the antiphospholipid syndrome as a model for b cell-induced autoimmune diseases

intl.elsevierhealth.com/journals/thre

Thrombosis Research (2004) 114, 363--369

The antiphospholipid syndrome as a model for Bcell-induced autoimmune diseases$

Pierre Youinou*, Yves Renaudineau

Laboratory of Immunology, Brest University Medical School Hospital, 5 av Foch CHU Morvan BP 824, F 29609Brest Cedex, France

Received 1 June 2004; received in revised form 11 June 2004; accepted 15 June 2004
Available online 20 July 2004
0049-3848/$ - see front matter A 200doi:10.1016/j.thromres.2004.06.019

$ Contribution to the XIth Internat* Corresponding author. Tel.: +33-2E-mail address: youinou@univ-bres

KEYWORDS

Antiphospholipid syn-drome;

B cells;Autoimmune diseases

ABSTRACT

Growing awareness of the pathophysiological importance of B cells for antiphospho-lipid antibody syndrome (APS), particularly those expressing the Tcell marker CD5, hasrecently led to the proposal that their tolerance may be used as a method to reducespecific antibody (Ab) production. B cell tolerance has indeed become one of themostexciting developments in the treatment of this disease. Based on their production ofmultispecific Ab, these CD5+ B lymphocytes, also referred to as B-1 cells, are thoughtto account for most of the AutoAb in autoimmune murine models. Raised numbers ofcirculating CD5+ B cells correlate with high levels of anti-phospholipid (PL) Ab in someAPS patients, and participate in altered immunity of women with recurrentspontaneous abortion. These findings are not surprising in view of the cross-reactionwith PL of anti-bromelain-treated erythrocyte Ab secreted by these cells. Transgenicanimals have, however, shown that B lymphocytes contribute to such disordersthrough a variety of characteristics other than Ab production. Indeed, owing to therole of the CD5molecule in themaintenance of clonal anergy, increased proportions ofB-1 cells may merely reflect their defective regulation through CD5 itself. Various Bcell receptor (BCR)-associated transmembrane glycoproteins are also involved in thebehavior of the cells. These include CD19 which amplifies the message, and CD22which dampens down the BCR signaling. In addition, B lymphocytes may act as potentantigen-presenting cells for autoantigens, all themore because they secrete an excessof autocrine-acting interleukin-10 in autoimmune states. Furthermore, by modifyingthe specificity of their BCR, not only in the bone-marrow, but also in the secondarylymphoid organs, autoreactive B cells may initiate new immunoglobulin rearrange-ments. It is interesting that self-reactive Ab-making cells present with suchrearrangements. Finally, B cells have the capacity to polarize into B effector (Be)-1and Be-2, with different cytokine patterns that regulate the levels of T helper (Th)-1and Th-2, respectively. Such a cytokine might be defective in nonorgan-specificautoimmune diseases. In conclusion, B lymphocytes are required for the initiation of

4 Elsevier Ltd. All rights reserved.

ional Congress on Antiphospholipid Antibodies, 14--18th November 2004, Sydney, Australia.98-22-33-84; fax: +33-298-22-38-47.t.fr (P. Youinou).

P. Youinou, Y. Renaudineau364

anti-self Ab-associated disorders, such as APS. Their classical view in the biology ofimmune responses to self as autoAb secreting cells turns out to be rather naive, and anessential role for B lymphocytes may not be producing autoAb.A 2004 Elsevier Ltd. All rights reserved.

acterized by the recurrence of widespread throm- leukaemic B cells. It was later shown to be
The antiphospholipid (PL) syndrome (APS) is char-
botic events [1], and the production of aremarkably diverse family of antibodies (Abs).These autoAbs bind PLs in the presence or ab-sence of cofactors, most notably h2-glycoprotein I(h2GPI), but some of them are directed againsth2GPI itself [2]. Animal models have unequivocal-ly settled their pathogenicity [3], whilst theirproximal cause remains controversial.

How does APS arise? There is a consensus thatthe disease has a multifactorial etiology, dependingon both genetic and environmental factors. Owingto their predominance within the inflamed tissues,T cells have long occupied central stage of thisendless debate, and autoAbs restricted to a roleof disease markers. Recently, the significance of Bcells to the development of nonorgan-specific au-toimmune diseases has been emphasized [4--6].Numerous studies have, at least in part, decipheredthe biochemistry of the activation process [7]through surface immunoglobulins (Ig). Further-more, new insights into the functions of B cellshave been revealed over the past decade. Althoughour early concepts focused on their ability to pro-duce autoAbs, it has became apparent that thesecells accomplish various other tasks. Any of themmay be faulty, and any of the resulting inadequa-cies contribute to autoimmunity.

The APS may even be conceived as a model for Bcell-induced autoimmunity. Numerous studies havesparked off a great deal of interest in the possibilityof a crucial role for B cells in the pathophysiology ofsuch systemic diseases. Many rational therapeuticapproaches, including tolerance of B cells, may bederived from our better understanding of theirresponsibility. Modulation of B cell activity [8] onan antigen (Ag)-specific basis has indeed beenidentified as a means of pharmacological interven-tion in Ab-mediated pathologies.

Polyclonal or antigen-driven activation ofB cells

CD5-expressing B cells could be involved in thedevelopment of APS. The transmembrane glyco-protein CD5 was first described as a T cell marker

in the mouse, and then identified on human

expressed by a minute population of circulatingB cells, also termed B1 cells, compared withconventional B2 cells. However, the level of mes-senger RNA for this molecule was much lower in Bthan T lymphocytes [9]. Advances in leukocytephenotyping allowed the B1 population to bedivided into B1a and B1b cells. The latter sub-population lacks cell surface CD5, but shares allthe other attributes of B1a cells. High numbers ofthese B1a cells have been reported in patientswith rheumatoid arthritis, systemic lupus erythe-matosus (SLE) or primary Sjogren’s syndrome (SS),but also shown to correlate with levels of IgManti-PL Abs [10]. This is not surprising, given thelarge proportion of B1 lymphocytes that recognizephosphatidylcholine. Interestingly, raised titers ofthese autoAbs paralleled the augmentation inCD5+ B cells following blood transfusion ofpatients who underwent orthopedic surgical pro-cedures [11]. These results support the view thatsome anti-PL Abs are natural.

It is unclear whether CD5 is induced by acti-vation, or signifies a separate lineage. The acti-vation concept implies that B1a cells could begenerated from B2 cells, while the lineage para-digm posits the existence of separate progenitorsfor each of the two subsets of B cells. Our earlierstudies had approached this question through thescreening of Epstein--Barr virus (EBV)-immortal-ized cord blood B cell clones [12]. Twenty-six ofthem were derived from CD5-expressing B lym-phocytes, of which 5 displayed PL reactivity, and32 from CD5-nonexpressing B lymphocytes, ofwhich only 1 did so. It was, however, clear, fromsubsequent experiments, that, when the B cellculture supernatants were tested with 20 Ags,including PLs and/or h2GPI, Abs bound to eithernone, one or several of them, irrespective of theCD5 expression. This data suggests that the pro-duction of polyspecific Abs to self components isnot confined to those B cells that carry the CD5marker. In this respect, it is worth noting thatelevated levels of circulating B1a cells, whichhave been described in EBV-caused infectiousmononucleosis [13], have never been confirmed,based on the fact that this virus abrogates thesynthesis of the CD5 protein.

The antiphospholipid syndrome as a model for B cell-induced autoimmune diseases 365

The synthesis of pathogenic anti-PL Abs could,alternatively, be driven by exogenous Ags. Con-sistent with this view is the recent report that aproportion of anti-PL Ab-secreting cells from EBV-infected patients does express CD27, which is thehallmark of memory B lymphocytes [14]. Assum-ing that EBV picks out memory B cells, thisobservation implies that other microorganismshave been encountered in the past by thepatients. One step further, these viral or bacte-rial epitopes would mimic h2GPI, and, thereby,favor the emergence of the related memory Bcells. As a consequence, some of them are pre-disposed to the production of anti-PL Abs follow-ing reactivation with EBV. This interpretationpoints to the importance of microbial peptidesat the onset of the autoimmune process. It haslong been suspected that infection may be re-sponsible for the induction or precipitation ofAPS [15], but the most convincing evidenceexpanding this hypothesis has recently came fromthe extensive usage of combinatorial peptidelibraries. Bacterial induction of autoAbs toh2GPI may indeed account for the infectiousetiology of APS [16], and molecular mimicry beone of the most efficient mechanisms for break-ing the tolerance [17], and triggering this Ag-driven process. In support of this interpretation,variable regions of some anti-PL Abs appear to behighly mutated [18], the major epitotic regionson the h2GPI molecule targeted by these autoAbsto be exquisitely circumscribed [19], and dimer-ization of h2GPI to induce a significant increasein anti-h2GPI Ab affinity in the ELISA [20]. Thefinding that reactivity in anti-PL Ab enzyme-linked immunosorbent assay (ELISA) is largelydue to IgG2 in high-titer patients [21] is consis-tent with a polysaccharide Ag, and would makeone suspicious of a bacterial mimetope. This isreminiscent of the fact that infection-relatedanti-PL Abs differ from genuine APS Abs [22],be it primary or secondary to SLE, with respectto IgG subclass, activity and epitope specificity.

Other functions of B lymphocytes

B lymphocytes coordinate T lymphocytemigration

Pathways of T lymphocyte activation and differ-entiation into effector T helper (Th)1 and Th2cells have been the focus of intensive research[23]. This has brought about evidence that the Tlymphocyte differentiation along these two path-

ways affects appropriate immunity to deal withinfectious pathogens, and also modulates suscep-tibility to autoimmune pathology. Lowered pro-duction of interleukin (IL)-3 might thus favorfetal loss in APS [24]. This multilineage growthfactor induces the maturation of precursor Tcells, as well as the proliferation of B cells. Inthis respect, the role of different costimulatorymolecules CD28--B7 and CD40--CD40 ligand (L)interactions, and, more recently, the OX-40/OX-40L interaction have also been extensively inves-tigated. CD28 whose engagement by its ligandsCD80 and CD86, constitutively expressed by ma-ture dendritic cells (DC), is an absolute requisitefor T cell-dependent germinal centre (GC) forma-tion, whereas this has a limited role in the T cellsubset differentiation. The CD40L, which is upre-gulated following the initial stimulation by Agand the CD28--CD80/86 juncture, is suspected tostimulate DCs release interleukin (IL)-12. This, inturn, influences the differentiation of effector Tcells into Th1 cells. Studies on the expression ofchemokines, such as CXCR5, and the migration ofthe activated T cells to GCs have also highlightedthe importance of the bridge between OX-40 andOX-40L that is instigated in a CD28-dependentmanner. It appears that Ag-specific concertedaction of B and T lymphocytes, such as theengagement of OX-40 on activated T cells byOX-40L on activated B cells induces IL-4 synthe-sis, suppresses interferon (IFN)-g production andresults in the differentiation of Th2 lymphocytesand plasma cells under the influence of IL-4 thathas just been released.

B cells produce specific cytokines

Beyond the paradigm that, at least with respectto T cell-dependent responses, T lymphocytesmaintain strict control over B lymphocytes, it isnow acknowledged that the latter cells have thecapacity to solicit their own help from the for-mer [25], and thus to produce a flurry of cyto-kines [26]. It is even more important that naıve Bcells differentiate themselves into polarized Bcells with different cytokine profiles, followingstimulation with Ag and polarized effector Th1and Th2. These newly polarized B cells have beentermed B effector (Be)1 and Be2, respectively[27], inasmuch as, once induced, they regulatethe level of Th1 and Th2 cells. Consequently, Be1cells, by virtue of their production of IFN-g andpresentation of specific Ag to T lymphocyte,promote the expansion of Th1 cells. In contrast,


IL-4 produced by Be2 cells encourage the devel-opment of Th2 cells. Therefore, Be1 and Be2cells behave as classical antigen-presenting cellsthat regulate the profile of the immune re-sponse. This is particularly true for polyspecificIg-bearing B cells, i.e. CD5+ B lymphocytes. Theobservation that Th1/Th2 and Be1/Be2 cells cancross-regulate the differentiation of naıve B andT cells leads to the possibility of a significantamplification of immune responses. Such positivefeedback regulation, if out of control, couldpotentially initiate autoimmunity [25--27]. Forexample, assessment of circulating Be1 and Be2cells in SLE [28] has indicated elevated IL-10-producing CD5+ B cells. The precise mechanismfor generating Be1 and Be2 cells remains, how-ever, to be deciphered.

B cells influence dendritic cells

In addition to the increased awareness of the roleof B lymphocytes in affecting effector T-lympho-cyte differentiation, and by inference their abilityto induce Th1- or Th2-type cytokines, it has beendemonstrated that B cells regulate the productionof cytokines by DCs, as a result of their ability toproduce IL-10. The latter cytokine, produced by Bcells among other cells, has indeed been shown toinhibit the production of IL-12 by DCs, which, inturn, induces IFN-g production by Th1 cells, andacts to suppress Th2 development, as well asfavor Th1-cell expansion. Hence, not only do Bcells directly affect the differentiation of Th cellsin an immune response, but they can also exertan effect on Th differentiation by DCs through anIL-10-dependent pathway.

Intrinsic abnormalities of B cells inautoimmunity

Altered signaling of B lymphocytes

The B cell Ag receptor (BCR) is a multiproteincomplex [29]. In the vicinity of this complex, anumber of surface molecules, most notably CD19[30] and CD22 [31], play a role in maintenance oftolerance. On the one hand, CD19 is physicallycoupled and functionally associated with surfaceIg. This coreceptor amplifies the BCR transduc-tion through several mechanisms among which isthe increase of the protein tyrosine kinase activ-ity. B lymphocytes from CD19-defective micebecome indeed hyporesponsive to transmembrane

signals, whereas those from mice that overex-press this molecule turn out to be hyperrespon-sive. It is of great interest that modest changesin its expression or function may shift the bal-ance between tolerance and immunity to auto-immunity, as established in systemic sclerosis[32]. On the other hand, CD22 acts to dampensignals generated through the BCR, so that itsdeficiency correlates with the development ofautoAbs. Such abnormalities could generate dys-function of the BCR-associated lipid rafts (LR) ofB cells in autoimmune states. These membranemicrodomains act as platforms for signaling andtrafficking from the BCR [33]. Owing to theprolonged residency of the BCR into LRs, CD19retards its internalization, a function that maycontribute to enhance the B lymphocyte responsein autoimmunity [34]. Using confocal microscopy,it was seen that the delivery of the BCRs fromthe LRs was delayed in CD19bright than in CD19dim

B cells from autoimmune patients.

Disturbances in peripheral blood B cellhomeostasis

A handful of membrane markers have been usefulin classifying developmental stages from naıve,Bm1, to memory B cells, Bm5 [35]. Based on theexpression of such markers, the distribution ofperipheral B cell subpopulations has been shownto be profoundly impaired in nonorgan-specificautoimmune diseases. For example, an excess ofGC founder cells has been revealed in the circu-lation of patients with SLE [36], while the propor-tion of memory B cells appeared to be diminishedin that of patients with primary SS [37]. There isevery likelihood that similar abnormalities do existin APS (Pers, d’Arbonneau and Youinou, unpub-lished results). Still, the reasons for these biasesare elusive.

The protection of autoreactive B cells fromdeletion may, alternatively, be ascribed to exter-nal interferences [38]. One such signal comes fromthe B cell-activating factor of the tumor necrosisfactor family (BAFF). This survival factor inhibitsnegative selection by autoAgs via the BCR. As aresult, transfection with the relevant cDNA inducesthe synthesis of membrane-bound and cell-freeBAFF. Serum levels of this B cell-specific cytokinehave been found to be elevated in SLE, primary SS,and APS, and its overexpression claimed to beassociated with an SLE-like condition in transgenic(Tg) mice, and with primary SS, as they age. Atleast three receptors have hitherto been identi-fied. Two of them are shared by BAFF and a


proliferation-inducing ligand (APRIL): the trans-membrane activator and cyclophilin ligand inter-actor (TACI) and the B cell maturation Ag. Thethird receptor is termed BAFF-R, specific for BAFFand restricted to B cells. Once inserted into thisthird receptor, BAFF acts to rescue B cells fromapoptosis. Given the increasing complexity of thissystem, it is understandable that loss of APRIL-binding TACI causes autoimmune glomerulonephri-tis, whereas BAFF costimulates B cells to secreteautoAbs, depending on its ability to delay B cellapoptosis.

CD5+ B cells revisited

It does not appear that CD5+ B cells have theexclusive rights to the production of pathogenicautoAbs [39]. Rather, increased numbers of CD5+B cells might reflect defective regulation of B cellfunction through the CD5 molecule itself. There isnow a growing body of evidence that its isessential in modulating signals downstream ofthe BCR. We have shown that ligation of CD5 orIgM on tonsillar B but not blood T cells resulted inapoptosis. In addition, anti-CD5 Ab maintainedthe proliferation of tonsillar B cells, providedthese had been pre-activated with anti-IgM andIL-2. This result was in contrast to cord bloodCD5+ B cells which did not apoptose in responseto anti-CD5 (Pers and Youinou, unpublishedresults) but reflected the fact that the cord bloodB cells were continuously exposed to autoAgs invivo. CD5 has been demonstrated to be associat-ed, both physically and functionally, with the BCRand it is important to note that the src-homology2 domain-containing protein tyrosine phosphatase(SHP-1) is constitutively linked with the transduc-ing Iga/Igh chains of the BCR through the immu-noregulatory tyrosine based inhibitory motif ofCD5. It has thus been suggested that such aninteraction with CD5 sequesters the SHP-1, andlimits its role with important molecules in posi-tive signaling through the BCR. Furthermore, therole of CD5 in the maintenance of clonal anergyhas recently been addressed by elegant experi-ments [40] using the hen egg lysosyme (HEL)-Ig Tgmouse. In this model, mice Tg for HEL-Ig and themembrane-bound form of the self Ag HEL mountapoptosis of HEL-specific B cells, while those Tgfor HEL-Ig and the soluble form of HEL initiateanergy via the SHP-1. Breeding of the latter Tgmice onto a CD5� /� background results in lossof tolerance. These data indicate that the pres-ence of CD5 raises the threshold required foractivation of self-reactive B cells and therebydetermines their ultimate fate. Consistent with

this novel role for CD5 is another recent model inwhich CD5� /� spleen cells from mice madetransgenic for anti-RNP Ab (a common autoAb inSLE and other connective tissue diseases) wereinjected into irradiated naıve mice. They migrat-ed to the peritoneal cavity (the site where mostof the CD5+ B cells are found) and began toexpress CD5 which prevented their production ofautoAbs.

Furthermore, we have recently identified anovel exon 1 for the CD5 gene (Renaudineau,Mageed and Youinou, submitted for publication).This is exclusively transcribed in B lymphocytes.Intriguingly, its existence is due to a defectivehuman endogenous retrovirus. Our data providesattractive evidence for a reciprocal expression ofthis alternative exon 1, designated exon 1B, withthe conventional exon 1, referred to as exon 1A.Exon 1B-type transcripts are translated into atruncated variant of the CD5 molecule devoid ofleader peptide. Consequently, whereas exon 1Apromotes expression of membrane CD5 protein inT and a subset of B cells, exon 1B tends to reduceCD5 protein expression in other B cells and,therefore, possibly, to prevent the signaling func-tions of CD5, the translocation of SHP-1, andthereby the production of Abs against infectiousagents and autoAbs, e.g., anti-PL Abs. This bal-ance between the two exons 1 might be impor-tant in the regulation of membrane expression ofCD5. By raising the threshold of the BCR for theresponse, CD5 should trigger the expression ofrecombination activating genes (Hillion, Youinouand Jamin, submitted for publication), and thusfacilitate the revision of variable genes which hasbeen shown to be instrumental in the preventionof autoimmune states.

To conclude, interest revives B lymphocytes ascontributors to the cause of APS. Although there isstill a long way before setting up a unifying modelfor how such a B cell hyperactivity leads to thedisease, we are rather close to understand theway these lymphocytes operate in autoimmunity.Paradoxically, in the light of recent findings of themodulation of B cell signaling by CD5, this andcompanion molecules play a crucial role in pre-venting autoimmunity, instead of inducing autoAb-related diseases. Evidence is, therefore, accumu-lating that aberrations of the transduction throughCD5 do exist. They could lead to autoimmunedisorders. Hence, the present views on the poten-tial functions of CD5+ B cells in autoimmunity arequite different from the previous and rather naıveinterpretation that the increased levels of CD5+ Bcells represented a direct source of autoAb lead-ing to pathogenesis. Finally, B cell tolerance


entails administering synthetic mimotope pepti-des, and it has became likely that B cell ablativetreatments of APS will be completed in the nextfew years.

Acknowledgements

The authors gratefully acknowledge Cindy Sene forher secretarial assistance.

References

[1] Hughes GRV. The antiphospholipid syndrome: ten years on.Lancet 1993;342:341--4.

[2] Li Z, Krilis SA. Anti-beta(2)-glycoprotein I antibodies andthe antiphospholipid syndrome. Autoimmun Rev 2003;2:229--34.

[3] Blank M, Cohen J, Toder V, Shoenfeld Y. Induction of anti-phospholipid syndrome by passive transfer of anticardioli-pin antibodies. Proc Natl Acad Sci U S A 1991;88:3069--73.

[4] Lipsky PE. Systemic lupus erythematosus: an autoimmunedisease of B cell hyperactivity. Nat Immunol 2001;2:764--6.

[5] Youinou P, Lydyard PM, Mageed RA. B cells underpin lupusimmunopathology. Lupus 2002;11:1--3.

[6] Tuscano JM, Harris GS, Tedder TF. B lymphocytes contributeto autoimmune disease pathogenesis: current trends andclinical implications. Autoimmun Rev 2003;2:101--8.

[7] Wienands J. The B-cell antigen receptor: formation of sig-naling complexes and the function of adaptor proteins.Curr Top Microbiol Immunol 2000;245:53--76.

[8] Coutts S, Plunkett ML, Iverson GM, Barstad PA, Berner CM.Pharmacological intervention in antibody-mediated dis-ease. Lupus 1996;5:158--9.

[9] Youinou P, Jamin C, Lydyard PM. CD5 expression in human Bcell populations. Immunol Today 1999;20:312--6.

[10] Velasquillo MC, Alocer-Varela J, Alarcon-Segovia D, Cab-iedes J, Sanchez-Guerrero J. Some patients with primaryantiphospholipid syndrome have increased circulating CD5+B cells that correlate with levels of IgM antiphospholipidantibodies. Clin Exp Rheumatol 1991;9:501--5.

[11] Paglieroni TG, Ward J, Holland PV. Changes in blood CD5(B1a) B-cell populations and autoantibodies following bloodtransfusion. Transfusion 1995;35:189--98.

[12] Lydyard PM, Youinou P. Human CD5+ (B1) B cells in healthand disease. Fundam Clin Immunol 1994;4:9--16.

[13] Hassan J, Feighery C, Bresnihan B, Whelan A. IncreasedCD5+ B cells in infectious mononucleosis. Br J Haematol1990;74:375--6.

[14] Lieby P, Soley A, Knapp AM, Cerutti M, Freyssinet JM, Pas-quali JL, et al. Memory B cells producing somatically mu-tated antiphospholipid antibodies are present in healthyindividuals. Blood 2003;102:2459--65.

[15] Blank M, Asherson RA, Cervera R, Shoenfeld Y. Antiphospho-lipid syndrome infectious origin. J Clin Immunol 2004;24:12--23.

[16] Blank M, Krause I, Fridkin M, Keller N, Kopolovic J, Gold-berg I, et al. Bacterial induction of autoantibodies to beta-2 glycoprotein I accounts for the infectious etiology of anti-phospholipid syndrome. J Clin Invest 2002;109:797--804.

[17] Rose NR. Infection, mimics and autoimmune diseases.J Clin Invest 2001;107:943--4.

[18] Lieby P, Soley A, Levallois H, Hugel B, Freyssinet JM, CeruttiM, et al. The clonal analysis of anticardiolipin antibodies ina single patient with primary antiphospholipid syndromereveals an extreme antibody heterogeneity. Blood2001;97:3820--8.

[19] Iverson GM, Reddel S, Victoria EJ, Cockerill KA, Wang YX,Marti-Renom MA, et al. Use of single point mutations indomain I of h2-glycoprotein I to determine fine specificityof antiphospholipid autoantibodies. J Immunol 2002;169:7097--103.

[20] Sheng Y, Kandiah DA, Krilis SA. Anti-beta 2-glycoprotein Ifrom patients with the ‘‘antiphospholipid’’ syndrome tobeta 2-glycoprotein I with low affinity: dimerization ofbeta 2-glycoprotein I induces a significant increase inanti-beta 2-glycoprotein I antibody affinity. J Immunol1998;161:2038--43.

[21] Sammaritano LR, Ng S, Sobel R, Lo SK, Simantov R, Furie R,et al. Anticardiolipin IgG subclasses: association of IgG2with arterial and venous thrombosis. Arthritis Rheum1997;40:1998--2006.

[22] Arvieux J, Renaudineau Y, Mane I, Perraut R, Krilis SA, Youi-nou P. Distinguishing features of anti-beta 2-glycoprotein Iantibodies between patients with leprosy and the antiphos-pholipid syndrome. Thromb Haemost 2002;87:599--605.

[23] Mosman TR, Sad S. The expanding universe of T-cell sub-sets: Th1, Th2 and more. Immunol Today 1996;17:138--46.

[24] Fishman P, Bakimer R, Blank M, Sredi D, Djaldetti M, Shoe-nfeld Y. The putative role of cytokines in the induction ofprimary anti-phospholipid syndrome in mice. Clin Exp Im-munol 1992;90:266--70.

[25] Stockinger B, Zal T, Zal A, Gray D. B cells solicit their ownhelp from T cells. J Exp Med 1996;183:891--9.

[26] Mosman TR. Complexity or coherence? Cytokine secretionby B cells. Nat Immunol 2000;1:465--6.

[27] Harris DP, Haynes L, Sayles PC, Duso DK, Eaton SM, LepakNM, et al. Reciprocal regulation of polarized cytokineproduction by effector B and T cells. Nat Immunol2000;1:475--82.

[28] Amel Kashipaz MR, Huggings ML, Lanyon P, Robins A, PowellI, Todd I. Assessment of Be1 and Be2 cells in systemic lupuserythematosus indicates elevated interleukin-10-producingCD5+ B cells. Lupus 2003;12:356--63.

[29] Reth M, Wienands J. Initiation and processing of signalsfrom the B cell antigen receptor. Annu Rev Immunol1997;15:453--79.

[30] Pesando JM, Bouchard LS, McMaster BE. CD19 is function-ally and physically associated with surface immunoglobu-lin. J Exp Med 1989;170:2159--64.

[31] Doody GM, Justement LB, Delkrias CC, Matthews RJ, Lin J,Thomas ML, et al. A role in B cell activation for CD22 andthe protein tyrosine phosphatase SHP. Science 1995;269:242--4.

[32] Sato S, Hasegawa M, Fujimoto M, Tedder TF, Takehara K.Quantitative genetic variation in CD19 expression correlatewith autoimmunity. J Immunol 2000;165:6635--43.

[33] Pierce SK. Lipid rafts and B-cell activation. Nat Rev., Im-munol 2002;2:96--105.

[34] Cherukwri A, Cheng PC, Sohn HW, Pierce SK. The CD19/CD21 complex functions to prolong B cell antigen receptorsignaling from lipid rafts. Immunity 2001;14:169--79.

[35] Frazer JK, Legros J, de Bouteiller O, Liu YJ, Banchereau J,Pascual V, et al. Identification and cloning of genes ex-pressed by human tonsillar B lymphocyte subsets. Ann NY Acad Sci 1997;815:316--8.

[36] Arce E, Jackson DG, Gill MA, Bennett LB, Banchereau J,Pascual V. Increased frequency of pre-germinal center B


cells and plasma cell precursors in the blood of childrenwith systemic lupus erythematosus. J Immunol 2001;167:2361--9.

[37] Hansen A, Odenthal M, Reiter K, Jacobi AM, Feist E, ScholzeJ, et al. Diminished peripheral blood memory B cells andaccumulation of memory B cells in the salivary glands ofpatients with primary Sjogren’s syndrome. Arthritis Rheum2002;46:2160--71.

[38] Mackay F, Browning JL. BAFF: a fundamental survival factorfor B cells. Nat Rev., Immunol 2002;2:465--75.

[39] Youinou P, Lydyard PM. CD5+ B cells in nonorgan-specificautoimmune diseases: a fresh look. Lupus 2001;10:523--5.

[40] Hippen K, Tze LE, Behrens TW. CD5 maintains tolerance inanergic B cells. J Exp Med 2000;191:883--9.

the antiphospholipid syndrome as a model for b cell-induced autoimmune diseases

Documents