Placental immunomodulator ferritin, a novel immunoregulator, suppresses experimental arthritis

ARTHRITIS & RHEUMATISM Vol. 48, No. 3, March 2003, pp 846–853 DOI 10.1002/art.10850 © 2003, American College of Rheumatology

Placental Immunomodulator Ferritin, a Novel Immunoregulator, Suppresses Experimental Arthritis Abraham Weinberger,1 Marisa Halpern,1 Muayad A. Zahalka,1 Francisco Quintana,2 Leonid Traub,1 and Chaya Moroz1 Objective. To determine the effect of treatment with C48, the recombinant cytokine-like domain of the novel human placental immunomodulator ferritin (PLIF) immunoregulator, on zymosan-induced arthritis (ZIA) in mice and on adjuvant-induced arthritis (AIA) in rats. Methods. The in vitro effect of PLIF/C48 was tested in mixed lymphocyte cultures (MLCs) of allogeneic mouse splenocytes. Arthritis was induced by intraarticular injection of zymosan into naive mice and by subcutaneous injection of Mycobacterium tuberculosis into rats. C48 was injected intraperitoneally daily from day 3 to day 9 or from day 7 to day 13 after induction of synovitis by zymosan, and every other day from day 2 to day 14 after induction of AIA. Swelling of the joints and histologic features of the synovium were assessed. Th1 and Th2 cytokines were quantified by enzyme-linked immunosorbent assay. Results. Both PLIF and C48 significantly inhibited the in vitro immunoreactivity of mouse splenocytes in MLCs. Treatment of ZIA mice and AIA rats with C48 effectively reduced joint swelling. C48 treatment reduced synovial lining thickening, numbers of mononuclear cells and histiocytes, as well as cartilage destruction and bone erosions. In vitro, activated splenocytes from C48-treated ZIA and AIA animals produced significantly higher levels of interleukin-10 (IL-10). In

animals with ZIA, this was accompanied by lower levels of tumor necrosis factor and IL-2. Conclusion. Human PLIF and C48 were shown to exert cross-species immunosuppressive activity in vitro. The in vivo suppression of articular inflammation in the experimental models of ZIA and AIA was the result of treatment with the antiinflammatory human C48. These results suggest that treatment with C48 may offer an effective immunotherapeutic means of controlling inflammatory polyarthritis. Several reports have described the high frequency of remission of rheumatoid arthritis (RA) during pregnancy (1–7). Similarly, most studies of pregnancy in patients with multiple sclerosis showed tendencies toward remissions during gestation and relapses during the postpartum period. Indeed, most pregnant women with RA have reported improvement of joint pain and/or swelling. A postpartum flare, however, can occur in the majority of women up to 8 weeks after delivery (7). Therefore, the state of pregnancy serves as a powerful and natural research means of identifying potential pathophysiologic mechanisms that modify disease activity and of generating innovative strategies for the development of new therapies for use in the nonpregnant state. The phenomenon of remission in RA can be attributed to several factors, one of which, an increase in estrogen secretion, occurs in pregnancy (8–10). In a collagen-induced arthritis animal model, administration of estrogen in the postpartum period prevented the postpartum flare (9). Another possible explanation for the decrease in arthritis during pregnancy and the postpartum flare of RA is the finding that glycosylation of IgG decreases during pregnancy and increases during the postpartum period (11). A similar explanation could apply to the postpartum flares seen in animal models of human arthritis (11,12).

1

Abraham Weinberger, MD, Marisa Halpern, MD, Muayad A. Zahalka, PhD, Leonid Traub, PhD, Chaya Moroz, PhD: Felsenstein Medical Research Center, Rabin Medical Center, Petah Tikva, Israel, and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel; 2 Francisco Quintana, PhD: The Weizmann Institute of Science, Rehovot, Israel. Address correspondence and reprint requests to Chaya Moroz, PhD, Laboratory of Molecular Immunology, Felsenstein Medical Research Center, Beilinson Campus, Rabin Medical Center, Petah Tikva 49100, Israel. E-mail: [email protected]. Submitted for publication July 12, 2002; accepted in revised form December 9, 2002. 846

PLIF SUPPRESSES EXPERIMENTAL ARTHRITIS

Recent studies by Sirota et al demonstrated that a unique isoform of placental isoferritin (PLF), composed of 43-kd protein (p43) and ferritin light chain, acts as a regulatory cytokine that down-regulates the immune response of the mother to her embryo (13). This protein is highly expressed in syncytiotrophoblast cells, in Hofbauer cells, and in decidual macrophages (14). Furthermore, studies by our group showed that elevation of p43 (PLF) begins as early as day 11 of pregnancy, continues throughout gestation, and declines at term (15). Recently, the gene coding for p43 has been cloned and designated placental immunomodulator ferritin (PLIF). A recombinant form of its bioactive cytokinelike domain (CLD), called C48, is composed of 48 amino acids. Both PLIF and C48 were subcloned and expressed in Escherichia coli (16). PLIF has been localized in the syncytiotrophoblast cells of human placenta at 8 weeks. PLIF and C48 were shown to down-modulate the human allogeneic mixed lymphocyte reaction and to inhibit the proliferation of peripheral blood mononuclear cells stimulated with OKT3 monoclonal antibody (16). The present study was designed to evaluate the effect of treatment with C48 in zymosan-induced arthritis (ZIA) in mice and in adjuvant-induced arthritis (AIA) in rats. Our investigation included histopathologic evaluation of knee joint sections, measurement of joint swelling, and elucidation of the change in cytokine production. We found that C48 was effective in suppressing the arthritis process in an animal model of experimental arthritis. Analysis of cytokine secretion by spleen cells following polyclonal activation revealed a shift toward the Th2 pathway following treatment. MATERIALS AND METHODS Animals. Six-week-old female ICR and C57BL mice and female Lewis rats were fed a standard laboratory diet and tap water ad libitum. The study was approved by the Animal Review Board of the Rabin Medical Center. Induction of ZIA. ZIA was induced as previously described (17). Briefly, a homogeneous suspension of 15 mg zymosan A (Saccharomyces cerevisiae; Sigma, St. Louis, MO) was dissolved in 1 ml saline and sterilized by autoclaving. Arthritis was induced in both hind knees by injecting 50 ␮g of zymosan solution into each joint. Measurement of the joint thickness was performed using a Mitutoyo caliper (Mitutoyo, Kanagawa, Japan). C48 administration after ZIA induction. C48, a recombinant form of the CLD of the human PLIF gene, was produced in E coli as previously reported (16). This PLIF gene codes for a protein with partial homology to human ferritin heavy chain, but contains 48 C-terminal amino acids of a novel sequence with no homology to any known protein.

847

In one group of mice, 50 ␮g of C48 in 0.5 ml of phosphate buffered saline (PBS), pH 7.2, was injected intraperitoneally (IP) daily from day 3 to day 9 following the intraarticular zymosan injection. Similarly, C48 was injected into another group of mice from day 7 to day 13 (delayedinjection group). The control group of mice received PBS only. Induction of AIA. AIA was induced as previously described (18). Briefly, Lewis rats were vaccinated with 1 mg of Mycobacterium tuberculosis H37Ra (Difco, Detroit, MI) in Freund’s complete adjuvant (Difco) subcutaneously at the base of the tail. Severity of disease (arthritis index) was scored by 2 blinded observers (MAZ and LT) according to National Institutes of Health protocols for AIA (19), where 0 ⫽ no arthritis, 1 ⫽ redness of the joint, or 2 ⫽ redness and swelling of the joint. The ankle and tarsometatarsal joints of each foot were scored. A maximum score of 16 can be obtained, but a score ⬎8 indicates severe disease. C48 administration after AIA induction. Rats were injected IP with 180 ␮g of C48 in 0.5 ml of PBS every other day from day 2 to day 14 following AIA induction. A control group of rats was injected IP with 0.5 ml of PBS. Mixed lymphocyte cultures (MLCs). Responder splenocytes were obtained from C57BL mice, and allogeneic stimulator spleen cells were obtained from ICR mice (Harlan, Rehovot, Israel). Mouse splenocytes were cultured alone (LCs) or were mixed with irradiated allogeneic stimulator cells at a ratio of 1:1 (MLCs). For culture treatments, 1 ␮g/ml of purified PLIF or C48 was added to cultures immediately after responder cell plating. These experiments were performed in 96-well flat-bottomed tissue culture plates in a culture medium containing 10% fetal calf serum. On day 4 of the experiment, cells were pulsed with 1 ␮Ci/well of 3H-thymidine and harvested 16 hours later. The proliferation index (PI) of lymphocyte cultures was calculated as the counts per minute in the PLIF- or C48-treated cultures divided by the cpm in the untreated culture. The PI in MLCs was calculated as the cpm in stimulated MLCs divided by the cpm in unstimulated LCs for each specific treatment. Cytokine assay. Mouse spleen cells were harvested from ZIA mice treated daily with IP injections of C48 (30 ␮g) or PBS as indicated. Splenocytes (2 ⫻ 106/ml) were cultured with concanavalin A (Con A; 1 ␮g/ml). Supernatants were collected after 24, 48, or 72 hours of stimulation with Con A or medium alone. Murine cytokines were quantitated in the culture supernatants with an enzyme-linked immunosorbent assay (ELISA) as previously described (20). Paired antibodies from BD PharMingen (San Diego, CA) were used for the detection of interleukin-4 (IL-4), IL-10, and IL-12. Paired antibodies from Endogen (Cambridge, MA) were used for the quantitation of IL-2, tumor necrosis factor ␣ (TNF␣), and interferon-␥ (IFN␥). Briefly, ELISA plates (Maxisorp; Nunc, Roskilde, Denmark) were coated overnight at 4°C with capture antibodies. Nonspecific binding was blocked by incubation with 1% bovine serum albumin (BSA) for 1 hour at room temperature (RT), and culture supernatants or recombinant cytokines were incubated overnight at 4°C. Cytokine binding was detected by biotinylated detection antibodies and streptavidin conjugated to alkaline phosphatase (Jackson ImmunoResearch, West Grove, PA). Cytokine levels in supernatants are expressed as pg/ml based on calibration curves constructed using recombi-

848

WEINBERGER ET AL

nant cytokines as standards. The lower limits of detection for these experiments were 15 pg/ml for IL-2, IL-10, IL-12, and IFN␥, 5 pg/ml for TNF␣, and 30 pg/ml for IL-4. Rat spleen cells were harvested on day 16 from AIA rats treated every other day with IP injections of C48 (180 ␮g) or PBS on days 2–14. Splenocytes (2 ⫻ 106/ml) were cultured with and without Con A (2 ␮g/ml), and supernatants were collected after 24, 48, and 72 hours of stimulation with Con A or medium alone. Rat IL-10, IFN␥, and TNF in the culture supernatants were quantified by ELISA using the OptEIA kit (BD PharMingen) with some modifications. Briefly, ELISA plates (Maxisorp; Nunc) were coated overnight at 4°C with anti-rat cytokine monoclonal capture antibodies. Nonspecific binding was blocked by incubation with 1% BSA for 1 hour at RT, and culture supernatants or recombinant cytokines were added for 2 hours at RT. After washing the plates, biotinylated detection antibodies were added for 1 hour at RT, then extensively washed and incubated with streptavidin conjugated to alkaline phosphatase (Jackson ImmunoResearch) for 30 minutes at RT. The plates were washed, alkaline phosphatase substrate (Sigma) was added, and after 30 minutes of incubation at RT, samples were read at 405 nm. Cytokine levels in supernatants are expressed as pg/ml based on calibration curves constructed using recombinant cytokines as standards. The lower limits of detection for these experiments were 15 pg/ml for TNF, IL-10, and IFN␥. Histologic processing. Animals were killed on day 14, and the knees were removed in toto, fixed, decalcified, dehydrated, embedded in paraffin, and cut into 4-␮m sections. All sections were stained with Harris’ hematoxylin and eosin. Assessment of joint inflammation. The assessment of all pathologic features was performed by a blinded observer (MH) using a grading scale of 0–3 according to the proportion of areolar tissue that was densely infiltrated with mononuclear

Figure 2. Effect of C48 treatment on hind joint swelling in zymosaninduced arthritis. Mice were injected with intraarticular zymosan and then treated with C48 on days 3–9 (n ⫽ 10) or days 7–13 (delayed treatment; n ⫽ 10) or with phosphate buffered saline (PBS; n ⫽ 10) on days 3–9 following zymosan injection. A, Swelling expressed as the percentage of change in joint diameter. B, Histologic score. Values are the mean ⫾ SEM. ⴱ ⫽ P ⬍ 0.05 versus control values, by Student’s t-test.

Figure 1. Effect of placental immunomodulator ferritin (PLIF) and C48 on allogeneic mouse mixed lymphocyte cultures. Shown is the proliferation index (PI) of stimulated mouse spleen cells either untreated or treated with 1 ␮g/ml PLIF or C48. Values are the mean ⫾ SEM of 3 experiments. ⴱ ⫽ P ⬍ 0.01 versus untreated cells, by Student’s t-test.

cells. Synovial lining cell hyperplasia was graded similarly on the same scale. A score of 0–3 was used for grading the number of histiocytes observed in synovial tissue. Cartilage destruction was scored on a scale of 0–3, ranging from no damage to fully destroyed cartilage layers. Bone erosion scores were obtained in a blinded analysis using semiquantitative grading scales: 0 ⫽

PLIF SUPPRESSES EXPERIMENTAL ARTHRITIS

849

Figure 3. Histologic features of joints from C48-treated or control mice with zymosan-induced arthritis and from C48-treated or control rats with adjuvant-induced arthritis. A, Joint from a control mouse injected intraarticularly with zymosan, showing extensive cellular infiltration in the synovium and erosion of cartilage. B, Corresponding specimen from a mouse injected intraarticularly with zymosan and treated intraperitoneally (IP) with C48, showing mild inflammation with primarily synovial hypertrophy. C, Joint from a control rat injected subcutaneously with Mycobacterium tuberculosis, showing massive synovitis. D, Joint from a rat injected with M tuberculosis and treated IP with C48, showing mild inflammation. (Hematoxylin and eosin stained; original magnification ⫻ 10 in A and B; ⫻ 20 in C and D.)

normal, 1 ⫽ minimal loss of cortical bone at a few sites, 2 ⫽ mild loss of cortical trabecular bone, 3 ⫽ moderate loss of bone at many sites, 4 ⫽ marked loss of bone at many sites, and

5 ⫽ marked loss of bone at many sites with fragmenting and full-thickness penetration of cortical bone. Statistical analysis. Student’s t-test was used to determine significant differences between experimental and control groups. P values less than 0.05 were considered significant.

RESULTS

Figure 4. Suppression of disease progression in rats with recent-onset adjuvant-induced arthritis (AIA). Twenty rats were vaccinated with Mycobacterium tuberculosis and randomly assigned to receive treatment with C48 (n ⫽ 10) or phosphate buffered saline (PBS; n ⫽ 10) at the indicated times following AIA induction (see Materials and Methods). Values are the mean ⫾ SEM. ⴱ ⫽ P ⬍ 0.001 versus control AIA rats, by Student’s t-test.

Effect of PLIF and C48 on the mouse immune response in vitro. To test whether human PLIF has a cross-species immunosuppressive activity, PLIF and C48, cleaved and purified from their glutathione Stransferase fusion protein, were tested for their effect on allogeneic mouse MLCs in vitro. Both PLIF and C48 significantly reduced the allogeneic stimulation (P ⬍ 0.01), as indicated by the decreased PIs of the treated cultures (6.8 and 9.8, respectively) compared with the PI of the untreated controls (25.4) (Figure 1). These results indicate a cross-species immunosuppressive activity of human PLIF and C48. Effect of C48 on ZIA. Mice that received C48 from day 3 to day 9 developed significantly milder disease than did the control mice. Joint swelling in these animals was significantly reduced (P ⬍ 0.05) compared

850

WEINBERGER ET AL

Figure 5. Production of cytokines in concanavalin A (Con A)–activated spleen cells from mice with zymosan-induced arthritis (ZIA) treated with C48 or phosphate buffered saline (PBS). Spleen cells were harvested from 6 mice on day 14 following zymosan inoculation. Mice were injected intraperitoneally with either C48 (n ⫽ 3) or PBS (n ⫽ 3) from day 5 to day 13. Cells were cultured with Con A (1 ␮g/ml) for 24, 48, and 72 hours, and their supernatants were tested for levels of interleukin-10 (IL-10) (A), IL-2 (B), tumor necrosis factor (TNF) (C), IL-4 (D), interferon-␥ (IFN␥) (E), and IL-12 (F). Values are the mean ⫾ SEM. ⴱ ⫽ P ⬍ 0.05 versus control ZIA mice, by Student’s t-test.

with that in PBS-treated controls (Figure 2A). Delayed treatment with C48 (days 7–13) also yielded significantly milder swelling, and this effect persisted longer (Figure 2A). Histologically, joints from all of the untreated mice were severely damaged by rapidly expanding synovial pannus. Mononuclear cell infiltration and thickening

of the synovial membrane were apparent (Figures 2B and 3A). In contrast, IP injection of C48 on either day 3 or day 7 (delayed treatment) suppressed inflammation in murine hind joints, as shown by a reduced number of synovial histiocytes, a low amount of mononuclear cell infiltration, and less synovial membrane thickening (Figures 2B and 3B). Cartilage destruction and bone

PLIF SUPPRESSES EXPERIMENTAL ARTHRITIS

Figure 6. Production of cytokines by Con A–activated spleen cells from rats with adjuvant-induced arthritis (AIA) treated with C48 or PBS. Spleen cells were harvested on day 16 from AIA rats that had been treated from day 2 to day 14 with PBS (n ⫽ 3) or C48 (n ⫽ 3). Cells were cultured with Con A (2 ␮g/ml) for 24, 48, and 72 hours, and their supernatants were tested for levels of IL-10 (A), TNF␣ (B), and IFN␥ (C). Values are the mean ⫾ SEM of 3 animals. ⴱ ⫽ P ⫽ 0.03; ⴱⴱ ⫽ P ⫽ 0.003 versus control AIA rats, by Student’s t-test. See Figure 5 for other definitions.

erosion were significantly reduced (P ⫽ 0.03 and P ⫽ 0.003, respectively) in ZIA mice treated early with C48, while the reduction in these parameters did not reach

851

statistical significance in the delayed-treatment group (Figure 2B). Effect of C48 on AIA. Based on the results obtained from the early- and delayed-treatment protocols for ZIA mice, we chose to treat AIA rats from day 2 (early) up to the peak of active disease on day 14 covering both periods. As shown in Figure 4, from day 14 through day 39, the arthritis index in AIA rats receiving treatment with C48 was significantly lower than that in control rats (P ⬍ 0.001). Histologic examination of joints chosen randomly from each group revealed less inflammation and less cartilage destruction and bone damage in animals injected with C48 compared with control animals (Figures 3C and D). Levels of Th1/Th2 cytokines in activated splenocytes from C48-treated ZIA mice and AIA rats. As shown in Figure 5, the time courses of the accumulation of the different cytokines produced by splenocytes from PBS- and C48-treated ZIA mice had similar patterns. However, significant differences were observed in the levels of cytokines between the two groups. In 48- and 72-hour cultures, there was a significant increase in the mean level of IL-10 (250% at the peak; P ⫽ 0.049) accompanied by a significant decrease in the mean levels of IL-2 and TNF (50%; P ⫽ 0.012 for each comparison) in C48-treated ZIA mice compared with PBS-treated control mice (Figures 5A–C). The differences in the mean levels of IFN␥ and IL-12 in C48-treated mice compared with PBS-treated mice did not reach significance (P ⫽ 0.09 and P ⫽ 0.06, respectively), while IL-4 levels were undetectable in C48-treated ZIA mice. Elevation of IL-10 levels was also observed in AIA rats treated with C48 (Figure 6A). A significant increase (100%) in the IL-10 level was measured in 48and 72-hour cultures of Con A–activated splenocytes from C48-treated AIA rats (P ⫽ 0.03 and P ⫽ 0.003, respectively) compared with those from PBS-treated AIA rats. Although IFN␥ levels in C48-treated AIA rats might appear to be lower, the levels of both TNF␣ and IFN␥ in C48-treated AIA rats did not differ significantly from those in control AIA rats (Figures 6B and C). It is noteworthy that in parallel nonactivated cultures, all cytokines were undetectable. DISCUSSION The data presented in this report indicate that human PLIF and its C48 peptide exert cross-species antiproliferative effects in vitro. They also show, as in human tissue cultures (16), that C48 represents the bioactive domain of PLIF. C48 is a highly effective form

852

of therapy for ZIA in mice and for AIA in rats. Accordingly, it was demonstrated that during the induction phase of arthritis, as well as during established disease, treatment of animals with C48 led to a significant reduction in both the clinical and histologic severity of arthritis. It is noteworthy that this effect was more pronounced when the treatment was started early during the induction phase of the disease. This finding is similar to the effect achieved in RA patients, for whom treatment with disease-modifying medication is recommended early in the disease course. There is direct evidence that TNF, IL-1, and IL-6 play a role in the pathogenesis of experimental arthritis (21–23). Similarly, TNF, IL-1, and IL-6, which are produced by Th1 cells, are clearly involved in the disease process of RA (24,25), whereas IL-4 and IL-10 (produced by Th2 cells) have inhibitory effects. The balance between Th1 and Th2 cell activity seems crucial in controlling the proinflammatory immune response. Therefore, stimulation of Th2 cell activity might ameliorate the RA disease process, either directly by inhibition of macrophage activity or indirectly by suppression of Th1 cell proliferation (26–28). Our findings in the current study and in human systems (16) have shown a direct inhibition of T cell proliferation by PLIF and C48. Furthermore, in the current study, it was shown that C48 treatment of arthritis caused a systemic shift of the immune response of the host to a lower proliferation and inflammatory response, as indicated by the increased IL-10 level both in ZIA mice and in AIA rats. In ZIA mice, this was also accompanied by decreased IL-2 and TNF production following polyclonal immune activation in vitro. Recently, we obtained evidence that PLIF/C48 induces high levels of IL-10 production by human monocyte/macrophages in vitro via a calcium/calmodulin p38 mitogenactivated protein kinase pathway, and that it decreases levels of inflammatory cytokines, such as IL-1, TNF, and IL-6 (29). Accordingly, the mode of action of C48 in decreasing synovitis would be appropriate with a Th2 response. PLIF is the gene that codes for p43 of PLF (16). Indeed, the role of p43 (PLF) as an immunoregulatory cytokine has been studied in pregnant women (15,30). Some investigators believe that cytokines such as IL-2 and TNF, which are harmful to the maintenance of pregnancy, may be secreted by the activated maternal immune system. However, the fetus protects itself by producing and secreting various cytokines, such as IL-4 and IL-10, which are normally produced by a subset of CD4⫹ T cells (Th2) and antigen-presenting cells (31).

WEINBERGER ET AL

PLIF has been shown to be produced by syncytiotrophoblast cells (14) and is elevated in the circulation during pregnancy (15). It may therefore act as the inducer of these Th2 cytokines in the placenta as well as systemically, leading to suppression of activated clones at different sites of disease. This mechanism may be involved in the remissions observed during pregnancy in patients with RA (5) and multiple sclerosis (32,33). The role of PLIF in pregnancy-associated clinical improvement in RA has not been directly exhibited. However, some circumstantial observations may shed light on its significance. The increased serum level of PLIF (p43) during pregnancy declines at term. This may be linked to the observed postpartum change in the profile of cytokines secreted by macrophages (to IL-12 and TNF), which results in RA (34). Further, it was found that pregnant women with low serum levels of PLIF (p43) have infants who are small for their gestational age (35). Recently, it was reported that pregnant RA patients whose arthritis was in remission had significantly heavier babies than did those with active disease (36). The lack of remission associated with small babies might be linked to low levels of PLIF (p43) in these patients. In summary, this study demonstrates that C48 is a candidate therapeutic agent for RA. Treatment with C48, an inducer of an antiinflammatory cytokine cascade, was effective in modifying synovitis in two models of arthritis. Its antiproliferative effect suggests that C48 has pleiotropic effects that, in the context of autoimmunity, are likely to be beneficial. REFERENCES 1. Smith WD, West HF. Pregnancy in rheumatoid arthritis. Scand J Rheumatol 1960;6:189–201. 2. Felbo M, Snorrason E. Pregnancy and the place of therapeutic abortion in rheumatoid arthritis. Acta Obstet Gynecol Scand 1961;40:116–26. 3. Neely NT, Persellin RH. Activity of rheumatoid arthritis during pregnancy. Tex Med 1977;73:59–63. 4. Ostensen M, Aune B, Husby G. Effect of pregnancy and hormonal changes on the activity of rheumatoid arthritis. Scand J Rheumatol 1983;12:69–72. 5. Østensen M, Husby G. A prospective clinical study of the effect of pregnancy on rheumatoid arthritis and ankylosing spondylitis. Arthritis Rheum 1983;26:1155–9. 6. Unger A, Kay A, Griffin AJ, Panayi GS. Disease activity and pregnancy associated alpha 2-glycoprotein in rheumatoid arthritis during pregnancy. Br Med J (Clin Res Ed) 1983;286:750–2. 7. Barrett JH, Brennan P, Fiddler M, Silman AJ. Does rheumatoid arthritis remit during pregnancy and relapse postpartum? Results from a nationwide study in the United Kingdom performed prospectively from late pregnancy. Arthritis Rheum 1999;42: 1219–27. 8. Vorgensen C, Sany J. Modulation of the immune response by the

PLIF SUPPRESSES EXPERIMENTAL ARTHRITIS

9.

10. 11. 12. 13. 14. 15. 16. 17.

18. 19.

20.

21. 22. 23.

neuroendocrine axis in rheumatoid arthritis. Clin Exp Rheumatol 1994;12:435–41. Mattsson R, Mattsson A, Holmdahl R, Whyte A, Rook GA. Maintained pregnancy levels of oestrogen afford complete protection from postpartum exacerbation of collagen-induced arthritis. Clin Exp Immunol 1991;85:41–7. Jara LJ, Lavalle C, Fraga A, Gomez-Sanchez C, Silviera LH, Martinez-Osuna P, et al. Prolactin, immunoregulation and autoimmune diseases. Semin Arthritis Rheum 1991;20:273–84. Rook GA, Steele J, Brenley R, Whyte A, Isenberg D, Sumar N, et al. Changes in IgG glycoform levels are associated with remission of arthritis during pregnancy. J Autoimmun 1991;4:779–94. Bond A, Ratkay LG, Warterfield JD, Hay FC. Postpartum flare in MRL-lpr/lpr mice is associated with a parallel increase of Nacetylglucosamine on serum IgG. Br J Rheumatol 1997;36:174–7. Sirota L, Kupfer B, Moroz C. Placental isoferritin as a physiological down regulator of cellular immunoreactivity during pregnancy. Clin Exp Immunol 1989;77:257–62. Maymon R, Jauniaux E, Greenwold N, Moroz C. Localization of p43 placental isoferritin in human feto-maternal tissue interface. Am J Obstet Gynecol 2000;182:670–4. Maymon R, Moroz C. Placental isoferritin: a new biomarker from conception to delivery. Br J Obstet Gynaecol 1996;103:301–5. Moroz C, Traub L, Maymon R, Zahalka MA. PLIF: a novel human ferritin subunit from placenta with immunosuppressive activity. J Biol Chem 2002;277:12901–5. Keystone EC, Schorlemmer HU, Pope C, Allison AC. Zymosaninduced arthritis: a model of chronic proliferative arthritis following activation of the alternative pathway of complement. Arthritis Rheum 1977;20:1396–401. Pearson CM. Development of arthritis periarthritis periostitis in rats given adjuvants. Proc Soc Exp Biol Med 1956;91:95–101. Coligon JE, Kruisbeek AM, Margulies DH, Shevach EM, Strober W. Adjuvant arthritis in the rat. In: Coligon JE, Kruisbeek AM, Margulies DH, Shevach EM, Strober W, editors. Current Protocols in Immunology. Suppl 19. New York: John Wiley & Sons; 1996. p. 15.4.1–7. Quintana FJ, Rotem A, Carmi P, Cohen IR. Vaccination with empty plasmid DNA or CpG oligonucleotide inhibits diabetes in nonobese diabetic mice: modulation of spontaneous 60-kDa heat shock protein autoimmunity. J Immunol 2000;165:6148–55. Dingle JT, Page Thomas DP, Hazleman B. The role of cytokines in arthritic diseases: in vitro and in vivo measurements of cartilage degradation. Int J Tissue React 1987;9:349–54. Von Asmuth EJ, Maessen JG, van der Linden CJ, Buurman WA. Tumor necrosis factor alpha (TNF-␣) and interleukin 6 in a zymosan-induced shock model. Scand J Immunol 1990;32:313–9. Erdo F, Torok K, Szekely JI. Measurement of interleukin-1 liberation in zymosan air-pouch exudate in mice. Agents Actions 1994;41:93–5.

853

24. Deleuran BW, Chu CQ, Field M, Brennan FM, Katsiki P, Feldmann M, et al. Localization of interleukin-1␣, type 1 interleukin-1 receptor and interleukin-1 receptor antagonist in the synovial membrane and cartilage/pannus junction in rheumatoid arthritis. Br J Rheumatol 1992;31:801–9. 25. Farahat MN, Yanni G, Poston R, Panayi GS. Cytokine expression in synovial membranes of patients with rheumatoid arthritis and osteoarthritis. Ann Rheum Dis 1993;52:870–5. 26. Van Roon JAG, van Roy JLAM, Gmelig-Meyling FHJ, Lafeber FPJG, Bijlsma JWJ. Prevention and reversal of cartilage degradation in rheumatoid arthritis by interleukin-10 and interleukin-4. Arthritis Rheum 1996;39:829–35. 27. Joosten LAB, Lubberts E, Durez P, Helsen MMA, Jacobs MJM, Goldman M, et al. Role of interleukin-4 and interleukin-10 in murine collagen-induced arthritis: protective effect of interleukin-4 and interleukin-10 treatment on cartilage destruction. Arthritis Rheum 1997;40:249–60. 28. Verhoef CM, van Roon JAG, Vianen ME, Bruinzeel-Koomen CAFM, Lafeber FPJG, Bijlsma JWJ. Mutual antagonism of rheumatoid arthritis and hayfever: a role for type I and type 2 T cell balance. Ann Rheum Dis 1998;57:275–80. 29. Zahalka MA, Barak V, Traub L, Moroz C. The bioactive domain of PLIF is a direct inducer of IL-10 production in monocytes: a calcium/calmodulin p38 MAP kinase dependent pathway. FASEB J. In press. 30. Maymon R, Bahary C, Moroz C. Placental isoferritin measured by a specific monoclonal antibody as a predictive marker for preterm contraction outcome. Obstet Gynecol 1989;74:1–3. 31. Wegmann TG, Lin H, Guilbert L, Mosmann TR. Bidirectional cytokine interactions in the maternal-fetal relationship: is successful pregnancy a TH2 phenomenon? Immunol Today 1993;14: 353–6. 32. Abramsky O. Pregnancy and multiple sclerosis. Ann Neurol 1994;36:S38–S41. 33. Confavreux C, Hutchinson M, Hours MM, Cortinovis-Tourniaire P, Moreau T, the Pregnancy in Multiple Sclerosis Group. Rate of pregnancy-related relapse in multiple sclerosis. N Engl J Med 1998;339:285–91. 34. Elenkov I, Wilder R, Bakalov V, Link A, Dimitrov M, Fisher S, et al. IL-12, TNF-alpha, and hormonal changes during late pregnancy and early postpartum: implications for autoimmune disease activity during these times. J Clin Endocrinol Metab 2001;86:4933–8. 35. Rosen AC, Hafner E, Auerbach L, Rosen HR, Schuchter K, Huber K, et al. Placental isoferritin in pregnancies with small-forgestational age fetuses. Prenat Diagn 1996;16:641–6. 36. Bowden A, Barrett J, Fallow W, Silma A. Women with inflammatory polyarthritis have babies of lower birth weight. J Rheumatol 2001;28:355–9.