© 1997 American Heart Association, Inc.
Articles |
Complement C5b-9 Increases Plasminogen Binding and Activation on Human Endothelial Cells
the Department of Medicine and William K. Warren Medical Research Institute, University of Oklahoma Health Sciences Center, Oklahoma City, and the Blood Research Institute of the Blood Center of Southeastern Wisconsin (P.J.S.), Milwaukee.
Correspondence to Victoria J. Christiansen, PhD, BSEB Rm 306, University of Oklahoma Health Sciences Center, PO Box 26901, Oklahoma City, OK 73190. E-mail vicki.christiansen@cclink.net.UOKHSC.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Deposition of the terminal complement proteins (C5b-9) on human endothelial cells can result in cell lysis or nonlytic alterations of cell function including procoagulant responses. Because regulation of fibrinolysis is a central endothelial function and because C9 contains a carboxyl-terminal lysine similar to other proteins that bind and facilitate activation of plasminogen (PG), the effects of complement injury on PG binding and activation on these cells were investigated. Activation of complement through deposition of C5b67 complexes on endothelial cells resulted in a small increase (
20%) in PG binding. Incorporation of C8 into C5b-8 resulted in no further increase in binding; however, specific 125I-PG binding was increased by 100% after C5b-9 deposition. Moreover, PG was found to bind specifically to C7 and C9. The PG bound to endothelial cells after C5b-9 deposition was readily activated by tissue-type plasminogen activator (TPA). In a cell-free system, complement C9 and a synthetic peptide composed of the 20 carboxyl-terminal amino acids of C9 enhanced PG activation by TPA. Removal of the carboxyl-terminal lysine of C9 abolished the enhancement of PG activation without diminishing PG binding. We conclude that membrane C9 may comprise a binding site for PG and serve to enhance activation of this zymogen by TPA. These findings suggest that immune injury to the endothelium may enhance both the fibrin-generating and fibrinolytic capacity of the vessel wall.
Key Words: fibrinolysis • complement 9 • plasmin • von Willebrand factor • carboxypeptidase
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The zymogen form of PM, PG, is a serine protease whose major function is the degradation of fibrin. This enzyme has also been implicated in various physiological and pathological processes requiring the proteolysis of the extracellular matrix, including inflammation, wound healing, and tumor metastasis.1 2 3 4 PG is a single-chain glycoprotein containing five homologous regions known as kringles. These kringles contain binding sites for lysine, lysine analogues, and lysine-containing peptides and proteins and play key roles in the interaction of PG with cell surfaces, PG activators, PM substrates, and other components of the fibrinolytic system. PG that is bound to lysine undergoes a conformational change and is more readily activated to PM by PG activators such as TPA and urokinase.5
PG binds to cell-surface receptors on a variety of circulating blood cells, including monocytes, lymphocytes, and granulocytes, and on cultured cells, including adherent cells such as fibroblasts and ECs and nonadherent cells such as monocytoid U937 and B lymphoblastoid RPMI 1788 cells.6 7 8 These receptors are believed to be important for enhancement of PG activation, localization of active PM on cell surfaces, and protection of cell-associated PM from inhibition by 
2-antiplasmin.7 8 9
The endothelium serves as an important regulator of coagulation. ECs express thrombomodulin, which binds thrombin, changing the preferred substrate of this enzyme from fibrinogen to protein C. These cells also express binding sites for anticoagulant proteins (activated protein C and protein S)10 11 and fibrinolytic proteins (PG and TPA).8 12 Deposition of the terminal complement proteins (C5b-9) in sublytic concentrations on ECs produces procoagulant changes in these cells. These include influx of extracellular calcium, plasma membrane vesiculation with expression of binding sites for factor Va and exposure of a catalytic surface for the assembly of the prothrombinase enzyme complex, activation of secretion of the platelet adhesive protein vWF, and transient expression of P-selectin, a binding site for monocytes and neutrophils.13 14 Many disorders associated with complement activation (eg, septicemia, trauma, burns, and hemolytic transfusion reactions) may be associated with disseminated intravascular coagulation15 ; we therefore hypothesized that, in addition to the observed increase in fibrin-generating potential, the fibrinolytic capacity of vascular ECs might be enhanced by complement. The purpose of the current investigation was to evaluate the binding and activation of PG on ECs after exposure to activated complement.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Cell Culture and Preparation of Polyclonal Rabbit Anti-EC Antibody
ECs were harvested from human umbilical veins and cultured.14 First-passage cells grown in 0.32-cm2 wells in 96-well plates were used in all experiments described here.
Plasma membrane vesicles from cultured human ECs were prepared by using the method of Moldovan et al16 and used to immunize rabbits.14 IgG was purified by using caprylic acid.17
Purification of Human Complement Proteins
Human C7, C8, and C9 were purified,18 19 and human serum deficient in complement component C8 was prepared by absorption of normal serum against rabbit antibody to C8 coupled to agarose.14 C9 was polymerized by incubation for 48 to 72 hours at 37°C in the presence of zinc chloride 100 µmol/L.20 21
Exposure of ECs to Activated Complement
The buffer solution used for all EC experiments consisted of Hanks' balanced salt solution containing 10 mmol/L HEPES and 1% BSA. Cultured ECs were exposed to various stages of complement activation by sequential incubation at 37°C with antiendothelial IgG (6 mg/mL) for 20 minutes, 25% C8-deficient serum for 10 minutes, and lastly human C8 (3.2 µg/mL) and C9 (16 µg/mL) for 10 minutes. This concentration of C8 has been verified to produce EC secretion with minimal cell lysis.14
CPB Treatment of C9
Removal of the carboxyl-terminal lysine of C9 was performed by incubation of 0.2 mg/mL C9 with 200 U/mL CPB (Sigma Chemical Co) for 30 minutes at room temperature. CPB was inactivated by using carboxypeptidase inhibitor from potato tubers (Sigma) at a ratio of 1 µg inhibitor/2 µg CPB. CPB activity was confirmed by cleavage of hippuryl-L-arginine (Bachem) as detected by spectrophotometric measurement at 254 nm.22 After treatment by CPB, the lytic activity of C9 was unchanged as determined by a hemolysis assay using sheep erythrocytes. Briefly, sheep erythrocytes were washed and incubated sequentially with rabbit antibody to sheep red blood cells (Accurate Chemical and Scientific Corp), human C8–deficient serum, and human C8 (90 ng/mL) and C9 or CPB-treated C9 from 0.05 to 400 ng/mL or no C9. Hemolysis was detected by measuring absorbance at 415 nm.
Assay for vWF Secretion
After exposure of ECs for 15 minutes to histamine (Sigma), thrombin (provided by Dr David Aronson, Bureau of Biologics, Federal Drug Administration, Bethesda, Md), PMA (Sigma), A23187 (Sigma), or terminal complement proteins, supernatants were removed and assayed for released vWF by using an enzyme-linked immunosorbent assay.23 Pooled normal human plasma was used as the standard, and results were expressed in units per milliliter, where normal pooled plasma is defined as containing 1 U/mL.
PG Binding to Human ECs
PG was purified by using lysine agarose chromatography24 and was labeled with 125I by using immobilized lactoperoxidase–glucose oxidase (Enzymobeads radioiodination reagent, Bio-Rad). Protein was assayed by using the Bio-Rad protein assay with BSA as the standard (specific activity, 529 to 1770 cpm/ng). The cells were exposed to IgG, various stages of activated complement, or other agonists (as indicated in the figures). The cells were then washed and incubated with labeled PG (50 or 100 µg/mL) for 1 hour at 4°C. PG binding to ECs has been reported to saturate at 1 µmol/L (80 µg/mL),8 and plasma concentrations of PG are 120 µg/mL. Therefore, PG concentrations in the physiological range (50 or 100 µg/mL) were selected and verified in preliminary experiments to produce consistent binding results. All binding experiments were performed at 4°C to prevent the release of TPA by ECs and subsequent generation of PM and to prevent internalization of binding sites. Cells were then washed five times with chilled buffer and dissolved with 10% sodium dodecyl sulfate, and bound PG was quantified by using a gamma counter. Nonspecific binding was determined in the presence of 10 mmol/L 
-ACA (Sigma).8 25 26 Our preliminary experiments confirmed that specific binding as determined by using -ACA was qualitatively similar to that determined by using a 50-fold excess of unlabeled PG. In some experiments, the PM inhibitor aprotinin (10 U/mL) was included during PG binding to prevent generation of new PG binding sites by traces of active PM.26
PG Binding to Immobilized C9
C7, C9, CPB-treated C9, or polyC9 (10 µg/mL each) in buffer (0.06 mol/L NaHCO3 and 0.03 mol/L Na2CO3, pH 9.6) was coated on microtiter plates by incubating for 2 hours at 22°C on a rotary shaker. The plates were washed, blocked with 1% BSA for 1 hour at 22°C, and incubated with 125I-PG (25 to 275 µg/mL or 50 µg/mL for polyC9) for 2 hours at 22°C. The wells were washed five times with 0.15 mol/L NaCl, 8.8 mmol/L Na2HPO4, 1.6 mmol/L KH2PO4, and 0.05% Tween 20 and then separated, and the bound radioactivity was determined by using a gamma counter. Nonspecific binding was determined in the presence of 10 mmol/L -ACA. In some experiments, C9 was coated on microtiter plates as described above and treated with PM for 1 hour at 22°C. PM was prepared by incubating 4 nmol/L PG with 250 U/mL TPA for 15 minutes at 22°C. The plates were then washed, incubated with 125I-PG (50 µg/mL) and treated as described above.
PM Cleavage of C9
C9 was labeled with 125I by using immobilized lactoperoxidase–glucose oxidase (specific activity, 555 cpm/ng). Labeled C9 (50 µg/mL) in buffer (0.06 mol/L NaHCO3 and 0.03 mol/L Na2CO3, pH 9.6) was coated on microtiter plates by incubating overnight at 4°C on a rotary shaker. Radiolabeled C9 was deposited on ECs by exposing the ECs to activated complement as described above. Soluble 125I-C9 (16 µg/mL), immobilized 125I-C9, and 125I-C9 incorporated into C5b-9 on ECs were exposed to PM for 1 hour at 22°C. Cleavage was stopped, and immobilized protein and cells were removed from the plates by adding 10% sodium dodecyl sulfate, 10% glycerol, 0.125 mol/L Tris, and 0.005% Bromphenol blue. Cleavage of 125I-C9 was demonstrated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis under reducing conditions on a 4% stacking and 10% polyacrylamide separating gel according to the method of Laemmli, followed by autoradiography.27
PG Activation
Coupled assays were used to evaluate the initial rate of PG activation by TPA (Genentech) by monitoring the amidolytic activity of generated PM. To evaluate activation of PG in the absence of cells, PG in 0.05 mol/L Tris and 0.15 mol/L NaCl, pH 7.4, was incubated with the PM substrate S-2251(Pharmacia Hepar Inc) and either lysine, C9, polyC9, CPB-treated C9, synthetic peptide, or buffer (final concentrations are given in the figure legends). Activation was initiated by the addition of TPA (final concentration, 5 U/mL) for a final volume of 100 µL. Experiments omitting either TPA or PG showed a lack of substrate cleavage, confirming the absence of PM in any of the protein preparations. CPB and inactivated CPB had no effect in this assay. To evaluate the activation of cell-bound PG, control and complement-treated cells were incubated with 50 µg/mL PG for 1 hour at 4°C. After washing, the PM substrate S-2251 was added, and activation was initiated by the addition of TPA (final concentration, 1 U/mL). The PM hydrolysis of S-2251 at 22°C was monitored continuously at 405 nm by using a Molecular Devices kinetic microplate reader, and the Vi of S-2251 hydrolysis, as a measure of PM generation, was calculated.28
Synthesis of Peptides
The five peptides described in the Table
were synthesized by using the N-9-fluorenylmethyloxycarbonyl solid-phase synthesis strategy.29 The peptides were purified by using reversed-phase C18 high-performance liquid chromatography. The major peak was analyzed for amino acid content after hydrolysis in 6N HCl for 20 to 24 hours under vacuum. Peptide 1 is the 20–amino acid carboxyl terminus of C9. Lysines 20, 6, and 4 were replaced by arginine in peptides 2, 3, and 4, respectively. In peptide 5, the amino acids of peptide 1 were scrambled.
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Effects of Activated Complement on PG Binding to ECs
Assembly of the terminal attack complex of complement on ECs results in secretion of vWF and vesiculation of plasma membrane with expression of prothrombinase activity,13 14 potentially leading to platelet aggregation and fibrin formation. To determine whether these procoagulant responses were accompanied by any alteration in the endothelial capacity to lyse fibrin, we investigated the effects of complement on PG binding to ECs. We observed a significant (P<.01) C5b-9–dependent increase in PG binding after complement exposure (Fig 1
). A small increase in PG binding (20%) followed deposition of C5b67; there was no further increase after deposition of C5b-8 on the cell surface. However, the addition of C9 to this complex resulted in 100% increase in PG binding to the cell surface. The C5b-9–dependent increase in PG binding to ECs suggested either a direct interaction of PG with a component(s) of the C5b-9 complex or a C5b-9–induced change in EC surface properties that promoted association of PG with one or more plasma membrane constituents. One potentially important C5b-9–induced change in the EC is the influx of extracellular calcium, which is required for the stimulation of storage granule secretion and plasma membrane vesiculation. To determine the importance of an increase in cytosolic calcium in the observed increase in PG binding, we performed experiments in which the final stages of complement assembly (addition of C8 and C9) and PG binding were accomplished in buffer containing 0.1 mmol/L EGTA. To our surprise, PG binding to ECs increased approximately fivefold in the presence of EGTA, both in control and complement-treated cells (data not shown). This increase in PG binding observed in untreated cells incubated at low calcium concentrations precluded further investigation of the contribution of increased cytosolic calcium due to influx of Ca2+ across the C5b-9 pore to the generation of new PG binding sites. Stack et al,30 who have investigated the effects of divalent cations on the activation of PG by streptokinase, TPA, and urokinase (in the absence of lysine analogues or fibrin fragments), report that divalent cations inhibit activation of glu-PG. The effects of divalent cations on PG binding to cell-surface receptors and enhancement of PG activation by lysine analogues and lysine-containing peptides have not been previously reported.
|
C5b-9 is one of a limited number of agonists known to activate rapid secretion by ECs. To determine whether the increased PG binding induced by C5b-9 (Fig 1
) is specific to complement among known EC activators, the effects of four other agonists on PG binding were evaluated. Fig 2 shows that although all agonists caused a release of vWF, not all resulted in increased PG binding. Deposition of the C5b-9 complex resulted in a significant increase in PG binding (P<.01), as did histamine treatment (P<.05). Notably, after A23187 (a calcium ionophore), PG binding was unchanged.
|
Activation of PG Bound to Complement-Treated Cells
To determine whether the observed increase in binding of PG to ECs after complement deposition would be associated with an increase in cell-surface fibrinolytic capacity, we evaluated the ability of the cell-bound PG to be activated to PM by TPA. The Vi of the cleavage of S-2251 by generated PM is a measure of the initial rate of PG activation by TPA. Even though ECs produce TPA, no activation of PG in the absence of added TPA was detectable during the 1-hour incubation. In the presence of added TPA, the additional PG bound to ECs after deposition of the terminal attack complex of complement was readily activated to PM, as shown by the increased Vi observed in those cells containing C5b-9 compared with control cells (control, 1.17±0.01 and C5b-9, 1.54±0.01 pmol PM/min [mean±SE; n=3]). Of note, the enhancement of PG activation by C5b-9 treatment is rather small compared with the increase in PG binding. This discrepancy may have several explanations. The PG binding experiments were performed at 4°C, whereas the PG activation experiments were performed at room temperature. At room temperature the cells may be expected to shed vesicles containing the C5b-9 complexes13 and any PG bound to C9, thus reducing the observed PG activation. In addition, ECs secrete PG activator inhibitors (eg, PAI-1) that would reduce activation of PG by TPA but probably would not affect binding.
PG Binds to Immobilized C7, C9, and PolyC9
PG binds to proteins and peptides containing carboxyl-terminal lysines. This binding is enhanced by the presence of additional lysines at amino acid positions 12 through 17 from the carboxyl terminus.31 Because C9 has a carboxyl-terminal lysine and has lysines at amino acid positions 15 and 17 from the carboxyl terminus and since the greatest increase in PG binding occurred after the incorporation of C9 into the terminal attack complex, we hypothesized that PG bound to C9 monomer or polymer. Also, because of a report that PG binds to C732 and the small increase in PG binding observed after deposition of C5b67 on cells (Fig 1), we examined binding of PG to human C7 for comparison with binding to C9. Fig 3 illustrates dose-dependent, saturable, and specific binding of PG to C9 and C7 immobilized on microtiter plates.
|
When C9 is activated by C5b-8 on cell membranes, it undergoes a conformational change, becomes amphiphilic, and assembles into a polymer containing up to 12 to 18 C9 monomers.20 33 PolyC9 can also be formed by incubation of a C9 solution at 37°C in the presence of zinc. To determine whether PG binds to C9 in its polymerized form, as would occur in cell-deposited C5b-9, we examined binding of radiolabeled PG (50 µg/mL) to polyC9 immobilized on microtiter plates and found that PG does bind specifically to polyC9 (0.68 ng/well above binding to BSA-coated wells). This is comparable with PG binding to C9 monomer (Fig 3). Other experiments showed that binding of PG to immobilized polyC9, like binding to immobilized C7, C9, and complement-treated cells, was reduced by >90% in the presence of 10 mmol/L -ACA (data not shown). By contrast, we were unable to detect direct binding between PG and either C9 or polyC9 in solution using gel filtration (data not shown), suggesting either a lack of interaction or a low-affinity interaction that could not be detected by this method.
PG Binding to PM-Cleaved C9
Traces of PM generated during PG binding experiments have been reported to cleave substrate proteins, creating new C terminal lysines and thus new PG binding sites.26 34 The presence of the PM inhibitor aprotinin did not alter the increased PG binding after complement deposition (data not shown), demonstrating that the C5b-9–dependent increase in PG binding shown in Fig 1 was not mediated by PM activation. We next wished to determine whether C9, like fibrin and IgG, is a substrate that can be cleaved by PM to enhance subsequent PG binding. For these experiments, C9 was immobilized on microtiter plates and treated with PM; the plates were then washed and incubated with radiolabeled PG. Pretreatment of C9 with PM increased specific binding of PG (untreated C9, 8.73±1.8 and PM-treated C9, 14.89±1.0 ng bound PG; n=2), suggesting that cleavage of C9 generates new PG binding sites. However, soluble C9 proved to be a poor substrate for PM under the conditions of our experiments. When soluble C9 was exposed to PM for 24 hours at 22°C, most of the C9 monomer remained intact, and faint bands of three fragments with molecular weights ranging from 14 to 31 kD were generated. These bands may represent cleavage products of C9 monomer or some trace contaminant of the C9 preparation. By contrast, incubation of C9 immobilized on plastic with PM for 1 hour resulted in almost complete disappearance of the C9 monomer. Similar results were obtained for PM degradation of 125I-C9 deposited on the surface of ECs exposed to activated complement (data not shown). That bound C9 appears to be more readily cleaved by PM may be explained by an increase in local concentration of the substrate C9 or by a conformational change in bound C9 that increases affinity for PM or exposes a cleavage site.
C9 Enhances PG Activation
Lysine and certain lysine-containing peptides and proteins induce a conformational change in PG, thereby facilitating its activation to PM by TPA. To determine whether the cell-bound polymerized form of C9 could potentiate PG activation by TPA, we evaluated the effects of polyC9 and found that it reliably enhanced TPA activation of PG (control Vi, 0.57±0.04 and polyC9 Vi, 0.98±0.01 pmol PM/min; n=2). Purified C9 monomer, however, was more variable in its ability to enhance PG activation. With most C9 preparations, C9 markedly enhanced the activation of PG by TPA; the magnitude of this enhancement was often greater than that observed with 10 mmol/L lysine (Fig 4). By contrast, freshly prepared C9, which was active in inducing complement-mediated hemolysis or endothelial vWF release, often did not result in significant enhancement of PG activation. From previous experiments, we knew that extended storage of C9 at 4°C may result in spontaneous polymerization of C9, and we speculated that a conformational change in C9, such as that which occurs during polymerization on cell membranes, might be important for the interaction with PG that subsequently makes PG more readily activated. We learned that the activity of C9 to enhance PG activation could be augmented by prior C9 polymerization (prolonged incubation at 37oC in the presence of zinc), by brief (1-hour) incubation at 37°C, or by immobilizing C9 on plastic; these results were all consistent with the requirement for a conformational change in C9. To determine whether all the activity of C9 in our PG activation assay resulted from the adherence of C9 to the microtiter plates during the assay, we compared PG activation rates in the presence of C9 that had been preincubated in the microtiter plates followed by removal of unbound C9 by washing with C9 that had been preincubated in the microtiter plates without removal of unbound C9. The ability to enhance PG activation appears to reside entirely with the C9 adherent to the microtiter plates, as removal of unbound C9 did not diminish activation (Fig 5). In fact, removing the soluble C9 by washing resulted in more activity than when both adherent and soluble C9 were present during the assay. The data in Fig 5 raise the possibility that soluble C9 does interact with PG and competes with or inhibits PG binding to adherent C9 but that it does not induce the change in PG necessary to enhance activation.
|
|
CPB Treatment of C9 Abolishes C9 Enhancement of PG Activation
Our data demonstrating enhanced PG activation by C9 (Fig 4) are consistent with reports that proteins or peptides containing carboxyl-terminal lysines bind PG and in some cases potentiate activation of PG. To determine whether the carboxyl-terminal lysine of C9 is important for the enhanced activation of PG by C9, we evaluated the effects of removal of the carboxyl-terminal lysine of C9 by CPB. CPB treatment of C9 reduced PG activation to control levels (control Vi, 0.27±0.03; C9 Vi, 6.92±0.02; and CPB-treated C9 Vi, 0.26±0.01 pmol PM/min; n=2). These data suggest that the carboxyl-terminal lysine of C9 is necessary for the C9-induced enhancement of PG activation by TPA.
CPB Treatment of C9 Does Not Alter Binding of PG
We next determined whether the carboxyl-terminal lysine of C9 was also essential for binding of PG to immobilized C9 or complement-treated cells. In contrast to PG activation, binding to immobilized C9 was not altered by CPB treatment of C9; PG binding to CPB-treated C9 was 102%±8.6% of PG binding to C9 (n=11). These data suggest that binding between PG and C9 does not require the C-terminal lysine. In addition, CPB treatment of C9 did not alter the increased binding of PG to ECs containing C5b-9 complexes (C9, 178%±14% and CPB-treated C9, 163%±14%, mean of four experiments expressed as percent of control). It appears that the increased binding of PG to cells after deposition of C5b-9 is not mediated by the carboxyl-terminal lysine of C9 alone. Our data are most consistent with PG binding sites on other regions of C9, but these data do not exclude the possible contribution of PG binding sites on the cell surface distinct from C5b-9 or PG binding sites on C5b-8 that are not accessible until the incorporation of C9.
Peptides Mimic C9 in Enhancement of PG Activation
Since our data demonstrated that the carboxyl-terminal portion of C9 was important for the enhanced activation of PG by TPA, we synthesized the 20–amino acid C terminus of C9. This peptide enhanced TPA activation of PG (Fig 6). Replacement of the carboxyl-terminal lysine by arginine (peptide 2) diminished the enhancement of PG activation almost to control levels. Replacement of the two internal lysines by arginine (peptides 3 and 4) also reduced the enhancement, but not to control levels. The scramble peptide (peptide 5), which contained all three lysines, exhibited the same amount of enhancement of PG activation as the carboxyl-terminal peptide (peptide 1).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Deposition of the terminal complement proteins on the EC surface alters the plasma membrane, potentially leading to accelerated platelet adhesion and thrombin generation. We now demonstrate that these procoagulant responses are accompanied by an increase in binding of PG to the cell surface, probably resulting from PG binding to an epitope of C9 exposed upon incorporation into C5b-9. The bound PG can be activated to PM, suggesting that C5b-9 deposition would enhance the degradation of fibrin by these cells. Moreover, cell-bound C9 is a substrate for PM and may be cleaved to generate further PG binding sites, potentially amplifying the fibrinolytic process. Since PM has also been implicated in wound healing,3 the increase in endothelial PM may facilitate repair processes after an inflammatory response. In addition, PM in vitro is known to cleave C1s to C1s esterase35 and to cleave C5 to generate C5a and C5b activity36 37 ; that PM may indeed activate complement in vivo is suggested by several reports documenting fluid-phase complement activation in patients treated with thrombolytic agents.38 39 Thus, PM bound to endothelium could serve to amplify complement activation and consequent vascular injury.
In addition to the C5b-9–induced increase in PG binding to ECs, we have also shown that other agonists that cause EC secretion may induce a rapid increase in PG binding. Thrombin has been reported to decrease PG binding to ECs, but binding was assessed after only 18 hours of thrombin treatment.40 Of note, stimulation with PMA for 6 to 18 hours increases PG binding sites on monocytes, and rapid upregulation of PG binding sites has been observed on THP-1 monocytic cells after treatment with PMA for only 30 minutes.41 To our knowledge, we are the first to observe acute (within 15 minutes) modulation of PG binding to ECs. This increased binding is not directly linked to secretion of vWF or cell-surface expression of P-selectin, since A23187 (a calcium ionophore) did not cause any increase in PG binding. Histamine, thrombin, PMA, and complement activate protein kinase C, suggesting involvement of this intracellular mediator in triggering the expression of new PG binding sites by these agonists. Generation of new PG binding sites after exposure to activated complement required incorporation of C9 into membrane C5b-9, with little increase after C5b-8. In an attempt to better define the nature of these newly generated PG binding sites, we looked for direct interaction between C9 and PG as a possible component; we in fact discovered that PG binds directly to immobilized C9 and polyC9. -ACA competed for this binding, which suggested the participation of the lysine-binding sites of PG and possibly the carboxyl-terminal lysine of C9. The binding, however, was not significantly abrogated by the removal of the carboxyl-terminal lysine, suggesting that some other region, probably lysine containing, was involved in the binding of PG to C9. Other proteins such as laminin, fibronectin, vitronectin, and thrombospondin are reported to bind PG,42 43 44 yet these proteins do not contain carboxyl-terminal lysines. Our data suggest that PG may bind directly to C9 after its incorporation into the cell membrane, although we have not excluded other sites on the C5b-9 complex or newly exposed cell-surface sites. If membrane-bound C9 comprises a binding site for PG, the number of such sites generated during complement activation is likely much greater than demonstrated in our experiments, which examined only cell-surface binding sites. Because complement triggers vesiculation of the EC membrane with shedding of membrane fragments that contain C5b-9 complexes,13 it is probable that many C9-dependent PG binding sites are released from the cell surface and could potentially facilitate PM generation in the circulation.
Our data support a low-affinity interaction between PG and complement C9. By contrast, Reinartz et al32 report that PG binds to C7 but not C9. However, these investigators used dot-blotting to examine PG binding to complement components. We were also unable to demonstrate PG binding to C9 by using this method, presumably because of the long washing times required and probable rapid dissociation of PG-C9. It would not be unexpected that PG might bind to both C7 and C9 since they share a 23% homology in protein sequence.45 These investigators also did not test C9 in its physiological form, either on cell surfaces or polymerized in vitro. Although our results demonstrate that PG binds to a greater extent to immobilized C7 than C9, when C7 is incorporated into C5b67 complexes on cell surfaces, only a small increase in PG binding ensues (Fig 1). By contrast, after the addition of C9 to C5b-9 complexes, there was twofold increase in binding of PG on ECs (Fig 1). This is likely due to the greater number of C9 monomers (12 to 18) relative to C7 molecules (1) incorporated into each C5b-9 complex.20 33
Our work implicates C9 incorporated into membrane C5b-9 as a possible PG binding site. The binding interactions demonstrated in our experiments were between PG and conformationally altered forms of C9 (polyC9 or C9 immobilized on plastic). In addition, the ability of C9 to enhance activation of PG by TPA seemed to require, or be augmented by, a conformational change in C9 most reliably produced by heat, adherence to plastic, or polymerization. The transition of C9 from the globular, hydrophilic native monomer to an extended and amphiphilic form is essential for membrane insertion, C9 polymerization, and target lysis.46 47 48 CD59, a membrane inhibitor of complement-mediated lysis, limits incorporation of C9 into C5b-9; this inhibitor binds to adherent C9 but not native C9 monomer, suggesting that it recognizes C9 only after its transition to its amphiphilic form.49 We speculate that PG likewise recognizes a region(s) of C9, perhaps the carboxyl terminus, that is not exposed in the native C9 monomer but becomes accessible in immobilized C9, C9 polymer, or C5b-9.
The enhancement of PG activation by immobilized C9 and polyC9 is analogous to that observed when PG binds to lysine, lysine analogues, or lysine-containing proteins. The enhancement due to C9 was abolished by the removal of the carboxyl-terminal lysine by CPB, indicating that the terminal lysine is critical for enhanced activation. A synthetic peptide corresponding to the 20–amino acid carboxyl terminus of C9 also facilitated the activation of PG by TPA, although a much higher concentration was required, suggesting that the peptide does not contain all the components of C9 important for enhancement of activation. Sequential replacement of each of the three lysines in the peptide by arginine resulted in a diminished ability to enhance PG activation, with replacement of the carboxyl-terminal lysine reducing the activation to control levels. However, a scramble of the peptide (containing all three lysines) enhanced the activation of PG to the same extent as the peptide containing the native C9 carboxyl-terminal sequence. The requirements for ligand interaction with a lysine-binding site and for induction of the conformational change in PG are only partially understood.31 50 51 The five kringles of PG each have lysine-binding sites with differing affinities and fine specificities.51 In response to occupation of one or more lysine-binding sites, PG can assume at least three conformations that determine its ease of activation to PM.50 Peptide ligands have been identified which, like our scramble peptide, accelerate TPA-catalyzed PG activation but contain only closely spaced internal lysines or similar positively charged amino acids.52 Identification of the site(s) on PG with which C9 interacts will be important for understanding the mechanism by which C9 facilitates conversion of this zymogen to its active form.
Our work to date on the effects of vascular complement injury has demonstrated only procoagulant responses.13 14 However, this study demonstrates a mechanism for increased clearance of fibrin by ECs exposed in vitro to antibody-initiated complement activation. The increased binding sites for PG after an immune injury to ECs may function to balance the procoagulant effects elicited by complement deposition. This enhancement of endothelial fibrinolytic potential may explain why mild complement activation is not usually associated with overt thrombosis, whereas diseases in which marked complement consumption occurs may lead to thrombosis or to both fibrin generation and degradation, eg, disseminated intravascular coagulation.
Our data demonstrate for the first time rapid exposure of endothelial PG binding sites in response to activation of these cells. This increase in PG binding was observed after activation both by the terminal complement proteins C5b-9 and by agonists such as thrombin and histamine. Characterization of these newly exposed binding sites will be important to fully understand the endothelial regulation of fibrinolysis and its acute modulation by local mediators of inflammation.
|
Selected Abbreviations and Acronyms |
|---|
|
|
Acknowledgments |
|---|
This work was supported by National Institutes of Health grants HL48688 (Dr Hamilton) and HL36061 (Dr Sims). The peptides were synthesized by the Molecular Biology Resource of the William K. Warren Medical Research Institute, Oklahoma City, Okla, under the direction of Dr Kenneth Jackson. The authors wish to thank Janet Heuser and Darlene Schwartzott for their excellent technical assistance and Tommie Howard for her assistance with preparation of the manuscript.
Received September 13, 1995;
revision received September 19, 1996;
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Strickland S, Reich E, Sherman MI. Plasminogen activator in early embryogenesis: enzyme production by trophoblast and parietal endoderm. Cell. 1976;9:231-240.[Medline] [Order article via Infotrieve]
2. Reich E. Activation of plasminogen: a widespread mechanism for generating localized extracellular proteolysis. In: Ruddon RW, ed. Biological Markers of Neoplasia: Basic and Applied Aspects. Amsterdam, Netherlands: Elsevier; 1978:506-512.
3.
Highsmith RF. Isolation and properties of a plasminogen activator derived from canine vascular tissue. J Biol Chem. 1981;256:6788-6795.
4. Ossowski L, Vassalli JD. Plasminogen activation in normal and malignant cells: a comparison of enzyme levels and of hormonal responses. In: Ruddon RW, ed. Biological Markers of Neoplasia: Basic and Applied Aspects. Amsterdam, Netherlands: Elsevier; 1978:473-480.
5.
Mangel WF, Lin B, Ramakrishnan V. Characterization of an extremely large, ligand-induced conformational change in plasminogen. Science. 1990;248:69-73.
6. Miles LA, Plow EF. Receptor mediated binding of the fibrinolytic components, plasminogen and urokinase, to peripheral blood cells. Thromb Haemost. 1987;58:936-942.[Medline] [Order article via Infotrieve]
7.
Plow EF, Freaney DE, Plescia J, Miles LA. The plasminogen system and cell surfaces: evidence for plasminogen and urokinase receptors on the same cell type. J Cell Biol. 1986;103:2411-2420.
8.
Hajjar KA, Harpel PC, Jaffe EA, Nachman RL. Binding of plasminogen to cultured human endothelial cells. J Biol Chem. 1986;261:11656-11662.
9.
Miles LA, Plow EF. Binding and activation of plasminogen on the platelet surface. J Biol Chem. 1985;260:4303-4311.
10.
Hackeng TM, Hessing M, van't Veer C, Meijer-Huizinga F, Meijers JCM, de Groot PG, van Mourik JA, Bouma BN. Protein S binding to human endothelial cells is required for expression of cofactor activity for activated protein C. J Biol Chem. 1993;268:3993-4000.
11.
Esmon CT. The regulation of natural anticoagulant pathways. Science. 1987;235:1348-1352.
12.
Barnathan ES, Kuo A, Van der Keyl H, McCrae KR, Larsen GR, Cines DB. Tissue-type plasminogen activator binding to human endothelial cells. J Biol Chem. 1988;263:7792-7799.
13.
Hamilton KK, Hattori R, Esmon CT, Sims PJ. Complement proteins C5b-9 induce vesiculation of the endothelial plasma membrane and expose catalytic surface for assembly of the prothrombinase enzyme complex. J Biol Chem. 1990;265:3809-3814.
14.
Hattori R, Hamilton KK, McEver RP, Sims PJ. Complement proteins C5b-9 induce secretion of high molecular weight multimers of endothelial von Willebrand factor and translocation of granule membrane protein GMP-140 to the cell surface. J Biol Chem. 1989;264:9053-9060.
15. Schrier SL. Hematology, VI: disorders of hemostasis and coagulation (disseminated intravascular coagulation). In: Rubenstein E, Federman DD, eds. Scientific American Medicine. New York, NY: Scientific American, Inc; 1993:43-46.
16. Moldovan NI, Radu AN, Simionescu N. Endothelial cell plasma membrane obtained by chemically induced vesiculation. Exp Cell Res. 1987;170:499-510.[Medline] [Order article via Infotrieve]
17. Steinbuch M, Audran R. The isolation of IgG from mammalian sera with the addition of caprylic acid. Arch Biochem Biophys. 1969;135:279-284.
18. Van Der Meer BW, Fugate RD, Sims PJ. Complement proteins C5b-9 induce transbilayer migration of phospholipids. Biophys J. 1989;56:935-946.[Medline] [Order article via Infotrieve]
19.
Wiedmer T, Sims PJ. Effect of complement proteins C5b-9 on blood platelets: evidence for reversible depolarization of membrane potential. J Biol Chem. 1985;260:8014-8019.
20.
Podack ER, Tschopp J. Circular polymerization of the ninth component of complement. J Biol Chem. 1982;257:15204-15212.
21.
Tschopp J. Circular polymerization of membranolytic ninth component of complement: dependence on metal ions. J Biol Chem. 1984;259:10569-10573.
22. Fricker LD, Devi L. Comparison of a spectrophotometric, a fluorometric, and a novel radiometric assay for carboxypeptidase E (EC 3.4.17.10) and other carboxypeptidase B-like enzymes. Anal Biochem. 1990;184:21-27.[Medline] [Order article via Infotrieve]
23. Hamilton KK, Sims PJ. Changes in cytosolic Ca2+ associated with von Willebrand factor release in human endothelial cells exposed to histamine: study of microcarrier cell monolayers using the fluorescent probe indo-1. J Clin Invest. 1987;79:600-608.
24.
Deutsch DG, Mertz ET. Plasminogen purification from human plasma by affinity chromatography. Science. 1970;170:1095-1096.
25.
Gonzalez-Gronow M, Edelberg JM, Pizzo SV. Further characterization of the cellular plasminogen binding site: evidence that plasminogen 2 and lipoprotein compete for the same site. Biochemistry. 1989;28:2374-2377.[Medline]
[Order article via Infotrieve]
26.
Harpel PC, Sullivan R, Chang T. Binding and activation of plasminogen on immobilized immunoglobulin G. J Biol Chem. 1989;264:616-624.
27. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680-685.[Medline] [Order article via Infotrieve]
28. Stack S, Gonzalez-Gronow M, Pizzo SV. Regulation of plasminogen activation by components of the extracellular matrix. Biochemistry. 1990;29:4966-4970.[Medline] [Order article via Infotrieve]
29. Chang CD, Meienhofer J. Solid-phase peptide synthesis using mild base cleavage of N-alpha-fluorenylmethyloxy carbonylamino acids exemplified by a synthesis of dihydrosomatostatin. Int J Pept Protein Res. 1978;11:246-249.[Medline] [Order article via Infotrieve]
30. Stack S, Gonzalez-Gronow M, Pizzo SV. The effect of divalent cations on the conformation and function of human plasminogen. Arch Biochem Biophys. 1991;284:58-62.[Medline] [Order article via Infotrieve]
31.
Miles LA, Dahlberg CM, Plescia J, Felez J, Kato K, Plow EF. Role of cell-surface lysines in plasminogen binding to cells: identification of -enolase as a candidate plasminogen receptor. Biochemistry. 1991;30:1682-1691.[Medline]
[Order article via Infotrieve]
32. Reinartz J, Hansch GM, Kramer MD. Complement component C7 is a plasminogen-binding protein. J Immunol. 1995;154:844-850.[Abstract]
33.
Tschopp J, Engel A, Podack ER. Molecular weight of poly (C9). J Biol Chem. 1984;259:1922-1928.
34. Fleury V, Angles-Cano E. Characterization of the binding of plasminogen to fibrin surfaces: the role of carboxy-terminal lysines. Biochemistry. 1991;30:7630-7638.[Medline] [Order article via Infotrieve]
35. Ratnoff OD, Naff GB. The conversion of C'1S to C'1 estrase by plasmin and trypsin. J Exp Med. 1967;125:337-358.[Abstract]
36.
Wetsel RA, Kolb WP. Expression of C5a-like biological activities by the fifth component of human complement (C5) upon limited digestion with noncomplement enzymes without release of polypeptide fragments. J Exp Med. 1983;157:2029-2048.
37.
Arroyave CM, Muller-Eberhard HJ. Interactions between human C5, C6, and C7 and their functional significance in complement-dependent cytolysis. J Immunol. 1973;111:536-545.
38. Bennett WR, Yawn DH, Migliore PJ, Young JB, Pratt CM, Raizner AE, Roberts R, Bolli R. Activation of the complement system by recombinant tissue plasminogen activator. J Am Coll Cardiol. 1987;10:627-632.[Abstract]
39.
Frangi D, Gardinali M, Conciato L, Cafaro C, Pozzoni L, Agostoni A. Abrupt complement activation and transient neutropenia in patients with acute myocardial infarction treated with streptokinase. Circulation. 1994;89:76-80.
40.
Miles LA, Levin EG, Plescia J, Collen D, Plow EF. Plasminogen receptors, urokinase receptors, and their modulation on human endothelial cells. Blood. 1988;72:628-635.
41.
Felez J, Miles LA, Plescia J, Plow EF. Regulation of plasminogen receptor expression on human monocytes and monocytoid cell lines. J Cell Biol. 1990;111:1673-1683.
42.
DePoli P, Bacon-Baguley T, Kendra-Franczak S, Cederholm MT, Walz DA. Thrombospondin interaction with plasminogen: evidence for binding to a specific region of the kringle structure of plasminogen. Blood. 1989;73:976-982.
43. Preissner KT. Specific binding of plasminogen to vitronectin: evidence for a modulatory role of vitronectin on fibrin(ogen)-induced plasmin formation by tissue plasminogen activator. Biochem Biophys Res Commun. 1990;168:966-971.[Medline] [Order article via Infotrieve]
44.
Moser TL, Enghild JJ, Pizzo SV, Stack MS. The extracellular matrix proteins laminin and fibronectin contain binding domains for human plasminogen and tissue plasminogen activator. J Biol Chem. 1993;268:18917-18923.
45.
DiScipio RG, Chakravarti DN, Muller-Eberhard HJ, Fey GH. The structure of human complement component C7 and the C5b-7 complex. J Biol Chem. 1988;263:549-560.
46. Tschopp J, Muller-Eberhard HJ, Podack ER. Formation of transmembrane tubules by spontaneous polymerization of the hydrophilic complement protein C9. Nature. 1982;298:534-538.[Medline] [Order article via Infotrieve]
47. Laine RO, Esser AF. Detection of refolding conformers of complement protein C9 during insertion into membranes. Nature. 1989;341:63-65.[Medline] [Order article via Infotrieve]
48. DiScipio RG. The relationship between polymerization of complement component C9 and membrane channel formation. J Immunol. 1991;147:4239-4247.[Abstract]
49.
Ninomiya H, Sims PJ. The human complement regulatory protein CD59 binds to the -chain of C8 and to the `b' domain of C9. J Biol Chem. 1992;267:13675-13680.
50. Marshall JM, Brown AJ, Ponting CP. Conformational studies of human plasminogen and plasminogen fragments: evidence for a novel third conformation of plasminogen. Biochemistry. 1994;33:3599-3606.[Medline] [Order article via Infotrieve]
51. Hoover-Plow JL, Miles LA, Fless GM, Scanu AM, Plow EF. Comparison of the lysine binding functions of lipoprotein(a) and plasminogen. Biochemistry. 1993;32:13681-13687.[Medline] [Order article via Infotrieve]
52.
Stack MS, Pizzo SV. Modulation of tissue plasminogen activator-catalyzed plasminogen activation by synthetic peptides derived from the amino-terminal heparin binding domain of fibronectin. J Biol Chem. 1993;268:18924-18928.
This article has been cited by other articles:
 |
|
 |
|
  S. Ran, A. Downes, and P. E. Thorpe Increased Exposure of Anionic Phospholipids on the Surface of Tumor Blood Vessels Cancer Res., November 1, 2002; 62(21): 6132 - 6140. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Cines, E. S. Pollak, C. A. Buck, J. Loscalzo, G. A. Zimmerman, R. P. McEver, J. S. Pober, T. M. Wick, B. A. Konkle, B. S. Schwartz, et al. Endothelial Cells in Physiology and in the Pathophysiology of Vascular Disorders Blood, May 15, 1998; 91(10): 3527 - 3561. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||