British Journal of Haematology. 1990, 76, 380-386
Studies of natural anticoagulant proteins and anticardiolipin antibodies in patients with the lupus anticoagulant SAMUELC. L. Lo, HATEMH. SALEM,MARGARET A. HOWARD,MICHAELJ. OLDMEADOW AND BARRYG . FIRKINDepartment of Medicine, Monash Medical School, Alfred Hospital. Victoria, Australia
Received 5 March 1990; accepted for publication 18 June 1990
Summary. Components of the natural anticoagulant system (NAS) and anticardiolipin antibodies were examined in 2 1 patients with lupus anticoagulant (LA), 1 3 of whom had past histories of thrombotic episodes. No relationship could be shown between the antigenic levels of protein C and S (PC, PS) and a history of thrombosis. Inhibition of the anticoagulant activity of activated protein C (APC) was observed using plasma from 20/2 1patients when phospholipid vesicles were used as the surface for the coagulation reaction. This effect was not affected by the addition of PS. When platelet membranes were employed only 2/2 1 patients demonstrated inhibition of APC. Under the latter condition, PS functional activity was inhibited in 7/21 patients, six of whom had a past history of thrombosis.
Reduced antithrombin I11 or heparin cofactor I1 levels were observed in a total of 4/2 1 patients and may have contributed to the development of thrombosis in three of these patients. Antibodies specifically directed against these proteins were not detected suggesting the possibility of an associated constitutional deficiency. Anticardiolipin antibodies, though elevated in 17/2 1 patients, did not serve as a useful marker for an increased risk of thrombosis, and the level did not correlate with inhibition of the activity of APC or PS. We conclude that the mechanism of thrombosis in patients with LA is multi-factorial. A subset of patients in whom LA specifically inhibits PS function may represent patients who are at significant risk from thrombosis.
Lupus anticoagulant (LA) is an immunoglobulin which results in prolongation of in vitro clotting times. Current evidence suggests that this anticoagulant activity is mediated by an effect on phospholipids (Thiagarajan et a!, 1980).This is supported by the frequent demonstration of antiphospholipid and anticardiolipin antibodies (ACA) in the plasma and/ or sera of patients with LA (Harris et a/, 1983, 1984: Petri et a/, 1987). Furthermore, coagulation tests used to detect LA are mostly phospholipid dependent. The association between LA and a predisposition to thrombosis in patients with or without systemic lupus erythematosus (SLE) has been widely reported (Bowie et al, 1963;Johansson & Lassus, 1974; Lechner, 1974; Gastineau et al, 1985; Petri et al, 1987).The pathogenesis of thrombosis in these patients is poorly understood with several mechanisms proposed to explain this association (Carreras et al, 1981; Cosgriff & Martin, 1981; Sanfelippo & Drayna. 1982; Freyssinet et al, 1986; Friedman et a/, 1986; Tobelem & Cariou. 1986). Most studies reported in the literature have been on small numbers of patients and it is not clear whether
the abnormalities occur selectively in patients who experienced a thrombotic event or are a feature of all patients with LA. In this study we investigated 21 patients with LA. Patients were grouped for the presence or absence of a history of thrombosis. Plasma from these patients was examined for its effect on the anticoagulant potential of activated protein C (APC) and protein S (PS).In addition, we measured the levels of four components of the natural anticoagulant system (NAS) namely protein C (PC), PS, antithrombin-111 (AT-111) and heparin cofactor I1 (HC-11).Qualitative assays on PC. PS and AT-I11 using two-dimensional crossed immunoelectrophoresis (CIEP) were also performed. Our results suggest that the activity of APC is inhibited by plasma from the majority of patients with LA, and bears no relation to a history of thrombosis. On the other hand, a strong correlation between thrombosis and inhibition of PS function was noted.
Correspondence: Dr Samuel C. L. LO, Department of Medicine, Monash Medical School, Alfred Hospital, Commercial Road, Prahran, VIC 3181, Australia.
Platelin (rabbit brain phospholipid) and simplastin (rabbit brain and lung thromboplastin) (General Diagnostics, Organon Teknika Co., Durham, U.S.A.), activated thrombofax
MATERIALS A N D METHODS Materials
3 80
Natural Anticoagulants and the Lupus Inhibitor (bovine brain phospholipid) (Ortho Diagnostics Systems, Raritan, U.S.A.), Russell viper venom (Wellcome Diagnostics, Sydney, Australia), IL5I-Na (Amersham, Sydney), iodogen and sodium dodecyl sulphate (Pierce Chemical Co., Illinois, U.S.A.), glycine, sodium chloride and EDTA (Ajax Chemical Co., Sydney), the chromogenic substrate S2238 (KabiVitrum, Stockholm, Sweden), Th-1 (Nycomed AS, Oslo, Norway), bovine thrombin (Parke-Davis Pty Ltd, Sydney), Sephadex G-150, Sepharose CL4B and Sephacryl S-400 superfine (Pharmacia Fine Chemicals, Sydney), Cellex D (DEAE-Cellulose), DEAE-Affi-gel blue and Affi-Gel 10 (Bio-Rad Lab., Hornsby, Australia), sodium heparin (Weddel Pharniaceuticals, Sydney), gelbond and agarose (FMC Co., Marine Colloids Division, BioProducts. Rockland, Maine). gelatin (BDH. Sydney), barbitone, sodium barbitone, para-nitro-phenylphosphate, Trizma base and alkaline phosphatase conjugated goat anti-human polyvalent immunoglobulins (a,y and p chain specific) (Sigma Chemical Co., St Louis, U.S.A.), cellulose acetate filters (0.22 pm) (Millipore Products Division, Bedford, U.S.A.), anti-human PC rabbit immunoglobulin (American Diagnostica, Sydney), goat anti-human transferrin antibodies, goat anti-mouse IgG antibodies and horseradish peroxidase conjugated goat anti-human immunoglobulin antibodies (Silenus, Melbourne, Australia), rabbit antiserum to AT-Ill (Calbiochem-Behring, Sydney) and YM30 ultra-filtration membranes (Amicon Scientific, Melbourne), were all obtained as indicated. Residual heparin was removed from dermatan sulphate (chondroitin sulphate Type B, Sigma Chemical Co., St Louis, U.S.A.) using the method described by Teien et a1 (1977). The monoclonal antibodies specific for PC (Hau & Salem, 1984) and the free form of PS (Mitchell et a]. 1988) were developed as previously described. Polyclonal antibodies against PS were developed in female white New Zealand rabbits by standard procedures using purified proteins as indicated below. All coagulation factors used were of human origin. PC (Salem et al, 1984a). PS (Dahlback, 1983), factor X and prothrombin (Miletich et al, 1980), thrombin (Miletech et al, 1978),AT-I11 (Owen, 1975),and thrombomodulin (Salem et ul, 1984b) were purified and/or activated according to methods previously described. All proteins were shown to be homogeneous as judged by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) performed according to Laemmli (19 70). Proteins were labelled with IZiI-Na using the Iodogen method (Pierce Chemical Co., Illinois). METHODS Plasma preparation. Blood was collected by venepuncture into 10% volume 3.8% trisodium citrate. Platelet-poorplasma (PPP) was prepared by centrifugation at 2000 g for 10 min at room temperature and stored a t - 70°C. Prior to freezing, an aliquot of the plasma was tested for the presence of LA. Assags and studies ofanticoagulant proteins. PC antigen level was measured using a solid phase competitive immunoradiometric assay which utilizes a monoclonal anti-PC antibody
381
and a radiolabelled rabbit anti-mouse IgG antibody (Hau & Salem, 1984). The amidolytic activity of APC was assessed using the chromogenic substrate Th-1. Briefly, samples containing APC were added to 2 0 mM Tris, pH 7.4, with 0 . 1 5 M NaCI and 0.1% gelatin to a final volume of 588 pl. Then 12 pl of Th-1(10 PM stock solution) was added to the mixture and the change of absorbance at 4 0 5 n m of the mixture over time was measured. The anticoagulant activity of APC, in the presence or absence of PS. was assessed using a factor Xa recalcification time (Xa-RT) as described by Mitchell et a1 (1986), with modifications. In the first step control and patient plasma was adsorbed with A1(OH)j to remove the vitamin K-dependent proteins. The assay was carried out using a n equal volume of frozen patient and fresh normal plasma. Control studies were performed using a similar mix of normal plasma. The assay buffer used in these studies was 2 0 mM Tris pH 7.4, containing 150 mM NaCl and 0.1%gelatin. The Xa-RT was adjusted to yield clotting times in the range of 42-48 s and samples were assayed in triplicate. In some studies human platelet membranes (prepared as described by Mitchell & Salem, 1987) were used instead ofplatelin as a surface for the reaction. The amount of platelet membrane was adjusted to produce the same baseline Xa-RT as platelin. PS antigen was quantified using a modified Laurell immuno-electrophoretic technique which employed a '"Ilabelled monoclonal anti-PS antibody (*MAb) and a precipitating rabbit polyclonal PS antibody (PAb) (Laurell, 1966). *MAb recognizes only the free form of PS and does not recognize the protein when complexed to C4b-binding protein (Mitchell et al, 1988). Plasma samples were electrophoresed into a first dimension agarose (0.9%) agarose containing the *MAb (4 x loi c.p.ni./ml agarose) and subsequently into a second dimension agarose containing the PAb. After the electrophoresis and washing, the precipitin peaks formed were visualized by autoradiography. The amount of PS in the sample was estimated by comparison to a standard curve constructed using normal pooled plasma. AT-III was determined using a heparin cofactor chromogenic assay (Abildgaard et al. 1985). HC-I1 activity was measured in a similar assay system except that dermatan sulphate was used as the glycosaminoglycan instead of heparin (Tollefsen & Pestka, 1985). CIEP studies of PC, s'1 and AT-I11 were performed as described by Marciniak (198 1 ) with modifications. In the case of PS, EDTA was added to the agarose in a concentration of 2 mbi in the first dimension and 7 mM in the second dimension. CIEP of AT-I11 was performed in the presence of 1 5 U/ml heparin in the first dimension. In addition, the agarose of the second dimension of the CIEP for PS and AT-I11 was adjusted to contain 2% polyethylene glycol 6000. Polyclonal antibodies against PC. PS and AT-Ill were used in performing the CIEP. Anticardiolipin antibodies (ACA) assay. ACA was assayed as previously described (Lo et al, 1989). The concentration of ACA was calculated from a dilution curve of a laboratory standard. A specific dilution of the standard was arbitrarily designated as lOOO/, and the concentration of ACA in test samples expressed as percentage of the standard.
382
Samuel C. L. Lo et a1
The normal range (mean f 2 SD) of the ACA assay was established using 117 normal sera obtained from the local Red Cross Blood Bank. Purification of the lupus anticoagulant. IgG containing LA activity was purified from 10 ml patient plasma. The plasma was dialysed against 0.02 M K2HP04 (pH 8.0). The small amount of precipitate formed was removed by centrifugation at 2000 g for 10 min. The supernatant was loaded onto a DEAE-Affi-Gel blue column (1.25 x 20 cm) equilibrated with the same buffer. The flow through containing IgG and transferrin was collected and concentrated by ultrafiltration using an Amicon YM-30 membrane. The material was then loaded onto an anti-human transferrin affinity column (1.25 x 12.5 cm) created by coupling the antibody to Affi-Gel 10. The affinity column contained 1 5 mg of antibody/ml of gel. The unbound protein contained LA activity and appeared as homogeneous IgG on SDS-PAGE. After concentration by ultrafiltration, it was stored at - 7OOC. TgG from normal plasma was purified in the same manner. For the purification of IgM-LA, 11 ml of plasma was adsorbed with AI(0H)l at a final concentration of 10%v/v. The plasma was then separated by centrifugation at 2 7 000 g for 30 min at 4OC and loaded onto a Sephadex G-150 column (2.5 x 84 cm) equilibrated with 20 mM Tris containing 0.15 M NaCI, 0.02% N a N I (pH 8.0). The void volume was concentrated by YM-30 ultrafiltration, dialysed against 20 mM (pH 7.4) containing 20 mM NaCl and applied onto a DEAE-cellulose column (1.2 5 x 16 cm) equilibrated in the dialysis buffer. After washing with the equilibration buffer, the column was developed using a 20-100 mM NaCl gradient in the same buffer system and 7 ml fractions collected. LA activity was monitored using the modified diluted activated partial thromboplastin time as described (Lo et aI, 1989). IgG and IgM eluting off the column were detected using a quantitative Laurel1immunoelectrophoretic technique (Laurell, 1966). The LA activity was concentrated by YM-30 ultrafiltration and the material applied to a Sepharose 4R-CL column equilibrated in 20 mM Tris, pH 8, containing 0.15 M NaC1. 0.02% NaN+ Fractions with LA activity were pooled, concentrated as described and finally applied to a Sephacryl S-400 superfine gel filtration column (2.5 x 86.5 cm) equilibrated in the same buffer. Purified IgM containing LA activity appeared homogeneous on SDS-PAGE. The protein was pooled and stored at - 70°C. IgM from normal plasma was purified in the same manner. Anti-pZateZet antibodies assay (APA). APA was identified using the enzyme-linked immunosorbent assay (ELISA) reported by Pfueller et al(1988).
Patients A total of 21 patients with LA were studied, 11 of whom satisfied the 1982 American Rheumatism Association criteria for SLE (Tan et al, 1982).Patients with thrombosis were studied at least 6 months after their last thrombotic event. LA was identified using a combination of eight different coagulation tests as previously detailed (Lo et a!, 1989). Patients were preselected by prolonged activated partial thromboplastin time (APTT) not corrected by a 1: 1 mixing with normal
plasma. Control studies were performed using plasma from eleven healthy volunteers. Tnformed consent was obtained from each patient and control in keeping with the policies of the Ethics and Research Committees of the Alfred Hospital and Monash University. RESULTS Thirteen of the 2 1 patients (Nos. 1-13) had a past history of thrombosis (Table I). The other eight patients (Nos. 14-2 1) had no previous history of thrombosis. Table 1. Summary of the clinical data of patients with a previous history of thrombosis
Patient 1 2 3
4
5 6 7 8 9 10 11 12 13
Clinical diagnosis
History of thrombosis
Anterior tibia1 arterial occlusion Recurrent DVT DVT and PTE Recurrent DVT Recurrent DVT and PTE CVA during pregnancy DVT DVT PTE CVA during pregnancy DVT and PTE DVT Widespread afferent arteriolar thrombosis and multiple small cerebral infarcts
Angiogram Venogram Venogram. V/Q scan Venogram Venogram, V/Q scan CT brain scan Venogram Venogram V/Q scan CT brain scan Venogram. V/Q scan Venogram Renal biopsy. CT brain scan
Abbreviations: DVT = deep venous thrombosis; PTE = pulmonary thromboembolism; V/Q scan = ventilation/perfusion lung scan; CVA = cerebrovascular accident; CT brain scan = computerized tomography brain scan.
As shown in Table 11, five of the 2 1 patients were on warfarin therapy when studied. The remaining 16 patients were not on anticoagulant therapy and did not have evidence of vitamin K deficiency as judged from the results of coagulation studies performed on their plasmas. Assays and studies of the natural anticoagulant system The antigenic levels of PC, PS and functional levels of AT-I11 and HC-I1 are shown in Table 11. In the five patients receiving warfarin, PC and PS levels were predictably low. AT-111 and HC-I1 levels did not demonstrate a fixed pattern of variability in these patients. Only 4/16 patients not receiving warfarin had normal levels of their natural anticoagulant proteins. The remaining 12 had one or more abnormal results (either increased or decreased). The effect of the patient plasma on the activity of APC was measured using a Xa-RT assay system. Fig 1 demonstrates the anticoagulant activity of APC in normal plasma when a rabbit brain phospholipid vesicle preparation (platelin) was used as the source of phospholipid. Increasing concentrations of APC resulted in a linear prolongation of the Xa-RT. When patients plasma was mixed with normal plasma in equal proportions, inhibition of APC activity was noted in most
383
Natural Anticoagulants and the Lupus Inhibitor Table 11. Levels of the natural anticoagulants and anticardiolipin antibodies in patients with LA.
240 0 9) v)
v
Patient
w 200
PCS (Pg/ml)
Normal range (mean 1 2 SD) 4-5-9.7
I i= 65-125
87-123
76-123
0-32
F Croup A: thrombotic It w 2 w 3 w 4t w 5 N* h N 7: 2.9 8 10.9 9: N 10 N 11t 4.4 121 1 1.4* 13t w
N 59 173* W
Group B: non-thrombotic 14t N 15; N Iht 1 1.4 17 N 18; 10.9 19 2.8 20 4.3 21 4.1'
N 52 N N N N N N
w W
w w N N N
N N
N 75 153 N 127
51 87 88 N 80 92 N N 9h 47 42 57 80
N N 128
116 50 64
182 145 N N
1S O 54 156 N
N
N
N N N 78 N N
N N N
130*
145 N N 156 N 136 84
N 138
N N N 133 139
86 86 55 N 62
Abbreviations: I:patients with SLE; W: patient taking warfarin: N: value within normal range: *: abnormal CIEP: $: antigenic levels; 4: functional levels: SD: standard deviation.
160
I0
6
120
LL
0
5
F
80
5
.
. 3
Q:
0
40
6K
n
0
2
4
6
,a
8 1012
CONC. OF APC (nM)
Fig 1 . The anticoagulant activity of APC in normal and patient plasma using platelin as the source of phospholipid ( A ,prolongation of clotting time of individual patient plasma; mean&2 SD of the response of normal plasma).
1,
h
0 z 301 9)
W
z 0 z
T
I-
instances. Using 5.7 nM APC, plasmas from 12/21 patients with the LA caused inhibition of the activity of APC (Student's t test, P < 0.001).At a higher concentration ofAPC (11.4 nM), significant decrease of the anticoagulant potential of APC ( P < O , O O l ) was observed in 20/21 patients. Purified IgG/MLA obtained from five different patients reproduced the inhibition observed using plasma. In general, the degree of APC inhibition observed did not correspond to the concentration of LA in the various diagnostic tests (results not shown). Furthermore, there was no correlation between APC inhibition and a previous history of thrombosis. We also examined the effect of the purified immunoglobulins on the amidolytic activity of APC. No effect was noted using either IgM or IgG concentrations as high as 0.7 mg/ml and 25 mg/ml respectively. The effect of purified PS on the inhibition of APC by patient plasma was also assessed. For this purpose, the plasma from five patients with strong APC-inhibitory activity was selected. The concentration of APC was adjusted to produce minimal prolongation of the clotting time. Under these conditions and using an equal mix of fresh and frozen normal plasma, the addition of increasing concentrations of PS (8.45-33.8 nM) resulted in prolongation of the clotting time
HORMAL
20
40
9
U
0
Gc
10
1.
6 5
a
6
0 z 0
10
30
20
CONC. OF PS (nM)
..
Fig 2 . The cofactor activity of PS in plasma of patients with the LA
(I.
meanf 2 SD; 0,0, using platelin as the source of phospholipid v. and + represent the response of plasma from patients 9, 14. 16, 5 and h respectively).
with apparent saturation at a PS concentration of approximately 20 nM (Fig 2). In the presence of plasma from patients with the LA, the cofactor activity of PS was significantly inhibited in all five instances (Fig 2).
384 80
Samuel C. L. Lo et a1
Prolongation of clotting time (sec )
60
40
Fig 3 . Thc anticoagulant activity of APC (42 n M ) with and without PS (1 6.9 nM) in normal and patients’ plasma using platelet membranes as the surface for the reaction. Each bar represents the response of different groups of samples and represents mean+ 2 SD. The numbers besides the bar reprcsent those patients’ plasma in which the responses to APC and PS were less than the mean - 2 SD of that of normal plasma. In the presence of PS. the difference between normal and patient plasma was statistically significant (0.01> P > 0 . 0 0 5 ) . In the absence of PS, there was no significant difference between normal and patient plasma samples.
I
3,5, 10 4, 73,79 20
0 Without PS
With 16.9 nM PS
In all coagulation studies cited above, platelin was used as the source of phospholipid. Fig 3 shows the effect on the anticoagulant potential of APC with and without PS when platelet membranes were used as the source of phospholipid. With the addition of 42 n M APC and no PS, only 2/2 1 patients demonstrated minimal degree of inhibition of APC activity. When 16.9 n M PS was added to the reaction,
inhibition of the anticoagulant activity by plasma from seven patients six of whom had previous thrombotic problems was noted. The patient who did not have a past history of thrombosis had detectable platelet antibodies which could have interfered with the function of the platelet membrane as a surface for the reaction. CIEP studies of PS and AT-TIT revealed one patient with an
Fig 4(a).Protein S-CIEP of normal and patient (No. 12) plasma. (i) Normal CIEP; (ii) patient CIEP; (iii) superimposing CIEP of normal and patient.
Fig 4(b). Antithrombin 111-CIEP of normal and patient (No. 12) plasma. ( i ) Normal CIEP: (ii) patient CIEP; (iii)superimposing CIEP of normal and patient.
Natural Anticoagulants and the Lupus Inhibitor abnormal pattern (Fig 4). This same patient and two others showed retarded mobility in their PC-CIEP (Table TI). All other patients had normal CIEP studies. Anticardiolipiti antibody (ACA) studies Elevated levels of ACA were noted in 17/2 1 patients (Table TI). There was no direct correlation between ACA positivity or titre and the inhibition of APC or PS activity. Four patients (Nos. 4, 7 , 8 and 2 0 ) who demonstrated APC inhibition when platelin was used had ACA levels within the normal range. It is ofinterest to note that three of these patients (Nos. 4. 7 and 8) had a previous history of thrombosis. On the other hand, one patient with a high ACA level (No. 1 7 ) failed to demonstrate any APC inhibitory activity. Furthermore, when platelet membranes were used to substitute platelin, two patients (Nos. 4 and 8) whose plasmas inhibited the function of PS had normal levels of ACA. Both of these patients had previous thrombotic episodes. DISCUSSION The lupus type inhibitor has long been known to be associated with a heightened predisposition for the development of thrombosis. The pathogenesis of this complication is not clear and numerous mechanisms have been proposed (Freyssinet Kr Cazenave. 1987). Patients with the LA commonly exhibit antiphospholipid activity in their plasmas (Harris et al. 1983. 1984; Petri et al, 1987). a finding confirmed by our results. Since the expression of APC-mediated anticoagulation depends on a phospholipid surface, we hypothesized that the presence of antiphospholipid antibodies could contribute to the development of thrombosis by inhibiting the anticoagulant function of APC. Our results suggest that plasma from patients with LA commonly inhibits APC in the presence or absence of PS. The effect was variable and primarily dependent on the type of phospholipid surface used in the reaction. IJsing rabbit brain phospholipid. APC activity was inhibited in 20/2 1 patients studied and the effect could not be overcome by the addition of PS. When platelet membranes were used as the surface for the reaction, different resufts were noted. Inhibition of APC activity was observed in only two of the 2 1 patients whereas the cofactor activity of PS was inhibited in 7/2 1 patients. It is important to note in the latter group that six patients had a past history of thrombosis. These results suggest that functional assays of PS using platelet membranes may detect a subgroup of patients who are at increased risk of thrombosis. The inhibition of APC and PS by LA is most probably secondary to interference by LA with the assembly of the anticoagulant proteins and substrates on the reaction surface. A similar mechanism has been proposed to explain the prolongation of the in-vitro clotting time produced by LA (Thiagarajan rt al, 1980: Kauch et nl. 1986: Pengo et al, 1987). The variable inhibition obtained using platelet membranes versus phospholipid vesicles probably reflects differences in the affinities of the proteins for the different surfaces. Contrary to the findings of Friedman et a1 (1986),we were unable to document a n association of reduced PS antigen in
3 85
patients with LA and a previous history of thrombosis. With the exception of patients on anticoaguiant therapy, a reduced level of any of the anticoagulant antigens was not a common finding in our study. The role of AT-I11 and HC-I1 deficiency in the pathogenesis of thrombosis in patients with LA is difficult to determine. In our study, only two patients demonstrated marginal reduction in the functional AT-I11 levels, one of whom had a past history of thrombosis. Another two had reduced functional levels of HC-11. both of whom had a previous history of thrombosis. AT-111 deficiency in association with LA and thrombosis has previously been noted (Cosgriff & Martin, 1981 ). but there has been no previous documentation of HC11 deficiency in this setting. In most of our patients with thrombosis, the AT-I11 and HC-11 levels were normal. It therefore appears that reduced levels of these proteins occurs in a minority of patients with LA. The relationship of antiphospholipid activity and inhibition of the APC and PS system is not straightforward. We did not observe a direct relation between the titre of ACA and the degree of inhibition of the anticoagulant activities of these proteins. This suggests that there are at least two distinct populations of antiphospholipid antibodies, a finding which is in agreement with the observations of Exner et ml(1988). The value of ACA assays as a marker for an increased risk of thrombosis is doubtful since three patients who had thrombosis had normal ACA values. In conclusion, abnormalities in the natural anticoagulant proteins were found to be variable in pateints with LA. An association with a thrombotic tendency was only consistently observed when a n inhibition of functional PS activity could be demonstrated. In other patients reduced antigen levels of various components of the NAS were observed and might have contributed to the development of thrombosis. ACKNOWLEDGMENTS We would like to thank Dr 1. Sloan for providing the dermatan sulphate and Ms L. Hau for providing the thrombin and thrombomodulin used in these studies. In addition. we would like to thank Dr D. Kandiah and Mr R. Fox for their technical assistance. This research was supported by grants from the National Health and Medical Research Council of Australia. REFERENCES Abildgaard. U.. Anderson, T.R.. Larsen, M.1,. & Handeland. G.F. (1985) In vivo and in vitro consumption of dermatan sulphatc cofactor (Heparin cofactor 11). Thrombosis and Haemostnsis. 54, abstract 1198. Bowie, E.J.W.. Thompson, J.H., Pascuzzi, C.R. & Owen, C.A. ( 1 9 h 3 ) Thrombosis in systemic lupus el-ythematosus despite circulating anticoagulants. journul of’ Luhoratory and Clirricd Medirinr, 6 2 , 41 6-430. Carreras, L.O., Defreyn. G.. Machin, S.J., Vermylen. 1.. Deman, K.. Spitz. B. & Van Assche. A. ( 198 1)Arterial thrombosis, intrauterine death and ‘lupus’ anticoagulant.: detection of immunoglobulin interfering with prostacyclin formation. Lnr7cct. i, 244-246. Cosgriff. T.M. & Martin, B.A. (1981) Low functional and high antigenic antithrombin 111 level in a patient with the lupus
386
Samuel C. L. Lo et a1
anticoagulant and recurrent thrombosis. Arthritis and Kheumatism, 24, 94-96. Dahlback, B. (1 983) Purification of human vitamin K-dependent protein S and its limited proteolysis by thrombin. Biochemical Journal, 209, 837-846. Exner, T., Sahman. N. & Trudinger. B. (1988) Separation of anticardiolipin antibodies from lupus anticoagulant on a phospholipid-coated polystyrene column. Biochemical and Biophysical Research Communications, 1 5 5, 1001-1007. Friedman. K.D.. Marlar. R.A., Gill, J.C.. Endres-Brooks, J. & Montgomery, R.R. (1986) Protein S deficiency in patients with the lupus anticoagulant. Blood, 68, (Suppl.), 33 3a. Freyssinet. J.-M. & Cazenave. J.-P.(19 8 7) Lupus-like anticoagulants, modulation of the protein C pathway and thrombosis. Thrombosis and Haemostasis. 58, 679-681. Freyssinet. J.-M., Wiesel, M.L.. Gauchy. 7.. Boneu, B. &Cazenave. J.P. (1 986) An IgM lupus anticoagulant that neutralizes the enhancing effect of phospholipid on purified endothelial thrombomodulin activity-a mechanism for thrombosis. Thrombosis and Huemostasis, 5 5 , 309-313. Gastineau,D.A., Kazmier. F.J.,Nichols, W.L.&Bowie, E.J.W. (1985) Lupus anticoagulant: a n analysis of the clinical and laboratory features of 219 cases. American Journal of Hematology, 19, 265-275. Harris, EN.. Gharavi, A.E., Boey, M.L., Patel, B.M., MackworthYoung, C.G.. Loizou, S. & Hughes, G.R.V. (1983) Anticardiolipin antibodies: detection by radioimmunoassay and association with thrombosis in systemic lupus erythematosus. Lancet. ii, 121 11214. Harris, E.N., Loizou. S.. Englert. H., Derue, G., Chan, J.K.. Gharavi, A.E. & Hughes, G.R.V. (1984) Anticardiolipin antibodies and lupus anticoagulant. Lancet, ii, 1099. Hau. L. &Salem, H.H. (1984) The characterization of a monoclonal antibody to human protein C and its use to establish a radioimmunoassay for protein C. Proceedings ofthe Australian Societyfor Medical Research, 17, 20. Johansson, E.A. & Lassus, A. (1974) The occurrence of circulating anticoagulants in patients with syphilitic and biologically false positive antilipoidal antibodies. Annals of Clinical Research, 6, 105108.
Laemmli. U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage Tq. Nature, 227, 680-685. Laurell, C.B. (1966) Quantitative estimation of proteins by electrophoresis in antibody-containing agarose gel. Analytical Biochemistry, 15, 14-52. Lechner, K. (1 974) Acquired inhibitors in non hemophiliac patients. Haemostasis, 3, 65-93. Lo. S.C.L.. Oldmeadow, M.J.. Howard, M.A. & Firkin, B.G. (1989) Comparison of laboratory tests used for identification of the lupus anticoagulant. American Journal of Hematology, 30, 21 3-220. Marciniak, E. (198 1) Thrombin-induced proteolysis of human antithrombin 111: a n outstanding contribution of heparin. British Journal of Haematology. 48, 325-336. Miletich, J.P., Broze. G.J. & Majerus. P.W. (1980) The synthesis of sulphated dextran beads for isolation of human plasma coagulation factors 11, IX and X. Analytical Biochemistry, 105, 304-3 10. Mi1etich.J.P.. Jackson, C.M. & Majerus, P.W. (1978) Properties of the factor X, binding site on human platelets. Journal of Biological Chemistry. 253,-6908-6916.
Mitchell, C.A.. Hau, L. & Salem, H.H. (1986) Control of thrombin mediated cleavage of protein S. Thrombosis and Haemostasis, 56, 151-1 54. Mitchell, C.A., Kelemen, S.M. &Salem, H.H. (1988) The anticoagulant properties of a modified form of protein S. Thrombosis and Haemostasis, 60, 298-304. Mitchell, C.A. & Salem, H.H. (1987) Cleavage of protein S by a platelet membrane protease. lournal of Clinical Investigation, 79, 374-379. Owen, W.G. (1 975) Evidence for the formation of an ester between thrombin and heparin cofactor. Biochimica et Biophysica Acta, 405, 380-387. Pengo, V.. Thiagarajan. P.. Shapiro. S.S. & Heine, M.J. (1987) Immunological specificity and mechanism of action of IgG lupus anticoagulants. Blood. 70, 69-76. Petri, M.. Rheinschmidt, M., Whiting-O'Keefe, Q.. Hellmann, D. & Corash, L. (1987) The frequency of lupus anticoagulant in systemic lupus erythematosus. A study of sixty consecutive patients by activated partial thromboplastin time, Russell Viper Venom time and anticardiolipin antibody level. Annals of Internal Medicine. 106, 524-53 1. Pfueller. S.L., Bilston, R.A., Logan, D., Gibson. J.M. & Firkin, B.G. (1988) Heterogeneity of drug-dependent platelet antigens and their antibodies in quinine- and quinidine-induced thrombocytopcnia: involvement of glycoproteins Ib, IIb, IIIa, and IX. Blood. 72, 1155-1162. Rauch.H.,Tannenbaum.M.. Tannenbaum,H.,Ramelson,H., Cullis, P.R.. Tilcock, C.P.S.. Hope, M.J. & Janoff. AS. (1986) Human hybridoma lupus anticoagulants distinguish between lamellar and hexagonal phase lipid systems. Journal of Biological Chemistry, 261, 9672-9677. Salem, H.H., Esmon, M L , Esmon. C.T. & Majerus, P.W. (1984a) Effects of thrombomodulin and coagulation factor Va-light chain on protein C activation in vitro. Journal of Clinical Investigation, 73, 968-972. Salem, H.H.. Maruyama. I.. Ishii. I. & Majerus. P.W. (1984b) Isolation and characterization of thrombomodulin from human placenta. Journal ofBiologica1 Chemistry. 259, 12246-12251. Sanfelippo. M.J. & Drayna. C.J. (1982) Prekallikrein inhibition associated with the lupus anticoagulant-a mechanism of thrombosis. American Journal of Clinical Pathology, 77, 275-279. Tan, E.M., Cohen, AS.. Fries, J.F., Masi. A.T.. McShane, D.J.. Rothfield, N.F.. Schaller. J.G., Talal, N. & Winchester, R.J. (1982) The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis and Rheumatism. 25. 1271-1 277. Teien, A.N., Abildgaard, U. & Hook, M. (1 977) The anticoagulant effect of heparin sulphate and dermatan sulphate. Thrombosis Research. 8, 859-867. Thiagarajan, P.. Shapiro. S.S. & De Marco, L. (1980) Monoclonal immunoglobulin MA coagulation inhibitor with phospholipid specificity: mechanism of a lupus anticoagulant. Journal of Clinical Investigation. 66, 397-405. Tobelem. G. & Cariou. R. (1986) The lupus anticoagulant induced a prethrombotic state by altering endothelial cell functions. Pruceedings XXZ Congress lnternational Society for Haematology, Sydney, p. 245 (SYM-F-8-1 abstract.) Tollefsen. D.M. & Petska, C.A. (1985) Heparin cofactor 11 activity in patients with disseminated intravascular coagulation and hepatic failure. Blood. 66, 769-774.
Studies of natural anticoagulant proteins and anticardiolipin antibodies in patients with the lupus anticoagulant SAMUELC. L. Lo, HATEMH. SALEM,MARGARET A. HOWARD,MICHAELJ. OLDMEADOW AND BARRYG . FIRKINDepartment of Medicine, Monash Medical School, Alfred Hospital. Victoria, Australia
Received 5 March 1990; accepted for publication 18 June 1990
Summary. Components of the natural anticoagulant system (NAS) and anticardiolipin antibodies were examined in 2 1 patients with lupus anticoagulant (LA), 1 3 of whom had past histories of thrombotic episodes. No relationship could be shown between the antigenic levels of protein C and S (PC, PS) and a history of thrombosis. Inhibition of the anticoagulant activity of activated protein C (APC) was observed using plasma from 20/2 1patients when phospholipid vesicles were used as the surface for the coagulation reaction. This effect was not affected by the addition of PS. When platelet membranes were employed only 2/2 1 patients demonstrated inhibition of APC. Under the latter condition, PS functional activity was inhibited in 7/21 patients, six of whom had a past history of thrombosis.
Reduced antithrombin I11 or heparin cofactor I1 levels were observed in a total of 4/2 1 patients and may have contributed to the development of thrombosis in three of these patients. Antibodies specifically directed against these proteins were not detected suggesting the possibility of an associated constitutional deficiency. Anticardiolipin antibodies, though elevated in 17/2 1 patients, did not serve as a useful marker for an increased risk of thrombosis, and the level did not correlate with inhibition of the activity of APC or PS. We conclude that the mechanism of thrombosis in patients with LA is multi-factorial. A subset of patients in whom LA specifically inhibits PS function may represent patients who are at significant risk from thrombosis.
Lupus anticoagulant (LA) is an immunoglobulin which results in prolongation of in vitro clotting times. Current evidence suggests that this anticoagulant activity is mediated by an effect on phospholipids (Thiagarajan et a!, 1980).This is supported by the frequent demonstration of antiphospholipid and anticardiolipin antibodies (ACA) in the plasma and/ or sera of patients with LA (Harris et a/, 1983, 1984: Petri et a/, 1987). Furthermore, coagulation tests used to detect LA are mostly phospholipid dependent. The association between LA and a predisposition to thrombosis in patients with or without systemic lupus erythematosus (SLE) has been widely reported (Bowie et al, 1963;Johansson & Lassus, 1974; Lechner, 1974; Gastineau et al, 1985; Petri et al, 1987).The pathogenesis of thrombosis in these patients is poorly understood with several mechanisms proposed to explain this association (Carreras et al, 1981; Cosgriff & Martin, 1981; Sanfelippo & Drayna. 1982; Freyssinet et al, 1986; Friedman et a/, 1986; Tobelem & Cariou. 1986). Most studies reported in the literature have been on small numbers of patients and it is not clear whether
the abnormalities occur selectively in patients who experienced a thrombotic event or are a feature of all patients with LA. In this study we investigated 21 patients with LA. Patients were grouped for the presence or absence of a history of thrombosis. Plasma from these patients was examined for its effect on the anticoagulant potential of activated protein C (APC) and protein S (PS).In addition, we measured the levels of four components of the natural anticoagulant system (NAS) namely protein C (PC), PS, antithrombin-111 (AT-111) and heparin cofactor I1 (HC-11).Qualitative assays on PC. PS and AT-I11 using two-dimensional crossed immunoelectrophoresis (CIEP) were also performed. Our results suggest that the activity of APC is inhibited by plasma from the majority of patients with LA, and bears no relation to a history of thrombosis. On the other hand, a strong correlation between thrombosis and inhibition of PS function was noted.
Correspondence: Dr Samuel C. L. LO, Department of Medicine, Monash Medical School, Alfred Hospital, Commercial Road, Prahran, VIC 3181, Australia.
Platelin (rabbit brain phospholipid) and simplastin (rabbit brain and lung thromboplastin) (General Diagnostics, Organon Teknika Co., Durham, U.S.A.), activated thrombofax
MATERIALS A N D METHODS Materials
3 80
Natural Anticoagulants and the Lupus Inhibitor (bovine brain phospholipid) (Ortho Diagnostics Systems, Raritan, U.S.A.), Russell viper venom (Wellcome Diagnostics, Sydney, Australia), IL5I-Na (Amersham, Sydney), iodogen and sodium dodecyl sulphate (Pierce Chemical Co., Illinois, U.S.A.), glycine, sodium chloride and EDTA (Ajax Chemical Co., Sydney), the chromogenic substrate S2238 (KabiVitrum, Stockholm, Sweden), Th-1 (Nycomed AS, Oslo, Norway), bovine thrombin (Parke-Davis Pty Ltd, Sydney), Sephadex G-150, Sepharose CL4B and Sephacryl S-400 superfine (Pharmacia Fine Chemicals, Sydney), Cellex D (DEAE-Cellulose), DEAE-Affi-gel blue and Affi-Gel 10 (Bio-Rad Lab., Hornsby, Australia), sodium heparin (Weddel Pharniaceuticals, Sydney), gelbond and agarose (FMC Co., Marine Colloids Division, BioProducts. Rockland, Maine). gelatin (BDH. Sydney), barbitone, sodium barbitone, para-nitro-phenylphosphate, Trizma base and alkaline phosphatase conjugated goat anti-human polyvalent immunoglobulins (a,y and p chain specific) (Sigma Chemical Co., St Louis, U.S.A.), cellulose acetate filters (0.22 pm) (Millipore Products Division, Bedford, U.S.A.), anti-human PC rabbit immunoglobulin (American Diagnostica, Sydney), goat anti-human transferrin antibodies, goat anti-mouse IgG antibodies and horseradish peroxidase conjugated goat anti-human immunoglobulin antibodies (Silenus, Melbourne, Australia), rabbit antiserum to AT-Ill (Calbiochem-Behring, Sydney) and YM30 ultra-filtration membranes (Amicon Scientific, Melbourne), were all obtained as indicated. Residual heparin was removed from dermatan sulphate (chondroitin sulphate Type B, Sigma Chemical Co., St Louis, U.S.A.) using the method described by Teien et a1 (1977). The monoclonal antibodies specific for PC (Hau & Salem, 1984) and the free form of PS (Mitchell et a]. 1988) were developed as previously described. Polyclonal antibodies against PS were developed in female white New Zealand rabbits by standard procedures using purified proteins as indicated below. All coagulation factors used were of human origin. PC (Salem et al, 1984a). PS (Dahlback, 1983), factor X and prothrombin (Miletich et al, 1980), thrombin (Miletech et al, 1978),AT-I11 (Owen, 1975),and thrombomodulin (Salem et ul, 1984b) were purified and/or activated according to methods previously described. All proteins were shown to be homogeneous as judged by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) performed according to Laemmli (19 70). Proteins were labelled with IZiI-Na using the Iodogen method (Pierce Chemical Co., Illinois). METHODS Plasma preparation. Blood was collected by venepuncture into 10% volume 3.8% trisodium citrate. Platelet-poorplasma (PPP) was prepared by centrifugation at 2000 g for 10 min at room temperature and stored a t - 70°C. Prior to freezing, an aliquot of the plasma was tested for the presence of LA. Assags and studies ofanticoagulant proteins. PC antigen level was measured using a solid phase competitive immunoradiometric assay which utilizes a monoclonal anti-PC antibody
381
and a radiolabelled rabbit anti-mouse IgG antibody (Hau & Salem, 1984). The amidolytic activity of APC was assessed using the chromogenic substrate Th-1. Briefly, samples containing APC were added to 2 0 mM Tris, pH 7.4, with 0 . 1 5 M NaCI and 0.1% gelatin to a final volume of 588 pl. Then 12 pl of Th-1(10 PM stock solution) was added to the mixture and the change of absorbance at 4 0 5 n m of the mixture over time was measured. The anticoagulant activity of APC, in the presence or absence of PS. was assessed using a factor Xa recalcification time (Xa-RT) as described by Mitchell et a1 (1986), with modifications. In the first step control and patient plasma was adsorbed with A1(OH)j to remove the vitamin K-dependent proteins. The assay was carried out using a n equal volume of frozen patient and fresh normal plasma. Control studies were performed using a similar mix of normal plasma. The assay buffer used in these studies was 2 0 mM Tris pH 7.4, containing 150 mM NaCl and 0.1%gelatin. The Xa-RT was adjusted to yield clotting times in the range of 42-48 s and samples were assayed in triplicate. In some studies human platelet membranes (prepared as described by Mitchell & Salem, 1987) were used instead ofplatelin as a surface for the reaction. The amount of platelet membrane was adjusted to produce the same baseline Xa-RT as platelin. PS antigen was quantified using a modified Laurell immuno-electrophoretic technique which employed a '"Ilabelled monoclonal anti-PS antibody (*MAb) and a precipitating rabbit polyclonal PS antibody (PAb) (Laurell, 1966). *MAb recognizes only the free form of PS and does not recognize the protein when complexed to C4b-binding protein (Mitchell et al, 1988). Plasma samples were electrophoresed into a first dimension agarose (0.9%) agarose containing the *MAb (4 x loi c.p.ni./ml agarose) and subsequently into a second dimension agarose containing the PAb. After the electrophoresis and washing, the precipitin peaks formed were visualized by autoradiography. The amount of PS in the sample was estimated by comparison to a standard curve constructed using normal pooled plasma. AT-III was determined using a heparin cofactor chromogenic assay (Abildgaard et al. 1985). HC-I1 activity was measured in a similar assay system except that dermatan sulphate was used as the glycosaminoglycan instead of heparin (Tollefsen & Pestka, 1985). CIEP studies of PC, s'1 and AT-I11 were performed as described by Marciniak (198 1 ) with modifications. In the case of PS, EDTA was added to the agarose in a concentration of 2 mbi in the first dimension and 7 mM in the second dimension. CIEP of AT-I11 was performed in the presence of 1 5 U/ml heparin in the first dimension. In addition, the agarose of the second dimension of the CIEP for PS and AT-I11 was adjusted to contain 2% polyethylene glycol 6000. Polyclonal antibodies against PC. PS and AT-Ill were used in performing the CIEP. Anticardiolipin antibodies (ACA) assay. ACA was assayed as previously described (Lo et al, 1989). The concentration of ACA was calculated from a dilution curve of a laboratory standard. A specific dilution of the standard was arbitrarily designated as lOOO/, and the concentration of ACA in test samples expressed as percentage of the standard.
382
Samuel C. L. Lo et a1
The normal range (mean f 2 SD) of the ACA assay was established using 117 normal sera obtained from the local Red Cross Blood Bank. Purification of the lupus anticoagulant. IgG containing LA activity was purified from 10 ml patient plasma. The plasma was dialysed against 0.02 M K2HP04 (pH 8.0). The small amount of precipitate formed was removed by centrifugation at 2000 g for 10 min. The supernatant was loaded onto a DEAE-Affi-Gel blue column (1.25 x 20 cm) equilibrated with the same buffer. The flow through containing IgG and transferrin was collected and concentrated by ultrafiltration using an Amicon YM-30 membrane. The material was then loaded onto an anti-human transferrin affinity column (1.25 x 12.5 cm) created by coupling the antibody to Affi-Gel 10. The affinity column contained 1 5 mg of antibody/ml of gel. The unbound protein contained LA activity and appeared as homogeneous IgG on SDS-PAGE. After concentration by ultrafiltration, it was stored at - 7OOC. TgG from normal plasma was purified in the same manner. For the purification of IgM-LA, 11 ml of plasma was adsorbed with AI(0H)l at a final concentration of 10%v/v. The plasma was then separated by centrifugation at 2 7 000 g for 30 min at 4OC and loaded onto a Sephadex G-150 column (2.5 x 84 cm) equilibrated with 20 mM Tris containing 0.15 M NaCI, 0.02% N a N I (pH 8.0). The void volume was concentrated by YM-30 ultrafiltration, dialysed against 20 mM (pH 7.4) containing 20 mM NaCl and applied onto a DEAE-cellulose column (1.2 5 x 16 cm) equilibrated in the dialysis buffer. After washing with the equilibration buffer, the column was developed using a 20-100 mM NaCl gradient in the same buffer system and 7 ml fractions collected. LA activity was monitored using the modified diluted activated partial thromboplastin time as described (Lo et aI, 1989). IgG and IgM eluting off the column were detected using a quantitative Laurel1immunoelectrophoretic technique (Laurell, 1966). The LA activity was concentrated by YM-30 ultrafiltration and the material applied to a Sepharose 4R-CL column equilibrated in 20 mM Tris, pH 8, containing 0.15 M NaC1. 0.02% NaN+ Fractions with LA activity were pooled, concentrated as described and finally applied to a Sephacryl S-400 superfine gel filtration column (2.5 x 86.5 cm) equilibrated in the same buffer. Purified IgM containing LA activity appeared homogeneous on SDS-PAGE. The protein was pooled and stored at - 70°C. IgM from normal plasma was purified in the same manner. Anti-pZateZet antibodies assay (APA). APA was identified using the enzyme-linked immunosorbent assay (ELISA) reported by Pfueller et al(1988).
Patients A total of 21 patients with LA were studied, 11 of whom satisfied the 1982 American Rheumatism Association criteria for SLE (Tan et al, 1982).Patients with thrombosis were studied at least 6 months after their last thrombotic event. LA was identified using a combination of eight different coagulation tests as previously detailed (Lo et a!, 1989). Patients were preselected by prolonged activated partial thromboplastin time (APTT) not corrected by a 1: 1 mixing with normal
plasma. Control studies were performed using plasma from eleven healthy volunteers. Tnformed consent was obtained from each patient and control in keeping with the policies of the Ethics and Research Committees of the Alfred Hospital and Monash University. RESULTS Thirteen of the 2 1 patients (Nos. 1-13) had a past history of thrombosis (Table I). The other eight patients (Nos. 14-2 1) had no previous history of thrombosis. Table 1. Summary of the clinical data of patients with a previous history of thrombosis
Patient 1 2 3
4
5 6 7 8 9 10 11 12 13
Clinical diagnosis
History of thrombosis
Anterior tibia1 arterial occlusion Recurrent DVT DVT and PTE Recurrent DVT Recurrent DVT and PTE CVA during pregnancy DVT DVT PTE CVA during pregnancy DVT and PTE DVT Widespread afferent arteriolar thrombosis and multiple small cerebral infarcts
Angiogram Venogram Venogram. V/Q scan Venogram Venogram, V/Q scan CT brain scan Venogram Venogram V/Q scan CT brain scan Venogram. V/Q scan Venogram Renal biopsy. CT brain scan
Abbreviations: DVT = deep venous thrombosis; PTE = pulmonary thromboembolism; V/Q scan = ventilation/perfusion lung scan; CVA = cerebrovascular accident; CT brain scan = computerized tomography brain scan.
As shown in Table 11, five of the 2 1 patients were on warfarin therapy when studied. The remaining 16 patients were not on anticoagulant therapy and did not have evidence of vitamin K deficiency as judged from the results of coagulation studies performed on their plasmas. Assays and studies of the natural anticoagulant system The antigenic levels of PC, PS and functional levels of AT-I11 and HC-I1 are shown in Table 11. In the five patients receiving warfarin, PC and PS levels were predictably low. AT-111 and HC-I1 levels did not demonstrate a fixed pattern of variability in these patients. Only 4/16 patients not receiving warfarin had normal levels of their natural anticoagulant proteins. The remaining 12 had one or more abnormal results (either increased or decreased). The effect of the patient plasma on the activity of APC was measured using a Xa-RT assay system. Fig 1 demonstrates the anticoagulant activity of APC in normal plasma when a rabbit brain phospholipid vesicle preparation (platelin) was used as the source of phospholipid. Increasing concentrations of APC resulted in a linear prolongation of the Xa-RT. When patients plasma was mixed with normal plasma in equal proportions, inhibition of APC activity was noted in most
383
Natural Anticoagulants and the Lupus Inhibitor Table 11. Levels of the natural anticoagulants and anticardiolipin antibodies in patients with LA.
240 0 9) v)
v
Patient
w 200
PCS (Pg/ml)
Normal range (mean 1 2 SD) 4-5-9.7
I i= 65-125
87-123
76-123
0-32
F Croup A: thrombotic It w 2 w 3 w 4t w 5 N* h N 7: 2.9 8 10.9 9: N 10 N 11t 4.4 121 1 1.4* 13t w
N 59 173* W
Group B: non-thrombotic 14t N 15; N Iht 1 1.4 17 N 18; 10.9 19 2.8 20 4.3 21 4.1'
N 52 N N N N N N
w W
w w N N N
N N
N 75 153 N 127
51 87 88 N 80 92 N N 9h 47 42 57 80
N N 128
116 50 64
182 145 N N
1S O 54 156 N
N
N
N N N 78 N N
N N N
130*
145 N N 156 N 136 84
N 138
N N N 133 139
86 86 55 N 62
Abbreviations: I:patients with SLE; W: patient taking warfarin: N: value within normal range: *: abnormal CIEP: $: antigenic levels; 4: functional levels: SD: standard deviation.
160
I0
6
120
LL
0
5
F
80
5
.
. 3
Q:
0
40
6K
n
0
2
4
6
,a
8 1012
CONC. OF APC (nM)
Fig 1 . The anticoagulant activity of APC in normal and patient plasma using platelin as the source of phospholipid ( A ,prolongation of clotting time of individual patient plasma; mean&2 SD of the response of normal plasma).
1,
h
0 z 301 9)
W
z 0 z
T
I-
instances. Using 5.7 nM APC, plasmas from 12/21 patients with the LA caused inhibition of the activity of APC (Student's t test, P < 0.001).At a higher concentration ofAPC (11.4 nM), significant decrease of the anticoagulant potential of APC ( P < O , O O l ) was observed in 20/21 patients. Purified IgG/MLA obtained from five different patients reproduced the inhibition observed using plasma. In general, the degree of APC inhibition observed did not correspond to the concentration of LA in the various diagnostic tests (results not shown). Furthermore, there was no correlation between APC inhibition and a previous history of thrombosis. We also examined the effect of the purified immunoglobulins on the amidolytic activity of APC. No effect was noted using either IgM or IgG concentrations as high as 0.7 mg/ml and 25 mg/ml respectively. The effect of purified PS on the inhibition of APC by patient plasma was also assessed. For this purpose, the plasma from five patients with strong APC-inhibitory activity was selected. The concentration of APC was adjusted to produce minimal prolongation of the clotting time. Under these conditions and using an equal mix of fresh and frozen normal plasma, the addition of increasing concentrations of PS (8.45-33.8 nM) resulted in prolongation of the clotting time
HORMAL
20
40
9
U
0
Gc
10
1.
6 5
a
6
0 z 0
10
30
20
CONC. OF PS (nM)
..
Fig 2 . The cofactor activity of PS in plasma of patients with the LA
(I.
meanf 2 SD; 0,0, using platelin as the source of phospholipid v. and + represent the response of plasma from patients 9, 14. 16, 5 and h respectively).
with apparent saturation at a PS concentration of approximately 20 nM (Fig 2). In the presence of plasma from patients with the LA, the cofactor activity of PS was significantly inhibited in all five instances (Fig 2).
384 80
Samuel C. L. Lo et a1
Prolongation of clotting time (sec )
60
40
Fig 3 . Thc anticoagulant activity of APC (42 n M ) with and without PS (1 6.9 nM) in normal and patients’ plasma using platelet membranes as the surface for the reaction. Each bar represents the response of different groups of samples and represents mean+ 2 SD. The numbers besides the bar reprcsent those patients’ plasma in which the responses to APC and PS were less than the mean - 2 SD of that of normal plasma. In the presence of PS. the difference between normal and patient plasma was statistically significant (0.01> P > 0 . 0 0 5 ) . In the absence of PS, there was no significant difference between normal and patient plasma samples.
I
3,5, 10 4, 73,79 20
0 Without PS
With 16.9 nM PS
In all coagulation studies cited above, platelin was used as the source of phospholipid. Fig 3 shows the effect on the anticoagulant potential of APC with and without PS when platelet membranes were used as the source of phospholipid. With the addition of 42 n M APC and no PS, only 2/2 1 patients demonstrated minimal degree of inhibition of APC activity. When 16.9 n M PS was added to the reaction,
inhibition of the anticoagulant activity by plasma from seven patients six of whom had previous thrombotic problems was noted. The patient who did not have a past history of thrombosis had detectable platelet antibodies which could have interfered with the function of the platelet membrane as a surface for the reaction. CIEP studies of PS and AT-TIT revealed one patient with an
Fig 4(a).Protein S-CIEP of normal and patient (No. 12) plasma. (i) Normal CIEP; (ii) patient CIEP; (iii) superimposing CIEP of normal and patient.
Fig 4(b). Antithrombin 111-CIEP of normal and patient (No. 12) plasma. ( i ) Normal CIEP: (ii) patient CIEP; (iii)superimposing CIEP of normal and patient.
Natural Anticoagulants and the Lupus Inhibitor abnormal pattern (Fig 4). This same patient and two others showed retarded mobility in their PC-CIEP (Table TI). All other patients had normal CIEP studies. Anticardiolipiti antibody (ACA) studies Elevated levels of ACA were noted in 17/2 1 patients (Table TI). There was no direct correlation between ACA positivity or titre and the inhibition of APC or PS activity. Four patients (Nos. 4, 7 , 8 and 2 0 ) who demonstrated APC inhibition when platelin was used had ACA levels within the normal range. It is ofinterest to note that three of these patients (Nos. 4. 7 and 8) had a previous history of thrombosis. On the other hand, one patient with a high ACA level (No. 1 7 ) failed to demonstrate any APC inhibitory activity. Furthermore, when platelet membranes were used to substitute platelin, two patients (Nos. 4 and 8) whose plasmas inhibited the function of PS had normal levels of ACA. Both of these patients had previous thrombotic episodes. DISCUSSION The lupus type inhibitor has long been known to be associated with a heightened predisposition for the development of thrombosis. The pathogenesis of this complication is not clear and numerous mechanisms have been proposed (Freyssinet Kr Cazenave. 1987). Patients with the LA commonly exhibit antiphospholipid activity in their plasmas (Harris et al. 1983. 1984; Petri et al, 1987). a finding confirmed by our results. Since the expression of APC-mediated anticoagulation depends on a phospholipid surface, we hypothesized that the presence of antiphospholipid antibodies could contribute to the development of thrombosis by inhibiting the anticoagulant function of APC. Our results suggest that plasma from patients with LA commonly inhibits APC in the presence or absence of PS. The effect was variable and primarily dependent on the type of phospholipid surface used in the reaction. IJsing rabbit brain phospholipid. APC activity was inhibited in 20/2 1 patients studied and the effect could not be overcome by the addition of PS. When platelet membranes were used as the surface for the reaction, different resufts were noted. Inhibition of APC activity was observed in only two of the 2 1 patients whereas the cofactor activity of PS was inhibited in 7/2 1 patients. It is important to note in the latter group that six patients had a past history of thrombosis. These results suggest that functional assays of PS using platelet membranes may detect a subgroup of patients who are at increased risk of thrombosis. The inhibition of APC and PS by LA is most probably secondary to interference by LA with the assembly of the anticoagulant proteins and substrates on the reaction surface. A similar mechanism has been proposed to explain the prolongation of the in-vitro clotting time produced by LA (Thiagarajan rt al, 1980: Kauch et nl. 1986: Pengo et al, 1987). The variable inhibition obtained using platelet membranes versus phospholipid vesicles probably reflects differences in the affinities of the proteins for the different surfaces. Contrary to the findings of Friedman et a1 (1986),we were unable to document a n association of reduced PS antigen in
3 85
patients with LA and a previous history of thrombosis. With the exception of patients on anticoaguiant therapy, a reduced level of any of the anticoagulant antigens was not a common finding in our study. The role of AT-I11 and HC-I1 deficiency in the pathogenesis of thrombosis in patients with LA is difficult to determine. In our study, only two patients demonstrated marginal reduction in the functional AT-I11 levels, one of whom had a past history of thrombosis. Another two had reduced functional levels of HC-11. both of whom had a previous history of thrombosis. AT-111 deficiency in association with LA and thrombosis has previously been noted (Cosgriff & Martin, 1981 ). but there has been no previous documentation of HC11 deficiency in this setting. In most of our patients with thrombosis, the AT-I11 and HC-11 levels were normal. It therefore appears that reduced levels of these proteins occurs in a minority of patients with LA. The relationship of antiphospholipid activity and inhibition of the APC and PS system is not straightforward. We did not observe a direct relation between the titre of ACA and the degree of inhibition of the anticoagulant activities of these proteins. This suggests that there are at least two distinct populations of antiphospholipid antibodies, a finding which is in agreement with the observations of Exner et ml(1988). The value of ACA assays as a marker for an increased risk of thrombosis is doubtful since three patients who had thrombosis had normal ACA values. In conclusion, abnormalities in the natural anticoagulant proteins were found to be variable in pateints with LA. An association with a thrombotic tendency was only consistently observed when a n inhibition of functional PS activity could be demonstrated. In other patients reduced antigen levels of various components of the NAS were observed and might have contributed to the development of thrombosis. ACKNOWLEDGMENTS We would like to thank Dr 1. Sloan for providing the dermatan sulphate and Ms L. Hau for providing the thrombin and thrombomodulin used in these studies. In addition. we would like to thank Dr D. Kandiah and Mr R. Fox for their technical assistance. This research was supported by grants from the National Health and Medical Research Council of Australia. REFERENCES Abildgaard. U.. Anderson, T.R.. Larsen, M.1,. & Handeland. G.F. (1985) In vivo and in vitro consumption of dermatan sulphatc cofactor (Heparin cofactor 11). Thrombosis and Haemostnsis. 54, abstract 1198. Bowie, E.J.W.. Thompson, J.H., Pascuzzi, C.R. & Owen, C.A. ( 1 9 h 3 ) Thrombosis in systemic lupus el-ythematosus despite circulating anticoagulants. journul of’ Luhoratory and Clirricd Medirinr, 6 2 , 41 6-430. Carreras, L.O., Defreyn. G.. Machin, S.J., Vermylen. 1.. Deman, K.. Spitz. B. & Van Assche. A. ( 198 1)Arterial thrombosis, intrauterine death and ‘lupus’ anticoagulant.: detection of immunoglobulin interfering with prostacyclin formation. Lnr7cct. i, 244-246. Cosgriff. T.M. & Martin, B.A. (1981) Low functional and high antigenic antithrombin 111 level in a patient with the lupus
386
Samuel C. L. Lo et a1
anticoagulant and recurrent thrombosis. Arthritis and Kheumatism, 24, 94-96. Dahlback, B. (1 983) Purification of human vitamin K-dependent protein S and its limited proteolysis by thrombin. Biochemical Journal, 209, 837-846. Exner, T., Sahman. N. & Trudinger. B. (1988) Separation of anticardiolipin antibodies from lupus anticoagulant on a phospholipid-coated polystyrene column. Biochemical and Biophysical Research Communications, 1 5 5, 1001-1007. Friedman. K.D.. Marlar. R.A., Gill, J.C.. Endres-Brooks, J. & Montgomery, R.R. (1986) Protein S deficiency in patients with the lupus anticoagulant. Blood, 68, (Suppl.), 33 3a. Freyssinet. J.-M. & Cazenave. J.-P.(19 8 7) Lupus-like anticoagulants, modulation of the protein C pathway and thrombosis. Thrombosis and Haemostasis. 58, 679-681. Freyssinet. J.-M., Wiesel, M.L.. Gauchy. 7.. Boneu, B. &Cazenave. J.P. (1 986) An IgM lupus anticoagulant that neutralizes the enhancing effect of phospholipid on purified endothelial thrombomodulin activity-a mechanism for thrombosis. Thrombosis and Huemostasis, 5 5 , 309-313. Gastineau,D.A., Kazmier. F.J.,Nichols, W.L.&Bowie, E.J.W. (1985) Lupus anticoagulant: a n analysis of the clinical and laboratory features of 219 cases. American Journal of Hematology, 19, 265-275. Harris, EN.. Gharavi, A.E., Boey, M.L., Patel, B.M., MackworthYoung, C.G.. Loizou, S. & Hughes, G.R.V. (1983) Anticardiolipin antibodies: detection by radioimmunoassay and association with thrombosis in systemic lupus erythematosus. Lancet. ii, 121 11214. Harris, E.N., Loizou. S.. Englert. H., Derue, G., Chan, J.K.. Gharavi, A.E. & Hughes, G.R.V. (1984) Anticardiolipin antibodies and lupus anticoagulant. Lancet, ii, 1099. Hau. L. &Salem, H.H. (1984) The characterization of a monoclonal antibody to human protein C and its use to establish a radioimmunoassay for protein C. Proceedings ofthe Australian Societyfor Medical Research, 17, 20. Johansson, E.A. & Lassus, A. (1974) The occurrence of circulating anticoagulants in patients with syphilitic and biologically false positive antilipoidal antibodies. Annals of Clinical Research, 6, 105108.
Laemmli. U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage Tq. Nature, 227, 680-685. Laurell, C.B. (1966) Quantitative estimation of proteins by electrophoresis in antibody-containing agarose gel. Analytical Biochemistry, 15, 14-52. Lechner, K. (1 974) Acquired inhibitors in non hemophiliac patients. Haemostasis, 3, 65-93. Lo. S.C.L.. Oldmeadow, M.J.. Howard, M.A. & Firkin, B.G. (1989) Comparison of laboratory tests used for identification of the lupus anticoagulant. American Journal of Hematology, 30, 21 3-220. Marciniak, E. (198 1) Thrombin-induced proteolysis of human antithrombin 111: a n outstanding contribution of heparin. British Journal of Haematology. 48, 325-336. Miletich, J.P., Broze. G.J. & Majerus. P.W. (1980) The synthesis of sulphated dextran beads for isolation of human plasma coagulation factors 11, IX and X. Analytical Biochemistry, 105, 304-3 10. Mi1etich.J.P.. Jackson, C.M. & Majerus, P.W. (1978) Properties of the factor X, binding site on human platelets. Journal of Biological Chemistry. 253,-6908-6916.
Mitchell, C.A.. Hau, L. & Salem, H.H. (1986) Control of thrombin mediated cleavage of protein S. Thrombosis and Haemostasis, 56, 151-1 54. Mitchell, C.A., Kelemen, S.M. &Salem, H.H. (1988) The anticoagulant properties of a modified form of protein S. Thrombosis and Haemostasis, 60, 298-304. Mitchell, C.A. & Salem, H.H. (1987) Cleavage of protein S by a platelet membrane protease. lournal of Clinical Investigation, 79, 374-379. Owen, W.G. (1 975) Evidence for the formation of an ester between thrombin and heparin cofactor. Biochimica et Biophysica Acta, 405, 380-387. Pengo, V.. Thiagarajan. P.. Shapiro. S.S. & Heine, M.J. (1987) Immunological specificity and mechanism of action of IgG lupus anticoagulants. Blood. 70, 69-76. Petri, M.. Rheinschmidt, M., Whiting-O'Keefe, Q.. Hellmann, D. & Corash, L. (1987) The frequency of lupus anticoagulant in systemic lupus erythematosus. A study of sixty consecutive patients by activated partial thromboplastin time, Russell Viper Venom time and anticardiolipin antibody level. Annals of Internal Medicine. 106, 524-53 1. Pfueller. S.L., Bilston, R.A., Logan, D., Gibson. J.M. & Firkin, B.G. (1988) Heterogeneity of drug-dependent platelet antigens and their antibodies in quinine- and quinidine-induced thrombocytopcnia: involvement of glycoproteins Ib, IIb, IIIa, and IX. Blood. 72, 1155-1162. Rauch.H.,Tannenbaum.M.. Tannenbaum,H.,Ramelson,H., Cullis, P.R.. Tilcock, C.P.S.. Hope, M.J. & Janoff. AS. (1986) Human hybridoma lupus anticoagulants distinguish between lamellar and hexagonal phase lipid systems. Journal of Biological Chemistry, 261, 9672-9677. Salem, H.H., Esmon, M L , Esmon. C.T. & Majerus, P.W. (1984a) Effects of thrombomodulin and coagulation factor Va-light chain on protein C activation in vitro. Journal of Clinical Investigation, 73, 968-972. Salem, H.H.. Maruyama. I.. Ishii. I. & Majerus. P.W. (1984b) Isolation and characterization of thrombomodulin from human placenta. Journal ofBiologica1 Chemistry. 259, 12246-12251. Sanfelippo. M.J. & Drayna. C.J. (1982) Prekallikrein inhibition associated with the lupus anticoagulant-a mechanism of thrombosis. American Journal of Clinical Pathology, 77, 275-279. Tan, E.M., Cohen, AS.. Fries, J.F., Masi. A.T.. McShane, D.J.. Rothfield, N.F.. Schaller. J.G., Talal, N. & Winchester, R.J. (1982) The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis and Rheumatism. 25. 1271-1 277. Teien, A.N., Abildgaard, U. & Hook, M. (1 977) The anticoagulant effect of heparin sulphate and dermatan sulphate. Thrombosis Research. 8, 859-867. Thiagarajan, P.. Shapiro. S.S. & De Marco, L. (1980) Monoclonal immunoglobulin MA coagulation inhibitor with phospholipid specificity: mechanism of a lupus anticoagulant. Journal of Clinical Investigation. 66, 397-405. Tobelem. G. & Cariou. R. (1986) The lupus anticoagulant induced a prethrombotic state by altering endothelial cell functions. Pruceedings XXZ Congress lnternational Society for Haematology, Sydney, p. 245 (SYM-F-8-1 abstract.) Tollefsen. D.M. & Petska, C.A. (1985) Heparin cofactor 11 activity in patients with disseminated intravascular coagulation and hepatic failure. Blood. 66, 769-774.
