Induction of proliferative synovitis in rabbits by intra-articular injection of immune complexes.код для вставкиСкачать
Induction of Proliferative Synovitis in Rabbits by Intra-articular Injection of Immune Complexes By ARNOLDJ. RAWSONAND TITO P. TORRALBA S of rheumatoid arthritis (RA) have implicated immunoglobulins in the genesis of the local inflammation. Rheumatoid factors (RF), IgG, and IgM have been shown to be present in, and presumably are produced by, synovial plasma ce1ls.l Polymorphonuclear leukocytes (PMN) of RA joint fluid have cytoplasmic granules,Z which contain IgG, IgM3, and, in many cases, BIG g l ~ b u l i n .Moreover, ~ joint fluid of RA has a relatively decreased content of several complement ( C ’ ) components when compared to paired serum samples, and to joint fluid from most other forms of joint inflammation5j6;material containing IgG or IgM with C’, and IgG with RF, has been demonstrated in synovial tissues.’ Even patients with very low levels of serum immunoglobulins may develop arthritis in which immunoglobulins are present in the joint fluid.* Although proliferative synovitis has been produced experimentally in hyperimmune animals by the intra-articular injection of specific antigen,O>l2there is no convincing evidence for a hyperimmune state in patients with RA. On the contrary, the observations seem to be consistent with the reaction of immunoglobulins within the joint space itself. The purpose of the present study is to determine if immunoglobulin reactions localized to the joint space of nonimmunized experimental animals reproduce the anatomic lesions of RA. Various materials were injected into the knee joints of normal rab- bits, and the resulting tissue changes were observed. From the Department of Pathology and the Arthritis Clinic, Uniuerssity of Pennsylvania School of Medicine, Philadelphia, Pa. Supported by RG6218, N I H , USPHS. ARNOLDJ. RAWSON,M.D.: Associate Professor of Surgical Pathology; TITOP. TOHRALBA, M.D.: Fellow, Arthritis Section; both of the University of Pennsylvania School of Medicine. TUDIES MATERIALS AND METHODS Experimental Animals. Eighteen female albino rabbits weighing from 3 to 4 Kg. were used. All injections were given into the knee joints, in most cases using the contralateral joint as control. The skin was shaved, sterilized with iodine, and the joint was entered anteriorly and slightly laterally with a #22 needle. Sterile technic was observed throughout. (The experimental plan is summarized in Table 1; 11 of the 18 animals were injected only once in acute experiments.) Injections were given twice weekly for the duration of the experiment. Knee joints were aspirated occasionally during the course of the experiment. Animals were sacrificed with intravenous toxital, 2 ml. At Autopsy, joint fluid and all synovial membranes were removed from the knee joints. Reagents. Egg albumin (EA) 5X crystalline (Nutritional Biochemicals Corp.) was titrated against rabbit antiovalbumin (Nutritional Biochemicals Corp.). It was found that equivalence required the use of a 0.025 per cent solution of antigen. One-half ml. of antigen solution, onehalf ml. of antiserum, and one-half ml. of fresh autologous serum (FS) as a source of C’ constituted the injection mixture for this system. The components were injected immediately after mixing. Pneumococcus type 111 pofysaccharide (PN) and its specific rabbit antiserum were obtained through the courtesy of Dr. Gerald Schiffman. The antiser i m (lot 699-17) contained 1.5 mg. antibody N per ml. The haptene, used a t equivalence, was at a concentration of 400 pg per ml. One-half ml. each of antiserum and haptene solution constituted the injection mixture. FS was not added because C‘ activity was present in the antiserum. The animals which were hyperiinrnnnizecl were given 1 ml. of pneumococcus type I11 vaccine (lot A66) intravenously 4 days a week for 2 weeks, following which 0.5 ml. of haptene solution 44 ARTHRITIS AND RHEUMATISM,VOL. 10, No. 1 ( FEBHUARY, 1967) 45 INDUCTION OF PROLIFERATIVE SYNOVITIS I N RABBmS Table 1.-Summaru Rabbit Number 10 13 2 8 9 11 14 19 7 18 Left Knee Injection AGG-serum PN-anti P N PN-anti P N PN-anti P N human albumin AGG-serum PN-anti PN-RF AGG* PN-anti P N AGG-serum 4 PN-anti P N 3 EA-anti E A 1 12 16 11 EA-anti EA AGG-saline PN-anti P N PN-anti P N PN PN 5 6 of Procedure and Results* R:ght Knee Lesions PMN PMN-G PMN-G, S P PMN-G, S P N PMN-G. S P PMN-G, S P see text PMN-G, SP, GC PMN-G, S P see text PMN-G, S P VIL, PC, F PMN-G, S P VIL, PC, F N see text N N PMN-G, SP, VAS PMN-G, SP, VAS Injection Lesions AGG-saline AGG-serum serum PN human albumin AGG-RF PN-anti P N AGG*-serum serum AGG-RF serum PMN PMN N N N PMN-G, S P PMN-G, S P see text N PMN-G, S P VIL, PC. F N serum N serum AGG-RF saline saline saline saline N see text N N N N General injected 1 wk. only zymosan nitrogen mustard nitrogen mustard hyperimmunized hyperimmunized Total Injections Sacrificed 1 1 1 1 1 1 1 1 2 4 8 hrs. 8 hrs. 3 days 3 days 3 days 3 days 3 days 3 days 5 days 2 wks. 8 4 wks. 12 6 wks. 2 1 1 1 2 2 6 wks. 8 hrs. 3 days 3 days 5 days 5 days *Except f o r serum or saline, the quantities of which were equal on each side, 0.5 ml. of each reagent was injected. Serum was fresh, isologous or autologous, each knee receiving the same a s its control. AGG* is labeled with fluorescein isothiocyanate. PMN-G = PMN with cytoplasmic granules: SP = synovial proliferation ; GC = giant cells ; VIL = villi ; PC = plasma cells : F = follicles : VAS = vasculitis ; N = normal. was given in one knee and 0.5 ml. saline in the other. As a source of rheumatoid factor (RF), serum from a patient with rheumatoid arthritis was employed. This serum was stored for 1 year, at the end of which time C' activity could not be demonstrated by the sensitized erythrocyte test. The titer of R F immediately before use was 15120 (latex fixation Difco F 11). One-half ml. of serum was used for each injection. Aggregated IgG (AGG) was prepared by heating human gamma globulin, fraction I1 (Pentex, Inc.) in 1 per cent solution in borate buffered saline (pH 8.0) at 65 C. for 20 minutes. One-half ml. was used for each injection. Controls. The knee opposite to that receiving the complete immunologic mixture received either serum alone or antigen alone in the acute experiments. These reagents were made up to comparable volume with isotonic sodium chloride solution. In the longer experiments, antigen controls were omitted to avoid sensitization. In no instance was a lesion seen in the control knee. Decomplementation. Boiled zymosan in isotonic sodium chloride solution (15 mg. per ml.) was administered as follows: At -45 minutes: 1.0 ml. intravenously (I.V.) and 1.5 ml. intraperitoneally (1.P.). At -15 minutes: 1.0 ml. I.V., 1.5 ml. I.P. The same dosages were repeated at +1 hour, +2 hours, +4 hours, +6 hours. Blood was drawn at +45 minutes, and again at $4 hours. Activity of B,, globulin (measured by Dr. Fletcher Taylor) was 50 per cent of normal in both specimens. The assay was performed by selective destruction of B,, globulin activity with hydrazine, after prior neutralization of C', esterase with anti-C', serum. Depletion of P M N . Nitrogen mustard was given intravenously (1.75 mg. per Kg.), and the blood leukocytes were examined daily thereafter. When the PMN count reached zero (1-2 days) the experiment was begun. The animals were maintained throughout on 125 mg. streptomycin and 150,000 units penicillin daily. Histologic Technic. Synovial membranes were fixed in neutral buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. Joint fluids were centrifuged, and the sediment washed 3 times in isotonic sodium chloride solution, and fixed 10 minutes in acetone. Fhosescence Studies. Fluorescein-labeled goat antirabbit globulin serum (Hyland) was employed for study of joint fluid cells. Specificity was controlled by inhibition with nonfluorescent antiserum. AGG was labeled with fluorescein isothiocyanate (1 mg. per 40 mg. protein). The mixture was kept at pH 9.2 to 9.5 for 1 hour, stored at 4 C. overnight, and taken off sephadex G50 with borate buffer pH 8.2 (AGG"). 46 RAWSOh' AND TORRALBA Fig. 1A.-Rabbit 13 (left knee). Joint fluid cells stained with fluorescein-labeled goat anti-rabbit IgG serum. Note fluorescent intracytoplasmic granules. x 750. Duration of experiment: 8 hours. in more PMN, but far fewer than in the RESULTS An acute inflammatory reaction was well nondecomplemented animals. The synovial established eight hours after the intra- membranes, following both injection mixarticular injection of PN-anti PN, EA-anti tures, showed a few PMN within or near EA, AGG-RF, AGG-FS, or AGG. The joint venules and a few at the synovial surface. space contained turbid fluid and many Other inflammatory features were lacking. PMN. Intracytoplasmic granules were not Three days after injection of either AGGprominent following the injection of AGG FS or AGG-RF into intact animals there or AGG-FS, but granules were numerous was minimal joint fluid with rare PMN after injection of other systems and were containing a few granules. The synovial specifically stained with goat antirabbit membranes showed a slight proliferative gamma globulin labeled with fluorescein reaction of the synovial cells. An animal (Fig. 1A). The synovial membrane was receiving fluorescein-labeled AGG in one edematous, the synovial cells showed nu- joint and labeled AGC together with FS in clear enlargement and rounding, and nu- the other showed only traces of fluorescence merous PMN were clustered along the in the synovial membrane of both joints after three days. synovial surface (Fig. 1B ). Three days after injection of PN-anti PN, The acute reaction was somewhat modified in the zymosan-treated animal. The however, there was prominent proliferation injection of AGG-FS was followed by mini- of the synovial cells. PMN were not seen mal joint fluid and only a rare PMN in among the proliferated cells, but were coneight hours. Injection of AGG-RF resulted centrated along the synovial surface. Small INDUCTION OF PROLIFERATIVE SYNOVITIS IN RABBITS 47 Fig. 1B.-Rabbit 13 (left knee). Synovial membrane. Note clustering of PMN on surface. H. and E., x 280. Duration of experiment: 8 hours. vessels of the synovia showed no evidence of vasculitis (Fig. 2). Five days after the initial injection of PN-anti PN, there was even more marked proliferation of synovial cells. Mitotic figures were numerous, and many multinucleated giant cells were seen (Fig. 3 ) . PMN containing granules were present in the synovial cavity, but were not seen in the synovial membrane. Two weeks after the initial injection of AGG-FS there was moderate proliferation of the synovial cells (Fig. 4A) with scattered clumps of fibrinous material on the surface. A few small inconspicuous villi were seen, and rare plasma cells were in the synovia. PMN containing granules were present in the joint space, but not in the synovial membrane. After four injections (in two weeks) of AGG-RF there was marked synovial cell proliferation. Villi were large and prominent. Throughout the proliferat- ing area were numerous plasma cells and early lymphoid follicles (Fig. 4 B ) . Multinucleated giant cells were also present. PMN with granules were numerous in the synovial fluid, but absent from the membrane. After 8 (in 4 weeks) injections with PN-anti PN, and after 12 (in 6 weeks) injections with EA-anti EA, there was gross enlargement of the joint with thickening of the soft tissues and granulomatous proliferation appearing microscopically as a pannus covering the cartilage (Fig. 5A). There were very large, prominent villi containing lymphoid follicles devoid of germinal centers, closely resembling those seen in human RA (Fig. 5B). The t'issues were heavily infiltrated by plasma cells. Although PMN were numerous in the joint space, they were absent from the synovial tissues. Animals made agranulocytic with nitrogen mustard received injections of PN-anti 48 RAWSON AND TORRALBA Fig. 2.-Rabbit 2 (left knee). Synovial membrane. Note marked proliferation of synovial cells, with surface clustering of PMN, but no inflammation of venule at lower left. H. and E., x 280. Duration of experiment: 3 days. Fig. 3.-Rabbit 7 (left knee). Synovial membrane. Note marked synovial cell proliferation and multinucleated giant cells. H. and E., x 280. Duration of experiment: 5 days. INDUCTION OF PROLIFERATIVE SYNOVITIS IN RABBITS Fig. 4A.-Rabbit 18 (right knee). Synovial membrane after injection of AGG-serum. Synovial cell proliferation. H. and E., x 280. Uuration of experiment: 2 weeks (compare with Fig. 4B). Fig. 4B.-Rabbit 18 (right knee). Synovial membrane after injection of AGG-RF. Villous synovitis with formation of follicles composed of plasma cells and lymphocytes. H. and E., x 280. Duration of experiment: 2 weeks (compare with Fig. 4 A ) . 49 50 RAWSON AND TORRALBA Fig. 5A.-Rabbit 4. Left knee (left) syiiovial membrane after injection of PN-anti PN. Shows marked thickening of synovium with some pannus formation. (Duration of experiment: 4 weeks). Fig. 5B.-Rabbit 4 (left knee) synovial membrane after injection of PN-anti PN. There is villous synovitis with lymphoid follicles and heavy plasma cell infiltrate. H. and E., x 280. Duration of experiment: 4 weeks. INDUCTION OF PROLIFERATIVE SYNOVITIS IN RABBITS 51 response; multiple injections induced a recurrent acute response and produced the typical anatomic lesions of RA. The injected material included a protein antigen, a polysaccharide haptene, and RF-AGG. AGG injections alone produced an arthritis, but changes were less typical of RA. Constant presence of the reactants appeared to be necessary, as there was a tendency, at least in the early stages, for the process to regress if injections were discontinued. Labeled AGG was almost completely removed from the joint cavity within three days. Although treatment with zymosan did not remove all the Blc globulin, infiltration of PMN into the injected joint was reduced. Interpretation of this finding is uncertain, since this treatment also may depress the level of circulating leukocytes and, unfortunately, blood counts were not obtained in these experiments. In view of the demonstration of chemotactic properties of C’oC6,l0J1 it is likely that the C’ system may be of importance in the initial attraction of PMN to the joint. The failure of synovial proliferation to develop in animals made leukopenic with nitrogen mustard suggests that the lesions are mediated by PMN2 rather than by a direct injury to the synovial cells themselves. The ability of streptolysin S to induce acute and chronic arthritis13 suggests that tissue damage may be mediated by substances released from damaged lysosomes. Here the postulated lysosomal damage may be from PMN or from synovial lining cells. The tissue changes appearing in the hyperimmune animal are very different from those appearing in the nonhyperimmune animal and involve necrotizing vasculitis, a process not seen in the joints of RA. It DISCUSSION is tempting to speculate that a hyperThese experiments demonstrate that sin- immune state may have developed in those gle injections of various antigen-antibody rheumatoid patients whose course is comsystems into normal rabbit joints are cap- plicated by vasculitis. able of inducing an acute inflammatory PN into the knee joints according to the same schedule as that of the intact animals. Three days after the initial injection the animals were sacraced: PMN were absent from both the joint cavity and synovial tissues, no significant joint fluid was present, and the synovial cells showed no evidence of proliferation. The tissues were indistinguishable from those of the normal control knee. The joints made hyperimmune to type 111 pneumococcus, of animals which received intra-articular injections of PN over a 5 day period, contained fluid and numerous PMN with intracytoplasmic granules, which showed specific staining with fluoresceinlabeled goat antirabbit globulin serum. There was marked synovial cell proliferation comparable with that noted after the injection of PN-anti PN into the joints of normal animals over a 5 day period. The hyperimmune animals differed from the nonhyperimmune ones, however, in that in the former PMN were present in large numbers throughout the synovial membrane instead of being limited to the synovial surface. In addition, there was severe vasculitis, with necrosis of venular walls, and concentration of PMN in these areas. Extensive fibrinoid necrosis of synovial membrane was also seen. In rabbit No. 14, an attempt was made to determine the effect of R F on the biologic activity of an antigen-antibody system. No significant alteration in response could be seen when RF was added to the freshly mixed reagents just prior to injection. An animal which received intra-articular human serum albumin (0.5 per cent in saline), aggregated by heating to 80 C., showed no inflammatory response after 3 days. 52 RAWSON AND TORRALBA SUMMARY Injection of various immune systems, including complexes of rheumatoid factoraggregated gamma globulin, into normal rabbit joints induces t h e anatomic lesions similar to those of rheumatoid arthritis when introduced into normal rabbit joints. The development of these lesions seems to depend upon t h e continued presence of t h e reactants and the availability of polymorphonuclear leukocytes. Injections into hyperimmune animals induced vasculitis in addition to an inflammatory reaction. SUMMARIO IN INTERLINGUA Varie systemas immun-incluse complexos de factor rheuinatoide con aggregate globulina gamma-induce le lesiones anatomic de arthritis rheumatoide quando illos es introducite ad in le articulationes de conilios normal. Le disveloppamento del lesiones pare depender del presentia continue del reactantes e del disponibilitate de leucocytos polymorphonucleari. Injectiones in animales hyperimmun induceva vasculitis a parte le reaction inflammatori. REFERENCES 1. Mellors, R. C., Nowoslawski, A., and Korngold, L.: Rheumatoid arthritis and the cellular origin of rheumatoid factors. Amer. J. Path. 39:533, 1961. 2. Hollander. J. L.,McCarty, Jr., D. J., Astorga, G., and Castro-Murillo, C.: Studies on the pathogenesis of rheumatoid joint inflammation. I. The “R. A.” cell and a working hypothesis. Ann. Int. Med. 62:271, 1965. 3. Rawson, A. J., Abelson, N. M., and Hollander, J. L.: Studies on the pathogenesis of rheumatoid joint inflammation. 11. Intracytoplasmic particulate complexes in rheumatoid synovial fluids. Ann. Int. Med. 62:281, 1965. 4. Rawson, A. J.: Unpublished data. 5. Pekin, T. J.. and Zvaifler, N. J.: Hemolytic complenient in synovial fluid. J. Clin. Invest. 43:1372, 1964. 6. Fostiropoulos. G., Austen, K. F., and Bloch, K. J.: Total hemolytic complement (CH,,) and second component of complement (C.211~)activity in serum and synovial fluid. Arthritis Rheum. 8:219, 1965. 7. Fish, A. J., &Tichael, A. F., Gewurz, H., and Good, R. A.: Immunopathologic changes in rheumatoid arthritis synovium. Arthritis Rheum. 9:267, 1966. 8. Collins, J. R., and Ellis, D. S.: Agammaglobulinemia, malabsorption, and rheumatoid-like arthritis. Amer. J. Med. 39:476, 1965. 9. Klinge, F.: Der Rheumatismus: Die experimentelle serumhyperergie. Ergebn. Allg. Path. 27:245, 1933. 10. Ward, P. A., and Cochrane, C. G.: Bound complement and immunologic injury of blood vessels. J. Exp. Med. 121:215, 1965. 11. Ward, P. A., Cochrane, C . G., and MiillerEberhard, H. J.: The role of serum complement in chemotaxis of leukocytes in uitro. J. Exper. Med. 122:327, 1965. 12. Glynn, L. E.: Arthropathy in man and animals. Proc. Roy. SOC.Med. 58:365, 1965. 13. Weissmann, G., Becher, B. S., Wiedermann, G., and Bernheimer, A. W.: Studies on lysosomes. VII. Acute and chronic arthritis produced by intra-articular injections of streptolysin S in rabbits. Amer. J. Path. 46:129, 1965.