translocation of ppl9/cofilin

4 downloads 84 Views 2MB Size Report
*Applied Immunology, German Cancer Research Center, Im Neuenheimer Feld 280, ... and Max Planck Institute for Biochemistry, Am Klopferspitz, D18033 ...
Proc. Natd. Acad. Sci. USA Vol. 91, pp. 4494-4498, May 1994 Immunology

Costimulatory signals for human T-cell activation induce nuclear translocation of ppl9/cofilin (CD2/sgnal transduction/comulaton/acln-bnding protein)

YVONNE SAMSTAG*t, CHRISTOPH ECKERSKORN4, SEBASTIAN WESSELBORG*, STEFAN HENNING*, REINHARD WALLICH*, AND STEFAN C. MEUER* *Applied Immunology, German Cancer Research Center, Im Neuenheimer Feld 280, D-69120 Heidelberg, Federal Republic of Germany; and Max Planck Institute for Biochemistry, Am Klopferspitz, D18033 Martinsried, Federal Republic of Germany Communicated by Stuart F. Schlossman, February 2, 1994

Here we have identified the 19-kDa phosphoprotein (ppl9), which is regulated by costimulatory signals, as a small actin-binding protein-namely, cofilin. Moreover, we have found that upon CD2 stimulation partially dephosphorylated pp19/cofilin translocates into the nucleus. Finally, dephosphorylation of ppl9/cofilin and its subsequent nuclear translocation occur spontaneously in autonomously proliferating T lymphoma cells, suggesting a relationship between a costimulation-dependent signaling pathway and clonal T-cell growth.

ABSTRACT Resting T lymphocytes that have recognized antigen bound to a major histocompatibility complex molecule with the T-cell receptor require ciniulatory through accessory receptors, includg CD2, CD4, CD8, and CD28, for their clonal growth and expression of their functional repertoires. Absence of costimulation, in contrast, can induce clonal anergy in vitro and selective tolerance in vivo. Here we have defined a potential intracellular messenger for T-cell activation which-is strictly regulated by costimulatory signals mediated through accessory receptors: ppl9/cofilin, a small actinbinding protein, undergoes dephosphorylation and subsequent translocation from the cytosol into the nucleus. In untransformed T cells this process correlates with functional responses essential for the induction of T-cell proliferation (i.e., production of interleukin 2). Moreover, spontaneous dephosphorylation as well as nuclear translocation of ppl9/cofllin occur in the autonomously proliferating T-lymphoma cell line Jurkat.

MATERIALS AND METHODS Cells and Antibodies. Peripheral human mononuclear cells (PBMCs) were obtained by Ficoll-Hypaque (Pharmacia) density gradient centrifugation of heparinized blood from healthy donors. Monocytes were removed by plastic adherence of PBMCs for 1-2 hr at 370C at a density of 3 x 106 cells per ml. Resting human T lymphocytes were purified by rosetting with 2-aminoethylisothiouronium bromide hydrobromide (AET)-treated sheep erythrocytes as described (9).

Engagement of the T-cell receptor (TCR)/CD3 complex by an antigen/major histocompatibility complex (MHC) molecule produces a "first signal" for T-cell activation which results in expression of the interleukin 2 (IL-2) receptor (1). However, whether encounter of antigen/MHC leads to clonal T-cell expansion or, alternatively, clonal anergy (unresponsiveness) depends on the availability of a costimulatory signal or signals. Costimulatory signals can originate from ligand attachment to accessory receptors like CD2, CD4, CD8, or CD28. They are required for the induction of responsiveness to monokines and production of the T-cell growth factor IL-2 (2-10). Intracellular messengers of costimulatory functions-i.e., components of signaling pathways selectively engaged through accessory receptors-have not yet been defined. We have recently observed a cytoplasmic phosphoprotein (ppl9) in human T lymphocytes whose phosphoserine residues are dephosphorylated in response to accessory receptor triggering (i.e., activation by CD2 or CD3 x CD4 or CD3 x CD8 crosslinking), but not after TCR/CD3 stimulation alone (2, 11). Dephosphorylation of ppl9 correlates with the induction of interleukin 6 responsiveness and secretion of IL-2. In contrast, interferon fy (IFN-y) production is inducible to a similar degree through TCR/CD3 triggering alone (2). The latter parameter was used to ensure that-despite their differential activities regarding the observed dephosphorylation event-stimulating monoclonal antibodies directed at TCR/CD3 and the accessory receptors CD2, CD4, or CD8, respectively, exhibited comparable intrinsic triggering capacities.

The human T-cell clone (TCR a/.B/CD3+; CD2+; CD4+; CD8-) was generated as described (12). The human T-lymphoma line Jurkat was maintained in RPMI 1640 medium containing 10%o fetal calf serum. Ascites fluids containing monoclonal antibodies against different epitopes of the human CD2 molecule [T111A (3T48B5), T112 (1OLD4C1), and T113 (lmono2A6)(4)] were kindly provided by S. F. Schlossman (Dana-Farber Cancer Institute, Boston). The rabbit antiserum against a synthetic pp19/cofilin oligopeptide was raised in our own laboratory. 32p Incorporation and Stimulation of T Lymphocytes. For in vivo labeling of cellular phosphoproteins, purified resting T lymphocytes (5 x 106 cells per ml) were incubated overnight (Jurkat cells for 4 hr) with [32P]orthophosphate (Amersham) at 100 puCi/ml (1 Ci = 37 GBq) in phosphate-free medium containing 0.2% bovine serum albumin. For CD2 simulation T cells were exposed to mitogenic concentrations of monoclonal antibodies against different epitopes ofthe human CD2 molecule [TlllA (3T48B5) at 1:3000 ascites fluid in medium, T112 (1OLD4C1) at 1:2000, T113 (lmono2A6) at 1:2000). Cells were then harvested and lysed with 1% Nonidet P-40 as described (2). Lysates were centrifuged at 4400 X g for 10 min. Postnuclear lysates comprising cytosol plus membrane were analyzed on two-dimensional (2D) gels after addition of 0.2% SDS. Nuclei were washed once in lysis buffer, centrifuged again, and then dissolved and extracted in lysis buffer containing 0.2% SDS for 20 min. Insoluble material was Abbreviations: TCR, T-cell receptor; IL-2, interleukin 2; IFN-'y, interferon y; 2D, two-dimensional; FITC, fluorescein isothiocya-

The publication costs of this article were defrayed in part by page charge

payment. This article must therefore be hereby marked "advertisement"

nate. tTo whom reprint requests should be addressed.

in accordance with 18 U.S.C. §1734 solely to indicate this fact.

4494

Immunology: Samstag et al.

Proc. Natl. Acad. Sci. USA 91 (1994)

pelleted at 13,000 x g for 20 min and nuclear proteins were separated by 2D PAGE. 2D PAGE of Radlolabeled Ceil Lysates. 2D PAGE was carried out as described (13), using isoelectric focusing (IEF) in the first dimension (LKB ampholytes: 1.45% pH 3.5-10, 0.1% pH 2.5-4, 0.2% pH 4-6, 0.2% pH 9-11) and SDS/ PAGE (12% polyacrylamide) in the second dimension. Equivalents of 2 x 106 32P-labeled cells were loaded on each gel. Dried gels were exposed to Kodak X-AR autoradiography films. Peptide Mapping. Peptide mapping employing Staphylococcus aureus V8 protease digestion was performed as described previously (14). Protein Sequencing. ppl9 spots obtained from 50 individual preparative 2D gels were stained with Coomassie blue, excised, and digested with trypsin directly in the gel matrix (15). The eluted peptides were separated by reversed-phase HPLC. Sequence analysis was performed as described (16). Immunoblots. 2D PAGE was carried out as described (17) using nonequilibrium pH gradient electrophoresis (NEPHGE) in the first dimension (LKB ampholytes: 1.45% pH 3.5-10, 0.1% pH 2.5-4, 0.2% pH 4-6, 0.2% pH 9-11) and SDS/PAGE (15% polyacrylamide) in the second dimension. Equivalents of 10 x 106 cells were loaded on each gel. Separated proteins were blotted onto Hybond C extra nitrocellulose (Amersham) by employing the Multiphor II Nova Blot electrophoretic transfer unit (Pharmacia LKB). Filters were blocked overnight at 4°C in Tris-buffered saline (TBS; 10 mM Tris HCl, pH 7.5/150 mM NaCl) containing 5% nonfat dry milk. They were rinsed three times with TBS containing 0.2% Tween-20 (TBST) and were incubated for 2 hr with affinity-purified rabbit anti-ppl9/cofilin antiserum diluted 1:500 in TBS/5% nonfat dry milk. After three

A

4495

washes in TBST, filters were incubated for 1 hr with goat anti-rabbit alkaline phosphatase-conjugated antibody (Dianova, Hamburg, F.R.G.) 1:1000 in TBS/5% nonfat dry milk. Filters were washed three times with TBST, rinsed twice with TBS, and incubated with the alkaline phosphatase substrate 5-bromo-4-chloro-3-indolyl phosphate (Sigma) at 330 mg/liter in nitro blue tetrazolium (Sigma) at 165 mg/liter in 9.7% (vol/vol) diethanolamine/5 mM MgCl2/ 0.02% NaN3 for 10 min. Confocal Laser Scanning Microscopy. Cloned cells were rearrested by culture in IL-2 for 10 days. Then they were stimulated with monoclonal antibodies as described above, washed three times with phosphate-buffered saline (PBS) and allowed to sediment on adhesion slides (Bio-Rad) for 20 min (4 x 106 cells per ml). Slides were rinsed once with PBS, and cells were fixed and permeabilized with 100%1 ethanol for 5 min. Cells were incubated for 1 hr with affinity-purified rabbit anti-ppl9/cofilin antiserum diluted 1:20 in PBS/0.1% gelatin/0.02% NaN3 (PBS/G). After three rinses with PBS they were incubated for 1 hr with swine anti-rabbit fluorescein isothiocyanate (FITC)-conjugated antibody (Dako) diluted 1:20 in PBS/G. Slides were washed three times with PBS and incubated for 15 min with propidium iodide (Sigma) at 0.2 ,ug/ml in PBS. After three washes in PBS slides were covered with Vectorshield (Vector Laboratories). Digitized images were generated by using a confocal laser scanning microscope (Zeiss LSM 10). For evaluation of two-color experiments digitized images were stored for each fluorochrome and then overlayed electronically.

RESULTS To investigate the role of ppl9 in the cascade of signals between cell surface and nucleus, nuclear and postnuclear preparations

nucleus

cytosol/mermbrane

Control d 3 2

CD2

..

0

(1 Omin) f

e

CD2

-.

(30min)

B

h

9 IC ( 2

(60min) O *.I Jurkat u n s ti m u -

i

3_

4

3 2 1

.0

ated

FIG. 1. Dephosphorylation of ppl9 upon CD2 stimulation is accompanied by nuclear translocation. (A) 2D gel electrophoresis of 32P-labeled subceliular fractions shows that activation of resting T lymphocytes through CD2 results within 10-30 min in a transient decrease of phosphorylated ppl9 in the cytoplasmic fraction (gels a, c, e, and g). Concomitantly, two 19-kDa phosphoproteins with more basic pI appear in the nuclear compartment (b) and become progressively dephosphorylated with time (d, f, and h). Dephosphorylation and translocation of ppl9 occurs spontaneously in the autonomously proliferating T-cell lymphoma line Jurkat (arrows in gels i andj). IEF, isoelectric focusing. (B) Peptide

map analysis of the cytosolic (spot 1 ofA) and nuclear (spots 2 and 3 ofA) 19-kDa phosphoproteins confirms their identity (lane 1, V8 protease digest of spot 1; lane 2, spot 2; lane 3, spot 3). Besides three identical phosphopeptides (triangles; the second open triangle marks a phosphopeptide which is faintly visible but consistently observed), a strongly phosphorylated 6-kDa peptide (arrow) of the cytosolic ppl9 is not detectable in digests of the nuclear ppl9 molecule. Peptide maps of the respective spots from Jurkat cells (arrows in A) are identical (not shown). (C) Besides the phosphorylated cytosolic ppl9 molecule (spot 1), partially dephosphorylated forms of ppl9 with more basic pI (spots 2 and 3) are visible in total lysates of untransformed resting human T lymphocytes upon triggering by CD2. The peptide maps of spots 1-3 are identical to those of spots 1-3 in A (not shown).

44%

Immunology: Samstag et al.

of human T lymphocytes were obtained at different times after stimulation with CD2 and compared in 2D gel electrophoresis. Within 10-30 min this mode of activation leads to a transient decrease of pp19 spot intensity in the cytoplasmic fraction (Fig. 1A, gels a, c, e, and g). Simultaneously, two 19-kDa phosphoproteins are detected in the nuclear fraction (Fig. 1A, gels b, d, f, and h). Peptide maps of the cytoplasmic spot 1 and the more basic nuclear spots 2 and 3, respectively (Fig. 1B), indicate that the two latter molecules represent the same protein as spot 1 at different levels of phosphorylation. Besides three identical phosphoprotein bands, one strongly phosphorylated 6-kDa fragment present in spot 1 is missing from peptide maps of spots 2 and 3. Since the molecular masses of the above three ppl9 species are identical (Fig. 1 A and C), disappearance of this 6-kDaphosphopeptide fragment must be interpreted as dephosphorylation rather than degradation. The more basic forms 2 and 3 of ppl9 are also detectable in total cellular lysates, comprising both nuclei and cytoplasm, of CD2-stimulated cells (Fig. 1C). Thus, CD2 stimulation induces both dephosphorylation of pp19 and its subsequent translocation into the nucleus (see also Fig. 4, which confirms its intranuclear localization by confocal laser scan microscopy). Given that the dephosphorylation of ppl9 occurs on serine residues, it was necessary to investigate whether the type 2B serine phosphatase calcineurin, which is inhibited by cyclosporin A (18), was involved in this regulatory process. Although not shown here, this is clearly not the case, since high concentrations of cyclosporin A that completely block lymphocyte proliferation are incapable of inhibiting the dephosphorylation of ppl9. In this regard, our previous experiments performed with the serine phosphatase inhibitor okadaic acid suggested instead that type 1 or 2A serine phosphatases are responsible for ppl9 dephosphorylation (11). In the autonomously growing T-lymphoma line Jurkat the constitutive phosphorylation state of ppl9 corresponds to that observed in untransformed T cells when stimulated through CD2 (Fig. 1A; compare gels i andj with c and d), indicating that ppl9 undergoes spontaneous dephosphorylation and nuclear translocation in this malignant cell type. Similarly, ppl9 is constitutively dephosphorylated in a series of additional leukemia lines (e.g., Molt4, HPB-ALL, the Epstein-Barr virustransformed B cell line LAZ 509, and K562). In marked contrast, some other protein alterations known to occur in T-cell activation-e.g., the molecular mass shifts of the CD2/CD45associated phosphoproteins pp29/32-do not spontaneously occur in Jurkat cells (19), suggesting a deregulated process which is distal to the CD2 receptor complex (20). ppl9 spots obtained from preparative 2D gels were excised and digested with trypsin. Resulting peptides were separated by reversed-phase HPLC. Four discrete peptide peaks underwent microsequencing and yielded distinct amino acid sequence information. As shown in Fig. 2, the ppl9-derived peptides are completely identical with four internal peptide sequences of cofilin derived from human placenta (21-23). Moreover, we have cloned cofilin from a human T-cell cDNA library and found that cofilin expressed by human T lymphocytes exhibits complete sequence homology with the previously identified placental molecule. Finally, cofilin that we have expressed in Escherichia coli (not shown) and the phosphorylated ppl9 molecule in T cells both react in immunoblot analysis with a rabbit antiserum raised against a synthetic cofilin oligopeptide (see Fig. 3). 2D-immunoblot analysis of cell lysates, employing the affinity-purified rabbit antiserum to ppl9/cofilin (Fig. 3) detected ppl9 in phosphorylated form as well as a substantial pool in apparently unphosphorylated form. The identity of the more acidic spot with the phosphorylated ppl9 molecule could be confirmed by autoradiography of immunoblots performed with 32P-labeled lysates (not shown). Exposure of resting T cells to mitogenic CD2 antibodies induces a shift

Proc. Natl. Acad. Sci. USA 91 (1994)

from phosphorylated to unphosphorylated ppl9. In preactivated cells (untransformed T-cell clones) the fraction of unphosphorylated ppl9 is greater than in resting T cells and the phosphorylated molecule disappears upon CD2 stimulation. In the transformed T-cell line Jurkat, ppl9 exists mainly in the unphosphorylated form. Immunofluorescence staining of untransformed T-cell clones employing the ppl9/cofilin antiserum, combined with confocal laser scan microscopy, shows that in unstimulated cells the majority of the total ppl9/cofilin is diffusely distributed in the cytoplasm. In contrast, in response to triggering of the CD2 receptor a large proportion of this protein appears in distinct intranuclear aggregates (Fig. 4). Further colocalization studies are required to identify the components which associate with cofilin after its translocation into the nuclear compartment.

DISCUSSION Activation of human T lymphocytes through CD2 or CD3 x CD4 or CD3 x CD8 crosslinking engages a cyclosporin A-resistant signaling pathway which leads to dephosphorylation of a 19-kDa cytoplasmic molecule. The latter has now been identified as the small actin-binding protein cofilin. The linkage of this signaling event to an accessory receptorinducible pathway appears to be selective, since triggering of T cells through the TCR/CD3 complex alone, despite inducing a similar degree of IFN-y secretion, does not alter the phosphorylation state of ppl9/cofilin (2, 11). ppl9/cofilin possesses a sequence (KKRKK) (see Fig. 2) which is similar to the nuclear translocation signal sequence of simian virus 40 T antigen (21, 22, 24, 25). Serine residues in the vicinity of such nuclear translocation sequences contribute to the control of the activation of these signals (26, 27). Note that ppl9/cofilin contains two serine residues that are located next to the nuclear translocation sequence and may, therefore, control its nuclear entry. Such a view is supported -uCi ed LraIlsiOlc-tilCi sequenice R

ip

K

~

E

E

V

K

K R

K

K

A-V__

-l * * * _

V-

KID K

D

E

F

.,

.

i.

Ar

FIG. 2. Comparison of sequences of four internal ppl9 peptides with the sequence of human placental cofilin. As indicated by *, the peptides are identical with cofilin sequences.

Immunology: Sarnstag et al.

Proc. Natl. Acad. Sci. USA 91 (1994)

4497

CD2 (30min)

Control

b a'

resting T cells

I0

c

untransformed T cell clone e T cell lymphoma

a' I.

(Jurkat) basic

4

NE PHGE

acidic

basic

-

N PH GF

-

:i' ciC.

FIG. 3. In contrast to the situation in resting T lymphocytes (gels a and b) or untransformed T-cell clones (gels c and d), ppl9 is constitutively dephosphorylated in the malignant T-cell lymphoma line Jurkat (gels e and f). 2D-immunoblot analysis revealed two distinct spots. The more acidic spot (closed arrows) corresponds to the phosphorylated form of ppl9/cofllin. The more basic spot (open arrows) represents an unphosphorylated form of the ppl9/cofllin molecule. Gels a, c, and e are from unstimulated cells; b, d, and f are from cells stimulated through CD2. NEPHGE, nonequilibrium pH gradient electrophoresis.

One actin-binding motif of ppl9/cofilin also interacts with phosphatidylinositol 4,5-bisphosphate (PIP2) but not with its cleavage products inositol 1,4,5-trisphosphate (1P3) and diacylglycerol. Moreover, PIP2 binding inhibits the capacity of ppl9/cofilin to associate with and depolymerize F-actin (31, 32). Since PIP2 is a substrate of phospholipase C, ppl9/cofilin might be involved in the earliest events known to occur in T-cell activation. Several points provide further support for the notion that ppl9/cofilin acts as a second messenger rather than exerting a downstream effect of a variety of stimuli. Dephosphorylation and nuclear translocation occur within 10 min (Fig. 1) and, thus, represent very early events in lymphocyte triggering. In contrast, expression of CD2-inducible genes essential for the induction ofT-cell proliferation-e.g., IL-2-is maximal at a much later time (5-6 hr), and, therefore, ppl9/cofilin translocates clearly before T-cell growth ensues. Moreover, the finding that cyclosporin A, despite abolishing DNA synthesis, does not prevent ppl9/cofilin dephosphorylation points toward an inductive function of this protein

by our finding that nuclear translocation of ppl9/cofilin was always accompanied by its partial dephosphorylation and that we have never observed the completely phosphorylated ppl9 molecule in the nuclear fraction. In addition, ppl9/cofilin contains two actin-binding motifs (28, 29). In unstimulated cells actin is found mainly in extranuclear filaments. Association of cofilin with actin induces a shift from filamentous (F) toward oligomeric or monomeric (G) actin (30). Moreover, upon heat shock or dimethyl sulfoxide treatment of a mouse fibroblast cell line, cofilin undergoes dephosphorylation and forms intranuclear actin/cofilin rods. Experiments with FITC-labeled actin showed that during this process cytoplasmic actin filaments are depolymerized and the resulting monomeric or oligomeric actin translocates into the nucleus (22, 23, 25). We could also colocalize actin and cofilin within nuclei of T lymphoblasts. Given that actin itself does not contain a nuclear translocation sequence, cofilin may depolymerize actin and transport it into the nucleus (25). a

Control

b

CD2

FIG. 4. Confocal laser scanning microscopy of a T-cell clone stained with FITC-conjugated antiserum to ppl9/cofilin (green and yellow) and counterstained for DNA with propidium iodide (red) shows intranuclear localization of ppl9/cofilin after CD2 stimulation (the yellow results from colocalization of green and red fluorescence). The vast majority of the total ppl9/cofilin is present in the cytosol in unstimulated cells (a) and translocates into the nucleus in response to CD2 stimulation (b). Note that some cells in the "resting" population show weak intranuclear staining with the anti-ppl9/cofilin antiserum. This may be due to the fact that clonal T-cell cultures are not synchronized, so not all cells exist in a true resting state.

4498

Immunology: Samstag et al.

rather than its involvement in a secondary event associated with cell proliferation. Functional consequences which result from accumulation of cofilin inside the nucleus are unknown at present. Cotranslocation of monomeric actin, however, suggests a potential influence on transcriptional processes, since monomeric actin is proposed to be required for accurate transcription by RNA polymerase II. Thus, injection of actin antibodies into oocyte nuclei drastically inhibits the transcriptional activity of RNA polymerase II and leads to changes in chromosome morphology (33, 34). In contrast, nucleolar transcription by RNA polymerase I is not influenced under these conditions. Perhaps more important, monomeric actin represents a potent inhibitor of DNase I (35). The latter was recently identified as a key enzyme involved in cell death through apoptosis (36). Interestingly, T-cell stimulation through CD2 prevents antibody- (Apo-1) induced apoptosis (37). Although it has not yet been formally proven, these findings strongly support the concept that dephosphorylation of ppl9/cofilin and the resulting nuclear accumulation of actin/cofilin complexes may exert an anti-apoptotic activity during T-cell activation. The constitutive presence of large quantities of dephosphorylated cofilin within nuclei of permanently proliferating transformed cells is in accordance with such a view. These potential functions ofintranuclear actin/cofilin complexes precisely represent intracellular effector mechanisms which one would expect to be regulated through a pathway transmitting costimulatory signals for T-cell activation. Accordingly, dephosphorylation of ppl9/cofllin should not be inducible in anergized T cells, which are incapable of responding to accessory receptor triggering with clonal expansion (4). We have investigated this point by employing TCR/CD3 down-modulated peripheral T lymphocytes (4), and indeed we found that, despite expression of CD2 on the cell surface at normal levels, triggering of these anergized cells with mitogenic monoclonal antibodies to CD2 does not result in pp19 dephosphorylation (not shown). In conclusion, dephosphorylation of pp19/cofilin and nuclear translocation of the actin/cofilin complex may play an important role for the "decision" following T-cell receptor triggering between T-cell activation toward clonal expansion or T-cell unresponsiveness (clonal anergy/apoptosis). This "decision" is believed to depend on the presence or absence of accessory receptor-mediated costimulatory signals (3, 10, 38). Moreover, constitutive dephosphorylation and nuclear translocation of ppl9/cofilin as observed in Jurkat cells may represent a mechanism supporting unregulated proliferation and the prolonged lifespan of malignant cells. The authors thank Dr. D. Schiller-Gramlich for expert advice in

protein purification procedures and Dr. P. Lichter and A. Kurz for their support in laser scanning microscopy. We are indebted to Prof. W. W. Franke and Prof. T. Stachelin for their critical comments on the present manuscript. We also thank C. Brenner, H. Holz, and M. Zummack for excellent technical assistance. Finally, we acknowledge expert photographic work by B. Engelhardt, J. Jung, and R. Kuhnl-Bontzol. This work is supported by a grant from the Deutsche Forschungsgemeinschaft (DFG Me-693/3-1). 1. Meuer, S. C. & Meyer zum Bfschenfelde, K.-H. (1986) J. Immunol. 136, 4106-4112. 2. Samstag, Y., Henning, S. W., Bader, A. & Meuer, S. C. (1992) Int. Immunol. 4, 1255-1262. 3. Mueller, D. L., Jenkins, M. K. & Schwartz, R. H. (1989) Annu. Rev. Immunol. 7, 445-480. 4. Meuer, S. C., Hussey, R. E., Fabbi, M., Fox, D., Acuto, O.,

Proc. Nad. Acad. Sci. USA 91

5.

6. 7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

(1994)

Fitzgerald, K. A., Hodgdon, J. C., Protentis, J. P., Schlossman, S. F. & Reinherz, E. L. (1984) Cell 36, 897-906. Emmrich, F., Strittmatter, U. & Eichmann, K. (1986) Proc. Natl. Acad. Sci. USA 83, 8298-8302. Samstag, Y., Emmrich, F. & Staehelin, T. (1988) Proc. Natl. Acad. Sci. USA 85, 9689-9693. Rudd, C. E., Trevillyan, J. M., Dasgupta, J. D., Wong, L. L. & Schlossman, S. F. (1988) Proc. Natl. Acad. Sci. USA 85, 5190-5194. Veillette, A., Bookman, M. A., Horak, E. M. & Bolen, J. B. (1988) Cell 55, 301-308. Endler-Jobst, B., Schraven, B., Hutmacher, B. & Meuer, S. C. (1991) J. Immunol. 146, 1736-1742. Gimmi, C. D., Freeman, G. J., Gribben, J. G., Gray, G. & Nadler, L. M. (1993) Proc. Natl. Acad. Sci. USA 90, 65866590. Samstag, Y., Bader, A. & Meuer, S. C. (1991) J. Immunol. 147, 788-794. Wesselborg, S., Janssen, O., Pechhold, K. & Kabelitz, D. (1991) J. Exp. Med. 173, 297-304. O'Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021. Schraven, B., Samstag, Y., Altevogt, P. & Meuer, S. C. (1990) Nature (London) 345, 71-74. Eckerskorn, C. & Lottspeich, F. (1990) Chromatographia 28, 92-94. Eckerskorn, C., Mewes, W., Goretzki, H. & Lottspeich, F. (1988) Eur. J. Biochem. 176, 509-519. O'Farrell, P. Z., Goodman, H. M. & O'Farrell, P. H. (1977) Cell 12, 1133-1142. Schreiber, S. L. & Crabtree, G. R. (1992) Immunol. Today 13, 136-142. Schraven, B., Schirren, A., Kirchgessner, H., Siebert, B. & Meuer, S. C. (1992) Eur. J. Immunol. 22, 1857-1863. Schraven, B., Wild, M., Kirchgessner, H., Siebert, B., Wallich, R., Henning, S., Samstag, Y. & Meuer, S. C. (1993) Eur. J. Immunol. 23, 119-123.

21. Ogawa, K., Tashima, M., Yumoto, Y., Okuda, T., Sawada, H., Okuma, M. & Maruyama, Y. (1990) Nucleic Acids Res. 18, 7169. 22. Ohta, Y., Nishida, E., Sakai, H. & Miyamoto, E. (1989) J. Biol. Chem. 264, 16143-16148. 23. Nishida, E., lida, K., Yonezawa, N., Koyasu, S., Yahara, I. & Sakai, H. (1987) Proc. Natl. Acad. Sci. USA 84, 5262-5266. 24. Kalderon, D., Roberts, B. L., Richardson, W. D. & Smith, A. E. (1984) Cell 39, 499-509. 25. Iida, K., Matsumoto, S. & Yahara, I. (1992) Cell Struct. Funct. 17, 39-46. 26. Rihs, H. P. & Peters, R. (1989) EMBO J. 8, 1479-1484. 27. Addison, C., Jenkins, J. R. & Sturzbecher, H. W. (1990) Oncogene 5, 423-426. 28. Yonezawa, N., Nishida, E., lida, K., Kumagai, H., Yahara, I. & Sakai, H. (1991) J. Biol. Chem. 266, 10485-10489. 29. Yonezawa, N., Nishida, E., Ohba, M., Seki, M., Kumagai, H. & Sakai, H. (1989) Eur. J. Biochem. 183, 235-238. 30. Nishida, E., Maekawa, S. & Sakai, H. (1984) Biochemistry 23, 5307-5313. 31. Yonezawa, N., Nishida, E., Iida, K., Yahara, I. & Sakai, H. (1990) J. Biol. Chem. 265, 8382-8386. 32. Yonezawa, N., Homma, Y., Yahara, I., Sakai, H. & Nishida, E. (1991) J. Biol. Chem. 266, 17218-17221. 33. Scheer, U., Hinssen, H., Franke, W. W. & Jockusch, B. (1984) Cell 39, 111-122. 34. Rungger, D., Rungger-Brandle, E., Chaponnier, C. & Gabbiani, G. (1979) Nature (London) 282, 320-321. 35. Blikstad, I., Markey, F., Carlsson, L., Persson, T. & Lindberg, U. (1978) Cell 15, 935-943. 36. Peitsch, M. C., Polzar, B., Stephan, H., Crompton, T., MacDonald, H. R., Mannherz, H. G. & Tschopp, J. (1993) EMBO J. 12, 371-377. 37. Klas, C., Debatin, K.-M., Jonker, R. R. & Krammer, P. H. (1992) Int. Immunol. 5, 625-630. 38. Zanders, E. D., Lamb, J. R., Green, N., Feldmann, M. & Beverley, P. C. (1983) Nature (London) 303, 456-458.