The diagnosis of rheumatoid arthritis (RA) remains imprecise, particularly early in the course of the disease. Up to now, rheumatoid factor (RF) has been used as a typical marker for RA; however, RF has a low specificity because it also occurs in many inflammatory diseases as well as in healthy people. It has recently been shown that autoantibodies directed toward cyclic citrullinated peptides (anti-CCP) are potentially important serological markers for diagnosis and prognosis in RA.¹

The proteolytic pathways associated with cellular interactions also seem to have an important role in joint cartilage destruction of RA. Matrix metalloproteinase-3 (MMP-3) is secreted by fibroblasts and synovial cells,² and it has been shown to be increased in RA serum, not only in the late stage but also in the early stage of the disease. Therefore, MMP-3 is a useful marker for predicting bone damage in early RA.³

In many early cases of RA, patients will not always fulfil the American College of Rheumatology criteria. Therefore, it is important to detect the point at which disease-specific autoantibodies and proteinases appear in RA serum. In this study we directly investigated the production of anti-CCP and MMP-3 using a severe combined immunodeficiency (SCID) mutant mouse, into which human RA synovial tissue was transferred.

SCID mice are a well known model for analysing the developmental mechanism of T and B cells. We previously developed SCID mice (CB.17/Icr; Charles River Japan, Tokyo, Japan) in which human RA synovial tissue was grafted (SCID-HuRAg), and we evaluated them as models for RA.^4,5 The histological features of RA were observed in all SCID-HuRAg mice and these mice induced the production of human antibodies. Proliferative synovial fibroblasts and infiltration of inflammatory cells were also seen in the pannus.

The serum levels of anti-CCP and human MMP-3 of SCID-HuRAg mice were examined weekly after implantation. Non-implanted SCID mice sera were used as controls. Figure 1 shows representative results of independent experiments from two patients. The mean (SD) anti-CCP and MMP-3 levels in the serum of the patients were 702 (26.9) U/ml and 321 (12.0) ng/ml, respectively. Human, not mouse, MMP-3 appeared in mouse serum shortly after implantation and decreased immediately. On the other hand, anti-CCP increased gradually and reached a plateau from the fifth week. It is reported that anti-CCP are produced locally in the inflamed synovial tissue from RA.^6,7 Therefore we think it is reasonable that the surviving, activated human B cells can produce autoantibodies, and the proliferative synovial fibroblasts around the engrafted tissue can produce MMP-3, and that producing autoantibodies spontaneously takes much longer than inducing enzyme. In our previous report, both human RF and interleukin 6 were also detected in these mice sera⁴ and not in mice with tissue implants from osteoarthritis synovia (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 Time course of the serum levels of anti-CCP and MMP-3 in SCID-HuRAg mice. The results represent two independent experiments. The tissues from each of two patients were transferred to each of eight mice and each point represents a mean (SD) value (n = 8). Anti-CCP antibodies were measured with a second generation anti-CCP enzyme linked immunosorbent assay (ELISA) kit (Immunoscan RA the anti-CCP Mark2 kit; Euro-Diagnostica, Arnhem, The Netherlands). The concentrations of MMP-3 in sera were determined by a one step sandwich ELISA system using a human MMP-3 kit (Panaclear MMP-3; Daiichi Pure Chemicals Co, Tokyo, Japan). Both assays were conducted according to the manufacturer’s instructions. Their cut off values in human serum levels are 25 U/ml for anti-CCP and 29.0 ng/ml for MMP-3. Non-implanted SCID mice sera were used as controls (week 0).

MRL-lpr/lpr mice develop an autoimmune syndrome resembling human systemic lupus erythematosus.⁸ Anti-CCP were present in these mice and (NZW × B6)F₁-hbcl-2 transgenic mice, which have defects in the regulation of B cell apoptosis.⁹ However, these animals showed not only antibody reactivity against the citrullinated peptide cfc1-cyc peptide, but also against the non-citrullinated control peptide; therefore, Vossenaar et al reported that the antibodies in these mice are not citrulline specific, and citrulline-specific autoantibodies are present only in human patients with RA and not in animal models of autoimmune disease.¹⁰

We think that the main problem is that citrulline-specific autoantibodies detected by both groups were of mouse origin. On the contrary, all the target cells within SCID-HuRAg mice which we used were of human origin, having migrated from the implanted tissue, and anti-CCP and MMP-3 were of human origin, too.

In conclusion, we showed the point at which disease-specific autoantibodies and proteinases appear in RA. SCID-HuRAg mice will be worth using in the study and development of new drugs for RA.