in vitro challenge with self peptide Aβq 30-60,while DBA/1 mice responded very vigorously (Δcpm52,571±6307). Mice expressing H-2b,f haplotypes exhibited similar T cell responses following in vitro challenge using peptide Aβq 60-90 (Δcpm’s 23,000-30,000). However, neither B10.Q nor DBA/1 mice were autoreactive to Aβq 60-90 peptide.

Overall these data showed strong T cell response generated by DBA/1 mice to self-derived peptides 30-60 and 60-90, both were non-responsive in B10.Q mice. All differences were statistically significant between the B10 and B10.Q and DBA/1 mice.


Fig. 1: DBA/1 mice do not generate response to self-peptides from Aαq

Draining lymph node cells from primed mice were challenged in vitro with peptides and T cell responses measured. Medium control values varied between 1,000 and 5,000 cpm. * p<0.01 DBA/1 compared with other strains.

DBA/1 mice have self-reactive T cells to Aβq peptides.

Fig. 2: DBA/1 mice have self-reactive T cells to Aβqpeptides.

Draining lymph node cells from primed mice were challenged in vitro with peptides and T cell responses measured. Medium control values varied between 1,000 and 5,000 cpm. *p<0.01 DBA/1 compared with other strains.tides and T cell responses measured. Medium control values varied between 1,000 and 5,000 cpm. * p<0.01 DBA/1 compared with other strains.

Aβq 30-60 peptide augments allo- and auto-reactivity of DBA/1 lymph node cells:

In light of the strong autoreactivity of peptide Aβq 30-60 in DBA/1 mice, a series of experiments were performed ascertaining the in vitro effects of adding self peptide to cultures of LNCs fromDBA/1 (Figure 03). The addition of immunogenic self Aβq 30-60 but not Aαq 30-60 peptide induced some activation of LNCs of naïve DBA/1 mice (cpm 1,977). Only after specific immunization could responses for Aβq 30-60 be detected. We then performed MLR using DBA/1 splenic cells as stimulators. The addition of Aβq 30-60 but not Aαq 30-60 peptide (cpms of 37,945 and 9,386 respectively) augmented an autologous mixed lymphocyte reaction (cpm 9,791) in DBA/1 mice. The immunopotentiating effect of Aβq 30-60 peptide was also observed when DBA/1 T cells were stimulated with allogeneic MLR assays where spleen cells from B10.D2 or B10 mice were used. The responses of DBA/1 cells to B10 or B10.D2 stimulator cells was augmented if Aβq 30-60 peptide was added during in vitro culture initiation. The addition of Aαq 30-60 peptide to cultures containing allogeneic B10 or B10.D2 was without effect.

Aβq peptide augments mixed lymphocyte reactions involving DBA/1 cells.

Fig. 3: Aβq peptide augments mixed lymphocyte reactions involving DBA/1 cells.

DBA/1 lymph node cells were cocultured either with medium or 1 x 106 irradiated (3300 rads) spleen cells, from self or other strains, and peptides and T cell proliferation measured.

Priming with human collagen peptide 250-270 confers reactivity to peptide Aβq 30-60:

DBA/1 mice develop chronic inflammatory arthritis following immunization using heterologous type II collagen. We and others have reported that one of the determinants recognized by DBA/1 T cells on human type II collagen is located within residues 250-270.14 We therefore determined the effects of adding either Aαq 30-60 or Aβq 30-60 peptide to cultures of DBA/1 T cells primed to human type II collagen (HII) peptide 250-270 (Figure 4). Under conditions in which DBA/1 cells responded to challenge using HII peptide 250-270 (Δcpm 29163), the addition of peptide Aβq 30-60 but not Aαq 30-60 increased the response to the collagen peptide. Interestingly, HII 250-270 peptide primed cells responded if challenged with Aβq 30-60 but not Aαq 30-60 peptides.

Priming with human collagen peptide 250-270 confers reactivity to peptide Aβq .

Fig. 4: Priming with human collagen peptide 250-270 confers reactivity to peptide Aβq .

Mice were primed with HII 250-270 peptide and draining lymph node cells were cultured in the presence of the immunizing peptide and self-derived peptides Aαq 30-60 and Aβq 30-60 separately or in combination and T cell responses was measured.


The studies reported here support previous observations that self-MHC derived peptides may be involved in tolerance or pathogenicity. 15,16 Our observations reveal that all self-MHC peptides elicited different patterns of allogeneic T cell responses among the inbred strains of mice tested. Some of the Aβq peptides elicited vigorous self reactivity in DBA/1 but not B10.Q mice. In addition, Aβq 30-60 peptide augmented self- and non-self responses of DBA/1 T cells. Both B10.Q and DBA/1 mice are susceptible to CIA and carry same class II allele. This data supports role of genetic factors other than the class II genes may also be involved in overall immune response. CIA-resistant B10.M (H-2f) mice exhibited the strongest propensity to respond, responding to five of the six peptides, while B10 (H-2b) mice were characterized by the weakest peptide responses. T regulatory cells have been shown to be selected on self-peptides.17 One can speculate that CIA-resistant mice harbor self-reactive T regulatory cells. Collectively, these findings suggest that self MHC peptide ligands may have potent influences on immune responses.

It was not possible to directly correlate the number of disparate residues with T cell reactivity. For example, H-2Ab and H-2Aq molecules differ at three residues corresponding to peptide Aαq 1-30 yet B10 mice do not respond. In contrast, H-2Ad and H-2Aq molecules differ at only two residues yet B10.D2 mice respond as well as B10.M mice which differ at three residues. Complex factors, including affinity and avidity of the peptides, in addition to sequence disparity probably contribute towards determining the magnitude of T cell responses to class II peptides.

An insight into the complexity of the peptide responses is further evidenced by the differing auto-responsiveness between DBA/1 and B10.Q mice. B10.Q mice were uniformly unresponsive to all six of the Aαq and Aβq peptides. In contrast, while DBA/1 mice responded to most of the Aβq peptides, they did not respond to any of the Aαq peptides. It has been reported that synthetic self Aαk 1-16 and Aβk 1-18 peptides elicit T cell responses in B10.A mice. 15 However, not all syngeneic class II peptides are immunogenic. We have reported that H-2E+ mice do not respond in vitro to synthetic Eαk peptides collectively spanning residues 90-150.11

The immunopotentiating effects of peptide Aβq 30-60 bears comment. We measured increase in LNCs responses from DBA/1 mice to allo- (B10.D2, B10) and auto- (DBA/1) challenge in vitro using irradiated spleen cells following the addition of Aβq 30-60 peptide. In contrast, the addition of Aαq 30-60 peptide was without effect. A study using rats with CIA showed that immunization with type II collagen lead to activation of autoreactive T cells which responded in autologous mixed lymphocyte reaction possibly due to T cell reactivity to self-MHC peptides.18 Similarly, some HLA-DR peptides can alter HLA-DR restricted allorecognition in vitro and in vivo. 19 Naturally processed class II peptides can be detected in the binding pockets of mouse and human class II molecules.20 These observations have led to theoretical concepts by us as well as by others that the physical binding of class II peptides to MHC molecules has functional consequences in autoimmunity.12 Human HLA-DR and DQ genes are homologous to mouse H2-E and H2-A genes. Patients with RA harbor type II collagen-related antibodies and reactive T and B cells 21,23 suggesting CIA in mice can serve to determine pathogenesis of RA.23

It has been hypothesized that some motifs of the third hypervariable region of HLA-DRB1 molecule play a role in the development of rheumatoid arthritis. Using human class II transgenic mice, we have been able to demonstrate a direct correlation between T cell responses to human class II peptides and resistance/susceptibility to reactive arthritis. 12,23 Our observation that the responses to human collagen peptide 250-270 is potentiated by the addition of Aβq30-60 peptide provides additional support for the role of class II peptides in modulating disease states. H-2A polymorphism has been associated with modulation of CIA via self MHC class II peptides.24 These observations are reminiscent of that observed in non-obese diabetes (NOD) mice where some H-2Ag7 peptides may alter progression in diabetes.10

The mechanism which confers autoreactivity to some Aβq peptides in DBA/1 mice is not known. A direct link between theT cells reactive to self-peptides is suggested by a tendency of DBA/1 mice primed to Aβq but not Aβq 30-60 peptide to have higher medium control values (not shown). However, since most autoimmune diseases are multigenic, an indirect role of the self-derived peptides is much more likely. It is possible that these self-derived MHC peptides are involved in positive selection of T cells in the thymus and the T cell repertoire selected by these peptides may have a different Th profile. In CIA susceptible mice, self-peptide reactive T cells may be producing pro-inflammatory cytokines, which under normal conditions may not be pathogenic. One can envisage that in genetically susceptible subjects harboring T cells reactive to self-class II molecules that infections or possibly superantigens, may cause proliferation of the self-reactive T cells contributing to inflammation and arthritis. These events need to be evaluated in future studies.