Complement Pathway Function

Transcription

1 INFECTION AND IMMUNrrY, Apr. 1977, p Copyright X) 1977 American Society for Microbiology Vol. 16, No. 1 Printed in U.S.A. Comparison of Ethyleneglycoltetraacetic Acid and Its Magnesium Salt as Reagents for Studying Alternative Complement Pathway Function DOUGLAS P. FINE Department ofmedicine, University oftexas Medical Branch, Galveston, Texas 77550*, and the Bacteriology Division, U. S. Army Medical Research Institute ofinfectious Diseases, Frederick, Maryland Received for publication 6 October 1976 The divalent cation chelators, ethyleneglycoltetraacetic acid (EGTA) and its magnesium salt, MgEGTA, were compared in studies of alternative complement pathway function. EGTA (0.01 M) inhibited both the rate and the amount of complement activation by zymosan whether compared to nonchelated serum or to serum chelated with MgEGTA (0.01 M). The rate of alternative pathway activation by zymosan was slightly slower in MgEGTA-chelated serum than in nonchelated serum, but the overall amount of complement consumed by a given amount of zymosan was not decreased. MgEGTA chelation spontaneously activated the alternative pathway, as reflected by lysis of erythrocytes from a patient with paroxysmal nocturnal hemoglobinuria. No evidence could be found that MgEGTA either spontaneously activated C2 or facilitated zymosan activation of C2. Suggested guidelines for the use of these chelators are advanced. Differential divalent cation chelation of serum using ethyleneglycoltetraacetic acid (EGTA) has provided a useful method for study of the alternative complement (C) pathway (5). The method depends upon the fact that, whereas the classical C pathway requires both calcium and magnesium, the alternative pathway requires only magnesium (13). EGTA, by chelating calcium considerably more avidly than magnesium, blocks the classical pathway selectively. The magnesium level in EGTAchelated serum is, however, lower than normal-enough lower that alternative pathway function is partially impaired (1, 3, 6). The magnesium salt of EGTA (MgEGTA) also binds calcium avidly but supplies normal (actually supranormal) levels of magnesium in chelated serum (2) and has been suggested as the preferable form of the chelator for alternative pathway study (1, 6). Enthusiasm for the use of MgEGTA has, however, been dampened by evidence that excessive magnesium per se, either as MgCl2 (9, 10) or as MgEGTA (3), activates the alternative pathway and perhaps even the classical pathway. This paper explores the kinetics of alternative pathway activation in the presence of the chelators EGTA and MgEGTA, reconfirms the findings of alternative pathway activation by MgEGTA, but fails to confirm evidence that MgEGTA, either alone or in concert with zymosan, activates the classical pathway at the level of the second component (C2). 124 MATERIALS AND METHODS Serum. Human blood from healthy adult male volunteers was collected by venipuncture; guinea pig blood from healthy outbred animals was obtained by cardiac puncture. After clotting for 2 h at room temperature, blood was centrifuged (2,500 x g) and serum was stored at -70 C in 1-ml aliquots until just prior to use. Once thawed, unused serum was discarded. Chelators and divalent cation salts. EGTA (Sigma Chemical Co., St. Louis, Mo.) and EDTA (Matheson Coleman and Bell Manufacturing Chemists, Norwood, Ohio) were prepared in 0.1 M stock solutions as previously described (5). MgEGTA was prepared in 0.1 M stock solution as described by Bryan (1). Serum was chelated by addition of 0.1 ml of stock solution to each 1 ml of serum for a final chelator concentration of approximately 0.01 M (5). Zymosan. Stock solutions of zymosan (Z, Nutritional Biochemicals Corp., Cleveland, Ohio) were prepared as previously described (5). Erythrocytes. Sheep erythrocytes were obtained from Cordis Laboratories, Miami, Fla. Erythrocytes from a man with paroxysmal nocturnal hemoglobinuria (E-PNH) were washed and suspended in saline and stored at 4 C. Hemolysin. Rabbit antiserum to sheep erythrocytes was supplied by Difco Laboratories, Detroit, Mich.; rabbit 19S antibodies to sheep erythrocytes were supplied by Cordis Laboratories, Miami, Fla. EA. Sheep erythrocytes (E) were sensitized with hemolysin (A) and brought to a final concentration of 109 cells/ml. A 3-ml volume of EA was then washed twice with distilled water, after which all supernatant fluid was aspirated and the erythrocyte

2 VOL. 16, 1977 membranes were used in experiments seen in Table 1. Complement components. Partially purified guinea pig components C1 and C2 were obtained from Cordis Laboratories, Miami, Fla. Complement titrations. Hemolytic measurements of C2 were performed by the method of Rapp and Borsos (12). Hemolytic titrations of the lateacting components C3 to C9 (i.e., C3-9) were performed using the following modification of methods of Rapp and Borsos (12) and Kabat and Mayer (8): 1- ml aliquots of sheep erythrocytes sensitized with 19S antibody, Cl, C4, and C2 by the method of Rapp and Borsos (12) were added to 6.0 ml of serum diluted serially in EDTA buffer. After 60 min of incubation at 37 C, tubes were centrifuged at 1,200 x g for 10 min, and optical density (412 nm) of the supernatant fluid was determined. Percentage of lysis was plotted against relative concentration of serum to determine the 50% end point. Statistical methods. Values were compared using Student's paired t test. RESULTS Effect of incubation time. Z was suspended at a concentration of 1 mg/ml in serum containing MgEGTA or EGTA (0.01 M) or an equivalent volume of saline. Aliquots of serum were withdrawn at 0, 15, 30, 45, and 60 min and diluted immediately in ice-cold EDTA buffer. When all samples had been obtained, C3-9 titrations were performed. Results were expressed as percentage of time 0 serum C3-9 titer under the particular chelation conditions. Figure 1 shows the results from four experiments in human serum. In nonchelated serum, essentially all C3-9 had been consumed in the reaction with Z by 15 min. C activation was slower in MgEGTA-chelated serum (03-9 at 15 min significantly higher than in nonchelated serum, P < 0.025) but was complete by 30 min. In EGTA-chelated serum, C3-9 values at 15 TABLE 1. Effect of MgEGTA on hemolytic C2 levels Chelatora Zymosanb EAc 1Eei/ml)d Saline Saline ± 0.1e Saline ± 0.1e MgEGTA ± 0.4f MgEGTA ± 0.3f MgEGTA ± 0.3f a MgEGTA (0.01 M) or an equivalent volume normal saline. btwo milligrams. c 3 x 109 cells washed twice in distilled water. d Mean ± standard error of five experiments. estatistically different from nonchelated serum without Z or EA (P < 0.05). f Not statistically different from nonchelated serum without Z or EA (P > 0.05). EGTA AND MgEGTA IN COMPLEMENT STUDIES 125 and 30 min were significantly higher than values in either MgEGTA-chelated (15 min, P < 0.05; 30 min, P < 0.01) or nonchelated (15 min, P < 0.001; 30 min, P < 0.025) serum. Although complete consumption did not seem to occur even at 60 min in EGTA-chelated serum, differences at 45 and 60 min did not reach statistical significance. Similar studies in guinea pig serum are not shown but gave similar results. Effect of Z concentration. To demonstrate differences at 60 min, varying concentrations of Z were incubated with chelated or nonchelated serum. Z was suspended in human serum at final concentrations ranging from 0.2 to 1.2 mg/ ml. After 60 min of incubation, serum was diluted in ice-cold EDTA buffer and residual C3-9 was measured (Fig. 2). Since there were no differences among the various baseline C3-9 titers in the experiments seen in Fig. 1 (i.e., chelation did not affect the baseline C3-9 titer), all values in this experiment were expressed as percentage of the C3-9 titer of nonchelated nonchallenged serum. Z consumed more C in Mg- EGTA-chelated serum than in nonchelated serum at concentrations of 0.2 (P < 0.02) and 0.4 mg/ml (P < 0.025). Less C was consumed in EGTA-chelated serum than in nonchelated serum at concentrations of 0.4 (P < 0.05) and 0.8 mg/ml (P < 0.01) and less than in MgEGTAchelated serum at concentrations of 0.2 (P < 0.025), 0.4 (P < 0.02), and 0.8 mg/ml (P < 0.02). a' f D C).' 50 n 0 JO min. FIG. 1. Sequential C3-9 measurements of human serum chelated with MgEGTA (0.01 M) or EGTA (0.01 M) or nonchelated (saline). After addition ofz (1 mg/ml) at time 0, aliquots of serum were withdrawn at indicated times and tested for residual C3-9. Results are expressed as mean + standard error of four experiments. Statistical significance is indicated for comparison with nonchelated serum by * (P < 0.01) and ** (P < 0.05), for comparison with Mg- EGTA-chelated serum by T (P < 0.01) and TT (P < 0.05).

3 126 FINE.~ Trr LLiI A zymoson (mg/ml) FIG. 2. C3-9 measurements in human serum chelated with MgEGTA (0.01 M) or EGTA (0.01 M) or nonchelated (saline). After addition of Z in concentrations ranging from 0.2 to 1.2 mg/ml, serum was incubated for 60 min, after which residual C3-9 was determined. Results are expressed as mean + standard error of four experiments. Statistical significance is indicated for comparison with nonchelated serum by * (P < 0.01) and ** (P < 0.05), for comparison with MgEGTA-chelated serum by T (P < 0.01) and TT (P < 0.05). Results with guinea pig serum (not shown) were similar. Effect of MgEGTA on C2. To determine whether chelation with MgEGTA spontaneously activated C2 or allowed Z-mediated C2 activation (3), hemolytic C2 levels in MgEGTAchelated serum (with or without Z) were compared to levels in nonchelated serum. Human serum, chelated with MgEGTA or nonchelated, with or without Z (2 mg) or EA (3 x 109 cells), was incubated at 370C for 30 min. After incubation, Z and EA were sedimented (1,200 x g) and discarded. The supernatant fluid was recalcified (CaCl2 [0.01 M]), and residual C2 was measured (Table 1). Z and EA appropriately depressed C2 levels in nonchelated serum. C2 levels in nonchelated and MgEGTA-chelated sera were identical (P > 0.05). The C2 levels in chelated serum incubated with Z and EA also were not statistically different (P > 0.05). In results not shown, EGTA chelation did not affect C2 levels. Effect of MgEGTA chelation on lysis of E- PNH. To confirm spontaneous alternative C pathway activation in MgEGTA-chelated serum (3), lysis of E-PNH in chelated serum was measured. E-PNH were washed and suspended in normal saline (109 cells/ml). Human serum was diluted 1:4 with saline and chelated with EGTA, EDTA, or MgEGTA or not chelated (saline added in an equivalent volume). E-PNH INEFECT. IMMUN. (0.3 ml) was added to 1 ml of serum and incubated at 370C for 60 min. Z (2 mg) was added to one tube containing nonchelated serum and E- PNH as a positive control for activation of the alternative pathway. After centrifugation (1,200 x g), optical density of the supernatant fluid was determined at 541 nm. Results (Table 2) demonstrate lysis of E-PNH only in Mg- EGTA-chelated serum and in nonchelated serum in which the alternative pathway had been activated by Z. DISCUSSION Activation of the classical C pathway requires both calcium and magnesium ions; the alternative pathway requires magnesium only (13). This differing requirement for calcium can be exploited experimentally using EGTA, a substance which chelates calcium avidly but magnesium weakly (2). Essentially, C consumption uninhibited by EGTA (i.e., calcium independent) is assumed to proceed through the alternative pathway. The original application of this technique to studies of the alternative pathway utilized undiluted human serum, Z in a concentration of 2 mg/ml of serum, 60 min of incubation at 370C, and measurement of whole hemolytic C (CH50) consumption as a reflection of C activation (5). Subsequent studies suggested that guinea pig C was more unstable in EGTA-chelated serum than was human C (4). Platts-Mills and Ishizaka demonstrated that the effect of EGTA on C was attributable to dissolution of Cl after chelation of calcium, the ligand which holds the Cl trimolecular complex together (11). In their studies, they used the magnesium salt of EGTA (MgEGTA). EGTA (0.01 M), although only a weak chelator of magnesium (log Ka = 5.4), will reduce the magnesium concentration of human serum from the TABLE 2. Lysis of E-PNH by chelation of human serum with MgEGTA Chelatora Lysis of E-PNH (%)b Saline 2 ± 2 Salinec d EGTA 2 2e EDTA 0 Oe MgEGTA 12 ± ld a EGTA, ethylenediaminetetraacetic acid (EDTA), or MgEGTA (0.01 M) or an equivalent volume of normal saline. b Mean ± standard error of four experiments. c Nonchelated serum challenged with zymosan (2 mg) to activate alternative pathway. d Statistically different from nonchelated serum (P < 0.05). e Not statistically different from nonchelated serum (P > 0.05).

4 VOL. 16, 1977 normal level of 1 x 10-3 M to approximately 1.6 x 10-6 M; it will reduce the calcium concentration from 3 x 10-3 M to 5 x M. In contrast, whereas calcium levels in serum chelated with MgEGTA (0.01 M) are also quite low (1.7 x 10-9 M), magnesium levels are four times normal (4 x 10-3 M) (2). It is reasonable to expect more nearly physiological function of the alternative pathway in serum with a more nearly normal concentration of magnesium, and many investigators have turned to MgEGTA in preference to EGTA. Bryan observed that MgEGTA was clearly preferable to EGTA in studying serum bactericidal activity, since the lower levels of magnesium in EGTA-chelated serum rendered some bacteria artifactually sensitive to lysozyme (1). Forsgren and Quie (6) and Des Prez et al. (3) subsequently confirmed impairrnent of a variety of alternative pathway functions by EGTA. This report provides further information regarding chelator inhibition of alternative pathway function. Depletion of C3-9 by Z was markedly slowed by EGTA (Fig. 1); and Z consumed less C in EGTA-chelated than in nonchelated serum, even when incubation was carried to 60 min (Fig. 2). More importantly, C activation by Z in EGTA-chelated serum was significantly impaired compared to activation in MgEGTAchelated serum. Snyderman and Pike (14) showed that agents (such as Z) which activate both C pathways generally require an intact classical pathway for maximum efficiency; the slower Z activation in MgEGTA-chelated serum compared to nonchelated serum (Fig. 1) very likely reflected this less efficient alternative pathway function and not any inhibition of the alternative pathway by MgEGTA. Thus, MgEGTA appeared to be a superior reagent to EGTA for alternative pathway studies. However, Des Prez et al. (3) reported the disturbing finding that MgEGTA chelation per se activated the alternative pathway, as reflected both by decreased CH50 in chelated serum and by immunochemical conversion of factor B of the alternative pathway. Furthermore, they demonstrated decreased hemolytic C2 activity in MgEGTA-chelated serum and a further decrease in C2 activity after addition of Z to chelated serum. Since C2 and alternative pathway activation both require magnesium, and since excess magnesium alone has been shown to activate the alternative pathway (9, 10), it seemed likely that the supraphysiological magnesium level of MgEGTA-chelated serum spontaneously activated both C2 and the alternative pathway. Data in Fig. 2 support the notion that Mg- EGTA activates the alternative pathway. At EGTA AND MgEGTA IN COMPLEMENT STUDIES and 0.4 mg of Z per ml, more C3-9 was consumed in MgEGTA-chelated serum than in nonchelated serum; differences were significant at the level but not at the 0.01 level. Table 2 provides stronger support. E-PNH, which have been shown to be lysed via the alternative pathway (7), underwent lysis in serum to which MgEGTA had been added. Bryant and Jenkins made this same observation some years ago (2). No evidence could be found to support the contention that MgEGTA either spontaneously activates C2 or facilitates C2 activation by Z. Hemolytic C2 levels in serum chelated with MgEGTA, with or without added Z, did not differ from C2 levels in nonchelated serum (Table 1). From the present study, one can make several conclusions regarding the use of chelators to dissect the pathways of complement activation. First, MgEGTA chelation provides a more nearly physiological milieu for alternative pathway function than does EGTA chelation, while still effectively inhibiting the classical pathway. MgEGTA has one important disadvantage, i.e., its tendency to activate spontaneously the alternative pathway. This effect may, for some purposes, be adequately counteracted by attention to a chelated serum control. In other circumstances it may be critical to avoid any spontaneous activation, in which case EGTA remains a useful chelator. Provided excess challenge material (e.g., 2 mg of Z) and prolonged incubation time (e.g., 60 min) are provided, complement activation in EGTA-chelated serum specifically reflects alternative pathway function. ACKNOWLEDGMENTS I thank Charles K. Ambrose and Roger C. Coble for technical assistance. Jack B. Alperin generously supplied PNH erythrocytes. James C. Guckian reviewed the manuscript. LITERATURE CITED 1. Bryan, C. S Sensitization ofe. coli to the serum bactericidal system and to lysozyme by ethyleneglycoltetraacetic acid. Proc. Soc. Exp. Biol. Med. 145: Bryant, R. E., and D. E. Jenkins, Jr Calcium requirements for complement dependent hemolytic reactions. J. Immunol. 101: Des Prez, R. M., C. S. Bryan, J. Hawiger, and D. G. Colley Function of the classical and alternate pathways of human complement in serum treated with ethylene glycol tetraacetic acid and MgCl2-ethylene glycol tetraacetic acid. Infect. Immun. 11: Fine, D. P Activation ofthe classic and alternate complement pathways by endotoxin. J. Immunol. 112: Fine, D. P., S. R. Marney, Jr., D. G. Colley, J. S. Sergent, and R. M. Des Prez C3 shunt activation in human serum chelated with EGTA. J. Immunol. 109:

5 128 FINE 6. Forsgren, A., and P. G. Quie Opsonic activity in human serum chelated with ethylene glycoltetraacetic acid. Immunology 26: Gotze, O., and H. J. Muller-Eberhard Paroxysmal nocturnal hemoglobinuria: hemolysis initiated by the C3 activator system. N. Engl. J. Med. 286: Kabat, E. A., and M. M. Mayer Experimental immunochemistry. Charles C Thomas, Springfield, Ill. 9. Lew, F. T., Y. Yukiyama, H. S. Waks, and A. G. Osler Activation of the alternative (properdin) pathway by divalent cations. J. Immunol. 115: May, J. E., W. Rosse, and M. M. Frank Paroxysmal nocturnal hemoglobinuria. Alternate-comple- INFECT. IMMUN. ment-pathway-mediated lysis induced by magnesium. N. Engl. J. Med. 289: Platts-Mills, T. A. E., and K. Ishizaka Activation of the alternate pathway of human complement by rabbit cells. J. Immunol. 113: Rapp, H. J., and T. Borsos Molecular basis of complement action. Appleton-Century-Crofts, New York. 13. Sandberg, A. L., and A. G. Osler Dual pathways of complement interaction with guinea pig immunoglobulins. J. Immunol. 107: Snyderman, R., and M. C. Pike Interaction of complex polysaccharides with the complement system: effect of calcium depletion on terminal component consumption. Infect. Immun. 11: Downloaded from on November 11, 2018 by guest