Activation of the alternative pathway of complement by monosodium urate crystals

CLINICAL

IMMUNOLOGY

Activation

THEODORE

AND

IMMUNOPATHOLOGY

of the Alternative Monosodium

26, 249-257

(1983)

Pathway of Complement Urate Crystals’

by

R. FIELDS,* STEVEN B. ABRAMSON,? GERALD WEIssMANN,t P. KAPLAN,* AND BERHANE GHEBREHIWET”,~

ALLEN

*Depurtment o,f Medicine, Division oj’ Aller~~y, Rheumatology, and Clinicul Immunology, Stute University ?f’ New* York, Stony Brook, New York 11794, and ?Deparfment of Medicine, Division qf Rhelrmutology, New York Unitaersity School of Medicine, New York. New York 10016 Monosodium urate crystals (MSIJ) have been shown to activate the alternative pathway of complement in a dose- and time-dependent fashion at 37°C. Activation was maximal upon addition of lo-20 m&ml monosodium mate crystals to C2-deficient human serum (C2D) or normal human serum containing 5 mM MgEGTA. Immunoelectrophoretic analysis of such treated sera demonstrated cleavage of C3 and factor 9. Incubation of highly purified C3 and factor B with 10 mgiml MSU did not, however, affect their immunoelectrophoretic pattern, suggesting that cleavage of either factor B or C3 in serum requires an intact alternative complement pathway. The fluid-phase control proteins, Factor H and Factor I, were not found to be diminished upon incubation of C2D serum or NHS containing MgEGTA with MSU. Thus activation appeared to be surface dependent and not a consequence of control protein depletion. It was also found, in agreement with earlier observations, that the classical complement pathway is activted, with concomitant depletion of Cl and C4. We conclude that MSU crystals activate both the classical and alternative pathways, and that such activation may participate in the pathogenesis of gouty arthritis.

INTRODUCTION

The pathogenesis of gouty arthritis has long been thought to be initiated by monosodium m-ate crystals (MSU)“. In support of this notion, injection of MSU into human joints reproduces the symptoms, physical findings, and joint-fluid changes characteristic of gout (1). The mechanism by which this inflammatory reaction is initiated has not been fully elucidated. One process generally regarded to be part of this mechanism is activation of the complement system. Webster et al. (2) have shown that depletion of complement in rats prior to the administration of MSU decreased the inflammation caused by these crystals. Although earlier ’ This investigation, supported by Grant 5ROl HL23714-02 from the National Heart, Lung and Blood Institute and Grant P50Al 16337-02 from National Institute of Allergy and Infectious Diseases, was presented in part at the VIII Pan-American Congress of Rheumatology, Washington, D.C., 1982. ’ To whom correspondence should be addressed: SUNY-Stony Brook, Division of Allergy, Health Sciences Center, T-16, 040, Stony Brook, N.Y. 11794. 3 Abbreviations used: MSU. monosodium urate crystals; VB, Verona]-buffered saline, pH 7.4. containing 0.15 mM CaCl, and 0.5 mM MgCl,; GVB, VB containing 0.1% gelatin; GVBE. GVB containing 0.01 M EDTA; C2D, C2-deficient serum; NHS, normal human serum; MgEGTA, 0.1 M MgCI,, 0.1 M EGTA adjusted to pH 7.0; MgGVB, VB containing 0.5 mM MgCI, and 0.1% gelatin: EA. antibody-sensitized sheep erythrocytes. 249 0090-1229/83/020249-09$01.50/O Copyright 0 1983 by Academic Press, Inc. All rights of reproduction in any form reserved

250

FIELDS

ET

AL.

studies indicated that Cl hemolytic function was minimally decreased in serum incubated with MSU (3), Giclas et ml. (4) have clearly demonstrated that MSU causes a marked decrease in the hemolytic activity of serum Cl. Furthermore, incubation of isolated macromolecular C 1 with MSU resulted in activation of C I that was independent of IgG, suggesting that MSU can directly activate the classical pathway of complement (4). Consistent with this observation, crossed immunoelectrophoretic analysis of serum incubated with various doses of MSU has revealed depletion of C3 which was both time and dose dependent (5). Previous studies have concluded that MSU do not activate the alternative pathway of complement if? \-itro (5), although immunoelectrophoretic analysis of synovial fluids from patients with gouty inflammation indicated the presence of cleavage products of factor B (6). In the following study, we present evidence that in addition to the classical pathway, the alternative pathway of complement is activated when MSU are incubated with normal human serum. Thus, both pathways may contribute to the inflammatory reaction initiated by MSU in patients with gout. MATERIALS

AND METHODS

Buglers. The following buffers were used in this study: VB, isotonic Veronalbuffered saline, pH 7.4: VB 2+, VB containing 0.15 mM CaCl, and 0.5 mM MgCl,; GVB”‘, VB?+ containing 0.1% gelatin; GVBE, VB containing 0.1% gelatin and 0.01 M EDTA; SGVB”+, GVB”+ containing 2.5% (w/v) sucrose; MgEGTA, 0.1 M MgCI, and 0.1 M EGTA adjusted to pH 7.0: and MgGVB, VB containing 0.5 m&Z MgC& and 0.1% gelatin. These buffers were prepared according to published procedures (7). Purjfied proteins. Highly purified C3 was prepared according to the method of Tack and Prahl (8) and factor B was isolated according to a previously published procedure (9). Antisera. Monospecific antisera to C3, factor B, Factor H, and Factor I were raised in rabbits. Serum from a patient with no detectable antigenic or functional C2 was the generous gift of Dr. Max Hamburger. Prepuration qf mute crystals. Microcrystalline urate was prepared according to McCarty and Faires (10). Samples were grown to produce particles I- 15 pm in length as estimated by a micrometer under light microscopy. For each preparation of MSU, half of the newly formed crystals were heated to 200°C for 2 hr to destroy possible contaminating pyrogens. Crystals were then divided into 30-mg aliquots and stored at 4°C in glass tubes that had been heated to 200°C for 2 hr. Crystals were shown to be free of bacterial contamination as assessed by a culture of the crystal suspension. In addition, neither the crystal suspension nor the assay reagents individually tested had detectable endotoxin when tested by the limulus test (ll), which is sensitive to as little as l-5 r&ml of endotoxin. Hemdvtic ussu~s. Activation of complement by urate crystals was assessed by incubating various amounts of crystals with sera and determining the percentage depletion of hemolytic complement. This determination reflects activation and then decay of critical components. Heated or unheated MSU were incubated at 10 mg/ml with normal human serum (NHS) for 60 min at 37°C. The reaction mixture

COMPLEMENT

ACTIVATION

BY

URATE

CRYSTALS

251

was centrifuged and lo-p1 aliquots were added to 0.1 ml of sensitized sheep erythrocytes (EA, 5 x lO*/ml), the volume adjusted to 0.4 ml with GVB’+, and further incubated for 60 min at 37°C. The reaction was stopped by addition of 1.6 ml of cold GVB’+, the mixture was centrifuged, and the specific hemoglobin released was measured in the supernate at 412 nm. Percentage inhibition was determined from the formula: { [A,,,NHS - A&NHS + MSU)]/A,,,NHS} x 100, and the percentage hemolysis was calculated by comparison against H,O lysis. Time-dependent study was done by incubation of NHS with MSU for varying time intervals from 5 to 120 min, removing lo-p1 aliquots after centrifugation, and adding to EA as described above. Dose-dependent study was done by adding varying doses of MSU from 2.5 to 20 mg/ml to NHS. After centrifugation lo-p1 aliquots were removed and further incubated with EA. The alternative pathway hemolytic activity was determined by incubating C2D serum or NHS containing 5 mM MgEGTA with various amounts (1 .O-20 mgiml) of heated or unheated MSU in GVB for 60 min at 37°C. Then 5 x lo7 rabbit erythrocytes (ER) were added to 25 ~1 of such treated serum and the mixture was further incubated for 60 min at 37°C in a total volume of 0.3 ml MgGVB. The reaction was stopped with 1 ml cold GVBE and after centrifugation, the hemoglobulin released into the supernatant was determined spectrophotometrically at 412 nm. Time-dependent analysis was performed by incubating 1 ml of C2D or MgEGTA serum with 10 mg of MSU. Then, O.l-ml fractions were removed at different time intervals and assayed for residual complement hemolytic activity as described above. Dose-dependent analysis was performed by incubating C2D or MgEGTA serum with doses of MSU from 2.5 to 20 mg. Samples of 0.1 ml were then removed from each tube after centrifugation, and assayed for alternative pathway hemolytic activity by the method described. Determination of CH50 (12) and individual complement titrations of C 1 (7) and C4 (13) were performed according to previously published procedures. Zmmunochemical analyses. Double radial immunodiffusion (Ouchterlony) and immunoelectrophoresis analyses were carried out in 1% agar in barbital buffer, pH 8.6, at 22°C. Immunoelectrophoresis was performed at 4.7 V/cm at 4°C for 2 hr. Quantitative single radial immunodiffusion analysis was performed according to the method of Mancini et ul. (14). RESULTS Consumption

of Serum Complement

by Monosodium

Urute Crystals

When NHS was incubated with various doses of heated or unheated MSU for 60 min at 37°C and assayed for residual complement activity by the hemolytic assay described, a significant reduction in complement hemolytic activity was observed. This reduction was dependent upon the amount of MSU added, with maximal reduction obtained at a concentration of 320 mg/ml (Table 1). However, a concentration of 10 mg/ml, which effects a significant reduction in serum hemolytic activity, was arbitrarily chosen throughout these studies. Table 2 shows the kinetics of depletion of hemolytic complement upon addition of 10 mg/ml MSU. After 60 min of incubation, 62% of serum complement was consumed. Although variability was

252

FIELDS

E-I

TABLE DEPLETION

OF COMPLEMENT OF MONOSODIUM

AL.

I

BY INCREASING CONCENTRATION URATE CRYSTALS

Sample I. 2. 3. 4. 5. 6.

NHS NHS NHS NHS NHS NHS

+ + + + +

Percentage

MSU MSU MSU MSU MSU

Norr. NHS was incubated incubation, residual hemolytic

1 mg/ml 2.5 mg/ml 5 mg/ml 10 mgiml 20 mg/ml

inhibition

0 20 29 66 74

with varying doses of MSU, from I to 20 mg/ml for 60 min at 37°C. After activity was assessed as described under Materials and Methods.

observed when the effects of different preparations of MSU were tested, no significant difference was obtained when one kind of MSU preparation was assayed using sera from various donors (Table 3). Using heated MSU, similar results, although generally of lesser magnitude, were obtained. Figure 1 shows a representative series of experiments in which Cl, C4, and CH50 determinations were made. The results were consistent and reproducible throughout these studies, and confirm the findings of Giclas et al. that MSU efficiently activate the complement system. This activation proceeds via the classical pathway since both Cl and C4 were found to be simultaneously depleted (4). In order to determine the effect of MSU on the alternative pathway, C2D or NHS containing 5 n&I MgEGTA was incubated with GVB or 10 mg/ml MSU for 60 min at 37°C. After centrifugation. the residual alternative pathway hemolytic activity was determined as described under Materials and Methods. As shown in Table 4, addition of MSU to either CZ-deficient serum or normal serum containing MgEGTA gave less lysis than the respective control when hemolytic assay was performed on the supernate. The results shown in Table 4 are representative of 13 similar experiments, performed with different preparations of MSU and NHS pooled from various donors. The mean percentage inhibition was 49% with a range of 22 to 95%. Use of heated crystals gave, on the average, 6% less inhibition than unheated crystals. Table 5 demonstrates the dose-dependent nature of this activation. MSU at 5 mg/ml gave only a minimally detectable complement conTABLE TIME-DEPENDENT

DEPLETION

2

OF COMPLEMENT

Sample I. 2. 3. 4. 5. 6.

NHS NHS NHS NHS NHS NHS

+ + + + +

Nore. NHS was incubated tween 5 and 120 min.

BY MSU

CRYS-PALS

Percentage

MSU MSU MSU MSU MSU with

5 min 15 min 30 min 60 min 120 min or without

inhibition

4 I8 39 62 80 10 mgiml

MSU

at 37°C for varying

lengths

of time

be-

COMPLEMENT

ACTIVATION

253

BY URATE CRYSTALS

TABLE 3 ASSESSMENT OF THE ABILITY OF DIFFERENT PREPARATIONS OF MONOSODIUM URATE CRYSTALS TO DEPLETE COMPLEMENT UTILIZING DIFFERENT SERA Donor I. 2. 3. 4. 5.

MSU preparation

JD TF Pool 2 TF GG

1 I I 2 3

Consumption of complement” (96) 66 62 61 SO 74

Note. Five different experiments, using different NHS sources, are represented. In each case, NHS (from various batches) was incubated with or without 10 mgiml MSU. for 60 min at 37”C, and after incubation, the residual hemolytic activity was assessed as described. ” Mean is 63.

sumption, while at doses of 7.5 mg/ml and 10.0 mg/ml a marked increase in percentage depletion was observed. Since cleavage of factor B and C3 are indicators of activation of the alternative pathway of complement, we also analyzed the effect of MSU on serum C3 and factor B by immunoelectrophoresis. Figure 2 demonstrates that both C3 and factor B were cleaved in NHS-MgEGTA treated with MSU as judged by the shift of electrophoretic mobility with appearance of cleavage products. However, when 100

90

80

70

z 0 F

60

$

50

zs

40

30

20

10

0 Cl %O FIG. 1. Reduction of the activity of total complement and complement components in serum. CH50, Cl. and C4 determinations were done as described under Materials and Methods.

254

FIELDS

EFFECT

OF MSU

ET AL.

TABLE 4 ON THE ALTERNATIVE PATHWAY

Sample

OF COMPLEMENT Percentage

inhibition -

C2D + buffer C2D + MSU IO mgml C2D + zymosan 5 mg/ml NHS-MgEGTA + buffer NHS-MgEGTA + MSU 10 mgfml NHS-MgEGTA + 5 mgiml zymosan

30 70 41 94

Nore. C2D or NHS containing 5 mM MgEGTA was preincubated with 10 mg/ml unheated monosodium urate crystals for 60 min at 37°C. Similarly, zymosan. at 5 mg/ml, was incubated with C2D or NHS containing 5 mM MgEGTA for 60 min at 37°C. Alternative pathway hemolytic activity was then determined spectrophotometrically.

isolated C3 or factor B, at physiological concentrations, was separately incubated with heated or unheated MSU crystals (not shown), no significant shift in electrophoretic mobility was evident, indicating that MSU do not directly activate isolated C3 or factor B. To investigate the mechanism by which MSU effect the activation of the alternative pathway, NHS-MgEGTA which had been incubated with MSU was tested for the levels of the two alternative pathway regulatory proteins Factor H and Factor 1. Neither Factor H nor Factor I was shown to be depleted, as assessed by the single radial immunodiffusion technique using monospecific antisera to Factor H and Factor I. DISCUSSION

Many reports from different laboratories indicated that MSU-dependent activation of several pathways, individually or in concert, may be responsible for the induction of the pathologic state of gouty inflammation. Ginsberg er ul. (15) have shown recently that urate crystals activate Hageman factor in human plasma and synovial fluid, and postulated that this activation may lead to pain, vasodilatation, and increased vascular permeability. An activation fragment of Hageman factor, HFf, can initiate the classical pathway of complement, as has been recently shown by Ghebrehiwet et al. (16) in our laboratory. Thus, the two pathways may be

DOSE-DEPENDENT

DEPLETION

Dose of MSU (mgiml) 0 5 7.5 10.0

TABLE 5 OF THE ALTERNATIVE

COMPLEMENT

PATHWAY

Percentage inhibition

15 39 95

Nclrr. NHS containing 5 n&I MgEGTA was preincubated with varying doses of MSU (from 5 to 10 mgiml) for 60 min at 37°C. The reaction mixture in each case was centrifuged and 25 ~1 of supernatant serum was further incubated with 2.5 x IO7 rabbit erythrocytes (ER) for 60 min at 37°C.

COMPLEMENT

ACTIVATION

BY URATE

CRYSTALS

255

FIG. 2. Effect of MSU on serum C3 and factor B. NHS, containing 5 mM MgEGTA, was incubated with or without MSU for 60 min at 37°C. Then the effect of MSU on serum C3 and factor B was analyzed immunoelectrophoretically on 1% agar in barbital buffer. pH 8.6. The plates were developed with monospecific antisera to C3 and factor B.

linked. In addition, Giclas et al. (4) have clearly demonstrated that MSU activate the classical pathway of complement in a manner that does not require immunoglobulin. Our present studies show that: (1) in accord with the findings of Giclas et ctl. (4), the classical pathway is activated as shown by the reduction of total hemolytic activity as well as by the sequential depletion of individual components such as Cl and C4; 2) in addition to the classical pathway, the alternative pathway of complement is activated by monosodium urate crystals. C2D serum or NHS containing 5 mM MgEGTA was incubated with GVB either alone or with 10 mgiml of heated or unheated MSU and the total alternative pathway hemolytic activity was found to be significantly depleted in the samples incubated with MSU. Results were within a narrow range for assays on the same serum and the same batch of MSU. When different batches of urate and different sera were used, the qualitative findings were highly reproducible, yet the quantitative results were quite variable. The wide range of our percentage inhibition reflects this variability, a variability which cannot be fully explained. We did not feel that endotoxin was at all responsible for this variability. Since certain bacterial lipopolysaccharides are known to activate the alternative pathway (17), the finding that heated crystals activate the alternative pathway (although of somewhat lesser degree than untreated crystals) makes a possible contribution by endotoxin unlikely. In addition, the limulus test for the presence of endotoxin as well as the culture for bacterial contamination were found to be negative when MSU suspensions were tested. There was considerably greater variability in the ability of MSU to activate the alternative complement pathway in different sera than that found when activation of the classical complement pathway was assessed. The reason for this variability is not known. Concomitant with the depletion of the alternative pathway hemolytic activity, it was noted that C3 and factor B in MSU-treated serum were activated, as assessed by immunoelectrophoretic analysis using monospecific antisera to C3 and factor B. However, when highly purified C3 or factor B at physiological concentrations

256

FIELDS

E I- AL

was incubated with 10 mg/ml MSU, no shift in electrophoretic mobility or the appearance of cleavage products was observed. Thus, the other plasma proteins required for factor B cleavage, such as D, must be present. Analyzing synovia] fluids from gouty joints, Hunder et al. (6) also found that C3 and factor B were activated. Since simultaneous depletion of C3 and factor B is indicative of alternative pathway activation, these findings, taken together, suggest the prticipation of this pathway in gouty arthritis. Hasselbacher (5) has previously suggested that the alternative pathway was not activated by MSU. He used MgEGTA-NHS, and attempted to deplete complement with 2.1 mg/ml MSU. Our findings are actually consistent with his, in that greater doses of MSU appear to be required for alternative pathway activation. The mechanism by which activation of the alternative pathway by MSU proceeds is not known. However, since no depletion of the two fluid-phase control proteins of the alternative pathway, Factor H or Factor I, was observed, the possibility that activation is a result of depletion of these proteins is unlikely. Rather, we feel that sodium urate crystals provide an effective surface that can bind C3 and factor B so that cleavage by factor b can proceed. This has been shown to require that the surface exclude or otherwise minimize the effect of the control proteins Factor I and Factor H so that activation can proceed (18). Although certain classes of immunoglobulin, such as IgA. may activate the alternative pathway, it has become clear that this pathway may be activated in the total absence of specific immunoglobulins (19). In other preliminary experiments (not shown), no difference was observed when MSU was incubated with either NHS-MgEGTA or agammaglobulinemic serum in the presence of 5 mM MgEGTA. This result suggests that immunoglobulin might not play a role in the activation of the alternative pathway by MSU. Activation of both the classical and alternative pathways was found to be both time and dose dependent. However. in most of our studies, a concentration of 10 mg/ml serum and an incubation time of 60 min were employed. It is unknown whether such a concentration of urate crystals would be present in V~VO, under conditions that initiate the pathologic state of gouty arthritis. However, when urate crystals form within joints, and are visible by light microscopy, the concentrations we are dealing with are clearly in the milligram range. Injection of 15 mg of MSU in 2-3 ml pyrogen-free saline into the joints of anesthesized dogs, for example, was found to induce an acute exudative response (20) which is characteristic of gouty inflammation. It is possible, therefore, that initiation of the alternative pathway with as little as 5 mg/ml could play a role in the pathogenesis of the disease. Activation of complement through either pathway (classical or alternative) generates a number of biological peptides (such as C3a, C4a, and C5a) that are known to play a role in inflammatory reactions (21, 22). It is conceivable that in the absence of a fully functional classical pathway, due either to genetic deficiencies in the early components, such as Cl, C2, or C4, or to depletion of one of these proteins as a result of other kinds of inflammatory processes, the alternative pathway could be triggered by MSU. In patients with an intact classical pathway, the complement response may be amplified by the contribution of the alternative pathway.

COMPLEMENT

ACTIVATION

BY

URATE

CRYSTALS

257

ACKNOWLEDGMENT The authors

wish

to thank

Karen

Randall

for typing

the manuscript.

REFERENCES 1. Seegmiller. .I. E., Howell. R. R., and Malawista, S. E., J. Amer. Med. Assoc. 180, 469, 1962. 2. Webster. M. E., Maling, H. M.. Zweig, M. H., Williams, M. A., and Anderson, W.. Zmmunol. Commun. 1, 185, 1972. 3. Naff, G. B., and Byers. P. M.. J. Lob. Clin. Med. 81, 747. 1973. 4. Ciclas, P. C.. Ginsberg, M. H., and Cooper. N. R.. J. Clir7. Invest. 63, 759, 1979. 5. Hasselbacher, P.. Arthritis Rheum. 22, 571. 1979. 6. Hunder, G. G., McDuffie, F. C., and Mullen. B. .I.. J. Lab. C/in. Med. 89, 160, 1977. 7. Borsos, T., and Rapp, H. J., .I. Immunol. 9, 851, 1963. 8. Tack, B. F., and Prahl, J. W., Biochemistry 15, 4513, 1976. 9. Gdtze, O., and Miiller-Eberhard, H. J.. J. Exp. Med. 134, 9Os, 1971. 10. McCarty, D. J., and Faires, J., Curr. Ther. RPS. 5, 284, 1963. 11. Levin, J., Tomasulo. P. A., and Oser, R. S.. J. Lub. Clin. Med. 75, 903. 1970. 12. Mayer. M. M., In Experimental Immunochemistry. (E. A. Kabat and M. M. Mayer, Eds.). p. 133. Springfield, Illinois, Charles C. Thomas, 1961. 13. Mtlller-Eberhard, H. J.. and Biro, C. E.. J. Exp. Med. 118, 447, 1963. 14. Mancini. G. A., Carbonara. O., and Heremans. J. F., Immunochemistry 2, 235, 1965. 15. Ginsberg, M. H., Jaques, B., Cochrane. C. G.. and Griffin, J. H.. J. Lab. C/in. Med. 95, 497. 1980. 16. Ghebrehiwet, B.. Silverberg, M., and Kaplan, A. P., J. Exp. Med. 153, 665, 1981. 17. Bladen, H. A.. Gewurz, H., and Mergenhagen. S. E., J. Exp. Med. 125, 767, 1967. 18. Fearon, D. T., and Austen, K. F., Proc. Nut. Acad. Sri. USA 74, 1683, 1977. 19. Miiller-Eberhard, H. J., and Schreiber, R. D., Adv. Zmmunol. 29, 1. 1980. 20. McCarty, D. J.. Phelps, P., and Pyenson J.. J. Exp. Med. 124, 99. 1966. 21. Hugli, T. E., p. 321. Critical Review in Immunology. CRC Press, Boca Raton, Fla., 1981. 22. Gorski, J. P., Hugli, T. E., and Miiller-Eberhard, H. J., f’roc. Nur. Acad. Sci. (IsA 76, 5299, 1979. Received

May

26, 1982; accepted

with

revisions

August

28. 1982.