Eur. J. Immunol. 1991. 21: 2887-2890
C5a and antibody in the release of heparan sulfate
2887
Short paper Jeffrey L. Plattoov, Agustin P.Dalmasso", Bonnie J. LindmanO, Nathan S. IhrckeO and Fritz H. Bath*+ Department of Pediatricso, Department of Cell and Developmental Biologyo, Department of Laboratory Medicine and PathologyAand Department of Surgery+, Immunobiology Research Center, University of Minnesota and the Veterans Administrative Medical Center, Minneapolis
The role of C5a and antibody in the release of heparan sulfate from endothelial cells* The activation of endothelial cells is thought to contribute to the host response to infection and to the pathogenesis of autoimmune disease. It was recently shown that antibody and complement can activate endothelial cells leading to cleavage and release of heparan sulfate from the cells.We show here that release of heparan sulfate from endothelial cells is mediated by antibody and the complement fragment C5a and that assembly of the membrane attack complex and lysis of endothelial cells is not necessarily involved. These data suggest that the generation of C5a in conditions such as autoimmunity and infection in which anti-endothelial cell antibodies may also be present, might amplify tissue injury by a novel mechanism involving endothelial cell activation and loss of heparan sulfate mediated by antibody and C5a.
1 Introduction Endothelial cells, when stimulated by agents such as LPS, IL 1 or other cytokines, become "activated" and as such play a role in generation of tissue injury: the activated endothelial cells promote intravascular coagulation and lose barrier functions, among other changes [l-41. We recently showed that heparan sulfate, which under normal conditions would counter these changes [5,6], is lost from endothelial cells activated by anti-endothelial cell antibodies and C [7, 81. The release of heparan sulfate was found to vary with the titer of anti-endothelial cell antibodies and to require an intact classical C pathway. For certain aspects of endothelial cell activation, the entire C cascade, including the membrane attack complex that would lead to lysis of cells, is needed [9,10]. However, the generation of xanthine oxidase by endothelial cells may be stimulated by C5a [ll]. The experiments reported below were designed to elucidate the role of C in mediating release of heparan sulfate.
C5b6 was purified from human serum as described by Yamamoto and Gewurz [12] and C7 by the method of Podak et a1 [13]. C5b6 was tested for the ability to initiate lysis of rabbit RBC and was shown to be capable of initiating lysis of porcine aortic endothelial cells as follows. Porcine endothelial cell monolayers were labeled with W r and then incubated with 20 Vgwell of C5b6 complexes and 10 pg/well of C7 for 30 min at 37 "C. The endothelial cells were washed and then incubated with 5 pg each of C8 and C9 for 4 h at 37°C. Specific release of W r was determined as previously described [7]. Where indicated, C-depleted sera were reconstituted with 75 pg/ml of C5, C6, or C7 or with 100 pg/ml of C8 or C9. Recombinant C5a was used at 4 pg/ml which was observed to be the optimum concentration in dose-response experiments (not shown).
2.2 Analysis of heparan sulfate release
Porcine aortic endothelial cells were explanted and cultured in 24-well tissue culture plates according to the method of Ryan and Maxwell [14]. Proteoglycans were 2 Materials and methods biosynthetically labeled with [35S]sulfate(100 pCi/ml = 3.7 MBq/ml) for 16 h as previo described [7]. The labeled 2.1 Sources of C-depleted serum and C components cell monolayerswere washe h culture medium and then heat-inactivated human Pooled human sera, immunodepleted of various C compo- incubated for 1h with nents, and purified C5, C6, C8 and C9 were from Quidel serum, 25% (v/v), as a source of natural anti-porcine (San Diego, CA). Recombinant C5a was €rom Sigma (St. endothelial cell antibodies [ 151. The cells were washed and then incubated for 1h in 0 pl medium containingone Louis, MO) . of various serum pools. The cell and medium fractions were then separated and the cell fraction extracted with 4 M [I 95141 guanidinium chloride with protease inhibitors for 24 h at 4"C.The fractions were then dialyzed against 0.5 M sodium * This work was supported in part by the March of Dimes and by acetate, 0.1M sodium sulfate, 0.01 M EDTA, 0.1 M the NIH (HL46810 andDK13083).Thisisreport=# 567 from the 6-aminohexanoicacid and 10 mMPMSFat 4"C.Thepercent IRC. glycosaminoglycan release was determined as: 100 [(cpm Established Investigator of the American Heart Association medium)/(cpm medium cpm cells)]. The 35S-labeled Correspondence: Jeffrey L. Platt , Immunobiology Research macromolecules released from the endothelial cells were Center, University of Minnesota, Box 724 UMHC, Minneapolis, sensitive to deaminative cleavage with nitrous acid and to digestion with heparitinase (EC 4.2.2.8) and are, therefore, h4N 55455, USA
+
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
+
0014-2980/91/1111-2887$3.50 .25/0
2888
Eur. J. Immunol. 1991. 21: 2887-2890
J. L. Platt, A. €? Dalmasso, B. J. Lindman et al.
considered heparan sulfate. Results shown are the mean of four determinations.
release is not a consequence of the action of LPS or enzymes in the serum.
The approximate size of heparan sulfate chains released from endothelial cells was determined as follows. SN 3.2 C components involved fractions were dialyzed into 6 M urea, 0.1 M NaCl, 0.05 M Tris, 0.2% CHAPS with protease inhibitors (pH 7.0), The components of C involved in the release of heparan applied to DEAE-Sephacel (Pharmacia,Uppsala, Sweden) sulfate were evaluated by testing the ability of various columnsandelutedwith0.1 M to2.0 MNaClgradients [16]. serum pools, each depleted of one of the components of C, Fractions containing heparan sulfate were chromato- to trigger heparan sulfate proteoglycan cleavage. Also graphed on a Sepharose CG6B column equilibrated and evaluated was the ability of the corresponding purified C eluted with 4.0 M guanidinium chloride, 0.05 M sodium components to reconstitute or affect the activity of the acetate, and 0.01% Chaps. The relative elution volume, depleted sera. Serum pools depleted of C6, C7,C8 or C9 or Kav, is indicated. Samples were also studied by electro- caused release of heparan sulfate to the same extent as did phoresis in 0.6% agarose-1.2% polyacrylamide gels from normal human serum (Fig. 1). Reconstitution of the which fluorograms were prepared as previously described depleted sera with purified C components did not signifi[16]. Gel filtration and gel electrophoresis standards cantly increase the ability of the sera to mediate heparan included heparan sulfate proteoglycan and heparan sulfate sulfate release. These results indicated that neither the activation of terminal C components nor the lysis of glycosaminoglycan chains (not shown). endothelial cells were necessary.
3 Results 3.1 Release of heparan sulfate by the action of antibody and C
As shown in Fig. 1, if a human serum or a pool of sera containing C and natural antibodies directed at porcine endothelial cells is applied to porcine endothelial cells that had been labeled with [35S]sulfate,22% to 65% of labeled heparan sulfate is released.Thisresult is consistent with our prior observations [7]. Heparan sulfate was not released from endothelial cells incubated with serum in which C was inactivatedby heating (Fig. 1) or from cells incubated in medium to which LPS was added (not shown). Among individual sera tested, heparan sulfate release varies with the concentration of anti-endothelial cell antibodies [8].These findings together with the results reported herein would indicate that the activation of endothelial cells leading to heparan sulfate
In contrast, incubation of labeled endothelial cells in C5-depleted serum did not cause significant loss of heparan sulfate from the cells (Fig. 1). The ability of the depleted serum to mediate heparan sulfate release was restored by the addition of purified C5 to the serum. Thus, the active moiety of C which triggered heparan sulfate release appeared to derive from C5. Since purified C5 with antibody did not cause heparan sulfate release, the active moiety was C5a or C5b. The relative contributionsof C5a, C5b and antibody to the cleavage of heparan sulfate were tested as shown in Fig. 2. The addition of C5a to cultured endothelial cells to which natural antibody was bound caused significant release of heparan sulfate. In contrast, the treatment of endothelial cells with C5a alone (without antibody) or with heatinactivated serum, which contains natural antibodies but not active C, did not cause substantial heparan sulfate release. Further, the application to antibody-treated endothelial cells of C5b6, C5b67, or C5b-9 did not mediate
Serum Cdepleted
25
80
1
60 cn 0
-
240 0
K
20
HI
0
S H I HI+ c5
C5
C6
C7
C8
C9
Figurel. The involvement of C components in the release of heparan sulfate from porcine aortic endothelial cell monolayers. Heparan sulfate is released from cultured endothelial cells by whole serum (S) and by serum depleted of C6,C7, C8 or C9 but not by beat-inactivated serum (HI), serum depleted of C5 or purified C5 plus HI. Sera depleted of C5, C6 or C7 were reconstituted with 75 pg/ml and sera depleted of C8 or C9 with 100 p g / d of the corresponding C components where indicated by the solid bars.
CSa
C5a +Ab
CSb6 C S b - 7 C S b - 9 +Ab +Ab +Ab
Figure 2. The role of fragments of C5 in the cleavage and release of heparan sulfate from cultured porcine aortic endothelial cells.The incubation of endothelial cells with human heat-inactivated serum (HI) as a source of natural antibodies (Ab) and then with recombinant C5a (4 pg/ml in C5-deficient serum) triggers release of heparan sulfate. The incubation of endothelial cells with heatinactivated serum or recombinant C5a alone or with heatinactivated serum followed by C5b6, C5b-7 or C5b-9 does not cause substantial heparan sulfate release.
Eur. J. Immunol. 1991. 21: 2887-2890
release of heparan sulfate. However, without antibody C5b-9 was able to mediate endothelial cell lysis (data not shown). Thus, it is the combination of C5a and natural antibody acting on cultured endothelial cells which triggers the very significant loss of heparan sulfate. 3.3 The mechanism of heparan sulfate release
In prior studies we had demonstrated that C-mediated cleavage of heparan sulfate proteoglycan involved the
-
HS Released
By C-Deplotd Sorum
' . r
1500
g
1
C6Dapl.t.d C7D.pl.t.d
1000
0
500 0
-0.2 0.0 0.2 0.4 0 . 6 0.8 1.0 Kav
' u
a a
HSPG Isolrtod from EC HS ReIerrd by N o m l SoNm
1000
0 500
0
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 Kav
C5a and antibody in the release of heparan sulfate
2889
generation of glycosaminoglycan fragments or glycopeptides [7].Gel filtration chromatography and gel electrophoresis of the labeled moieties released from endothelial cells by C6-deficient serum or by antibody plus C5a revealed moieties of the same size as those released by fresh human serum (Fig. 3). This result suggested that the same mechanism of cleavage had occurred when C activation did not proceed beyond C5.
4 Discussion Heparan sulfate is a biologically active, anionic saccharide which in normal blood vessels promotes the integrity of endothelium [6], inhibits thrombus formation [5, 171, tethers superoxide dismutase to the cell [18], and attaches the endotheliumto the underlying extracellular matrix.The release of endothelial cell-associated heparan sulfate mediated by the combined effects of C5a and antibody would have important pathogenetic implications. For example, a decrease in endothelial cell-associatedsuperoxide dismutase caused by the cleavage of its heparan sulfate anchor in concert with the activation of xanthine oxidase by C5a [111 would likely amplify oxidant-mediated tissue injury. The requirement for both anti-endothelial cell antibody and C5a contrast with other manifestations of endothelial cell activation which appear to be generated by only one type of signal [l-4,111. Our results suggest that the release of heparan sulfate from endothelial cells may represent a unique type of endothelialcell activation which results from delivery of two or more discrete signals. Although the experimental model we used was intended to test the situation in which naturally occurring antibody and C mediate endothelial cell activation as it might occur in hyperacute xenograft rejection [19],we believe our findings are more generally applicable.The tissue injury associated with infection, autoimmune vasculitis and allograft rejection have been attributed to endothelial cell activation caused by toxic agents or cytokines [20-22].We suggest that in some instances [23,24] the pathology observed in these conditions might also result from the combined action of antibodies and C activation leading to the generation of C5a. Received April 15, 1991; in revised form July 19, 1991.
5 References Figure 3. Characterization by gel filtration chromatography and gel electrophoresis of the cleavage products of heparan sulfate proteoglycan (HSPG) released from porcine endothelial cells by the action of human serum as a source of anti-endothelial cell antibody and C. Antibody plus C5a, sera depleted of C7 or of C6 and C5 depleted serum (C5D) reconstituted with C5 released from endothelial cells %-labeled macromolecules which were significantly smaller than the intact proteoglycans. Shown above are Sepharose CL-6B chromatograms and below autoradiograms prepared after agarose-polyacrylamide gel electrophoresis. The Kav and electrophofetic migration of labeled moieties released by C5D plus C5 or by antibody plus C5a are the same as those of %-labeled kacromolehles released by whole serum (not shown) [7].
1 Nawroth, I? €? and Stern, D. M., J. Exp. Med. 1986. 163: 740. 2 Schleef, R. R., Bevilacqua, M. P,Sawdey, M., Gimbrone, M. A. Jr. and Loskutoff, D. J., J. Biol. Chern. 1988. 263: 5797. 3 Brett, J., Gerlach, H., Nawroth, €?, Steinberg, S., Godman, G. and Stem, D., J. Exp. Med. 1989.169: 1977. 4 Bevilacqua, M. R,Pober, J. S., Mendrick, D. L., Cotran, R. S. and Gimbrone, M. A. Jr., Proc. Natl. Acad. Sci. USA 1987.84: 9238. 5 Marcum, J. A., Reilly, C. E and Rosenberg, R. D., in Wight,T. N. and Mecham, R. €? (Eds.), Biology of proteoglycans, Academic Press, Orlando 1987, p. 301. 6 Matzner, Y., Bar-Ner, M., Yahalom, J., Ishai-Michaeli, R., Fucks. Z. and Vladavskv. I.., J. Clin. Invest. 1985. 76: 1306.
2890
J. L. Platt, A. P. Dalmasso, B. J. Lindman et al.
7 Platt, J. L.,Vercellotti, G. M., Lindman, B., Oegema,T. R., Jr., Bach, F. H. and Dalmasso, A. l?, J. Exp. Med. 1990. 171: 1363. 8 Platt, J. L., Lindman, B. J., Geller, R. L., Noreen, H., Swanson, J., Dalmasso, A. P. and Bach, E H., Transplantation 1991, in press. 9 Suttorp, N., Seeger, W., Zinsky, S. and Bhakdi, S., Am. J. Physiol. 1987.253: C13. 10 Hamilton, K. K., Hattori, R., Esmon, C. T. and Sims, l? J., J. Biol. Chem. 1990. 265: 3809. 11 Friedl,H.F!,Till, G. O., Ryan,U.S. andWard,l? A., FASEBJ. 1989. 3: 2512. 12 Yamamoto, K. and Gewurz, H., J. Zmmunol. 1978. 120: 2008. 13 Podack, E. R., Kolb, W. P. and Miiller-Eberhard, H. J., J. Immunol. 1976. 116: 263. 14 Ryan, U. S. and Maxwell, G., J. Tim, Cult. 1986. 10: 3. 15 Platt, J. L., Turman, M. A., Noreen, H. J., Fischel, R. J., Bolman, R. M. and Bach, F. H., Transplantation 1990. 49: 1OOO.
Eur. J. Immunol. 1991.21: 2887-2890 16 Platt, J. L., Brown, D. M., Granlund, K., Oegema,T. R. and Klein, D. J., Devel. Biol. 1987. 123: 293. 17 Marcum, J. A., Atha, D. H., Fritze, L. M. S., Nawroth, E , Stern, D. and Rosenberg, R. D., J. Biol. Chem. 1986. 261: 7507. 18 Karlsson, K. and Marklund, S. L., Biochem. J. 1987. 242: 55. 19 Platt, J. L. ,Vercellotti, G. M., Dalmasso, A. P., Matas, A. J., Bolman, R. M., Najarian, J. S. and Bach, F. H., Zmrnunol. Today 1990.11: 450. 20 Stem, D., Nawroth, I?, Handley, D. and Kisiel,W., Proc. Natl. Acad. Sci. USA 1985. 82: 2523. 21 Pober, J. S., Am. J. Pathol. 1988. 133: 426. 22 Brasile, L., Kremer, J. M., Clarke, J. L. and Cerilli, J., Am. J. Med. 1989. 87: 74. 23 Fattorossi, A., Aurbach, G. D., Sakaguchi, K., Cama, A., Marx, S. J., Streeten, E. A., Fitzpatnck, L. A. and Brandi, M. L., Proc. Natl. Acad. Sci. USA 1988. 85: 4015. 24 Leung, D.Y., Moake, J. L., Havens, P. L., Kim, M. and Pober, J. S., Lancet 1988. ii: 183.
C5a and antibody in the release of heparan sulfate
2887
Short paper Jeffrey L. Plattoov, Agustin P.Dalmasso", Bonnie J. LindmanO, Nathan S. IhrckeO and Fritz H. Bath*+ Department of Pediatricso, Department of Cell and Developmental Biologyo, Department of Laboratory Medicine and PathologyAand Department of Surgery+, Immunobiology Research Center, University of Minnesota and the Veterans Administrative Medical Center, Minneapolis
The role of C5a and antibody in the release of heparan sulfate from endothelial cells* The activation of endothelial cells is thought to contribute to the host response to infection and to the pathogenesis of autoimmune disease. It was recently shown that antibody and complement can activate endothelial cells leading to cleavage and release of heparan sulfate from the cells.We show here that release of heparan sulfate from endothelial cells is mediated by antibody and the complement fragment C5a and that assembly of the membrane attack complex and lysis of endothelial cells is not necessarily involved. These data suggest that the generation of C5a in conditions such as autoimmunity and infection in which anti-endothelial cell antibodies may also be present, might amplify tissue injury by a novel mechanism involving endothelial cell activation and loss of heparan sulfate mediated by antibody and C5a.
1 Introduction Endothelial cells, when stimulated by agents such as LPS, IL 1 or other cytokines, become "activated" and as such play a role in generation of tissue injury: the activated endothelial cells promote intravascular coagulation and lose barrier functions, among other changes [l-41. We recently showed that heparan sulfate, which under normal conditions would counter these changes [5,6], is lost from endothelial cells activated by anti-endothelial cell antibodies and C [7, 81. The release of heparan sulfate was found to vary with the titer of anti-endothelial cell antibodies and to require an intact classical C pathway. For certain aspects of endothelial cell activation, the entire C cascade, including the membrane attack complex that would lead to lysis of cells, is needed [9,10]. However, the generation of xanthine oxidase by endothelial cells may be stimulated by C5a [ll]. The experiments reported below were designed to elucidate the role of C in mediating release of heparan sulfate.
C5b6 was purified from human serum as described by Yamamoto and Gewurz [12] and C7 by the method of Podak et a1 [13]. C5b6 was tested for the ability to initiate lysis of rabbit RBC and was shown to be capable of initiating lysis of porcine aortic endothelial cells as follows. Porcine endothelial cell monolayers were labeled with W r and then incubated with 20 Vgwell of C5b6 complexes and 10 pg/well of C7 for 30 min at 37 "C. The endothelial cells were washed and then incubated with 5 pg each of C8 and C9 for 4 h at 37°C. Specific release of W r was determined as previously described [7]. Where indicated, C-depleted sera were reconstituted with 75 pg/ml of C5, C6, or C7 or with 100 pg/ml of C8 or C9. Recombinant C5a was used at 4 pg/ml which was observed to be the optimum concentration in dose-response experiments (not shown).
2.2 Analysis of heparan sulfate release
Porcine aortic endothelial cells were explanted and cultured in 24-well tissue culture plates according to the method of Ryan and Maxwell [14]. Proteoglycans were 2 Materials and methods biosynthetically labeled with [35S]sulfate(100 pCi/ml = 3.7 MBq/ml) for 16 h as previo described [7]. The labeled 2.1 Sources of C-depleted serum and C components cell monolayerswere washe h culture medium and then heat-inactivated human Pooled human sera, immunodepleted of various C compo- incubated for 1h with nents, and purified C5, C6, C8 and C9 were from Quidel serum, 25% (v/v), as a source of natural anti-porcine (San Diego, CA). Recombinant C5a was €rom Sigma (St. endothelial cell antibodies [ 151. The cells were washed and then incubated for 1h in 0 pl medium containingone Louis, MO) . of various serum pools. The cell and medium fractions were then separated and the cell fraction extracted with 4 M [I 95141 guanidinium chloride with protease inhibitors for 24 h at 4"C.The fractions were then dialyzed against 0.5 M sodium * This work was supported in part by the March of Dimes and by acetate, 0.1M sodium sulfate, 0.01 M EDTA, 0.1 M the NIH (HL46810 andDK13083).Thisisreport=# 567 from the 6-aminohexanoicacid and 10 mMPMSFat 4"C.Thepercent IRC. glycosaminoglycan release was determined as: 100 [(cpm Established Investigator of the American Heart Association medium)/(cpm medium cpm cells)]. The 35S-labeled Correspondence: Jeffrey L. Platt , Immunobiology Research macromolecules released from the endothelial cells were Center, University of Minnesota, Box 724 UMHC, Minneapolis, sensitive to deaminative cleavage with nitrous acid and to digestion with heparitinase (EC 4.2.2.8) and are, therefore, h4N 55455, USA
+
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
+
0014-2980/91/1111-2887$3.50 .25/0
2888
Eur. J. Immunol. 1991. 21: 2887-2890
J. L. Platt, A. €? Dalmasso, B. J. Lindman et al.
considered heparan sulfate. Results shown are the mean of four determinations.
release is not a consequence of the action of LPS or enzymes in the serum.
The approximate size of heparan sulfate chains released from endothelial cells was determined as follows. SN 3.2 C components involved fractions were dialyzed into 6 M urea, 0.1 M NaCl, 0.05 M Tris, 0.2% CHAPS with protease inhibitors (pH 7.0), The components of C involved in the release of heparan applied to DEAE-Sephacel (Pharmacia,Uppsala, Sweden) sulfate were evaluated by testing the ability of various columnsandelutedwith0.1 M to2.0 MNaClgradients [16]. serum pools, each depleted of one of the components of C, Fractions containing heparan sulfate were chromato- to trigger heparan sulfate proteoglycan cleavage. Also graphed on a Sepharose CG6B column equilibrated and evaluated was the ability of the corresponding purified C eluted with 4.0 M guanidinium chloride, 0.05 M sodium components to reconstitute or affect the activity of the acetate, and 0.01% Chaps. The relative elution volume, depleted sera. Serum pools depleted of C6, C7,C8 or C9 or Kav, is indicated. Samples were also studied by electro- caused release of heparan sulfate to the same extent as did phoresis in 0.6% agarose-1.2% polyacrylamide gels from normal human serum (Fig. 1). Reconstitution of the which fluorograms were prepared as previously described depleted sera with purified C components did not signifi[16]. Gel filtration and gel electrophoresis standards cantly increase the ability of the sera to mediate heparan included heparan sulfate proteoglycan and heparan sulfate sulfate release. These results indicated that neither the activation of terminal C components nor the lysis of glycosaminoglycan chains (not shown). endothelial cells were necessary.
3 Results 3.1 Release of heparan sulfate by the action of antibody and C
As shown in Fig. 1, if a human serum or a pool of sera containing C and natural antibodies directed at porcine endothelial cells is applied to porcine endothelial cells that had been labeled with [35S]sulfate,22% to 65% of labeled heparan sulfate is released.Thisresult is consistent with our prior observations [7]. Heparan sulfate was not released from endothelial cells incubated with serum in which C was inactivatedby heating (Fig. 1) or from cells incubated in medium to which LPS was added (not shown). Among individual sera tested, heparan sulfate release varies with the concentration of anti-endothelial cell antibodies [8].These findings together with the results reported herein would indicate that the activation of endothelial cells leading to heparan sulfate
In contrast, incubation of labeled endothelial cells in C5-depleted serum did not cause significant loss of heparan sulfate from the cells (Fig. 1). The ability of the depleted serum to mediate heparan sulfate release was restored by the addition of purified C5 to the serum. Thus, the active moiety of C which triggered heparan sulfate release appeared to derive from C5. Since purified C5 with antibody did not cause heparan sulfate release, the active moiety was C5a or C5b. The relative contributionsof C5a, C5b and antibody to the cleavage of heparan sulfate were tested as shown in Fig. 2. The addition of C5a to cultured endothelial cells to which natural antibody was bound caused significant release of heparan sulfate. In contrast, the treatment of endothelial cells with C5a alone (without antibody) or with heatinactivated serum, which contains natural antibodies but not active C, did not cause substantial heparan sulfate release. Further, the application to antibody-treated endothelial cells of C5b6, C5b67, or C5b-9 did not mediate
Serum Cdepleted
25
80
1
60 cn 0
-
240 0
K
20
HI
0
S H I HI+ c5
C5
C6
C7
C8
C9
Figurel. The involvement of C components in the release of heparan sulfate from porcine aortic endothelial cell monolayers. Heparan sulfate is released from cultured endothelial cells by whole serum (S) and by serum depleted of C6,C7, C8 or C9 but not by beat-inactivated serum (HI), serum depleted of C5 or purified C5 plus HI. Sera depleted of C5, C6 or C7 were reconstituted with 75 pg/ml and sera depleted of C8 or C9 with 100 p g / d of the corresponding C components where indicated by the solid bars.
CSa
C5a +Ab
CSb6 C S b - 7 C S b - 9 +Ab +Ab +Ab
Figure 2. The role of fragments of C5 in the cleavage and release of heparan sulfate from cultured porcine aortic endothelial cells.The incubation of endothelial cells with human heat-inactivated serum (HI) as a source of natural antibodies (Ab) and then with recombinant C5a (4 pg/ml in C5-deficient serum) triggers release of heparan sulfate. The incubation of endothelial cells with heatinactivated serum or recombinant C5a alone or with heatinactivated serum followed by C5b6, C5b-7 or C5b-9 does not cause substantial heparan sulfate release.
Eur. J. Immunol. 1991. 21: 2887-2890
release of heparan sulfate. However, without antibody C5b-9 was able to mediate endothelial cell lysis (data not shown). Thus, it is the combination of C5a and natural antibody acting on cultured endothelial cells which triggers the very significant loss of heparan sulfate. 3.3 The mechanism of heparan sulfate release
In prior studies we had demonstrated that C-mediated cleavage of heparan sulfate proteoglycan involved the
-
HS Released
By C-Deplotd Sorum
' . r
1500
g
1
C6Dapl.t.d C7D.pl.t.d
1000
0
500 0
-0.2 0.0 0.2 0.4 0 . 6 0.8 1.0 Kav
' u
a a
HSPG Isolrtod from EC HS ReIerrd by N o m l SoNm
1000
0 500
0
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 Kav
C5a and antibody in the release of heparan sulfate
2889
generation of glycosaminoglycan fragments or glycopeptides [7].Gel filtration chromatography and gel electrophoresis of the labeled moieties released from endothelial cells by C6-deficient serum or by antibody plus C5a revealed moieties of the same size as those released by fresh human serum (Fig. 3). This result suggested that the same mechanism of cleavage had occurred when C activation did not proceed beyond C5.
4 Discussion Heparan sulfate is a biologically active, anionic saccharide which in normal blood vessels promotes the integrity of endothelium [6], inhibits thrombus formation [5, 171, tethers superoxide dismutase to the cell [18], and attaches the endotheliumto the underlying extracellular matrix.The release of endothelial cell-associated heparan sulfate mediated by the combined effects of C5a and antibody would have important pathogenetic implications. For example, a decrease in endothelial cell-associatedsuperoxide dismutase caused by the cleavage of its heparan sulfate anchor in concert with the activation of xanthine oxidase by C5a [111 would likely amplify oxidant-mediated tissue injury. The requirement for both anti-endothelial cell antibody and C5a contrast with other manifestations of endothelial cell activation which appear to be generated by only one type of signal [l-4,111. Our results suggest that the release of heparan sulfate from endothelial cells may represent a unique type of endothelialcell activation which results from delivery of two or more discrete signals. Although the experimental model we used was intended to test the situation in which naturally occurring antibody and C mediate endothelial cell activation as it might occur in hyperacute xenograft rejection [19],we believe our findings are more generally applicable.The tissue injury associated with infection, autoimmune vasculitis and allograft rejection have been attributed to endothelial cell activation caused by toxic agents or cytokines [20-22].We suggest that in some instances [23,24] the pathology observed in these conditions might also result from the combined action of antibodies and C activation leading to the generation of C5a. Received April 15, 1991; in revised form July 19, 1991.
5 References Figure 3. Characterization by gel filtration chromatography and gel electrophoresis of the cleavage products of heparan sulfate proteoglycan (HSPG) released from porcine endothelial cells by the action of human serum as a source of anti-endothelial cell antibody and C. Antibody plus C5a, sera depleted of C7 or of C6 and C5 depleted serum (C5D) reconstituted with C5 released from endothelial cells %-labeled macromolecules which were significantly smaller than the intact proteoglycans. Shown above are Sepharose CL-6B chromatograms and below autoradiograms prepared after agarose-polyacrylamide gel electrophoresis. The Kav and electrophofetic migration of labeled moieties released by C5D plus C5 or by antibody plus C5a are the same as those of %-labeled kacromolehles released by whole serum (not shown) [7].
1 Nawroth, I? €? and Stern, D. M., J. Exp. Med. 1986. 163: 740. 2 Schleef, R. R., Bevilacqua, M. P,Sawdey, M., Gimbrone, M. A. Jr. and Loskutoff, D. J., J. Biol. Chern. 1988. 263: 5797. 3 Brett, J., Gerlach, H., Nawroth, €?, Steinberg, S., Godman, G. and Stem, D., J. Exp. Med. 1989.169: 1977. 4 Bevilacqua, M. R,Pober, J. S., Mendrick, D. L., Cotran, R. S. and Gimbrone, M. A. Jr., Proc. Natl. Acad. Sci. USA 1987.84: 9238. 5 Marcum, J. A., Reilly, C. E and Rosenberg, R. D., in Wight,T. N. and Mecham, R. €? (Eds.), Biology of proteoglycans, Academic Press, Orlando 1987, p. 301. 6 Matzner, Y., Bar-Ner, M., Yahalom, J., Ishai-Michaeli, R., Fucks. Z. and Vladavskv. I.., J. Clin. Invest. 1985. 76: 1306.
2890
J. L. Platt, A. P. Dalmasso, B. J. Lindman et al.
7 Platt, J. L.,Vercellotti, G. M., Lindman, B., Oegema,T. R., Jr., Bach, F. H. and Dalmasso, A. l?, J. Exp. Med. 1990. 171: 1363. 8 Platt, J. L., Lindman, B. J., Geller, R. L., Noreen, H., Swanson, J., Dalmasso, A. P. and Bach, E H., Transplantation 1991, in press. 9 Suttorp, N., Seeger, W., Zinsky, S. and Bhakdi, S., Am. J. Physiol. 1987.253: C13. 10 Hamilton, K. K., Hattori, R., Esmon, C. T. and Sims, l? J., J. Biol. Chem. 1990. 265: 3809. 11 Friedl,H.F!,Till, G. O., Ryan,U.S. andWard,l? A., FASEBJ. 1989. 3: 2512. 12 Yamamoto, K. and Gewurz, H., J. Zmmunol. 1978. 120: 2008. 13 Podack, E. R., Kolb, W. P. and Miiller-Eberhard, H. J., J. Immunol. 1976. 116: 263. 14 Ryan, U. S. and Maxwell, G., J. Tim, Cult. 1986. 10: 3. 15 Platt, J. L., Turman, M. A., Noreen, H. J., Fischel, R. J., Bolman, R. M. and Bach, F. H., Transplantation 1990. 49: 1OOO.
Eur. J. Immunol. 1991.21: 2887-2890 16 Platt, J. L., Brown, D. M., Granlund, K., Oegema,T. R. and Klein, D. J., Devel. Biol. 1987. 123: 293. 17 Marcum, J. A., Atha, D. H., Fritze, L. M. S., Nawroth, E , Stern, D. and Rosenberg, R. D., J. Biol. Chem. 1986. 261: 7507. 18 Karlsson, K. and Marklund, S. L., Biochem. J. 1987. 242: 55. 19 Platt, J. L. ,Vercellotti, G. M., Dalmasso, A. P., Matas, A. J., Bolman, R. M., Najarian, J. S. and Bach, F. H., Zmrnunol. Today 1990.11: 450. 20 Stem, D., Nawroth, I?, Handley, D. and Kisiel,W., Proc. Natl. Acad. Sci. USA 1985. 82: 2523. 21 Pober, J. S., Am. J. Pathol. 1988. 133: 426. 22 Brasile, L., Kremer, J. M., Clarke, J. L. and Cerilli, J., Am. J. Med. 1989. 87: 74. 23 Fattorossi, A., Aurbach, G. D., Sakaguchi, K., Cama, A., Marx, S. J., Streeten, E. A., Fitzpatnck, L. A. and Brandi, M. L., Proc. Natl. Acad. Sci. USA 1988. 85: 4015. 24 Leung, D.Y., Moake, J. L., Havens, P. L., Kim, M. and Pober, J. S., Lancet 1988. ii: 183.
