Induction of primary antiphospholipid syndrome in mice by immunization with a human monoclonal anticardiolipin antibody (H-3)

The lupus anticoagulant/antiphospholipid syndrome

Lupus anticoagulants and anticardiolipin antibodies have been strongly associated with the risk of thrombosis, recurrent fetal loss, thrombocytopenia, and a number of other clinical manifestations that together have been referred to as the antiphospholipid syndrome. Despite growing evidence of the significance of this relationship, the pathogenetic mechanisms involved are largely unknown. Recent data suggest strongly that lupus anticoagulants (LACs) and anticardiolipin antibodies (ACAs) are antibodies to protein-phospholipid complexes rather than to phospholipids, as had originally been thought, and that other protein-phospholipid complexes, not recognized by standard assays for LACs or ACAs, may also exist in patients with the antiphospholipid syndrome. Although very recent experimental data may lead to new therapeutic approaches in this syndrome, at present we can only deal with the thrombotic risk by the use of long-term anticoagulation. This chapter reviews current methods of diagnosis, concepts of pathogenesis, and the basis for an approach to anticoagulation in patients at risk for thrombosis or other manifestations of the antiphospholipid syndrome.

Aspirin-interleukin-3 interrelationships in patients with anti-phospholipid syndrome

PROBLEM

Previously we reported on the generation of experimental anti-phospholipid syndrome (APS) in mice. These models were employed to evaluate the efficacy of various novel therapeutic modalities including interleukin-3 (IL-3) and low dose aspirin. The efficacy of the latter was found to be interrelated. Low dose aspirin is capable of inhibiting the activity of the enzyme cyclooxygenase which is responsible for the metabolism of arachidonic acid towards the production of prostaglandins. This shifts the metabolism of arachidonic acid to the other pathway and leads to an overproduction of leukotrienes. The leukotrienes act as stimulators of IL-3 production, a positive cytokine in pregnancy which enhances placental and fetal development. In the current study we evaluated the IL-3 levels in pregnant women with APS and expanded our knowledge on the interrelationships between aspirin, arachidonic acid metabolites and IL-3 in the human system.

METHODS

IL-3 levels were recorded in the serum of pregnant women with APS and compared to a control pregnant group. In addition peripheral blood mononuclear cells from healthy subjects were incubated with different concentrations of aspirin or with arachidonic acid metabolites (Leukotriene B4, C4 or PGE2), and IL-3 production in the culture fluids was evaluated.

RESULTS

Serum level of IL-3 in pregnant patients with primary APS, APS secondary to SLE and SLE was lower in comparison to the control group. The in vitro studies revealed that only low dose aspirin (10 mg/microliters) stimulated IL-3 production while higher doses of the drug failed to induce the cytokine generation. Leukotriene B4 and C4 were stimulatory whereas PGE2 acted as inhibitor of IL-3 production.

CONCLUSIONS

The serum level of IL-3 is decreased to pregnant women with primary or secondary APS. Low dose aspirin is capable of stimulating IL-3 production in vitro most probably through an elevation of leukotriene production, which may explain its beneficial activity in preventing the manifestations of APS.

The significance of anticardiolipin antibodies in patients with lupus nephritis

The objective of this study was to determine whether anticardiolipin antibodies (ACL) in SLE patients are associated with a specific pattern of lupus nephritis and/or with renal microvascular changes. Patients with SLE, followed prospectively between June 1991-May 1994 at The Wellesley Hospital Lupus Clinic, who underwent a renal biopsy were included. The ACL was measured by the ELISA according to international standardized method. Renal biopsy morphology was assessed using the WHO criteria for the classification of lupus nephritis. Renal vascular changes included glomerular hyaline thrombi, intimal fibrosis and intraluminal thrombi of renal arterioles. There were 23 SLE patients. The mean age at diagnosis of SLE was 28.2 years and the mean disease duration was 6.3 years. Of these 10 (43%) had high levels of ACL. No difference in the frequency of severe nephritis (Class III and IV) was identified amongst patients with and without ACL. Mesangial nephritis was more common in patients with ACL 40% vs 0, p = 0.02). Glomerular hyaline thrombi occurred in 3 (13%) patients. None of them had positive ACL. Renal vascular lesions included intimal proliferation in 4 (ACL + , 1) occluded lumens by thrombi in 2 (ACL + 1). Our data indicate that the development of glomerular and/or microvascular changes is not related to the presence of ACL.

Anticardiolipin, but not the 16/6 Id anti-DNA antibody induces pregnancy failure

The primary antiphospholipid syndrome or the antiphospholipid syndrome in systemic lupus erythematosus patients (defined as secondary antiphospholipid syndrome) are characterized by the presence of thrombosis, thrombocytopenia and recurrent fetal loss in association with anticardiolipin antibodies. To determine the causal role of these antibodies in the pathogenesis of pregnancy failure we studied the effects of immunization with monoclonal anti-DNA antibody (designated 16/6 Id; no cardiolipin reactivity) and anticardiolipin monoclonal antibody (designated 2C4C2; binds DNA as well) on the outcome of allogeneic pregnancies in BALB/c mice. Mating of BALB/c females 4 weeks after active immunization with the 16/6 Id, anti-DNA monoclonal antibody resulted in normal pregnancy outcome, similar to control mouse groups. In contrast to that, immunization with the 2C4C2 anticardiolipin antibodies resulted in severe gestational failure with low pregnancy rate, low numbers of fetuses and high rates of resorptions. The fertility index of those mice was extremely low as compared to the 16/6 Id-immunized mice or the control groups. Furthermore, a correlation was shown between the presence of anticardiolipin antibody levels in the sera of the mice at the time of gestation and the pregnancy fate. The 2C4C2-immunized mice which produced high levels of anticardiolipin antibodies demonstrated severe pregnancy failure, whereas normal gestations were observed in the 16/6 Id primed or the control mouse groups that did not produce measurable amounts of the latter antibodies. Thus, our studies demonstrate that anticardiolipin but not the 16/6 Id anti-DNA antibodies can induce severe gestational impairment.

Antiphospholipid Antibodies and Associated Clinical Manifestations

The Antiphospholipid Syndrome (APS) is defined by the association between antiphospholipid antibodies, i.e. anticardiolipin (aCL) and/or lupus anticoagulant (LA) antibodies, and one or more of the following clinical manifestations: arterial and venous thrombosis, recurrent abortions and thrombocytopenia (1). Among them, deep venous thromboses, pulmonary embolism and thrombosis of the cerebral arteries are the most frequent events, occurring in approximately 1/3 of the patients. Thrombosis of the placental vessels (2) is considered the cause of the obstetrical complications (recurrent spontaneous abortions, fetal deaths or fetal growth retardation) suffered by approximately 10% of the women with antiphospholipid antibodies (1), whereas a variable degree of thrombocytopenia is reported by about 20-25% of the patients (3). Less commonly, skin necrosis, livedo reticularis, hemolytic anemia, dementia or other neuropsychiatric events and the so-called "catastrophic" APS may also develop in the setting of APS (4, 5). Two types of APS have been described: the "Primary" APS, which occurs in the absence of an underlying disease (6), and the "Secondary" APS, which is related to Systemic Lupus Erythematosus (SLE), other autoimmune or neoplastic diseases or other pathological conditions (7). Noteworthy, a substantial amount of patients suffering from APS are young: 50% of the patients enrolled in the Italian Registry of Antiphospholipid Antibodies were aged less than 40 years (8).

Anticardiolipin antibodies in systemic lupus erythematosus: clinical and laboratory correlations

PURPOSE

To examine the link between anticardiolipin antibodies and the features of antiphospholipid syndrome in patients with systemic lupus erythematosus (SLE).

PATIENTS AND METHODS

In this prospective cohort study conducted in a single center, 390 SLE patients were followed up between June 1991 and 1994. At each assessment, a complete history, physical examination, and laboratory evaluation (including measurement of anticardiolipin antibodies) were performed according to a standard protocol.

RESULTS

Forty-seven percent of the patients had an elevated level of anticardiolipin antibodies. In the univariate analysis, elevated anticardiolipin antibody levels were found to correlate with thrombocytopenia (P = 0.006), prolonged activated partial thromboplastin time (aPTT) (P = 0.003), and positive direct and indirect Coombs' test result's (P < 0.001). No correlation was identified with any of the clinical features of antiphospholipid syndrome. In the multivariate analysis, anticardiolipin antibodies remained highly associated with thrombocytopenia (odds ratio [OR] 4.05, P = 0.02), positive direct Coombs' test (OR 2.31, P < 0.001), and prolonged aPTT (OR 1.73, P = 0.03). In the multivariate model using venous/arterial thrombosis as the outcome variable, only prolonged aPTT was associated with venous/arterial thrombosis (OR 7.9, P < 0.001). None of the laboratory variables were found to correlate with fetal loss.

CONCLUSIONS

The presence of anticardiolipin antibodies in patients with SLE is associated with prolonged aPTT, thrombocytopenia, and positive Coombs' test result, but not with antiphospholipid syndrome. Prolonged aPTT is strongly associated with venous/arterial thrombosis.

The antiphospholipid-protein syndrome

The pathogenesis of the antiphospholipid syndrome remains uncertain. Antibodies that react with phospholipids may not be directly responsible for cellular injury, but may be part of the immune network through which autoantibodies with pathogenic potential are generated. The latter may recognize proteins such as beta 2-glycoprotein I that form complexes with phospholipids, proteins whose functions depend upon interaction with phospholipids such as protein C and its cofactors, altered lipoproteins such as oxidized low-density lipoproteins, or other molecules that share only antigenic similarity. Thus, a spectrum of autoantibodies that recognize different lipid-protein complexes may develop in these patients and contribute to the observed clinical heterogeneity of the syndrome. Current techniques do not permit identification of the subset of patients with antiphospholipid antibodies at risk for thrombosis or abortion and there are no prospective, controlled trials addressing the prophylaxis or treatment of affected individuals. Identification of the cellular targets of antibodies to lipid-protein moieties is needed to identify patients at risk for these complications and as a means to monitor therapy.

Natural autoantibodies in sera of patients with Gaucher's disease

Gaucher's disease (GD) is associated with hyperactivity of the immune system, which manifests by polyclonal hypergamma-globulinemia and an increased incidence of monoclonal gammopathies in GD patients. We analyzed sera of 43 patients with GD for the presence of autoantibodies against 14 autoantigens. The results demonstrated a significant increase in the incidence of all autoantibodies tested, ranging from 11% for anti-RNP, pyruvate dehydrogenase (PDH), and DNA antibodies to 57% for rheumatoid factor. The autoantibodies were of all three isotypes, namely, IgG, IgM, and IgA. There was no correlation between the levels of immunoglobulins in the serum and the titer of autoantibodies found. Immunization of naive mice with a pool of purified anti-DNA antibodies form GD patients did not result in induction of experimental systemic lupus erythematosus (SLE), suggesting that they may represent natural autoantibodies that are not pathogenic. In conclusion, we found high titers of natural, polyspecific, nonpathogenic autoantibodies in the sera of GD patients.

Induction of high levels of anticardiolipin antibodies in mice by immunization with beta 2-glycoprotein I does not cause fetal death

OBJECTIVE

Our purpose was to determine whether anticardiolipin antibodies induced by immunization with beta 2-glycoprotein I cause fetal death in mice.

STUDY DESIGN

Female BALB/c mice were immunized with beta 2-glycoprotein I in a carbohydrate adjuvant or with carbohydrate adjuvant alone. The mice were mated with BALB/c males and killed on day 11 to 13 of pregnancy, and the fetal status was determined. Posttreatment blood samples were obtained for measurement of anticardiolipin and anti-beta 2-glycoprotein I antibodies and platelet counts.

RESULTS

Anticardiolipin and anti-beta 2-glycoprotein I antibodies developed in all mice immunized with beta 2-glycoprotein I. Fetal death occurred in 17 of 145 gestational sacs (12%) in 18 mice immunized with beta 2-glycoprotein I compared with 24 of 177 (14%) sacs in 21 control mice. There were no morphometric or histologic differences between gestational tissues, and platelet counts were similar for each group.

CONCLUSIONS

The induction of high levels of anticardiolipin antibodies in BALB/c mice by beta 2-glycoprotein I immunizations did not result in fetal death or thrombocytopenia. These nonpathogenic beta 2-glycoprotein I-induced anticardiolipin antibodies should prove useful in the characterization of clinically relevant epitopes for antiphospholipid syndrome.

Immunization with anti-neutrophil cytoplasmic antibody (ANCA) induces the production of mouse ANCA and perivascular lymphocyte infiltration

Wegener's granulomatosis (WG) is a granulomatous necrotizing vasculitis associated with the presence of ANCA, predominantly directed against proteinase 3 (PR3). The titres of ANCA correlate with disease activity and titre increases may precede disease exacerbations. Previously, we have shown that it is possible to induce autoimmune disease (systemic lupus erythematosus (SLE) and anti-phospholipid syndrome) in naive mice following active immunization with human autoantibodies, namely anti-DNA and anti-cardiolipin, respectively. The mice developed first anti-autoantibodies and, after about 4 months anti-anti-autoantibodies (Ab3), simulating auto-antibodies (Ab1) in their binding activities, and their presence was associated with the development of disease manifestations, characteristic of the human disease. So far, there is no good animal model for WG. In the current study we have immunized mice with human ANCA with the aim of inducing experimental WG. In two separate studies 30 mice were immunized in their footpads with autoantigen-purified IgG fraction (ANCA) from the sera of two patients with untreated WG, emulsified in Freund's complete adjuvant, followed 3 weeks later by ANCA injection in PBS. In the first experiment mice immunized with ANCA developed sterile microabscesses in the lungs after 8 months, and died after 8-15 months. In the second experiment, mice immunized with ANCA developed after 4 months mouse ANCA, with specificity both to PR3 and to myeloperoxidase, as well as anti-endothelial autoantibodies (AECA), as shown by radioimmunoprecipitation. Pathologically, the immunized mice developed proteinuria but not haematuria, and histological sections of the lungs demonstrated mononuclear perivascular infiltration, while diffuse granular deposition of immunoglobulins was noted in the kidneys. Our results point to a pathogenic role of ANCA in WG, and confirm the importance of the idiotypic network in the etiopathogenesis of autoimmune conditions.

Induction of tolerance to experimental anti-phospholipid syndrome (APS) by syngeneic bone marrow cell transplantation

Previously, we have shown the ability to induce experimental autoimmune conditions (e.g. SLE, APLS, Wegener's granulomatosis) following active immunization with the pathogenic autoantibody emulsified in adjuvant. The mice first develop anti-autoantibodies (Ab2: anti-id) and eventually generate anti-anti-autoantibodies (Ab3: anti-anti-id) which carry the same antigen binding characteristic as the autoantibody (Ab1). The appearance of the specific autoantibodies in mice sera was associated with the emergence of the compatible laboratory and with the clinical findings characteristic to the respective autoimmune disease. The effects of syngeneic bone marrow transplantation (BMT) on experimental anti-phospholipid syndrome (APS) were investigated. BALB/c mice were immunized with anti-cardiolipin monoclonal antibody (MoAb) named CAM and developed elevated serum titres of anti-phospholipid Abs accompanied by prolonged activated partial thromboplastin time, thrombocytopenia and a high percentage of fetal resorptions. These mice were then lethally irradiated and transfused with bone marrow (BM) cells (T cell depleted) from syngeneic naive mice. The titres of antiphospholipid antibodies were reduced in the recipients. The decrease in titre of autoantibodies was found to be related to depletion of antibody forming cells in vivo, associated with reduced proliferative response of lymph node cells to anti-cardiolipin MoAbs. The recipients showed improvement in clinical parameters following syngeneic BMT. The same recipients developed specific unresponsiveness to a second challenge with the anticardiolipin MoAb (CAM), but developed experimental systemic lupus erythematosus upon immunization with a monoclonal anti-DNA antibody. We conclude that acute myeloablative immunosuppression combined with syngeneic bone marrow transplantation may induce a state of tolerance to the pathogenic autoantibodies in mice with experimental APLS. Our results suggest that a similar approach may be useful in treating life-threatening autoimmune syndromes (e.g. catastrophic APLS) in clinical practice.

Demonstration of antiphospholipid antibody heterogeneity by phospholipid column chromatography and salt gradient elution techniques

Antiphospholipid (aPL) antibodies are associated with thrombosis, recurrent abortions and thrombocytopenia. Studies to determine the mechanisms of action of these antibodies have been hindered by their heterogeneity and limited availability of techniques to isolate and characterize subgroups of the antibodies. We report a new phospholipid affinity chromatography method which enables separation of antiphospholipid positive sera into more than one antibody subpopulation. Sera from five patients with complications of the antiphospholipid syndrome (APS) were studied. Each serum was applied to chromatography columns prepared by coating polystyrene beads (diameter 100 A) with phosphatidylserine (PS) or cardiolipin (CL). A linear salt gradient (0.03-1.0 M NaCl) was used for elution. Eluates were analyzed for phospholipid binding and for inhibition of the prothrombin-thrombin conversion reaction. Each sample yielded two to three peaks for CL and PS affinity columns. Molarities at which peaks were eluted differed between samples. For individual samples, molarities at which peaks were eluted differed between CL and PS columns. These data suggest that aPL antibodies are heterogenous, with differences existing between patients and even within single serum samples. Subpopulations differed in their avidities for CL and PS but generally all had prothrombinase inhibitory activity.

Early onset of autoimmunity in MRL/++ mice following immunization with beta 2 glycoprotein I

Antiphospholipid antibodies (aPL) are associated with thrombosis, thrombocytopenia and recurrent fetal loss in humans and in some animal models. Immunization with beta 2 glycoprotein I (beta 2GPI) induced aPL production in normal rabbits and mice. However, the association of these antibodies with disease manifestations remains controversial. To determine whether induction of aPL by beta 2GPI immunization in an autoimmune strain of mice (MRL/++) would result in acceleration of clinical and serological autoimmune disease manifestations, three groups of 8-week-old female mice were studied. One group was immunized with beta 2GPI, and one with ovalbumin (OVA); the third was not immunized. After two booster injections, sera were analysed for the presence of anticardiolipin (aCL) and anti-DNA by ELISA and anti-nuclear antibody (ANA) by immunofluorescence. Mice were studied for thrombocytopenia, proteinuria, fecundity rates, litter sizes and the development of central nervous system dysfunction. Elevated levels of aCL, anti-DNA and ANA were detected in all beta 2GPI-immunized, in three OVA-immunized, and in none of the unimmunized mice. The anti-DNA antibodies were inhibited by CL micelles, suggesting cross-reactivity between aCL and anti-DNA. Platelet counts, fecundity rates and litter size were reduced in beta 2GPI-immunized but not in OVA-immunized or unimmunized mice. None of the mice developed neurological dysfunction or significant proteinuria over a 10-week period post-immunization. These findings suggest that beta 2GPI immunization induces aPL in MRL/++ mice associated with accelerated autoimmune manifestations resembling the antiphospholipid syndrome.

Antiphospholipid syndrome and the idiotypic network

To study whether monoclonal anticardiolipin antibodies (aCL), derived from patients with antiphospholipid syndrome (APS), have similar pathogenic potential, we have employed an experimental model of antiphospholipid syndrome. Monoclonal aCL were produced by the combined method of EBV transformation and somatic cell hybridization of lymphocytes, derived from patients with APS. The monoclonal aCL were used to immunize mice at the footpads and the mice were followed for serological and clinical manifestations of APS. The monoclonal antibody EY2C9, was found to bind weakly to cardiolipin and other phospholipids (i.e. phosphatidyl-serine, phosphatidyl-ethanolamine and phosphatidyl-inositol). The antibody TM1B9, although derived from a patient with SLE and with secondary APS, did not react with phospholipids. Immunization of naive BALB/c mice with EY2C9 was followed by production of sustained high titers of antiphospholipid antibodies associated with prolonged activated partial thromboplastin time (APTT) (46.8 +/- 5.0 s vs. 22.4 +/- 1.7 s, in the non-immunized mice). Mice immunized with TM1B9 had a more moderate titer of antiphospholipid antibodies and did not show prolonged APTT. The pregnant mice, that were immunized with EY2C9, had increased fetal resorption rate (the equivalent of fetal loss in the human) of 36.8 +/- 10% (vs. 2 +/- 4% in mice immunized with TM1B9). Our results confirm that monoclonal aCL, derived from a patient with APS, can have a pathogenic potential, dysregulating the idiotypic network and leading to the development of characteristic signs of APS.(ABSTRACT TRUNCATED AT 250 WORDS)

Role of beta 2-glycoprotein I in the antiphospholipid antibody binding to endothelial cells

A group of anticardiolipin antibodies (aCL) require beta 2-glycoprotein I (beta 2GPI) to recognize their target, which might be located on endothelial cells (EC) and/or platelets. Following incubation with epithelial cells, 13 of 30 lupus sera retained EC-reactive antibodies of the IgG, IgA and IgM isotypes. Associated aCL and anti-phosphatidylethanolamine antibodies were partly absorbed on eC as well as EC. The former antibodies were more efficiently removed in the presence than in the absence of the latter. The presence of beta 2GPI in the affinity-purified aCL preparations may explain their binding to EC, as this cross-reaction was abrogated by the removal of the cofactor and restored by its re-introduction. Seventy four per cent of EC were faintly stained with polyclonal or monoclonal antibody directed to the cofactor. The beta 2GPI mediated aCL binding to EC membranes could this be influential in the development of thrombosis and/or thrombocytopenia in aCL-positive patients.

Cerebrovascular disease with antiphospholipid antibodies: immune mechanisms, significance, and therapeutic options

There has been a recent, dramatic surge in interest in antiphospholipid antibodies and associated clinical disorders, especially focal ischemic cerebrovascular disease. Antiphospholipid antibodies are a heterogeneous group of antibodies with varying specificities. Coagulation assays will detect lupus anticoagulants while enzyme-linked immunosorbent assays detect anticardiolipin antibodies. There are numerous potential links between antiphospholipid antibodies and coagulation disorders, including interaction of antiphospholipid antibodies and a cofactor, beta 2-glycoprotein I, which itself is involved in coagulation mechanisms. While the specific mechanism of antiphospholipid antibody-related coagulopathy is unknown, it is clear that antiphospholipid antibodies are associated with an immune-mediated prothrombotic state. Patients with the highest titers of IgG antiphospholipid antibodies have a relatively high risk of recurrent thrombotic events, especially stroke, deep venous thrombosis, and spontaneous abortion. Because of limited controlled, prospective data, current therapy remains empiric and directed at coagulation mechanisms, immune mechanisms, or both.

Nervous system lupus: pathogenesis and rationale for therapy

Several different pathogenic mechanisms appear to be involved in CNS lupus. These include: B-cell/autoantibody-mediated nervous system compromise; immune complex deposition and vasculitis; microthrombosis and vasculopathy; aberrant MHC Class II antigen expression with T-cell mediated disease (multiple-sclerosis model); and, cytokine-induced brain inflammation. These processes are not mutually exclusive: there exist in vitro and in vivo models for each of these. A number of autoantibodies, especially those with specificities for shared neuronal/lymphocyte antigens, are associated with certain forms of cognitive dysfunction or overt nervous system manifestations. In MRL/lpr mice, lymphoid infiltrates in the brain parenchyma are related to a neurobehavioural dysfunction which develops very early in the course of autoimmune disease. Recent results, both in animal models and in human studies on the therapeutic effects of corticosteroids, immunosuppressive drugs or anticoagulants on clinical and subclinical manifestations of CNS lupus are highlighted in an attempt to develop a rationale for intervention based upon presumed pathogenesis.

Immunoglobulin gene sequence analysis of anti-cardiolipin and anti-cardiolipin idiotype (H3) human monoclonal antibodies

Heavy and light chain variable region nucleotide sequences were derived from 6 human hybridoma antibodies which bear characteristics of antibodies associated with the phospholipid antibody syndrome. All antibodies originated from non-autoimmune individuals and were polyspecific. Four of these reacted with cardiolipin (and other antigens) and three carried the H3 idiotype which is expressed on a high percentage of disease-associated anti-cardiolipin antibodies. This idiotype was localized to the lambda light chain of the H3 monoclonal antibody and found on two other antibodies which like H3 expressed V lambda 4 or the related V lambda 3 subgroup light chains. The H3 idiotype however did not define these subgroups nor was it required or sufficient for anti-cardiolipin activity. Anti-cardiolipin binding was found in VH1, VH3 and VH4 heavy chain families and in a V kappa 1 light chain. The D region was diverse in both length and gene usage. Although all cardiolipin binding antibodies showed little deviation from germline variable (V) gene sequences, where mutations occurred they tended to be replacement mutations and clustered in complementarity determining regions (CDR) suggesting these B cells were derived from antigen-driven responses. These results from our panel of hybridomas and their comparison to other human antibodies provide extensive information on the diversity of genetic elements which can be used by cardiolipin-binding antibodies. We also show gene sequences which encode the disease-associated H3 idiotype and its location on lambda light chains, which imply that some labda light chains may be preferentially utilized in auto-reactive hybridomas.

Transfer of experimental antiphospholipid syndrome by bone marrow cell transplantation. The importance of the T cell

OBJECTIVE

To investigate the potential of bone marrow cells from mice with primary antiphospholipid syndrome (APS) to transfer the disease to naive mice, and to determine the importance of the role of T cells in the APS.

METHODS

Experimental primary APS was induced in naive mice following active immunization with anticardiolipin (aCL) monoclonal antibody (MAb). Whole-population or T cell-depleted bone marrow cells from mice with experimental primary APS were infused into total body-irradiated naive BALB/c recipients.

RESULTS

Bone marrow cells (in the presence of T cells) had the potential to induce experimental APS in naive mice, which resulted in high serum titers of aCL, antiphosphatidylserine, and antiphosphatidylinositol antibodies; an increased number of antibody-forming cells specific for each of the above phospholipids; a positive lymph node cell proliferative response to aCL MAb; and clinical features of primary APS, including thrombocytopenia, prolonged activated partial thromboplastin time (indicating the presence of lupus anticoagulant), and a high frequency of fetal resorptions (the equivalent of human fetal loss). T cell-depleted bone marrow cells did not transfer the disease.

CONCLUSION

This study demonstrates the important role of T cells in the development and transfer of experimental primary APS and raises the possibility of T cell manipulations in treatments to prevent this condition.

Antiphospholipid antibodies and the antiphospholipid syndrome

The study of aPL antibodies and the APS has appealed to large numbers of investigators over the last decade. This accounts, in part, for the great degree of apparently contradictory data being published. What seems certain is that these antibodies are associated with thrombosis and recurrent pregnancy loss, and animal data suggest a direct role in pathogenesis. In vitro studies demonstrate that these antibodies have a variety of functional effects on the hemostatic system, giving further credence to an antibody role in thrombosis. Further studies will doubtless give better insight into the mechanisms of antibody action and this will provide a more rational basis for treatment.

Anticardiolipin antibodies in infections are heterogenous in their dependency on beta 2-glycoprotein I: analysis of anticardiolipin antibodies in leprosy

We studied the effect of beta 2-GPI on binding of antibodies in sera from patients with leprosy and patients with the antiphospholipid syndrome (APS) to CL in enzyme-linked immunosorbent assays (ELISAs). Increased levels of IgG aCL were detected in 59 of 61 leprosy patients' sera by the standard aCL-ELISA in the presence of bovine beta 2-GPI and in 60 of the 61 leprosy patients' sera by the modified aCL-ELISA without beta 2-GPI. When tested by both aCL-ELISAs on the same plate, 10/31 leprosy sera and 9/10 APS sera bound better in the standard aCL-ELISA, 16/31 leprosy sera bound better in the modified aCL-ELISA and in five leprosy and one APS sera the difference was not significant. A dose-dependent enhancing effect of beta 2-GPI on the leprosy and APS sera binding to CL was confirmed using purified human beta 2-GPI. Enhanced binding was seen if beta 2-GPI was added either before or together with the test serum. In 11/61 leprosy sera increased levels of IgG antibodies against beta 2-GPI were found by ELISA. Leprosy anti-beta 2-GPI antibodies appear to be a separate antibody population recognizing only beta 2-GPI adsorbed on the ELISA plate. These results demonstrate heterogeneity of leprosy aCL with respect to their beta 2-GPI requirement for binding to CL.

Effect of long-acting thromboxane receptor antagonist (BMS 180,291) on experimental antiphospholipid syndrome

Antiphospholipid syndrome (APS) is characterized by recurrent thromboembolic phenomena, recurrent fetal loss and thrombocytopenia associated with high titers of IgG anticardiolipin antibodies and/or lupus anticoagulant. There is an increased platelet aggregation in these patients and thus aspirin was found to be effective in abrogating some of the clinical findings. The purpose of this study was to employ the experimentally induced APS in mice infused with anticardiolipin antibodies, to study the effect of a thromboxane receptor antagonist (BMS, 180, 291) on the various overt manifestations of APS. Experimental APS was induced in pregnant female mice by iv infusion of a pathogenic anticardiolipin antibody (CAM). The mice were then treated daily with 300 micrograms/mouse of BMS. The study group and the untreated group were killed on day 17 of pregnancy. Live and absorbed fetuses and the mean weight of the placentae, fetuses and platelet counts were recorded. BMS treated mice had a significant reduction in fetal resorption rate from 45% to 19.8% and an increase in mean placental and embryo weights (182 vs 104, 1043 vs 721 mg, respectively). In parallel, an increase in platelet count (from 597,100 to 1075,000 platelets/mm3) and decrease in activated thromboplastin time (95 to 44s) was seen. It seems that thromboxane receptor antagonist may be effective in abrogating the diverse manifestations seen in APLS. Increased platelet aggregation may be one of the pathogenetic mechanisms in APS.

Primary antiphospholipid syndrome and self-limited renal vasculitis during pregnancy: case report and review of the literature

A case of self-limited renal vasculitis during pregnancy in a patient with a history of recurrent fetal loss and with a positive cardiolipin immunoglobulin G antibody test is described. The renal disease manifested as an acute renal failure in the second trimester of the patient's third pregnancy, concurrent with severe pre-eclampsia. Vasculitis is a rare manifestation of antiphospholipid syndrome that has been described mostly in peripheral arteries. Renal vasculitis, however, has not yet been reported in association with this syndrome. The full spectrum of renal involvement in antiphospholipid syndrome is presently being determined, and we suggest that renal vasculitis be included in that spectrum.

Anticardiolipin antibodies in an unselected stroke population

A pathogenetic role in thrombotic disease, particularly in young people, has been postulated for anticardiolipin antibody (ACA). We have carried out a prospective controlled study of 262 unselected patients with acute stroke and 226 controls to assess the prevalence and relation to age and vascular risk factors of ACA. Titres of IgG, IgA, or IgM ACA were above the upper normal limit in 38% of patients. The proportions of patients and controls with raised titres did not differ significantly (13 vs 8% IgG, 22 vs 29% IgA, 11 vs 7% IgM). IgG titres were higher among patients than among controls (mean 3.88 vs 2.86 u/mL [95% CI for difference 0.62-0.87], p = 0.0004), whereas IgA and IgM titres were lower in patients than in controls (IgA 4.82 vs 5.98 u/mL [1.12-1.60], p = 0.01; IgM 3.00 vs 3.64 u/mL [1.01-1.45], p = 0.04). However, within age tertiles the only significant difference between patients and controls for IgG ACA was in the oldest tertile. Analysis by number of risk factors for stroke showed a significant difference between the groups only for subjects with one risk factor. IgA and IgM ACA titres were higher among controls only in those with no vascular risk factors. We found no evidence to support the hypothesis that ACA is an independent risk factor for stroke in young people. The increase in IgG titre with age and number of vascular risk factors in stroke patients suggests that ACA may be a non-specific accompaniment of vascular disease. Routine testing for ACA in stroke patients is not justified.

Pathogenic natural anti-cardiolipin antibodies: the experience from monoclonal gammopathy

Anti-cardiolipin antibodies (ACA) were detected in 19% of sera from patients with monoclonal gammopathies (MG). ACA were purified from the sera of patients with MG. One of the IgG-ACA was found to be monospecific with high affinity for cardiolipin, and to carry a pathogenic ACA Id (1.10). Active immunization of naive BALB/c mice with the purified IgG-ACA was followed by production in the mice of sustained high titres of ACA, associated with prolonged activated partial thromboplastin time (APTT) (61 +/- 14s versus 31 +/- 2s in control mice; P < 0.001) and thrombocytopenia (468,000 +/- 224,000/mm3 versus 994,000 +/- 92,000/mm3 in controls; P < 0.001). The titres of other autoantibodies (e.g. anti-DNA, anti-histones), although being high after immunization, decreased rapidly and were undetected after 1 month following the boost injection. The mice immunized with the IgG-ACA exhibited low fecundity (36% of mice became pregnant versus 62% observed in the group immunized with control IgG). The pregnant mice had increased resorption rate (the equivalent of fetal loss in the human) of 52 +/- 8% (versus 5 +/- 4% in the control group). The mean (+/- s.d.) embryo and placental weights in mice with anti-phospholipid syndrome (APLS) were significantly lower compared with the mice injected with control IgG (682 +/- 304 mg and 102 +/- 12 mg versus 1303 +/- 105 mg and 145 +/- 8 mg, respectively; P < 0.001). Serum monoclonal immunoglobulins having autoantibody activity may be regarded as an expansion of clones producing natural autoantibodies. Our results confirm the pathogenic role of natural ACA in the pathogenesis of the anti-phospholipid syndrome.

Antiphospholipid antibodies in an in vivo thrombosis model in mice

The mouse model described in this study offers a unique method of determining the characteristics and mechanism(s) of action of aCL antibodies in thrombosis in vivo. In addition, this animal model enables the study of the kinetics of formation and dissolution of thrombus, as well as clot area, to be studied in a dynamic fashion. Other models for evaluation of thrombus formation rely on measurements of thrombus size and weight in ligated vessel segments where flow may be interrupted artificially. In addition, two important findings can be extracted from the study. (1) The size of the thrombi were significantly larger in mice that were passively immunized with IgG-APS (four patient samples examined) and with IgM-APS (two patient samples examined) compared with mice injected with saline or with immunoglobulin from control patients. (2) The clot persisted significantly for longer periods of time (total time) in animals injected with IgG-APS or IgM-APS when compared with control animals. Based on in vitro experiments, it is possible that these antibodies may inhibit protein C activation, neutralize the inhibitory activity action of beta 2 glycoprotein I (beta 2GPI), or activate platelets at the site of the femoral vein injury. Because this model enables to study the dynamics of thrombus formation, it is possible that these hypotheses and other mechanisms by which aPL antibodies are thrombogenic be investigated. Future studies will also include the effects of different levels of antibodies, as well as effects of affinity purified and monoclonal aPL antibodies on thrombus formation.

Role of IL-3 in the antiphospholipid syndrome

Antiphospholipid syndrome (APLS) is characterized by recurrent thromboembolic phenomena, thrombocytopenia and fetal loss. We describe various methods of induction of experimental APLS. These models were employed to study a variety of therapeutic agents including low dose aspirin, low molecular weight heparin, IVIG and thromboxane receptor antagonist. Because interleukin-3 (IL-3) is a multilineage cytokine affecting also megakaryocytes, is regarded as a 'good' cytokine in various stages of pregnancy and as low levels of IL-3 were recorded in APLS, it was logical to employ IL-3 as a therapy for APLS. Indeed, this treatment completely abrogated all the manifestations of experimental APLS. Furthermore, it was found that low dose aspirin most probably affect positively APLS via inducing an increased production of IL-3 by monocytes.

A follow-up study of immunoglobulin levels and autoantibodies in an unselected pregnant population

PROBLEM

To obtain a systematic view of changes in the levels of immunoglobulins (Igs), other serum proteins, and autoantibodies during pregnancy and postpartum.

METHOD

A series of 220 women were followed throughout pregnancy and four to six months postpartum.

RESULTS

Immunoglobulin G (IgG) concentration clearly decreased toward term. The concentrations of IgM and IgA decreased only slightly. In most instances autoantibody levels paralleled changes in the corresponding immunoglobulin class levels. In few cases, however, there were clear deviations from this. With respect to IgG class autoantibodies, the highest autoantibody levels were found in the postpartum specimen. IgM class autoantibody levels remained almost constant throughout the follow-up period.

CONCLUSIONS

The intrapregnancy decrease of IgG is mainly due to hemodilution, but when the effect of hemodilution is taken into account, total amounts of IgM and IgA are increased. The results suggests that compared to IgG the regulatory mechanisms of IgM and IgA are altered during pregnancy.

Thrombosis and antiphospholipid syndrome: a preliminary assessment of three antithrombotic treatments

To assess retrospectively three antithrombotic treatments in the secondary prevention of thrombosis in the antiphospholipid syndrome (APS), 23 patients (six systemic lupus erythematosus, seven lupus-like disease and 10 primary antiphospholipid syndrome) were included in this study. Treatments assessed were: (1) aspirin 75 mg daily, (2) warfarin (international normalised ratios (INRs) 2.0-2.9) +/- aspirin 75 mg daily, and (3) warfarin (INRs > 2.9) +/- aspirin 75 mg daily. Where patients had received two or three of these treatments successively, the periods of time on each treatment were added and the number of patients with recurrence(s) on each treatment were compared by Fisher's exact probability test. 'High' anticoagulation (INRs > 2.9) +/- aspirin 75 mg daily was more effective than aspirin 75 mg daily, there was a trend in favour of 'high' anticoagulation (P = 0.066). No statistically significant difference could be demonstrated when comparing 'low' anticoagulation +/- aspirin 75 mg daily with aspirin 75 mg daily (P = 0.092). These results suggest that aggressive anticoagulation with or without low-dose aspirin is effective in preventing further thromboembolic events in APS.

Pathogenic serum IgG anticardiolipin antibodies and the idiotypic network

OBJECTIVES

To determine whether active immunisation of mice with pathogenic anticardiolipin antibodies (IgG and IgM), derived from the serum of a patient with the antiphospholipid syndrome, could dysregulate the idiotypic cascade and induce the production of anti-anti-anti-cardiolipin (Ab3) with anticardiolipin activity by the mice with the association of overt antiphospholipid syndrome.

METHODS

Anticardiolipin antibodies were purified from the serum of a patient with the antiphospholipid syndrome. The purified anticardiolipin antibodies were used to immunise mice at the footpads and the mice were then followed up for serological and clinical manifestations of the antiphospholipid syndrome.

RESULTS

The IgG anticardiolipin antibody was found to be monospecific and to bind cardiolipin with high affinity. Immunisation of naive BALB/c mice with the purified IgG anticardiolipin antibody was followed by production in the mice of sustained high titres of IgG anticardiolipin antibody, associated with a prolonged activated partial thromboplastin time (64.5 (9.7) v 30.1 (1.7) seconds in control mice) and thrombocytopenia (0.4 (0.06) x 10(9) v 1.0 (0.09) x 10(9)/l platelets in controls). The titres of other autoantibodies (for example, antibodies to DNA, histone), though high after the immunisation, decreased rapidly and were almost undetected one month after the boost injection. The mice immunised with the IgG anticardiolipin antibody showed low fecundity (36% of mice became pregnant v 62% in the group immunised with control IgG). The pregnant mice had an increased resorption rate (the equivalent of fetal loss in the human) of 61 (9)% v 5 (4)% in the control group. The mean (SD) embryo and placental weights in mice with the antiphospholipid syndrome were significantly lower than in the mice injected with control IgG (641 (210) and 103 (14) mg v 1303 (105) and 145 (8) mg respectively. The IgM anticardiolipin antibodies purified from the same patient were found to be polyspecific, binding with low affinity to anticardiolipin antibodies and double stranded DNA, and carried the anti-DNA idiotype 16/6. Mice immunised with the purified IgM anticardiolipin antibodies, though showing reduced fecundity (30%), had only a slightly increased resorption rate (12 (9) v 3 (5)% in controls) and only a slight and statistically non-significant decrease in mean (SD) embryo and placental weights (1134 (188) and 136 (11) mg respectively).

CONCLUSIONS

The results confirm the induction of pathogenic anticardiolipin antibodies by immunisation with serum anticardiolipin, dysregulating the idiotypic network, and point to the higher pathogenic potential of serum IgG v IgM anticardiolipin antibodies.

Prevention of fetal loss in experimental antiphospholipid syndrome by low-molecular-weight heparin

OBJECTIVE

The purpose of this study was to compare the effectiveness of low-molecular-weight heparin with regular heparin in the prevention of fetal resorption in mice with the antiphospholipid syndrome.

STUDY DESIGN

Antiphospholipid syndrome was passively induced in ICR mice by injecting them with anticardiolipin antibodies on the first day of pregnancy. Subsequently, these mice were treated with low-molecular-weight heparin in two different doses, with regular heparin, and with a placebo. On gestational day 17 the mice were killed by cervical dislocation, and the pregnancy outcome was evaluated. Statistical analysis was performed by means of a one-way analysis of variance using Bonferroni's t test.

RESULTS

Treatment with low-molecular-weight heparin resulted in a resorption rate of 22.4% as opposed to 41.4% in mice with antiphospholipid syndrome that were given regular heparin and 51.7% in nontreated controls.

CONCLUSION

We conclude that low-molecular-weight heparin can prevent fetal resorptions in mice with antiphospholipid syndrome.

Induction of phospholipid-binding antibodies in mice and rabbits by immunization with human beta 2 glycoprotein 1 or anticardiolipin antibodies alone

Anticardiolipin (aCL) antibodies are autoantibodies present in high concentrations in patients with the antiphospholipid syndrome (APS), a disorder of recurrent thrombosis and pregnancy loss. What induces aCL antibodies is uncertain, but a recent report suggested that immunization of mice with beta 2 glycoprotein 1 (beta 2 GP1) in Freund's complete adjuvant (FCA) resulted in aCL antibody production in the recipient mice. Since this observation might explain how autoantibodies might be induced by poor immunogens, such as phospholipids, we decided to explore the question further. In our first series of experiments, we found that aCL antibodies were induced in mice by beta 2GP1 mixed with adjuvants that did not contain lipids (Adju-Prime or aluminium hydroxide). This excluded the possibility that antibody induction occurred because beta 2GP1 formed complexes with lipids in FCA. We also found that aCL antibodies always appeared before anti-beta 2GP1 antibodies, excluding the possibility that aCL antibodies were directed to beta 2GP1 or were induced by formation of anti-idiotypic antibodies (to anti-beta 2GP1). In experiments, we found that immunization of mice with human IgG antibodies from patients with the APS (IgG-APS), also induced aCL antibodies. Immunization with pure bovine serum albumin (BSA) did not induce aCL antibodies. We propose that aCL antibodies are induced by proteins with high avidity for phospholipids. These proteins may be bound to phospholipids when introduced, or may bind circulating phospholipids, so transforming phospholipid molecules into immunogens. Similar mechanisms might explain autoantibody induction to other poor immunogens.

Detection of anti-phospholipid and anti-DNA antibodies and their idiotypes in newborns of mothers with anti-phospholipid syndrome and SLE

The titers, isotypes and idiotypes of antiphospholipid and anti-dsDNA antibodies were determined in seven pairs of mothers with antiphospholipid syndrome (APLS) and their offspring, in 11 pairs of SLE mothers and their matched infants and in seven respective pairs of healthy subjects. In addition, maternal as well as fetal sera were evaluated for the presence of anti-SSA (Ro), anti-SSB (La) and anti-70 kd RNP autoantibodies. In the sera from APLS patients, as well as in the sera from their offspring, the mean antibody titer of IgG aCL was found to be significantly higher then the corresponding value in the control group (P < 0.01). Highly significant increased titers of IgG anti-DNA antibodies were found in the sera of SLE mothers and their matched offspring (P < 0.0008). The prevalence of anti-SSA, anti-SSA, and anti-70Kd RNP antibodies was lower then that of antiphospholipid and anti-dsDNA antibodies. Only one of the respective offspring had increased levels of these antibodies. The quantity of maternal antibodies transferred to the fetus was depended on their concentration in the maternal circulation, as well as on their type and specificity. Follow-up of newborn sera showed a progressive decrease in the antiphospholipid antibody titers during 3 months. After 6 months it was undetected. Our results point to a transplacental transfer of aCL and anti-DNA antibodies, a phenomenon which is not necessarily associated with respective clinical manifestations, in contrast to the classical humoral mediated autoimmune diseases (e.g. myasthenia gravis).

Hepatic infarction in a pregnant patient with the 'primary' antiphospholipid syndrome

We describe a patient with previous venous thrombosis while using oral contraceptives and recurrent pregnancy loss, who presented with massive hepatic infarction in the last trimester of the fourth gestation. Thrombocytopenia, the lupus anticoagulant (LA) and the anticardiolipin antibody (aCL) were detected and a diagnosis of a 'primary' antiphospholipid syndrome (APS) was made. The clinical and histological manifestations and the differential diagnosis, especially with DIC and pre-eclampsia, are discussed.

Anti-cardiolipin antibodies induce pregnancy failure by impairing embryonic implantation

The antiphospholipid syndrome is characterized by thrombocytopenia, thrombosis, and recurrent fetal loss in association with anti-cardiolipin antibodies (ACAs) or lupus anti-coagulants. However, the causal role of these antibodies in the disease and the mechanisms by which the ACA may induce the syndrome are not clear. Recently, we have established an experimental mouse antiphospholipid syndrome induced by the mouse IgM monoclonal ACA designated 2C4C2. In the present study, we focused on the effects of immunization with the monoclonal ACA 2C4C2 on the outcome of pregnancies in BALB/c female mice. Four weeks after active immunization with the monoclonal ACA, a severe gestational failure with low pregnancy rates, low number of fetuses, and a high rate of resorptions was observed. Moreover, embryos obtained from the ACA-immunized females on day 3.5 of pregnancy were severely impaired, demonstrating developmental delay and abnormal morphology. These abnormal embryos failed also to develop in an in vitro implantation model. Furthermore, specific binding of the 2C4C2 ACA to the trophectoderm cell lineage of in vitro implanting normal embryos was observed. Thus, our studies demonstrate that the severe ACA-induced gestational failure results from an impairment of implantation and suggest that the ACA may react directly with the preimplantation embryos.

In vitro effect of anticardiolipin autoantibodies upon total and pulsatile placental hCG secretion during early pregnancy

BACKGROUND

The anticardiolipin syndrome is characterized among other features by recurrent thromboembolic events, thrombocytopenia, and recurrent fetal loss associated with high IgG titers of anticardiolipin antibodies and/or the presence of lupus anticoagulant.

AIMS

The mechanisms for the fetal loss in this syndrome have not yet been clearly elucidated, although several hypothesis based on experimental data have been put forward. We wanted to evaluated the effect in vitro of anticardiolipin antibodies on the secretion of human chorionic gonadotropin.

METHODS

Employing our previous experience with placental explants, we studied the effect of several mouse monoclonal and human polyclonal purified anticardiolipin antibodies (ACA), which were shown by us to induce experimental antiphospholipid syndrome (APLS), to affect the pulsatile secretion of beta human choriogonadotropin.

RESULTS

The mouse monoclonal ACA antibodies caused an increase in the pulsatility of beta human choriogonadotropin, while human polyclonal ACA derived from patients with ACA had an inhibitory effect.

CONCLUSIONS

These studies with placental explants show that ACA may have an additional effect on placental hormone secretion and thus affect the fate of the embryo.

The effect of aspirin on recurrent fetal loss in experimental antiphospholipid syndrome

PURPOSE

To evaluate the effect of aspirin treatment upon fetal loss in mice with experimental antiphospholipid syndrome (APLS).

MATERIALS AND METHODS

Experimental APLS was induced in pregnant mice by passive transfer of mouse monoclonal anticardiolipin antibody. The mice were treated with high (100 micrograms/d) or low (10 micrograms/d) dose of aspirin, using vitamin C (100 micrograms/d or 10 micrograms/d) as a control. The mice were assessed for the presence of lupus anticoagulants (prolonged aPTT), thrombocytopenia, degree of fetal resorption rate and mean embryo and placental weights.

RESULTS

The mice with APLS had a higher fetal resorption rate (45.7 +/- 12.2% vs 2.5 +/- 0.4%, P < 0.001), reduced placenta mean weight (104 +/- 8 mg vs 169 +/- 7 mg, P < 0.001), prolonged aPTT (94 +/- 14 sec vs 39 +/- 4 sec, P < 0.001), and reduced mean platelet count (597 +/- 186 x 10(3)/mm3 vs 847 +/- 51 x 10(3)/mm3, P < 0.001). The group of mice with APLS, who were treated with low-dose aspirin, had a lower resorption rate (11.1 +/- 9.3% vs 45.7 +/- 12.2%, P < 0.001), a higher placenta mean weight (178 +/- 8 mg vs 104 +/- 8 mg, P < 0.001), a higher mean embryo weight (1042 +/- 134 mg vs 721 +/- 91 mg, P < 0.001), and a lower aPTT (58 +/- 15 sec vs 94 +/- 14 sec, P < 0.001). Mice who were treated with high-dose aspirin also had a lower resorption rate, although not as much as in the low-dose aspirin group (34.2 +/- 12.7% vs 45.7 +/- 12.2%, P < 0.001).

CONCLUSION

Aspirin, especially in low dose, has a protective effect against obstetrical complications associated with experimental APLS.

Prevention of fetal loss in experimental antiphospholipid syndrome by in vivo administration of recombinant interleukin-3

Antiphospholipid antibodies are strongly associated with arterial and venous thrombosis and with fetal loss. Recently an experimental model for antiphospholipid syndrome (APLS) was established in our laboratory. In this model, mice are immunized passively or actively with anticardiolipin antibodies and acquire the syndrome, which is characterized by prolonged activated partial thromboplastin time (APTT), thrombocytopenia, low fecundity rate, and fetal loss. In a normal process of pregnancy, lymphokines affect fetal implantation and development. Cytokines from the colony stimulating factor family, like GM-CSF and IL-3, were shown to be positive signals for implantation and to promote placental development and fetal growth. Given our preliminary findings of low IL-3 in mice with APLS and the efficacy of IL-3 in preventing fetal loss in a strain of mice prone to fetal resorption, our aim in the present study was to examine the effect of murine recombinant IL-3 (mrIL-3) on pregnant mice induced with experimental APLS. Mice were passively transfused to the tail vein, 24 h following mating, with anticardiolipin antibodies. The mice were divided into two groups: one group was injected intraperitoneally with mrIL-3 on days 6.5, 8.5, and 10.5 after mating, while the control group was injected with PBS. When the mice were killed on day 15 of pregnancy a 32% +/- 4.2 resorption rate was observed in the anti-cardiolipin-immunized group, which was reduced to 4% +/- 0.3 following treatment with mrIL-3. The thrombocytopenia associated with the experimental APLS was also corrected following lymphokine administration. IL-3 may be effective in prevention of recurrent fetal loss in APLS.