Apoptosis and antiphospholipid antibodies

Apoptosis and Antiphospholipid Antibodies Valerio Pittoni and David Isenberg Objective: To analyze the potential links between antiphospholipid antibodies (aPL) and apoptosis in the pathogenesis of the antiphospholipid antibody syndrome (APS). Methods: A review was undertaken of the most relevant scientific literature on apoptosis and autoimmune phenomena. Experimental and human pathology were reviewed to substantiate the hypothesis that apoptosis is involved in the generation of aPL. Results: Several considerations suggest that exposure of phospholipids (PL) during apoptosis may be a driving antigenic stimulus to the production of aPL. Furthermore, the molecular PL-protein complexes formed during apoptosis are targeted by "pathogenic" aPL. The binding and the clearance of apoptotic ceils by these autoantibodies likely further enhances the aPL immune response. Experimental models and human pathology suggest that a restricted genetic background is key to the development of this immune response. Conclusions: Abnormalities of apoptosis observed in the course of autoimmune conditions likely provide an antigenic stimulus to the production of aPL. Semin Arthritis Rheum 28:163-178. Copyright © 1998 by W.B. Saunders Company

INDEX WORDS: Apoptosis; phospholipid; autoantibodies, ABBREVIATIONS

aPL: antiphospholipid antibodies ALPS: Autoimmune Lymphoproliferative Syndrome APS: Antiphospholipid Antibody Syndrome [32-GPI: [32-glycoprotein I CL: cardiolipin LDL: low-density lipoproteins oxLDL: oxidized low-density lipoproteins PCD: programmed cell death PE: phosphatidylethanolamine PL: phospholipids PS: phosphatidylserine ROS: reactive oxygen species Sph: sphingomyelin SLE: systemic lupus erythematosus LTHOUGH the precise mechanisms by which autoantibodies recognizing protein structures are generated are still unclear, even less is known about the production of autoantibodies directed against lipid components. Of particular interest is the fact that autoantibodies originally thought to be directed exclusively against negatively charged phospholipids (PL), a specific subset of lipids playing a crucial role in the coagulation cascade, are associated with the antiphospholipid syndrome (APS), which is characterized by thrombotic events,

A

recurrent miscarriage, and thrombocytopenia (1, 2). In the last few years, convincing evidence has been provided that antiphospholipid antibodies (aPL) represent a heterogeneous population of autoantibodies, and most attention has been paid to the involvement of PL-binding proteins such as [32-glycoprotein I ([32-GPI), prothrombin, and possibly others as the antigens or coantigens for aPL (3, 4). In support of this theory is the observation that [32-GPI seems to be a more specific antigen for APS (5), whereas aPL recognizing PL molecules occur in both APS and other conditions not associated with a thrombotic risk (6). However, the high

From the Centre for Rheumatology, Bloomsbury Rheumatology Unit, Department of Medicine, University College of London, London, UK. Valerio Pittoni, MD: Visiting Research Fellow; David Isenberg, MD, FRCP: Arthritis Research Campaign's Diamond Jubilee Professor of Rheumatology. Supported by Universit# degli Studi di Roma, "La Sapienza" Rome, Italy (V.R). Address reprint requests' to David Isenberg, MD, FRCP, Centre for Rheumatology Research, Bloomsbury Rheumatology Unit, Arthur Stanley House, 40-50 Tottenham St, London W1P 9PG. Copyright © 1998 by W.B. Saunders Company 0049-0172/98/2803-000358. 00/0

Seminars in Arthritis and Rheumatism, Vo128, N o 3 (December), 1998: pp 163-178

163

164

density of antigen (7), and the conformational change of [32-GPI (8) observed in vitro after interaction with PL, suggest that a crucial generation of new [32-GPI epitopes may also occur in vivo as a result of this interaction. Casciola-Rosen et al (9) proposed that programmed cell death (PCD), also known as apoptosis, is one of the mechanisms responsible for the generation of aPL (9). The authors support the hypothesis that non-bilayer PL structures and associated PL-binding proteins, such as [~2-GPI and annexin V, are exposed to other autoantigens on the surface of cells undergoing apoptosis. These clusters of autoantigens, including DNA and nucleoproreins, are targeted in systemic lupus erythematosus (SLE), and the immune response against these molecular complexes is often characterized by epitope spreading. The exposure of phosphatidylserine (PS), and the additional procoagulant properties expressed by apoptotic cells, may also contribute to the thrombotic diathesis observed in the APS inasmuch as aPL were found to interact with membrane PS. The imbalance of coagulation homeostasis resulting from this interaction could account for the thrombotic events (9). PROGRAMMED CELL DEATH AND REGULATION OF THE IMMUNE RESPONSES: PHYSIOLOGICAL ROLE OF PHOSPHOLIPIDS

PCD seems to play a pivotal role in the negative selection of autoreactive cells and in the maintenance of homeostasis in the lymphoid system (10). This general mechanism of cell disruption has been recognized as a distinct process characterized by sequential predetermined events leading to the self-elimination of the senescent cells of the organism (11). The genetic and biochemical machinery underlying this cell "suicide" has only been partially clarified (12). In most conditions, PCD occurs with characteristic changes of cell morphology known as "apoptosis." This process takes place through a sequential involvement of cell compartments, including the plasma membrane, cytoplasm, and nucleus. "Ruffling" and "blebbing" of the cell membrane, shrinkage of the cytoplasm with no associated morphological changes of cytoplasmic organelles, and nuclear and chromatin condensation leading to the formation of "dense bodies" are the main morphological cell changes occurring during apoptosis, regardless of the initiating stimulus (13-15).

PITTONI AND ISENBERG

It is now widely accepted that the sudden loss of the normal PL distribution in plasma membrane and the consequent exposure of PS, a negatively charged PL, is a specific and early sign of the cell commitment to apoptosis. Unlike necrosis, one of the hallmarks of apoptosis is a subtle progression of events meant to eliminate the cell without release of its internal components, thus preventing the development of an inflammatory reaction. Phagocytosis of apoptotic cells or bodies in solid tissue presumably protects the surrounding tissue from the damaging effects of released intracellular contents. In fact, the change of the membrane physiological PL asymmetry and the exposure of PS is one of the signals that makes phagocytic cells capable of recognizing a cell undergoing apoptosis (16). The investigations on the molecular sequences of events leading to this ordered "cell suicide" have highlighted the fundamental role of two genes that are also primarily involved in the regulation of immune response homeostasis. The product of Fas gene (CD95) interacts with its specific tigand and thereby triggers an intracellular signal leading to cell activation or PCD, depending on the specific status of the cell or other concomitant stimuli (18). Bcl-2 belongs to a family of genes capable of ensuring cell survival fundamentally through a negative control on apoptotic mechanisms (18). Even if apoptotic death occurs as a consequence of "external" harmful stimuli, this mechanism seems to be of vital importance in those systems characterized by highly selective physiological regulation of cell growth. The maturation of both B and T compartments of the immune system offers a remarkable example of the importance of such a mechanism. In thymic as well as bone marrow maturation, a hugh negative clonal selection takes place to eliminate high-affinity autoreactive cells. Morphological and biochemical studies have confirmed that apoptosis, probably caused by an immature signal transduction, is the preferential means by which the thymic and bone marrow microenvironments carry out this selection (19, 20). Nevertheless, in both T and B cell compartments, apoptotic death represents a mechanism of selection not only in the early development of lymphocytes but also in the latter stages, after the interaction with the antigen. Indeed, in the periphery, it is of vital importance to control the proliferation of activated T lymphocytes to avoid the accumulation of high

APOPTOSIS AND ANTIPHOSPHOLIPIDANTIBODIES

levels of cytokines, which are potentially harmful to the organism (21, 22). Some evidence suggests that B cell homeostasis is also maintained by apoptosis induction of activated cells (23, 24). However, no matter what the initiating stimulus, the exposure of PS on the cell membrane represents a general signal, operating in the immune system to "label" a cell committed to the apoptotic process. At this very early stage of commitment, the fundamental functional properties of the cell membrane, such as the exclusion of viability dyes, are maintained (25). During thymic selection, apoptosis may be not only the general mechanism by which high-affinity autoreactive clones are deleted, but also an efficient way to make the "inner self-molecules" of the cell available for such a negative selection. The disruption of this mechanism, attributable to a lack of apoptosis, may allow some autoreactive clones to survive~ Some of these clones may be specifically directed against molecules generated or modified in their cell distribution during apoptosis. PHOSPHOLIPID MEMBRANE DISTRIBUTION AND REGULATION IN PHYSIOLOGICAL CONDITIONS

In normal circumstances, the lipid composition of the two leaflets of the plasma membrane differs. Although the amino-PL, principally PS and phosphatidylethanolamine (PE), are located almost exclusively in the cytoplasmic leaflet, the outer leaflet comprises mostly choline-PL, essentially sphingomyelin and phosphatidylcholine. Maintenance and modification of this membrane asymmetric distribution must be of vital importance for cell functions (26).

PhosphoIipid Transporters The ability to generate and maintain this transbilayer lipid composition requires at least two types of cell activities. An amino-PL-specific transport was first described in erythrocyte membranes. This activity, also referred to as "translocase," is adenosine triphosphate (ATP)-dependent and consists of a rapid inward movement of PS and PE. The "translocase" is able to continuously shuttle PS and, less efficiently, PE toward the inner side of the plasma membrane, continuously removing this powerful signal from the cell surface (27). Although the precise molecular structures responsible for this ubiquitous cell function are unknown, it

165

was recently suggested that both an ATPase II and a 32-kDa transmembrane polypeptide may be involved (28). Another transporter acting in concert with the "translocase," although not specific for some PL and less efficient, was also first described in red blood cells and then recognized as a general and ubiquitous mechanism. This transporter, referred to as "floppase," is a slower system promoting the outward movement of both amino- and choline-PL. The "floppase activity" also requires energy in the form of ATP molecules, and there are some experimental clues strongly suggesting its protein nature (29). Although some reports have highlighted the potential importance of a selective interaction between amino-PL and cytoskeleton proteins in controlling the asymmetric "steady state" of membrane PL distribution, these interactions are considered of minor importance (30). Thus, the lipid asymmetry of the plasma membrane, observed in quiescent cells, seems to be maintained mainly by a complex, energy-dependent lipidtransport machinery, resulting from the combined activity of two independent systems, one shuttling all the PL outward, the other "pushing back" only the amino-PL to prevent the biological effects associated with the exposure of this specific group of PL. However, in some particular conditions, a process referred to as lipid scrambling can rapidly modify this "steady state." Although initially observed in platelets, this process seems to be a general mechanism operating in several other cell types (31). It is bidirectional and involves all maj or PL classes moving at comparable rates. A slightly slower inward movement was reported for sphingomyelin when compared with the rate of movement of the glycero-PL. Recently, Confurius et al (32) succeeded in isolating a heterogeneous protein fraction, "scramblase," reproducing the °'lipid scrambling" in specific experimental conditions (33). Membrane fusion events occurring during platelet "microvesiculation" as a result of several stimuli, seem somehow to be related to the loss of lipid asymmetry. However, these two phenomena, albeit often associated, can also be uncoupled in specific experimental conditions (34, 35). With respect to the mechanism believed to be operating in the course of apoptosis, evidence was found that nonspecific "lipid scrambling" and the inhibition of amino-PL-translocase are both responsible for

166

the loss of membrane asymmetry. The combined action of these two mechanisms seems to be involved both in the exposure of PS in the very early stage of apoptosis and in the "microvesciculation" observed later on. Thus, cells "committed" to PCD share several similarities with activated platelets, at least with respect to the early membrane apoptotic events (36, 37). Unlike other negatively charged PL, cardiolipin (CL), which has for practical reasons been the main antigen used to detect aPL, is almost exclusively located in the mitochondrial membrane. Although mitochondria play an important and early role in the apoptotic process (38), it is not known whether any rearrangement of CL molecules is associated with these events. Thus, the sudden modification of the plasma membrane PL distribution represents a general phenomenon providing a membrane signal that is not specific for a single cell function, but closely related to general mechanisms occurring in different pathophysiological processes such as the activation of the coagulation cascade as well as the recognition and removal of apoptotic cells. Although platelet "lipid scrambling" has been the most intensively studied, a similar "nonapoptotic" modification of PL membrane composition is an example of a general mechanism operating in several other cell types (Fig. 1). APOPTOSIS AS A POTENTIAL STIMULUS TO ANTIPHOSPHOLIPID ANTIBODY GENERATION

The observation that PL represent a ubiquitous component of cell membranes that undergo a functional redistribution in different pathophysiological conditions raises the fundamental question of whether, and to what extent, PL exposure occurring in course of apoptosis might be a key point in the pathogenesis of the APS. Indeed, apoptosis may represent a common path by which autoantibodies against several self-particles are produced. In a sequence analysis of four human monoclonal immunoglobulin G (IgG) anticardiolipin antibodies, it was evident that the formation of these antibodies was almost certainly antigen driven and characterized by an affinity maturation (39). Clustering and rearrangement of the autoantigens targeted in SLE in newly formed structures, as described by Casciola-Rosen et al (9), may explain the multiple and heterogeneous humoral immune responses observed in this disease. Molecular structures restricted normally to nuclear, or cytoplasmic

PITTONI AND ISENBERG

compartments, or the inner side of plasma membranes such as Ro, La, Sin, U1-RNR DNA, and PS, were shown to agglomerate on the outer surface of newly generated particles such as the apoptotic bodies (9, 40). Thus, PS, a potent surface procoagulant that is restricted normally to the inner surface of the bilayer, may represent, during apoptosis, an antigenic stimulus to aPL production. The exciting hypothesis of a common stimulus to autoantibody generation could explain why the autoantibody response observed in course of SLE, although heterogenous, is quite often characterized by crossreactivity (41) and epitope-spreading (42). This model also implies that self-molecules, fundamentally preserved, undergo some immunogenic modification in the setting of this "apoptotic" microenvironment. For instance, a specific proteolytic cleavage of the 70-kD protein U1 RNP has been described as a characteristic apoptotic biochemical feature (43).

Reactive Oxygen Species Reactive oxygen species (ROS) production is another biochemical event possibly related to the generation of newly modified epitopes. Many of the stimuli capable of triggering PCD can generate ROS, and cells exposed to ROS precursors can undergo death with an apoptotic morphology (44). Although there is increasing evidence showing that ROS can activate the apoptotic process through ROS-dependent signal transduction pathways, it is still unclear whether and to what extent these highly reactive species may contribute to the common final events leading to PCD (45). It is interesting that the autoantigens clustering in the apoptotic bodies are associated with subcellular compartments that are well-known sites of ROS generation, such as nuclear membranes or endoplasmic reticulum membranes and mitochondria (46). It is known that PL oxidation is crucial to unveil PL epitopes, against which most aPL are directed. In a recent paper, it was shown that aPL are actually directed to neoepitopes of oxidized PL or neoepitopes generated by adduct formation between breakdown products of oxidized PL and associated proteins (47). Thus, a CL molecule contains four fatty acids that undergo a progressive peroxidation as a result of simple air exposure, as happens during a standard anti-CL enzyme-linked immunosorbent assy (ELISA) (48). However, it is possible that a crucial in vivo

APOPTOSIS AND ANTIPHOSPHOLIPID ANTIBODIES

167

Phos|

Tran

Floppase

Scramblase

STEADY STATE

APOPTOSIS

Fig 1. PL membrane distribution and regulation in "steady state" and in the course of apoptosis, in "steady state," the lipid composition of the two leaflets of the plasma membrane differs. PS, an amino-PL, is located almost exclusively in the cytoplasmic leaflet, whereas the choline-PL are mainly located in the outer leaflet. This transbilayer composition is maintained by the combined activity of two independent systems, one shuttling both amino- and choline-PL outward ("floppase"), the other "pushing back" only the amino-PL ("translocase'). During apoptosis, the activity of a third system ('scramblase"), which moves all classes of PL in both directions, and the inhibition of the "translocase" modifies the "steady state" and consequently leads to PS exposure. PS exposed on apoptotic cells contributes to the activation of the coagulation factors (IX, VIIi, X, V, IlL which results in thrombin generation. The expression of tissue factor on apoptotic membrane provides a powerful procoagutant stimulus. PS-binding proteins such as I~2-GPI, annexin V, and protein C (APC} are activated on apoptotic membranes and thereby display anticoagulant activity. Furthermore, PS exposed during apoptosis provides a recognition signal for phagocytosis by a PS-receptor on macrophages.

[32-GPI-PL interaction (an oxidative stress due to the adjuvant-induced inflammatory reaction?) induces the formation of highly immunogenic PL or protein epitopes. In support of this view, it also was shown that APS sera or affinity-purified IgG do not bind efficiently to a CL analog that is incapable of undergoing lipid peroxidation (47). It is possible that there are different classes of aPL directed against "oxidation-dependent" and "oxidationindependent" neoepitopes. This hypothesis is also in accordance with the finding that some aPL recognize oxidized low-density lipoproteins

(oxLDL). This partial cross-reactivity could be explained by the observation that oxLDL represent a source of both PL and protein oxidized derivatives (49).

Clearance of Apoptotic Cells Sambrano and Steinberg (50) recently reported that a common mouse macrophage "scavenger receptor" is responsible for the removal of both oxLDL and apoptotic cells (50). With respect to the mechanism involved in apoptotic cell clearance, although tissue macrophages are the cells mainly

168

involved in phagocytosis of apoptotic cells in mammals, other cell types, such as hepatocytes (51), endothelial cells (52), glomerular mesangial cells (53), and fibroblasts (54), also take part in their removal. The mechanisms by which phagocytes recognize and engulf apoptotic cells have been partially clarified. Experimental evidence suggests that lectin-like molecules (55) as well as a sialoglycoprotein receptor/mannose receptor (51, 52) and a mannose/fucose receptor (54) are responsible for macrophage recognition of changes in the surface carbohydrate on apoptotic cells. The integrin av[33 (vitronectrin receptor) also has been identified as one of the molecular structures involved in the phagocytosis of apoptotic cells inasmuch as this molecule is able to recognize a sequence of aminoacids, arginin-glycin-aspartic acid (56). A protein bearing this sequence, thrombospondin, functions as a molecular bridge between the apoptotic cell and the phagocyte inasmuch as it also interacts with the macrophage cell receptor CD36 (57). The 61D3 antigen is a monocyte lineagespecific molecule involved in this mechanism, because it was demonstrated that antibodies against this molecule are able to inhibit the phagocytosis of apoptotic cells (58). With respect to the PS-specific receptor, one possible candidate for this molecular structure is a member of the scavenger receptor family. The original scavenger receptor (acetylLDL receptor), which was involved in the generation of atherosclerotic lesions, binds with high affinity to different ligands including modified LDL, several proteins, polysaccharides, and anionic PL (59). However, another scavenger receptor has been identified with the property of binding selectively to oxidatively damaged erythrocytes as well as oxLDL but not acetylated LDL, and further studies showed that macrophage binding to apoptotic mouse thymocytes is specifically but incompletely inhibited by oxLDL and PS liposomes (50, 60). Ramprasad et al (61) isolated a 94- to 97-kD protein from mouse peritoneal macrophages that binds selectively to oxLDL- and PS-rich liposomes and is identical to macrosialin, the mouse homologous of human CD68 (61). Although macrosialin may be implicated in recognition of both oxLDL and apoptotic cells through the interaction with PS, there may be additional complexity in receptor interactions as suggested by inhibition experiments (50, 60). A common clearance system would suggest that

PITTONI AND ISENBERG

the oxidative modification represents a general pathway for rapidly clearing cells and macromolecules severely damaged during the oxidative processes that continuously occur in the organism.

Tumor Necrosis Factor-Alpha (TNF-a) Some stimuli, for example, TNF-oL, are capable of inducing both cell activation and apoptosis or necrosis depending on the state of the cell or other concomitant environmental factors (62, 63). This cytokine also has the ability to activate oxidative pathways at least in some cell types (64). The possible role played by this cytokine in inducing the externalizations of inner "self" antigens has already been proposed and related to the induction of PCD in vivo (65). The importance of this cytokine in inducing apoptosis is also emphasized by the observation that its receptor belongs to a family of receptors including the Fas molecule (66, 67). The TNF-oL gene position in the context of the histocompatibility antigens could be related to a possible involvement of this cytokine in controlling central or peripheral self-tolerance mechanisms (68). Another interesting clue derives from the observation that the NZBW and BXSB crossed mouse strain (NZW × BXSB), which is both a lupus-prone and an excellent experimental model for aPL generation (69), is characterized by the presence of a defect in the TNF-o~ gene responsible for a reduced production of this cytokine (70). Monestier et al (71) recently were able to show that monoclonal antibodies generated from this animal model display either an anti-CL or an anti-[32-GPI reactivity, which may be responsible for the cardiovascular lesions observed in this mouse (71). Furthermore, it was shown that an increased rate of apoptosis may represent a potential source of immunogenic material. For instance, because specific cleavage of DNA is a well-defined biochemical feature of this mode of cell death, the abundant generation and release of nucleosomes in cells undergoing apoptosis may provide an antigenic stimulus for autoantibody production in a genetically susceptible individual (72, 73). APOPTOSIS AND PROCOAGULANT PROPERTIES OF ANTIPHOSPHOLIPID ANTIBODIES: POSSIBLE INVOLVEMENT OF A N N E X I N V AND 132-GLYCOPROTEIN I

Another potentially relevant point in understanding the role of PCD in APS pathogenesis comes

APOPTOSIS AND ANTIPHOSPHOLIPID ANTIBODIES

from the observation that apoptotic cells display procoagulant activity, at least in vitro. The production of active thrombin from its inactive precursor prothrombin requires the assembly of a prothrombinase complex, including activated coagulation factor X, on an anionic PL surface. Although this surface is classically provided by platelets, recent investigations have shown that different types of apoptotic cells play a similar role (9, 74). Given the functional effects of aPL on the coagulant and fibrinolytic properties of endothelial cells and monocytes, these cells are the most likely to be candidates as a cellular target of aPL. However, in both resting and cytokine-activated conditions, these cells do not seem to be directly targeted by aPL (75), although at least one study reported that TNF-oL treatment enhances the binding of antiendothelial cell antibodies from SLE patients to endothelial epitopes (76). More recently, the possible role played by this cytokine in endothelial dysfunction observed in the course of secondary APS was described (77). Although there is no definitive evidence supporting the hypothesis that apoptotic cells are procoagulant in vivo, several recent in vitro observations suggest this possibility. Thus, the apoptotic death of different cell types is accompanied by the appearance of several procoagulant activities other than the mere PS exposure, such as increased tissue factor activity and the loss of anticoagulant membrane properties (78-80). Annexin V

Annexin V is a protein that displays the wellknown property of binding with high affinity to PS exposed on "early" apoptotic cells (81). In this respect, it is reasonable to postulate that annexin V, as well as other proteins involved in aPL reactivity ([32-glycoprotein I, prothrombin), could also be associated with the relatively restricted group of autoantigens clustering in apoptotic bodies. Annexin V belongs to a family of structurally related proteins sharing the biological property of binding PL in a Ca2+-dependent manner (82). Although their homology is located in a conserved "core" domain including four 65- to 70-aminoacid-long homologous repeat sequences, the "fingerprint" of each annexin is found in the highly variable regions, that is, in the amino-terminal ends (83). In cultured endothelial cells, epithelial cells, and fibroblasts, this protein seems to be specifically concentrated in the cytoplasm and in the nucleolus (84).

169

Because the only nucleolar function is the synthesis and the assembly of ribosomal RNA, it is likely that annexin V is somehow involved in ribosomal formation transport and control. Although the annexins are generally considered to have a predominant intracellular localization, as suggested by the absence of classical "signal sequences" in their polypeptide chains, endothelial membranes exhibit a peculiar affinity for annexin V in resting conditions. This finding suggests that the in vitro anticoagulant protein of this protein may mirror a fundamental in vivo contribution to the endothelial coagulation homeostasis (85). Two recent investigations provided evidence that the ability to interfere with annexin V-endothelial and annexin V-trophoblast interactions (the interactions responsible for an impairment of the anticoagulant properties of these cells) could explain the mechanism of action of aPL. In these reports, the demonstration of specific effects of aPL on annexin V distribution on the surface of endothelial and trophoblast cells is accompanied by the observation that this interaction reduces the coagulation times displayed by these cells in an in vitro system. The finding of an in situ reduction of annexin V expression on the surface of placental villi from APS patients, as against placental villi from normal subjects or unl'elated miscarriage-associated conditions, provides complementary information about the potential pathogenic mechanism that may occur in vivo (86, 87). This model sheds new light on the placental effects of aPL. However, the questions of which antigen aPL recognize on endothelial cells, and the role possibly played by [32-GPI in this interaction, are not addressed in these investigations. Induction of a typical endothelial inflammatory-procoagulant phenotype does not modify annexin V-endothelium binding (88). The exposure of PS after an apoptotic stimulus may involve this membrane-bound annexin V reservoir. Although it is widely recognized that annexin V is a useful marker of apoptosis, in that as it specifically identifies early apoptosis as represented by membrane-exposed PS, the physiological significance of this binding and its potential implication for coagulation homeostasis are totally unknown. Likewise, although several in vitro anticoagulant properties of this protein have been described (89, 90) and the potential in vivo utility of these properties has been reported (91, 92), the physiological function of the low concentration of the circulating form of this protein is unclear (93).

170

Provided that apoptotic cells exhibit a significant procoagulant activity in vivo, we can only postulate that extracellular annexin V might well play a role in the modulation of this activity (94). One possible role of the intracellular form of this protein is the ability to inhibit annexin I phosphorylation by competing with protein kinase C for PS binding (95). A recent report proposed a direct role for aPL antibodies in inducing endothelial cells to undergo apoptosis. The authors claimed that some of these antibodies with lupus anticoagulant activity are able to crossreact with annexin V and that this property is responsible for inducing PCD (96). This finding needs to be confirmed, although it is now widely believed that the interaction of aPL with endothelial cells can trigger an inflammatory response essentially characterized by expression of adhesion molecules and cytokine production (97). It is likely that the "swing" from an activationinflammation cell program to an apoptotic one is simply attributable to the particular culture conditions, which may include an unrecognized concomitant stimulus, or to the restricted specificity of the population of aPL involved (Fig 1).

[32-Glycoprotein I During the past 7 years, much attention has been paid to the involvement of [32-GPI as a cofactor or coantigen of aPL (98). Unlike annexin V, this protein is found in human normal plasma at a relatively high concentration (200 pg/mL), 40% of which is associated with various lipoprotein structures (99). [32-GPI is a protein of approximately 50 kD molecular weight, consisting of five repeating motif domains of approximately 60 aminoacids (100). The fifth domain shows a slightly different structure characterized by two additional cysteine residues and a long terminal tail. A short peptide identified in the fifth domain represents the site for PL-protein interaction and, given the presence of lysine residues, is also a site of potential protein oxidation (101). It has been proved that [32-GPI can be specifically targeted by autoantibodies distinct from those directed to "pure" PL molecules.l°~ A crucial modification of [32-GPI structure occurring after PL interaction seems to be required to unveil "cryptic" epitopes (8) or to increase antigen density (7), because both of these conditions are believed to enhance anti-[32-GPI antibody binding. Although anti-[32-GPI antibodies have been dem-

PITTONI AND ISENBERG

onstrated with different techniques, including ELISA with nonoxidized (103), chemically treated (104), or polyvinylchloride plates (8), Western blot (102), and dot blot (105), an ELISA with oxidized ('v-irradiated) plates, seems to be the most suitable method of detecting these antibodies (106). Price et al (107), in a recent article, emphasized that aPL bind specifically to apoptotic but not viable mouse thymocytes. The interesting observation that this aPL binding is [32-GPI-dependent raises the possibility that circulating [32-GPI would complex with newly expressed anionic PL on the surface of apoptotic cells and thereby generate an epitope targeted by aPL (107). This finding could be highly relevant given the general consensus that the antibodies directed against 132-GPI represent the subset of aPL more closely correlated to the clinical manifestations of APS (108, 109). However, the importance of [32-GPI in APS pathogenesis remains to be determined. Even the physiological function of this protein, which has been highly conserved throughout evolution (110), is uncertain, although several anticoagulant properties have been described in vitro (111-115). However, Chonn et al (116) provided evidence that in vivo ~32-GPI is the major protein involved in the rapid clearance of liposomes containing negatively charged PL (116). By analogy, the authors of this paper postulated that this function may be crucial in the clearance of other "nonself" particles (oxidatively damaged structures, apoptotic cells exposing PS?). In the same study, the potential role played by [32-GPI as "opsonin" for the removal of these particles by the phagocyte system is also supported by the observation that the administration of an anti-~2-GPI polyclonal antibody is able to reduce the rate of clearance of these structures. However, in apparent contrast to these observations, it was recently reported that [32-GPI is able to inhibit macrophage uptake of oxLDL, a property that could underline a possible anti-atherosclerotic function of this protein. Conversely, as shown in the same investigation, certain polyclonal or monoclonal anti-[32-GPI antibodies modulate this uptake, a finding suggesting a potential role of these antibodies in the evolution of vascular lesions in patients with SLE (117). Of particular interest is the observation that, in contrast to the ingestion of microbes and other foreign particles, apoptotic cells are normally phagocytosed independently of antibody and comple-

APOPTOSISAND ANTIPHOSPHOLIPIDANTIBODIES

ment. Although the phagocytosis of antibodycoated particles results in the acquisition of an inflammatory phenotype with release of mediators such as arachidonic acid metabolites, the ingestion of apoptotic cells is not coupled with these events (118). Conflicting and confusing results have emerged from recent investigations on [~2-GPI plasma concentration in APS and SLE, with normal (119), decreased (120), or elevated levels (121) reported. These discrepancies may reflect genetic or environmental differences in the populations studied. At least one report indicated that both the presence of anti-[32-GPI and high [32-GPI levels are associated with a history of thrombosis in patients with APS and SLE (122). Although this study did not show a direct relationship between these two pai'ameters, further investigation is needed to establish whether anti-[32-GPI antibodies interfere with the rate or the mechanism of clearance (antibody-mediated versus receptor-mediated) of"[32-GPI-binding particles" (apoptotic cells?) and the possible consequences on inflammatory and coagulation homeostasis (Fig 1).

IS THERE ANY EVIDENCE OF A RELATIONSHIP BETWEEN APOPTOSIS AND THE PRESENCE OF ANTIPHOSPHOLIPID ANTIBODIES?

£as

Although little is known about the possible link between aPL and any abnormality of apoptosis in vivo, some intriguing clues come from experimental models of SLE. It is now accepted that the lpr and gld traits, both responsible for the acceleration of the disease in MLR mouse, are due to the genetic defects, respectively, of Fas molecule and its ligand (123,124). MRL-lpr/lpr mice develop a polyclonal accumulation of abnormal peripheral T lymphocytes characterized by an alpha/beta TCR but lacking CD4 and CD8 molecules (125). However, aPL in common with other autoantibody specificities can be detected in the MRL lpr/lpr mouse (126, 127). Thrombotic lesions in the central nervous system reminiscent of those observed in APS patients were reported in this mouse model and associated with the presence of aPL. The concomitant presence of thrombocytopenia in these mice supports the view of a noncoincidental pathogenic relationship with aPL (126, 128). Although these findings need confirmation (previous investigations

171

in this mouse strain failed to provide any histological evidence of clear ischemic lesions) (129, 130), the MRL lpr/Ipr mouse is a valuable in vivo model to clarify the factors controlling aPL generation and effects of these antibodies in the setting of a molecular defective apoptotic mechanism. The precise function of the Fas molecule in human pathology is still under investigation. Although a soluble form of this molecule has been described and characterized as being able to inhibit Fas-mediated PCD (131), elevated levels of this protein found in around 30% of SLE sera were not correlated to disease activity or any particular organ involvement but were significantly related to a particular HLA profile (A 1, B 8, DR3) (132).While a genetic defect of Fas molecule has not been described in SLE patients, it has been in a few cases of the human autoimmune lymphoproliferative syndrome (ALPS). Some of these young patients show clinical autoimmune manifestations such as hemolytic anemia, thrombocytopenia, neutropenia, glomerulonephritis, and vasculitis, aPL have not been described in course of ALPS, but few patients have been identified and/or their genetic background may be distinct. The possibility that Fas molecule abnormalities are not the only determinant of ALPS is also suggested by the observation that some patients have a lymphoproliferative syndrome but no autoimmune phenomena. The pattern of these manifestations differs from patient to patient, and does not affect the parents of heterozygous patients (133-135). With respect to human SLE, Emlen et al described an increased total lymphocyte apoptosis in peripheral blood and correlated it to active disease as judged by the Systemic Lupus Activity Measure and the rate of apoptosis in vitro. 136 The increased rate of peripheral blood cell apoptosis, which is not restricted to the lymphocyte (137), observed in the course of SLE, may represent a paradox. By analogy, studies on the MRL lpr/lpr model have reported a lack of apoptosis and lymphoproliferative phenotype in vivo (138) but increased apoptosis of lymph node lymphocytes in vitro (139). The former phenomenon is related to the genetic defect of Fas, and the latter could only be due to the presence of cells primed to undergo apoptosis but prevented from dying in vivo. On the contrm'y, the in vitro environment would not be able to block the apoptotic fate of these cells. However, in apparent contrast with this observation, decreased in vitro

172

apoptosis of splenic lymphocytes has been reported in C57BL/6-1pr/lpr and C57BL/6-gld/gld mice. 140 The basis of this discrepancy is unclear, although it may due to several factors such as the different origin (spleen or lymphonodes) of the lymphocyte population studied, the methods of cell separation, or the different genetic background of MLR and C57BL/6 strains. In the article by Emlen et al, it also was shown that lymphocytes from patients with SLE release nucleosomal material into the extracellular space in direct proportion to the rate of apoptosis, but no correlation of this phenomenon with patients' serology, including aPL, was discussed (136).

Bcl-2 Overexpression of Bcl-2 in the B cells of a transgenic mouse leads to the development of a lupus-like syndrome. In Strasser's transgenic mouse model, the expansion of a population of small nonneoplastic B lymphocytes is associated with the production of autoantibodies against nuclear structures, including ds-DNA, and the development of an immune complex glomerulonephritis of possible autoimmune origin. 141However, other Bcl-2 B cell transgenic models failed to show clear signs of autoimmune disease (142, 143). These findings would suggest that Bcl-2 overexpression in B cells is not sufficient per se to induce autoimmunity and that the different genetic background of the strains used in these studies is a highly relevant factor contributing to the development and the pattern of the autoimmune phenomena observed. However, in lupus patients, the overexpression of this protein has generally been observed in T rather than B lymphocytes and related to the disease activity (144). Of particular interest is the recent observation that an even closer association can be found if the analysis of circulating lymphocytes from SLE patients is restricted to the specific subset lacking CD4 and CD8 molecules (doublenegative T lymphocytes), a finding that requires further investigation (145, 146). It is known that an augmented synthesis of intracellular Bcl-2 blocks the increase of ROS levels associated with apoptosis, although this protein also can prevent apoptosis through a non-anti-oxidant mechanism (147). Furthermore, PS externalization and oxidation may occur on the surface of apoptotic cells or apoptotic vesicles at least as a consequence of an oxidative

PITTONI AND ISENBERG

apoptotic stimulus. The finding that these effects can be specifically reversed by Bcl-2 expression suggests that PL oxidation is an apoptosis-related event and that the rate of Bcl-2 expression could affect the availability of oxidized PL molecules on cell membrane (148). Provided that these events are somehow related to a selective elimination of autoreactive clones, we might postulate that decreased oxidized-PL exposure, due to Bcl-2 overexpression, could affect the establishment of appropriate tolerance to these antigens.

Functional Properties of Phospholipid-Protein Complexes Formed During Apoptosis Definitive evidence supporting overt links between abnormalities of apoptosis and the general mechanism of autoantibody generation in humans is awaited. However, two considerations strongly suggest that the link between aPL generation and apoptosis is particularly intriguing. PS are externalized amongst other autoantigens on the surface of cells undergoing apoptosis. However, unlike other autoantigens, the exposure of PS seems to have two precise functional properties. First, it is likely that PS, once externalized, contribute to the coagulation homeostasis of the apoptotic cell membrane by binding to coagulation-related proteins such as [32-GPI and annexin V. This observation suggests a link between the antigenic structure recognized by "pathogenic" aPL, complexes formed by PL, and these proteins or PL-dependent epitopes of these proteins, and what seems to be one of the most relevant functional effect of these autoantibodies, ie, the modulation of the procoagulant properties of cell membranes. Secondly, PS exposed on the cell membrane act as a recognition signal for macrophages, which display a receptor-mediated clearance of apoptotic cells. Preliminary studies suggest that this clearance requires [32-GPI and that this mechanism entails the formation of PL-[32-GPI complexes (149). This type of clearance may be impaired by aPL and an alternative, antibodymediated, clearance favored. Two important consequences may result from this shift of clearance. The antibody-mediated phagocytosis of apoptotic cells provides a nonspecific trigger for the immune response inasmuch as it enhances inflammation. Furthermore, highly specialized antigen presenting cells are particularly efficient at taking up antibodycoated structures and the following antigen process-

9db

ANTIGEN PROCESSING A DSENTATION PN RE

V

~/10

A

~

SU__Z.~V NL CLONESDT

PHAGOCYTOSI S ~ 1

GENERATIONOF NEW EPITOPES'. PL-PROTEINCOMPLEXES OXIDATION

It--

6 Ip SExPosoRElr t 5~ APOPTOSIS~~ ~1 1 Fig 2. Apoptotic PL-protein complexes and the stimulus to aPL generation. Altered rates of apoptosis during the development of immune reactions may induce an immune response against PL-protein complexes. Lack of apoptosis (1), as observed in the experimental models of autoimmune diseases, could account for less exposure of PS {2). During the establishment of tolerance to self-antigens, a reduced generation of PL-protein epitopes and oxidation on apoptotic bodies (3) leads to the survival of those B and T clones specifically directed against these epitopes. Increased apoptosis, as observed in peripheral blood of patients with SLE, may be responsible for augmented exposure of PS (6), resulting in more abundant generation of PL-protein complexes and in more oxidation of these structures. The exposure and oxidation of PS provides a molecular signal for phagocytosis by macrophages (8). The following antigen processing and presentation by macrophages or other antigen-presenting-cells (9) provides an antigenic stimulus to specific T and B clones leading to aPL production (10). aPL are responsible for the procoagulant phenomena observed in course of APS (11). aPL enhance the immune response against PL-protein complexes by promoting an antibody-mediated rather than a receptor-mediated phagocytosis (12). The antibody-mediated clearance of apoptotic cells results in a more efficient stimulus to the activity of antigen-presenting cells.

174

PI]-i-ONI AND ISENBERG

ing and presentation by these cells may result in an additional stimulus to aPL generation (Fig 2). CONCLUSIONS

The precise role of apoptosis in the generation of aPL as well other autoantibodies remains to be determined. It is probable that the generation of "pathogenic aPL" requh'es the coexistence of several factors. Given that apoptosis is a physiological phenomenon, it may be postulated that in the presence of a defined genetic background (eg, HLA, other autoimmunity-related genes, appropriate hormonal "milieu") or a particular cytokine

environment (TNF-o0, altered rates or unusual sites or abnormal processing of apoptotic cells could lead to the generation of autoantibodies against the "oxidation-dependent" epitopes clustered in the apoptotic bodies. Intriguing clues from experimental models and human pathology strongly suggest that this meaningful relationship does exist. Specific events occurring in the course of apoptosis such as PL oxidation and the potential in vivo interactions between cell membrane PL and plasma proteins, which could represent a key point in APS pathogenesis, are likely to be involved in the generation of the aPL immune response.

REFERENCES 1. Harris EN. A reassesment of the antiphospholipid syndrome. J Rheumatol 1990;17:733-35. 2. Levy RA. Clinical manifestations of the aPL syndrome. Lupus 1996;5:393-7. 3. Oosting JD, Derksen RHWM, Bobbink IWG, Hackengt TM, Bouma BN, De Groot PG. Antiphospholipid antibodies directed against a combination of phospholipids with prothrombin, protein C and protein S: an explanation for their pathogenic mechanism? Blood 1993;81:2618-25. 4. McNeil HR Simpson RJ, Chesterman CN, Krilis SA. Antiphospholipid antibodies are directed against a complex antigen that includes a lipid binding inhibitor of coagulation: beta2-glycoprotein I (apolipoprotein H). Proc Natl Acad Sci U S A 1990;87:4120-24. 5. Amengual O, Atsumi T, Khamashta MA, Koike T, Hughes GRV. Specificity of ELISA for antibody to [32-glycoprotein I in patients with antiphospholipid syndrome. Br J Rheumatol 1996;35:1239-43. 6. Sorice M, Griggi T, Circella A, et al. Detection of antiphospholipid antibodies by immunostaining on thin layer chromatography plates. J Immunol Methods 1994; 173:49-54. 7. Roubey RAS, Eisenberg R, Harper MF, Winfield JB. "Anticardiolipin" antibodies recognize [32-glycoprotein I in the absence of phospholipid: importance of ag density and bivalent binding. J Immunol 1995;154:954-60. 8. Pengo V, Basiolo A, Fior MG. Autoimmune antiphospholipid antibodies are directed against a cryptic epitope expressed when [32-glycoprotein I is bound to a suitable surface. Thromb Haemost 1995;73:29-34. 9. Casciola-Rosen L, Rosen A, Petri M, Schlissel M. Surface blebs on apoptotic cells are sites of enhanced procoagulant activity: implication for coagulation events and antigenic spread in systemic lupus erythematosus. Proc Natl Acad Sci U S A 1996;93:1624-29. 10. Osborne BA. Apoptosis and the maintenance of homeostasis in the immune system. Curr Opin Immunol 1996;8: 245-54. 11. Evan GI. Old cells never die, they just apoptose. Trends Cell Biol 1994;4:191-2. 12. Hale AJ, Smith CA, Sutherland LC, Stoneman VE, Longthorne V, Culhane AC, et al. Apoptosis: molecular regulation of cell death. Eur J Biochem 1996;237:884. 13. Thomas N, Bell PA. Glucocorticoid induced cell size

changes and nuclear fragility in rat thymocytes. Mol Cell Endocrinol 1981 ;22:41-84. 14. Ohyama H, Yamada T, Watabane I. Cell volume reduction associated with interphase death in rat thymocytes. Radiat Res 1981;85:333-9. 15. Godman GC, Miranda AF, Deitch AD, Tanenbaum SW. Action of cytochalasin D on cells of established lines: zeiosis and movement of cell surface. J Cell Biol 1975;64:644-67. 16. Fadok V, Voelker DR, Campbell PA, Cohen JJ, Bratton DL, Henson PM. Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J Immunol t 992; 148:2207-16. 17. Lynch DA, Ramsdell F, Alderson MR. Fas and FasL in the homeostatic regulation of immune responses. Immunol Today 1995;16:569-74. 18. Nunez G, Merino R, Grillot D, Gonzales Garcia M. Bcl-2 and Bcl-x: regulatory switches for lymphoid death and survival. Immunol Today 1994;15:582-8. 19. Smith CA, Williams GT, Kingston R, Jenkinson EJ, Owen JT. Antibodies to CD3/T cell receptor complex induce death by apoptosis in immature T cells in thymic cultures. Nature 1989;337:181-4. 20. Osmond DG, Rico-Vargas S, Valenzona H, Fauteux L, Janari R, Jacobsen LK. Apoptosis and macrophage-mediated cell deletion in the regulation of B lymphopoiesis in mouse bone marrow. Immunol Rew 1994;142:209-30. 21. Lenardo MJ, Boeheme S, Chen L, et al. Autocrine feedback death and the regulation of mature T lymphocyte antigen responses. Int Rev Immunol 1995;13:115-34. 22. Akbar AN, Salmon N. Cellular environment and apoptosis: tissue microenvironments control activated T-cell death. Immunol Today 1997; 18:72-6. 23. Basten A, Brink R, Peake P, et al. Self-tolerance in the B -ceil repertoir. Immunol Rev 1991; 122:5-19. 24. Pulendran B, Van Driel, Nossal GJV. Immunological tolerance in germinal centres. Immunol Today 1997;18:27-32. 25. Mower DA, Peckman DW, Illera VA, Fishbangh JK, Stunz LL, Ashman RF. Decreased membrane phospholipid packing and decreased cell size precede DNA cleavage in mature mouse B cell apoptosis. J Immunol 1994;152:4832-42. 26. Devaux FP, Zachowski A. Maintenance and consequences of membrane phospholipid asymmetry. Chem Phys Lipids 1994;73:107-20.

APOPTOSIS AND ANTIPHOSPHOLIPID ANTIBODIES

27. Williamson R Schlegel RA. Back and forth: the regulation and function of transbilayer phospholipid movement in eukaryotic cells. Mol Membrane Biol 1994; 11:199-216. 28. Schroit AJ, Zwaal RFA. Transbilayer movement of phospholipids in red cells and platelet membranes. Biochim Biophys Acta 1991; 1071:313-29. 29. Connor J, Pack CH, Zwaal RFA, Schroit AJ. Bidirectional transbilayer movement of phospholipid analogs in human red blood cells. J Biol Chem 1992;267:19412-17. 30. Haest CVM, Deutike B. Possible relationship between membrane proteins and phospholipid asymmetry in human erythrocyte membranes. Biochim Biophys Acta 1976;436: 353-65. 31. Stuart MCA, Bevers EM, Confurius P, Zwaal RFA, Reutetigsperger CPM, Frederik PM. Ultrastructural detection of surface exposed phosphatidylserine on activated blood platelets. Thromb Haemost 1995;74:1145-51. 32. Confurius R Smeets EF, Schlegel RA, Bevers EM, Zwaal RE Reconstitution of phospholipid scramblase activity from human blood platelets. Biochemistry 1996;35:7631-4. 33. Williamson R Kulick A, Zachowski A, Schlegel RA, Devaux PE Ca 2+ induces transbilayer redistribution of all major phospholipids in human erythrocytes. Biochemistry 1992;31: 6355-60. 34. Bevers EM, Confurius R Zwaal RFA. Regulatory mechanisms in maintenance and modulation of transmembrane lipid asymmetry: pathophysiological implications. Lupus 1996;5: 480-7. 35. Dachary-Prigent J, Pasquet JM, Freyssinet JM, Nurden AT. Calcium involvement in aminophospholipid exposure and microparticle formation during platelet activation: a study using Ca2÷-ATPase inhibitors. Biochemistry 1995;34:11625-34. 36. Zwaat FA, Schroit AJ. Pathophysiological implication of membrane phospholipid asymmetry in blood cells. Blood 1997; 89:1121-32. 37. Verhoven B, Schlegel RA, Williamson R Mechanism of phosphatidylserine exposure, a phagocyte recognition signal on apoptofic T lymphocytes. J Exp Med 1995; 182:1597-601. 38. Zazami N, Susin SA, Marchetti P, et al. Mitochondrial control of nuclear apoptosis. J Exp Med 1996;183:1533-44. 39. Menon SH, Rahman MAA, Ravirajan CT, et al. The production, binding characteristic and sequence analysis of four human IgG monoclonal antiphospholipid antibodies. J Autoimmun 1997;10:43-57. 40. Casciola-Rosen LA, Anhalt G, Rosen A. Autoantigens targeted in systemic lupus erythematosus are clustered in two populations of surface structures on apoptotic keratinocytes. J Exp Med 1994;179:1317-30. 41. Ravirajan CT, Harmer I, McNally T, Hohmann A, Mackworth-Young CG, Isenberg DA. Phospholipid binding specificities and idiotype expression of hybridoma derived monoclonal autoantibodies from splenic cells of patients with systemic lupus erythematosus. Ann Rheum Dis 1995;54:471-6. 42. Topfer F, Gordon T, McCluskey J. Intra- and intermolecular spreading of autoimmunity involving the nuclear self-antigens LA(SSB) and Ro (SSA). Proc Natl Acad Sci U S A 1995;92:875-9. 43. Casciola-Rosen LA, Miller DK, Anhalt GJ, Rosen A. Specific cleavage of the 70kD protein component of the U1 small nuclear ribonucleoprotein is a characteristic biochemical feature of apoptotic cell death. J Biol Chem 1994;269:30757-60.

175

44. Buttke TA, Sandstrom PA. Oxidative stress as a mediator of apoptosis. Immunol Today 1994; 15:7-10. 45. Jacobson DJ. Reactive oxygen species and programmed cell death. Trends Biochem Sci 1996;21:83-6. 46. Cross AR, Jones OTG. Enzymatic mechanism of superoxide production. Biochim Biophys Acta 1991; 1057:281-98. 47. Hokko S, Miller E, Dudl E, et al. Antiphospholipid antibodies are directed against epitopes of oxidized phospholipids. J Clin Invest 1996;98:815-25. 48. Harris EN, Pierangeli S, Birch D. Anticardiolipin wet workshop report: Fifth international symposium on antiphospholipid antibodies. Am J Clin Pathol 1994;101:616-24. 49. Vaarala O, Puurunen M, Lukka M, et al. Affinity-purified anticardiolipin antibodies show heterogeneity in their binding to oxidized low-density lipoprotein. Clin Exp Immunol 1996;104: 269-74. 50. Sambrano GR, Steinberg D. Recognition of oxidativelydamaged and apoptotic cells by an oxidized low density lipoprotein receptor on mouse peritoneal macrophages: role of membrane phosphatidylserine. Proc Natl Acad Sci U S A 1995;92:1396-400. 51. Dini L, Autuori Lentini A, Oliviero S, Piacentini M. The clearance of apoptotic cells in the liver is mediated by asialog!ycoprotein receptor. FEBS Lett 1992;296:174-8. 52. Dini L, Lentini A, Diez Diez G, et al. Phagocytosis of apoptotic bodies by liver endothelial cells. J Cell Sci 1995;108: 967-73. 53. Savill JS, Smith J, Sarraf C, Ren 5I, Abbot E Rees A. Glomerular mesangial cells and inflammatory macrophages ingest neutrophils undergoing apoptosis. Kidney Int 1992;42: 924-36. 54. Hall SE, Savill JS, Henson PM, Haslett C. Apoptotic neutrophils are phagocytosed by fibroblasts with participation of the fibroblast vitronectin receptor and involvement of a mannose! fucose-specific lectin. J Immunol 1994;153:3218-27. 55. Duvall E, Wyllie AH, Morris RG. Macrophage recognition of cells undergoing programmed cell death (apoptosis). Immunology 1985;56:351-8. 56. Savill JS, Dransfield I, Hogg N, HasIett C. Vitronectin receptor-mediated phagocytosis of cells undergoing apoptosis. Nature 1990;342:170-3. 57. Savill JS, Hogg N, Ren Y, Haslett C. Thrombospondin cooperates with CD36 and the vitronectin receptor in macrophage recognition of neutrophils undergoing apoptosis. J Clin hlvest 1992;90:1513-22. 58. Flora PK, Gregory GD. Recognition of apoptotic cells by human macrophages: inhibition by a monocyte/macrophage specific monoclonal antibody. Eur J Immunol 1994;24:2625-32. 59. Kreiger M, Herz J. Structure and function of multiligand lipoprotein receptors: macrophage scavenger receptor and LDL receptor-related protein (LPR). Annu Rev Biochem 1994;63: 601-37. 60. Sambrano GR, Parthasarathy S, Steinberg D. Recognition of oxidatively damaged erythrocytes by a macrophage receptor with specificity for oxidized low density lipoprotein. Proc Natl Acad Sci U S A 1994;91:3265-9. 61. Ramprasad MR Fischer W, Witztum JL, Sambrano GR, Quehenberger O, Steinberg D. The 94- to 97-kDa mouse macrophage membrane protein that recognizes oxidized low density lipoprotein and phosphatidylserine-rich liposomes is

176

identical to macrosialin, the mouse homologue of human CD68. Proc Natl Acad Sci U S A 1995;92:9580-4. 62. Polunowsky VA, Wendt CH, Ingbar DH, Peterson MS, Bitterman PB. Induction of endothelial cell apoptosis by TNFalpha: modulation of inhibitors of protein synthesis. Exp Cell Res 1994;214:584-94. 63. Laster SM, Wood JG, Gooding LR. Tumor necrosis factor can induce both apoptic and necrotic forms of cell lysis. J Immunol 1988; 141:2629-34. 64. Heller R, Kronke M. Tumor necrosis factor receptormediated signalling pathways. J Cell Biol 1994;126:5-9. 65. Dorner T, Hucko M, Mayet WJ, Trefzer U, Burmester GR, Hiepe E Enhanced membrane expression of the 52 kDa Ro(SSA) and La(SSB) antigens by human keratinocytes induced by TNF alpha. Ann Rheum Dis 1995;54:904-9. 66. Beutler B, van Huffel C. Unravelling function on the TNF ligand and receptor families. Science 1994;264:707-10. 67. Nagata S, Godstein E The Fas death factor. Science 1993 ;267:1449-56. 68. De Vries RRR Van Rood JJ. HLA and autoimmunity. In: Cohen I, editor. Perspectives on autoimmunity. Boca Raton, Florida: CRP Press, 1988. 69. Koike T. Anticardiolipin antibodies in NZW × BXSB F1 mice. Lupus 1994;3:241-6. 70. Jacob CO, Lee SK, Strassmann G. Mutational analysis of TNF-a gene reveals a regulatory role for the 3'-untranslated region in the genetic predisposition to lupus-like autoimmune disease. J Immunol 1996;156:3043-50. 71. Monestier M, Kandiah DA, Kouts S, et al. Monoclonal antibodies from NZW × BXSB F1 mice to [32-Glycoprotein I and cardiolipin. J Immunol 1996;156:2631-41. 72. Mohan C, Adams S, Stalik V, Datta SK. Nucleosome: a major immunogen for pathogenic autoantibody-inducing T cells of lupus. J Exp Med 1993;177:1367-81. 73. Bell DA, Morrison B, Vandenbygaard E Immunogenic DNA-related factor: nucleosome spontaneously released from normal human lymphoid cells stimulate proliferation and immunoglobulin synthesis of normal mouse lymphocytes. J Clin Invest 1990;85:1487-96. 74. Flynn PD, Byme CD, Baglin TR Weissberg PL, Bennett MR. Thrombin generation by apoptotic vascular smooth muscle cells. Blood 1997;89:4378-84. 75. Del Papa N, Meroni PL, Tincani A, Harris EN, Balestrieri G, Zanussi C. Relationship between antiphospholipid and antiendothelial antibodies: further characterization of the reactivity on resting and cytokine-activated endothelial cells. Clin Exp Rheumatol 1992;10:37-42. 76. Quadros NR Roberts Thompson PJ, Gallus AS. Sera from patients with systemic lupus erythematosus demonstrated enhanced IgG binding to endothelial cells pretreated with tumor necrosis factor alpha. Rheumatol Int 1995;15:99-105. 77. Ferro D, Pittoni V, Quintarelli C, et al. Coexistence of antiphospholipid antibodies and endothelial perturbation in systemic lupus erythematosus. Circulation 1997;95:1425-32. 78. Aupeix K, Toti F, Satta N, Bichoff P, Freyssinet JM. Oyxsterols induce membrane procoagulant activity in monocyte THP- 1 cells. Biochem J 1996;314:1027-33. 79. Bombeli T, Karsan A, Tait JF, Harlan JM. Apoptotic vascular endothelial cells become procoagulant. Blood 1997;89: 2429-42. 80. Greeno EW, Bach RR, Moldow CF. Apoptosis is associ-

PI-I-i-ONI AND ISENBERG

ated with increased cell surface tissue factor procoagulant activity. Lab Invest 1996;75:281-9. 81. Vermes I, Haanen C, Steffens-Nakken H, Reuteligsperger C. A novel assay for apoptosis-flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescent labelled annexin V. J Immunol Methods 1995; 184:3951. 82. Reutelingsperger CPM. The annexins. Lupus 1994;3: 213-6. 83. Van Heerde WL, De Groot PG, Reutelingsperger CPM. The complexity of the phospholipid binding protein annexin V. Thromb Haemost 1995;73:172-9. 84. Sun J, Salem HH, Bird E Nuclear and cytoplasmic localization of annexin V. FEBS Lett 1992;314:425-9. 85. Van Heerde WL, Sakariassen KS, Hemker HC, Sixma JJ, Reutlingsperger CR de Groot PG. Annexin V inhibits the procoagulant activity of matrices of TNF-stimulated endothelium under blood flow conditions. Arterioscler Thromb 1994; 14: 824-30. 86. Rand JH, Wu XX, Guller S, et al. Reduction of annexin V (placental anticoagulant protein I) on placental villi of women with antiphospholipid antibodies and recurrent spontaneous abortion. Am J Obstet Gynecol 1994;171:1566-72. 87. Rand JH, Wu XX, Andree HAM, Lockwood CJ, Guller S, Scher J, et al. Pregnancy loss in the antiphospholipidantibody syndrome: a possible thrombogenic mechanism. N Engl J Med 1997;337:154-60. 88. Van Herde WL, Poort S, van't Veer C, Reutlingsperger CR de Groot PG. Binding of recombinant annexin V to endothelial cells: effect of annexin V binding on endothelial cell mediated thrombin formation. Biochem J 1994;302:305-12. 89. Bombeli T, Mueller M, Haerbeli A. Anticoagulant properties of vascular endothelium. Thromb Haemost 1997;77:408-23. 90. Romisch J, Schorlemmer U, Fickenscher K, Paques EP, Heinburger N. Anticoagulant properties of placental protein 4 (annexin V). Thromb Res 1990;60:355-66. 91. Van Ryn McKenna J, Merk H, Muller TH, Buchanan MR, Eisert WG. The effects of heparin and annexin V on fibrin accretion after injury in the jugular vein of rabbits. Thromb Haemost 1993;69:227-30. 92. Chollet R Malecaze F, Hullin E et al. Inhibition of intraocular fibrin formation with annexin V. Br J Ophthalmol 1992;76:450-2. 93. Flaherty MJ, West S, Heimark RL, Fujikawa K, Tait JE Placental anticoagulant protein 1: measurement in extracellular fluids and cells of the hemostatic system. J Lab Clin Med 1990;115:174-81. 94. Reutelingsperger CR van Heerde WL. Annexin V, the regulator of phosphatidylserine-catalized inflammation and coagulation during apoptosis. Cell Mol Life Sci 1997;53:527-32. 95. Raynal R Hullin F, Rabag-Thomas JMF, Fauvel J, Chap H. Annexin 5 as a potential regulator of annexin 1 phosphorylation by protein kinase C. Biochem J 1993;292:759-65. 96. Nakamnra N, Shidara Y, Kawaguchi N, et al. Lupus anticoagulant autoantibody induces apoptosis in umbilical vein endothelial cells: involvement of annexin V. Biochem Biophys Res Commun 1994;205:1488-93. 97. Simantov R, La Sala JM, Lo SK, et al. Activation of cultured vascular endothelial cells by antiphospholipid antibodies. J Clin Invest 1995;96:2211-9. 98. McNeil HE Simpson RT, Chesterman CN, Krilis SA.

APOPTOSIS AND ANTIPHOSPHOLIPID ANTIBODIES

Antiphospholipid antibodies are directed against a complex antigen that includes a lipid-binding inhibitor of coagulation: beta2-glycoprotein 1 (apolipoprotein H). Proc Natl Acad Sci U S A 1990;87:4120-4. 99. Polz E, Kostner GM. The binding of [32-glycoprotein I to human serum lipoproteins: distribution among density fraction. FEBS Lett 1979;102:183-6. 100. Kandiah DA, Krilis SA. Beta2-glycoprotein I. Lupus 1994;3:207-13. i01. Hunt JE, Krilis S. The fifth domain of [32-glycoprotein I contains a phospholipid binding site (Cys281-Cys288) and a region recognized by anticardiolipin antibodies. J Immunol 1994;152:653-9. 102. Sorice M, Circella A, Griggi T, et al. Anticardiolipin and anti-beta2-GPI are two distinct populations of autoantibodies. Thromb Haemost 1996;75:303-8. 103. Cabral AR, Cabiedes J, Alarcon Segovia D. Antibodies to phospholipid-free [32-glycoprotein I in patients with primary antiphospholipid antibody syndrome. J Rheumatol 1995;22: 1894-8. 104. E1 Kadi HS, Keil LB, De Bail VA. Analytical and clinical relationship between human IgG antibodies to [32glycoprotein I and anticardiolipin antibodies. J Rheumatol 1995;22:2233-7. 105. Cabiedes J, Cabral AR, Alarcon Segovia D. Clinical manifestations of the antiphospholipid syndrome in patients with systemic lupus erythematosus. J Rheumatol t995 ;22:1899906. 106. Matsura E, Igarasfii J, Yasuda T, Triplett DA, Koike T. Anticardiolipin antibodies recognize [32-glycoprotein I structure altered by interacting with an oxygen modified solid phase surface. J Exp Med 1994;179:457-62. 107. Price BE, Raucfi J, Sfiia MA, et al. Anti-phospholipid antibodies bind to apoptotic but not viable thymocytes in a beta2-glycoprotein I-dependent manner. J Immunol 1996;157: 2201-8. 108. Roubey RAS, Maldonado MA, Byrd S. Comparison of an enzyme-linked immunosorbent assay for antibodies to [32glycoprotein I and a conventional anticardiolipin immunoassay. Arthritis Rheum 1996;39:1606-7. 109. Cabiedes J, Cabral AR, Alarcon Segovia D. Clinical manifestations of the antiphospholipid syndrome associate more strongly with anti-[32-glycoprotein I antibodies than with antiphospho!ipid antibodies. J Rheumatol 1995;22:1899-906. 110. Matsura E, Igarashi M, Igarashi Y, et ah Molecular definition of human [32-glycoprotein I ([32-GPI) by cDNA cloning and inter-species differences of [32-GPI in alternation of anticardiolipin binding. Int Immunol 1991;3:1217-21. 111. Schouboe I. Beta2-glycoprotein I: a plasma inhibitor of the contact activation of the intrinsic coagulation pathway. Blood 1985;66:1086-91. 112. Nimpf J, Wurm H, Kostner GM. Beta2-glycoprotein I (apo H) inhibits the release reaction of human platelets during ATP induced aggregation. Atherosclerosis 1987;63:109-14. 113. Henry M, Everson B, Ratnoff O. Inhibition of the activation Hageman factor (factor XII) by beta2-glycoproein I. J Lab CIin Med 1988;111:519-23. i14. Schousboe I. Inositotphospholipid-accelerated activation of prekallikrein by activated factor XH and its inhibition by [32-glycoprotein I. Eur J Biochem 1988;176:629-36. 115. Moil T, Takeya H, Nishioka J, Gabazza EC, Suzuki K.

177

[~2-Glycoprotein I modulates the anticoagulant activity of activated protein C on the phospholipid surface. Thromb Haemost 1996;75:49-55. 116. Cfionn A, Semple SC, Cullis PR. [32-Glyeoprotein I is a major protein associated with very rapidly cleared liposomes in vivo, suggesting a significant role in the immune clearance of "non self particles." J Biol Chem 1995;270:25845-9. 117. Hasunuma Y, Matsuura E, MaNta Z, Katahira T, Nishi S, Koike T. Involvement of [32-glycoprotein I and anticardiolipin antibodies in oxidatively modified low-density lipoprotein uptake by macrophages. Clin Exp Irrmaunol 1997;107:569-73. 118. Meagher LC, Savill JS, Baker A, Fuller RW, Haslett C. Phagocytosis of apoptotic neutrophils does not induce macrophage release of thromboxane B2. J Lenkoc Biol 1992;52: 269-73. 119. De Benedetti E, Reber G, Mischer PA, De Moerlose P. No increase of [32-glycoprotein I levels in patients with antiphospholipid antibodies. Thromb Haemost 1992;68:624. 120. Ichikawa K, Takamatsu K, Shimizu H, et al. Serum apolipoprotein H levels in systemic lupus erythematosus are not influenced by antiphospholipid antibodies. Lupus 1992;1:45-9. 121. Galli M, Cortellazzo S, Daldossi M, Barbui T. Increased levels of [32-GPI (aca-cofactor) in patients with lupus anticoagulant. Thromb Haemost 1992;67:386. 122. McNally T, Mackie IJ, Machin SJ, Isenberg DA. Increased levels of [~2-glycoprotein I antigen and [32-glycoprotein I binding antibodies are associated with a history of thromboembolic complications in patients with SLE and primary antiphospholipid syndrome. Br J Rheumatol 1995;34: 1031-6. 123. Nagata S, Suda T. Fas and Fas ligand: Ipr and gld mutations. Immunol Today 1995; 16:39-43. 124. Watanabe-Fukunaga R, Brannan CI, Copeland NG, Jenkis NA, Nagata S. Lymphoproliferation disorder in mice explained by defect in Fas antigen that mediates apoptosis. Nature 1992;356:314-7. 125. Sobel ES, Kakkanaiah VN, Rapoport RG, Eisenberg RA, Cohen PL. The abnormal lpr double-negative T ce!ls fail to proliferate in vivo. Clin Immunol Immunopathol 1995;74: 177-84. 126. Gharavi AE, Mellors RC, Elkon KB. IgG anticardiolipin antibodies in murine lupus. Clin Exp Immunol 1989;78: 233-8. 127. Murphy ED, Roths JB. Autoimmunity and lymphoproliferation: induction by gene lpr and acceleration by a maIeassociated factor in strain BXSB mice. In: Rose NR et al, editors. Genetic control of autoimmune disease. Amsterdam: Elsevier/North Holland, 1978. t28. Smith HR, Hanson CL, Rose R, Canoso RT. Autoimmune MRL-lpr/lpr mice are an animal model for the secondary antiphosphoiipid syndrome. J Rheumatol 1990; 17:911-5. 129. Accini A, Dixon FJ. Degenerative vascalar disease and myocardial infarction in mice with lupus-like syndrome. Am J Pathol 1979;96:477-92. 130. Alexander EL, Moyer C, Travlos GS, Roths JB, Murphy ED. Two histopathologic types of inflammatory vascular disease in MRLfMP autoimmune mice. Arthritis Rheum I985;28: 1146-55. 131. Cheng J, Zhou T, Liu C, et al. Protection from Fasmediated apoptosis by a soluble form of the Fas molectde. Science 1994;263:1759-62.

178

132. Rose LM, Latchman DS, Isenberg DA. Elevate soluble Fas production in SLE correlates with HLA status not with disease. Lupus 1997;6:717-22. 133. Le Deist E Emile JF, Rieux-Laucat F, et at. Clinical immunological and pathological consequences of Fas-deficient conditions. Lancet 1996;348:719-23. 134. Rieux Lacat E Le Deist E Hivroz C, et al. Mutation in Fas associated with human lymphoproliferative syndrome and autoimmunity. Science 1995;268:1347-9. 135. Fisher GH, Rosenberg FJ, Straus SE, et al. Dominant interfering Fas gene mutations impair apoptosis in a human autoimmune lymphoproliferative syndrome. Cell 1995;81: 935-46. 136. Emlen W, Niebur J, Kadera R. Accelerated in vitro apoptosis of lymphocytes from patients with systemic lupus erythematosus. J Immunol 1994;152:3685-92. 137. Richardson BC. Monocyte apoptosis in patients with active lupus. Arthritis Rheum 1996;39:1432-34. 138. Ravirajan CT, SarrafCE, AnilkumarTV, Golding MCH, Alison MR, Isenberg DA. An analysis of apoptosis in lymphoid organs and lupus disease in murine systemic lupus erythematosus (SLE). Clin Exp Immunol 1996;105:306-12. 139. Van Houten N, Budd RC. Accelerated programmed cell death of MRL-ipr/lpr T lymphocytes. J Immunoi 1992;149: 2513-7. 140. Reap EA, Leslie D, Abrahams M, Eisenberg RA, Cohen PL. Apoptosis abnormalities of splenic lymphocytes in autoimmune lpr and gld mice. J Immunol 1995;154:936-43. 141. Strasser A, Whittingham S, Vanx DL, et al. Enforced

PITTONI AND ISENBERG

Bcl-2 expression in B lymphoid cells prolongs antibody response and elicits autoimmune disease. Proc Natl Acad Sci U S A 1991;88:8661-5. 142. McDonnell TJ, Deane N, Platt FM, et al. Bcl-2immunoglobulin transgenic mice demonstrate extended B cell survival and follicular lymphoproliferation. Cell 1989;57:79-88. 143. McDonnell TJ, Nunez G, Platt FM, et al. Deregulated bcl-2-immunoglobulin transgene expands a resting but responsive immunoglobulin M and D-expressing B-cell population. Mol Cell Biol 1990;10:1901-7. 144. Aringer M, Wintersberger W, Steiner CW, et al. High levels of Bcl-2 protein in circulating T lymphocytes, but not B lymphocytes of patients with systemic lupus erythematosus. Arthritis Rheum 1994;37:1423-30. 145. Rose LM, Latchman DS, Isenberg DA. Bcl-2 and Fas, molecules which influence apoptosis: a possible role in systemic lupus erythematosus. Autoimmunity 1994;17:271-8. 146. Rose LM, Latchman DS, Isenberg DA. Apoptosis in peripheral lymphocytes in systemic lupus erythematosus: a review. Br J Rheumatol 1997;36:158-63. 147. Jacobson MD, Raft MC. Programmed-cell-death and Bcl-2 protection in very low oxygen. Nature 1995;374:814-6. 148. Fabisiak JP, Kagan VE, Ritov VB, Johnson DE, Lazo JS. Bcl-2 inhibits selective oxidation and externalization of phosphatidylserine during paraquat-induced apoptosis. Am J Physiol 1997;272:675-84. 149. Manfredi AA, Rovere R Heltai S, et al. Apoptotic cell clearance in systemic lupus erythematosus: role of [32glycoprotein I. Arthritis Rheum 1997;41:215-23.