Human T-cell activation deficiencies

REVIEW

Human T-cell activation deficiencies Antonio Arnaiz-Villena, Marcos Tim6n, Carlos Rodriguez-Gallego, Mercedes P&ez-Blas, Alfredo Corell, J. Manuel Martin-Villa and Jos R. Regueiro The increasing understanding of T-cell activation is paralleled by the recognition of a growing range of'experiments of nature' that cause T-cell activation deficiencies. Analysis of these deficiencies is, in turn, contributing to the understanding ofT-cell function in vivo. Here, Jose; Regueiro, Antonio ArnaizVillena and colleagues review current knowledge of structural and functional T-cell defects and the implications of these for T-cell biology. T-cell activation by antigen-presenting cells (APCs) involves the dynamic trans recognition and cis aggregation of several surface molecule pairs during cell-cell contact (Fig. 1 and Ref. 1). An initially weak adhesion stage through very prominent receptors, such as C D l l a CD18, and probably CD43 (Ref. 2) and CD45RO (Ref. 3), leads to intimate contact via CD2, CD5 and CD28 (Ref. 4). These increasingly stable, nonantigen-specific interactions allow the T cell to sample the particular peptides displayed on major histocompatibility complex (MHC) class I or class 1I molecules of the APC. The ensuing recognition step involves the T-cell receptor (TCR) complex, comprising CD8-CD3-TCR or CD4CD3-TCR in cytotoxic and helper T cells, respectively. If a peptide is bound, several signals are sent to the T-cell nucleus, where different groups of genes are sequentially activated or inhibited s. As a final consequence, an activated T cell exerts preprogrammed effector functions, measurable in vitro as proliferation, lymphokine secretion or cytotoxicity. In addition to their cell adhesion roles, all these surface T-cell molecules (plus CD7, for which no ligand has yet been identified) participate in the generation of intracellular signals during T-cell activation (Fig. 2) 6. The heterogeneous signals are integrated within the cell and modulate the behaviour of the responding T cell. Specific antibodies can be used in vitro to engage most T-cell molecules but, although great progress has been made ~, understanding of the relative biological role of the receptor molecules in vivo remains hazy. The precise molecular events that link receptor engagement to response gene activation are also unresolved. Nonetheless, certain universal transduction mechanisms are known to exist in T cells: protein phosphorylation/ dephosphorylation, G-protein-dependent hydrolysis of inositol phospholipids and, as a consequence, activation of protein kinase C (PKC) and an increase in cytosolic Ca -'+. Artificial means for transmembrane activation o f t cells have been devised, including the use of phorbol esters to activate PKC directly, calcium ionophores to increase free intracellular Ca 2+ concentrations or G protein activators. Subsequently, DNA-binding proteins that regulate a particular set of response genes are activated and/or induced s. Figure 2 shows some of the

regulatory proteins that control the interleukin 2 (IL-2) gene, a common response gene of T cells. These rapid advances in T-cell biology have been followed or, in some cases, preceded-' by the description of human structural and functional T-cell activation deficiencies that selectively affect the expression or function of one of the molecules involved. The molecular characterization of these defects offers interesting insights into the pathophysiology of T-cell function and ontogeny in vivo s,*, although the small number of cases sometimes requires cautious interpretation. Structural defects

CD 18 defects Leucocyte adhesion deficiency is a phagocyte defect that can give rise to recurrent bacterial and fungal infection l°. The deficiency can be caused by a range of mutations in the CD18 gene. CD18 is required for the expression of adhesive heterodimers by resting T cells (CDlla-CD18, also known as LFA-I) and by some activated T cells (CDIlc-CDI8). In vitro, patients' T cells exhibit impaired proliferation, cytotoxicity and help for antibody production (that is, cytokine secretion). In vivo, however, T-cell functions are largely normal: patients have few opportunistic or viral infections, and normal delayed antigen-specific skin tests, T-cell subsets and antibody responses. Thus, the CD1 la-CD18 interaction with CD54 may not be critical for the initial longrange adhesion step of T-cell activation (Fig. 1). Redundancy among the adhesion molecules that work at a similar (CD43, CD45) or closer (CD2, CD5, CD28) range may explain these findings. Alternatively, the patients may have highly selected T cells that can be activated without help from the CDI I a-CD 18 receptor.

CD3 defects Impaired expression and function of the T C R CD3y~e-~ complex can occur as a result of mutations in the y or • genes s. The clinical consequences are disparate: one of two brothers with y deficiency had severe combined immunodeficiency (SCID) symptoms (failure to thrive, intractable diarrhoea and fatal viral pneumonia) with certain organ-specific autoimmune features, whereas the other is presently healthyl~; the isolated

© 1992, Elsevier Science Publishers Ltd, UK.

ImmunologyTodav 2 5 9

vTot

No 7 99e

REVIEW

Antigen-presenting cell

z

T cell Fig. 1. The cell interface during T-cell activation. Molecules involved in adhesion, recognition and signal transduction in different cell types are depicted together for illustrative purposes only. Spiky molecules have kinase (P+) or phosphatase (P-) activity. mutant e case only showed mild respiratory symptoms 12. T cells from individuals lacking the CD3y chain (y-) express about half of the normal density of surface T C R CD3 complexes, whereas only 10% could be detected in the mutant e defect. These data suggest that e is more important than y for full receptor expression or conformation, and that the c43 TCR, ySe CD3 and ~ components do not need to be together in a single (xl3ySe~ complex to be expressed at the cell surface as a functional unit. Rather, the existence of subcomplexes or isoforms where CD3y and 8 are alternatives, is likely. The two independent isoforms are o~f3~/e~ and ~lBSe~ (Fig. 1), whose free association fulfils the requirements of two • chains per complex and most described chain interactions 13. T cells can be activated through a variety of surface receptors. The absence of y does not prevent signalling through CD2 or CD28, while mutant • cells have impaired CD2-mediated and preserved CD28-mediated activation pathways. CD2, therefore, may not require CD3y to activate T cells, and perhaps only ~, as in natural killer cells, is essential 14. CD28 also appears to be partly CD3 independent (Fig. 2). Lectin- as well as CD3-induced proliferation is impaired in both y - and mutant e peripheral T cells; antigen-induced proliferation, in contrast, is either low or normal, depending on the antigen tested. This suggests that antigens are better indicators of in vivo T-cell function than lectins or antibodies against surface receptors, and also that a very small number of TCR-CD3 complexes is sufficient for physiological T-cell function and Immunology Today

selection. Alternatively, positive selection of only highaffinity T cells may occur in these immunodeficiencies. Interestingly, T-cell subset ratios were opposite in the two types of CD3 defects: CD8 + T cells were underrepresented in the y - and normal in the mutant ~ defect, and CD4 + T cells were normal in the y - and depleted in the mutant c. The availability of T cells that selectively lack the y component (and, therefore, the o~13¥~~ isoform) has allowed analysis of the role of the ~/chain in signal transduction. Following TCR-CD3 triggering, a ~- CD4 + T-cell line showed normal proliferation, a low but significant calcium flux and very low levels of IL-2 synthesis Is. In contrast, CD8 + T cells lacking y proliferated poorly in response to lectin plus IL-2. These data suggest that y (or the oqBye{ isoform) is dispensable for certain CD4 +, but not CD8 +, T-cell functions in vitro. This is consistent with several in vivo observations. First, the antibody response to protein antigens, which is considered to be dependent on CD4 + T cells, was normal in y deficiency, while IgG2 levels and the antibody response to polysaccharides, which are not dependent on CD4 + T-cells, were defective. Secondly, the y - patient with severe disease displayed normal antibody responses but suffered a lethal viral infection. Thirdly, an abnormally low fraction of peripheral CD8 + T cells was consistently observed in the healthy y individual. Taken together, the evidence suggests that signalling through the CD8-CD3-TCR-{ ensemble is selectively impaired in the absence of y (Ref. 16 and authors' unpublished results). The local aggregation of molecules on the T-cell

260

rot 13 No. 7 1992

REVIEW Antibodies Lectins _ I G protein I

"~

I activat°rs I• •

•

•

•

~

(a)

•

(c)

PIP~

(g)

I Calcium I Phorbol I I i°n°P h°res II esters II •• DAG

ii~

IJ

(b)

ql NF-AT

( OCt I

FOS

ER Nucleus / k.

J

Cytoplas

Fig. 2. Signal transduction in T cells. Only those surface, transmembrane and nuclear events relevant to T-cell activation deficiencies are included. The role of G proteins in T-cell activation pathways has not been clearly established. Letters a-h denote proposed altered steps or interactions (see text). PL C: phospholipase C; PIPe: phospbatidylinositol 4,5-bisphosphate ; D A G : diacylglycerol; PK C: protein kinase C; I P ~: inositol 1,4,5-trisphosphat e. surface during antigen recognition involves at least CD4 or CD8 and the TCR-CD3 complex. This aggregation may be asymmetrical, with CD4 approaching the TCR through 8 and CD8 through y, as depicted in Fig. 1. CD38 and y chains may, like e, combine an interaction and a signal-transducing role ~. In the patient with the mutant e molecule, T-cell help was sufficient for a normal antibody response to most protein antigens, but not to polysaccharides. Thus, three different T-cell defects show a selective impairment in polysaccharide antigen handling in vivo~8: y deficiency, mutant ~ and the Wiskott-Aldrich syndrome (see below). This may indicate that these antigens are at least in part T-cell dependent, in contrast to current evidence in the mouse.

N F - A T defect

status of the patient (repeated opportunistic and viral infections), T-cell proliferation in response to surface (CD2, CD3, CD28) or transmembrane (phorbol ester plus calcium ionophore) activation is severely impaired, but the addition of exogenous IL-2 completely overcomes these defects, both in vivo and m vitro. The patient could not produce IL-2, -3, -4, -5 or gamma-interferon (IFN-y), whereas granulocyte-macrophage colonystimulating factor GM-CSF and IL-6 were synthesized normally, supporting the proposal that NF-AT coregulares a set of cvtokine genes (Fig. 2). CD4 + and CD8 + T-cell subset numbers were normal in the patient's peripheral blood, a finding that underlines the extraordinary redundancy of cytokine networks driving T-cell selection. Mice with single cytokine defects, for example 1L-2 or IL-4, support normal T-cell development (and, probably, function) in vivo 2°.'-I, but multiple cytokine gene disruption has not been attempted to date.

In a child with primary combined immunodeficiency, a T-cell defect was characterized, distal to the T C R - C D 3 complex, which resulted in a selective impairment of A D A and PNP defects cytokine gene transcription ~9. By analysing several reguAdenosine deaminase (ADA) and purine nucleotide latory factors of the IL-2 gene, a structural abnormality phosphorylase (PNP) are enzymes of the purine salvage was observed in the nuclear factor of activated T cells pathway. Although expressed in all tissues, their levels (NF-AT) s. The precise genetic lesion has not yet been are highest in the lymphoid system, perhaps explaining determined. As might be predicted from the clinical why different mutations in the structural genes of ADA Immunology Today

261

Vol. 15 No. 7

1992

REVIEW and PNP give rise to profound or milder T- and B-cell defects 22,23. The accumulation of toxic purine substrates and metabolites is thought to interfere with T-cell activation and proliferation, although the mechanism is presently far from clear. In a recent study, ADA defects strongly reduced TCR-CD3-mediated inositol phospholipid hydrolysis and calcium flux 24 (step (a) in Fig. 2). Inhibition of these early events and of later DNA synthesis strongly affects T-cell and B-cell development and function in vivo; the diseases are frequently lethal early in life as a result of overwhelming infection. Antibody responses are variable: patients with ADA defects have lower responses than those with PNP defects, perhaps due to a differential sensitivity of B cells to the respective toxic products. ADA deficiency has been the subject of considerable interest recently as it is the first human disease to be successfully treated by gene therapy 8.

Lack of CD8 + T cells has been observed in an infant with SCID familial background and an increased susceptibility to bacterial, viral and fungal infections 32. The patient was unable to reject an allogenic skin graft, but had normal antibody responses. The thymic cortex contains immature CD8 + cells, but these do not differentiate into mature medullary or peripheral CD8 + T cells. An isolated selective CD7 deficiency has been described in a SCID patient 33. Although the molecular basis was not determined, functional analysis suggested a crucial role for CD7 in T-cell and, interestingly, in B-cell development and function.

Other defects Defective expression of HLA class II molecules has been observed in patients with symptoms of combined immunodeficiency 2s. The undefined genetic lesion is not in the HLA region; rather, a heterogeneous range of defects involving the binding or activation of regulatory factors that control class II expression are thought to be responsible 26. CD4 + T cells, which require interaction with HLA class II molecules for their positive selection in the thymus, are frequently low in the patients, a finding consistent with mouse models 2v. Wiskott-Aldrich syndrome (WAS) is an X-linked immunodeficiency associated with severe eczema and thrombocytopenia 28. Descriptions of concomitant deficiency in CD43 expression on WAS lymphocytes and platelets spurred research on this molecule. CD43 is a major sialoglycoprotein with adhesion properties (Fig. 1)2. It is capable of intracellular T-cell signalling independently of the TCR-CD3 complex (Fig. 2). Recently, CD43-associated carbohydrate chains isolated from WAS T cells were shown to have an abnormal composition which may cause them to interfere with T-cell function 29. Alternatively (or additionally) the primary genetic lesion on the X chromosome may affect T-cell function through other pathways; other X-linked genes also affect T-cell (and B-cell) differentiation and function 3°. WAS T-cell proliferation and cytotoxicity are impaired in vitro and, probably, in vivo, as patients are susceptible to opportunistic and pyogenic infections. Interestingly, in vivo antibody responses to polysaccharide but not to protein antigens are severely deficient, in sharp contrast to CD18 defects 1°. Thus, engagement of CD43, but not CD11a-CD18, may be required for effective polysaccharide recognition by T (and/or B) cells. Defects in the CD18 and CD43 adhesion systems do not significantly affect peripheral CD4 + and CD8 + T-cell subset proportions, again suggesting functional redundancy of these molecules in vivo. A selective congenital deficiency of the CD4 + T-cell subset has been described in four unrelated patients with combined immunodeficiency of T and B cells 31. CD4 genes are present and no genetic defect has been reported. The peripheral T-cell compartment comprises mainly CD8 + T cells, but these are profoundly defective both in vitro and in vivo.

Primary signal transduction defects One early signal transduction defect is characterized by the selective failure of peripheral T cells to proliferate, synthesize IL-2 or express IL-2 receptor in response to the ligation of several structurally normal receptors, including TCR-CD3, CD2 and CD43 (Ref. 34). The defect inhibits early T-cell activation events, since transmembrane activators, such as phorbol ester plus calcium ionophore or G protein activators, elicited normal T-cell responses. This deficiency appears to result from defective coupling of receptors to signal-transducing proteins within the cell (steps (a) and (b) in Fig. 2). In vivo, T-cell function is also impaired (as evidenced by recurrent bacterial infections and weakened delayed antigenspecific skin tests and antibody responses), but T-cell subset levels and ratios are normal. We have recently identified three unrelated patients with a similar TCR-CD3-proximal defect. In these cases the TCR-CD3 complex is coupled to at least some signaltransduction pathways, because PKC activators, which are not mitogenic per se, completely rescue the impaired proliferation to TCR-CD3 ligands (C. RodrfguezGallego et al., unpublished). In these cases, step (c) in Fig. 2 may be affected. A later step, (d), may be altered in a SCID patient whose T cells could not proliferate, even in response to phorbol ester plus calcium ionophore 3s. Lastly, an infant with severe impairment in proliferation and IL-2 synthesis in response to costimulation through CD2 and CD28 has been described 36. The response to certain lectins and transmembrane activators was conserved, ruling out a general surface or intracellular signalling defect. Rather, a selective defect in CD28mediated T-cell activation was proposed, because a T-cell line derived from the patient showed normal proliferation after triggering via CD2 but not CD28. In vivo, T- and B-cell function was not severely impaired, suggesting that CD28-mediated activation may be yet another redundant adhesion/activation pathway for T cells.

Immunology Today 2 6 2

Functional defects In several reported primary and secondary defects of T-cell activation, the precise molecular lesions are unknown.

Primary gene induction defects Four unrelated immunodeficiency patients have membrane-distal defects that result in selective lymphokine disorders 37~°. These defects may result from abnormal cytokine (and other) gene regulation, analogous to the NF-AT defect (Fig. 2, step e), but the phenotypes r o t J3 No. 7 992

REVIEW vary, probably reflecting disparate biochemical bases. Most patients have impaired T-cell proliferation to mitogens and transmembrane activators (phorbol ester plus calcium ionophore), which is associated with reduced IL-2 production, but IFN-y and IL-2R gene expression is close to normal. Normal T-cell functional parameters can be achieved by the addition of exogenous IL-2. The existence of profound CD4 + and CD8 + T-cell subset imbalances in three of the four cases strongly suggests that multiple genes additional to IL-2 are affected, because mice selectively lacking only IL-2 or IL-4 showed normal T-cell development 2°,21. When fully characterized, these cases may help to define T-cell response gene sets and their common (and selective) regulatory factors. A subpopulation of patients with common variable immunodeficiency, a heterogeneous disease characterized by defects of immunoglobulins, have been thought to suffer primary T-cell lymphokine defects 41. Both membrane-distal and membrane-proximal defects have been described (c and e in Fig. 2). Further steps are required to assess the contribution of these T-cell defects to the immunoglobulin deficiency.

Secondary T-cell activation defects A range of secondary (or acquired) immunodeficiencies are associated with defects in T-cell activation.

A thorough review of these defects is beyond the scope of this paper. Briet/y, secondary defects in CD3-mediated, but not in CD2-mediated, T-cell activation have been reported in certain pathological situations (uraemia 42, human immunodeficiency virus (HIV) infection 43, Epstcin-Barr virus (EBV) infection44). Conversely, defects in the CD2 but not in the CD3 activation pathway are believed to be secondary to several disease states (Sj6gren's 4', lupus4% atopy 47, alcoholic liver disease 4s, Hodgkin's disease49). Lastly, in some patients both the CD2 and CD3 activation pathways show quantitative impairments, but a normal earn transme,nbrane activation biochemistry is preserved (Down's syndrome patients ~, healthy aged individuals s t bone marrow recipients'-'). The proposed altered steps in T-cell activation in each disorder are summarized in "Fable I and indicated in Fig. 2. Biological implications Defects have been defined at virtually all stages of T-cell activation (adhesion, recognition, signal transduction and gene induction, see Table 1). T-cell development can proceed normally in the absence of CDIS, normal CD43, or a large set of lymphokines, suggesting that these molecules are not essential for CD4+/CD8 + T-cell development. Furthermore, defects in the expression of the components of the TCR-CD3 complex and quanti-

Table 1. Human T-cell activation deficiencies

Defect

Adhesion molecule

Structural defects

CD18 CD43 CD7(?)

Recognition molecule

Signal transduction

Gene induction

CD3 CD4/CD8 MHC class 1I ADA PNP NF-AT Functional defects Primary

G-protein (a) and (b) CD28(?) (b) Pre-PKC (c) Post-PKC (d) Lymphokine defects (c) or (e)

Secondary

CD3(f) CD2 (g)

Associated diseases LAD WAS SCID SCID CID/SCID CID/SCID C1D/SCID CID/SC ID C1D Recurrent infections Thrombocytopenia Recurrent infections SCID SCID, CVID

Uraemia, HIV, EBV Slogren's syndrome, Lupus, atopy, alcoholemia, CD2(h) Hodgkin's disease CD2 + CD3 (a) and Down's syndrome, (g) age CD2+ CD3 (d) Bone marrow recipients Letters (a)-(h) indicate proposed altered steps or interactions (see text and Fig. 2); ADA: adenosine deaminase deficiency; PNP: purine nucleotide phosphorylase deficiency; EBV: Epstein-Barr virus; PKC: protein kinase C; HIV: human immunodeficiency virus; CVID: common variable immunodeficiency; LAD: leucocyte adhesion deficiency; WAS: Wiskott-Mdrich syndrome; SCID: severe combined immunodeficiency; CID: combined immunodeficiency.

Immu,,o;o y Vod,,?

263

Vo; 13 No. 7 1992

REVIEW tative primary defects in early T-cell signal transduction do not prevent T-cell differentiation, suggesting that T-cell signalling in the thymus may have less stringent functional requirements than in the periphery. Lack of CD3~/seems to impair CD8 + more than CD4 + T-cell selection and function, supporting an asymmetrical model for CD4+/CD8 + and TCR-CD3 interaction during T-cell selection and suggesting that TCR-CD3 isoforms lacking either ~ or 8 exist. Intriguingly, CD3 and CD43 defects correlate with low antibody responses to polysaccharides in vivo. One patient lacking CD3~ had severe organ-specific autoimmune features (autoimmune hemolytic anemia and gut epithelial cell autoantibodies), but no non-organspecific autoimmunity; this provides a molecular basis for the study of the frequent but unexplained association between immunodeficiency and autoimmunity 53. The diversity of early transduction defects reflects the complexity of the multitude of signals received by T cells through surface receptors. The preservation of CD28 signalling in primary and secondary CD3 defects provides support for the relative independence of CD28 from CD3 activation pathways, although the reverse is not always true. By contrast, the CD2 pathway seems to require at least some of the TCR-CD3-~ molecules. Redundancy and/or pleiotropy 54 in T-cell adhesion and activation molecules explain the viability of these natural defects, although the range of clinical symptoms is highly variable: from generally lethal in the first decade (defects in ADA/PNP, CD18, CD4/CD8, HLA class 1I, CD43) to relatively benign (defects in CD3 and NF-AT and most functional defects). Conclusion

Anderson, T. (1991) Cell 66, 1133-1144 4 DeFrancn, A.L. (1991) Nature 351,603-604 5 UIIman, K.S., Northrop, J.P.,Verweij, C.L. and Crabtree, G.R. (1990) Annu. Rev. Immunol. 8,421-452 6 Geppert, T.D., Davis, L.S., Gur, H., Wacholtz, M.C. and Lipsky, P.E. (1990) Immunol. Rev. 117, 5-66 7 Gallagher, R.B. (1990) lmmunol. Today 11, 187-189 8 Matsumoto, S., Sakiyama, Y., Ariga, T., Gallagher, R. and Taguchi, Y. (1992) hnmunol. Today 13, 4-5 9 Pdrez-Aciego, P. and Regueiro, J.R. (1991) Immunologia 10, 10-14 10 Fischer, A., Lisowska-Grospierre, B., Anderson, D.C. and Springer, T.A. (1988) Immunodef. Rev. 1, 39-54 11 Alarcon, B., Terhorst, C., Arnaiz-Villena, A., PerezAciego, P. and Regueiro, J.R. (1990) hnmunodef. Rev. 2, 1-16 12 Thoenes, G., Soudais, C., Le Deist, F. etal. (1992)J. Biol. Chem. 267, 487-493 13 Ashwell, J.D. and Klausner, R.D. (1990) Annu. Rev. hnmunol. 8, 139-167 14 Vivier, E., Morin, P.M., O'Brien, C., Schlossman, S.F. and Anderson, P. (1991) Eur. J. lmmunol. 21, 1077-1080 15 Perez-Aciego, P., Alarcon, B., Arnaiz-Villena, A. et al. (1991).1. Exp. Med. 174, 319-326 16 Regueiro, J.R., Tim6n, M., Perez-Aciego, P. et al. (1992) Scand. J. hnmunol. 36 17 Wegener, A.M.K., Letourner, F., Hoeveler, A. et al. (1992) Cell 68, 83-95 18 Regueiro, J.R., Pdrez-Aciego, P., Timon, M. et al. (I 991) Eur..l. hnmunol. 21, 2293-2296 19 Chatila, T., Castigli, E., Pahwa, R. etal. (1990) Proc. Natl Acad. Sci. USA 87, 10033-10037 20 Schorle, H., Holtschke, T., Hfinig, T., Schimpl, A. and Horak, I. (199l) Nature 352, 621-624 21 K/ihn, R., Ralewsky, K. and M/iller, W. (1991) Science 254, 707-710 22 Hirschhorn, R. (1990) lmmunodef. Rev. 2, 175-198 23 Markert, M.L. (1991) hnmunodef. Rev. 3, 45-81 24 Scharenberg, J.G., Rijkers, G.T., Akkerman, J.W., Staal, G.E. and Zegers, B.J. (1990) htt..1, hnmunopharmacol. 12,

T-cell activation is a complex, finely tuned chain of biochemical events that requires the participation of many molecules, both at the cell surface and within the 113-120 cell (Figs 1 and 2). T-cell activation deficiencies, primary 25 Griscelli, C., Lisowska-Grospierre, B. and Mach, B. as well as secondary, are only beginning to be charac- (1989) lmmunodef. Rev. 1, 135-153 terized, and further research is warranted; as progress is 26 Kara, C.J. and Glimcher, L.H. (199l) Science 252, made in understanding these defects, there will be infor- 709-712 mation flow in the opposite direction by unraveling the 27 Cosgrove, D., Gray, D., Dierich, A. et al. (1991) Cell 66, physiological relevance of the affected molecule or path- 1051-1066 way. This, in turn, will allow rational therapies to be 28 Remold-O'Donnell, E. and Rosen, F.S. (1990) developed for the rare primary and, importantly, the hnmunodef. Ref. 2, 151-174 29 Piller, F., Le Deist, F., Weinberg, K.I., Parkman, R. and more common secondary T-cell immunodeficiencies. Fukuda, M. (199l)1. Exp. Med. 173, 1501-1510 We thank A. Blanco, F.J.A. Guisasola, E. Barbosa, A. Garcia 30 De Saint Basile, G. and Fischer, A. (1991) lmmunol. and P. Iglesias-Casarrubios for their help. This work was sup- Today 12, 456-461 31 Sleasman, J.W., Ted&r, T.F. and Barrett, D.J. (1990) ported in part by FIS, C1CYT and NATO grants. Clin. Immunol. hnmunopathol. 55, 401-417 32 Roifman, C.M., Hummel, D., Martinez-Valdez, H. et al. Antonio Arnaiz- Villena, Marcos Tim6n, Carlos Rodriguez- (1989) .1. Exp. Med. 170, 2177-2182 Gallego, Mercedes P&ez-Blas, Alfredo Corell, J. Manuel 33 Jung, L.K., Fu, S.M., Hara, T., Kapoor, N. and Good, Martin-Villa and Josd R. Regueiro are at the Dept of Immu- R.A. (1986).l. Clin. Invest. 77, 940-946 nology, Hospital 12 de Octubre, Universidad Complutense, 34 Chatila, T., Wong, R., Young, M. etal. (1989) New Engl. J. Med. 320, 696-702 28041 Madrid, Spain. 35 Rijkers, G.T., Scharenberg, J.G., Van Dongen, J.J., Neijens, H.J. and Zegers, B.J. (1991) Pediatr. Res. 29, References 306-309 1 Altmann, A., Coggeshall, K.M. and Mustelin,T. (1990) 36 Perez-Blas, M., Arnaiz-Villena, A., Gongora, R. et al. Adv. hnmunol. 48,227-360 (1991) Clin. Exp. Immunol. 85,424-428 2 Rosenstein, Y., Park, J.K., Hahn, W.C. et al. (1991) Nature 37 Doi, S., Saiki, O., Tanaka, T. etal. (1988) Clin. lmmunol. 354, 233-235 Immunopathol. 46, 24-36 3 Stamenkovic, I., Sgroi, D., Aruffo, A., Sun Sy, M. and 38 Raziuddin, S. and Teklu, B. (1989)J. Clin. hnmunol. 9, Immunology Today

264

Vol. 13 No. 7 1992

REVIEW 44X-4.S.3

.Arthr~tra R/WHI. i-l, 56 I-57 I 47

Koni~no,

.\l.F..

Vnlerio,

G., Turco,

et ‘il. ( 1992

h1.C.

c:cd/. 1111111111101. 1.19, Y I-97 40

Disunto,

1. Exp. 41

J.i.,

Mctf.

Kecver,

171, G.P.,

Webster,

I~n~?zzirIoticq-. KU!.

2, 1YY-2

42

Spickctr,

CA.,

Mcuer,

&(I.,

Buschenfddc,

Small,

T.N.

it ‘11. ( 19%))

48

Hauer,

K.H.

A.D.B. and Farrant, j.

(I 9901

IY

,Lz., Kurz,

P., Meyer

zum

(1.H. 44

(;.I’.,

( IY8Sj

Hatzman,

.S&~ZCL~ 241.

I’crcz-Bias,

M.,

( 1992)

A.

4.5 b~uphinee,

121,1.? l’ovx,

( 19x9‘: 46

Fou,

Adnitrs I).;\.,

R.J., Ledbetter, J.R.,

(37.

and June,

Bertotto,

Fip.

/mm4nol.

J.A.,

J. ad 89, 83-88

A. nnd T~lal, J. et d.

(199 1)

Roux,

53

M.bl.

Y,~m+pmi,

hl..

hlartin-Villa.

c’t,I/.( I YY I I

F.

,tnd Schr,lven, s

K.

I_I YXY

i (,'/m.

I3.1-I.39

S., Arcingeli,

(1.

c’t~1.8,I YXY I .Siard.

i-586

I., D~mopoulos,

J. 1 I Y Y I )

X6. 13-l--135 N.

A., Rondoni,

I46

A., Crupi,

Raran,lrh,~r,lthorn.

420-429 Trelsman,

5.C..

I40-

/lllrlzzrr7o~Wthr,1. 50,

Beckman,

Kraciley, 52

Ruiz-Contreras,

Z., Bdlester,

Khew,~. 32, hlillard.

J.A.

573-576

Rrgueiro,

r\rn.lil-Vlllena,

k.,

/. //~m~r~d. i0, i8 51

Lincttc,

hlcuer,

SO Bcrrotto,

743--39 43

49

Irmzum~/

Kohler, H. (1987) 1. (&I. Itwest. X0,

and

Sp1110~71,

~wm~h~