CELLULAR
IMMUNOLOGY
144, 182- 189 ( 1992)
Involvement of a Protein Kinase C and Protein Phosphatases in Adhesion of CD4+ T Cells to and Detachment from Extracellular Matrix Proteins SHMUELMIRON,* RAMI HERSHKOVIZ,*IRIS TIROSH,* YORAM SHECHTER,t AVNER YAYUN,$ AND OFER LIDER*,’ Departments of *Cell Biology, $Chemical Immunology, and -fHormone Research, The Weizmann Institute of Science,Rehovot 76100, Israel Received April 16, 1992; acceptedJune 13, 1992 For immune surveillance and function to be effective, T lymphocytes constantly recirculate via lymph and blood between lymphoid organsand body tissues.To enable efficient cell movement and migration, cell adhesion to components of the basement membrane and the extracellular matrix (ECM) must be a rapid and transitory process. Whether phosphorylation and dephosphorylation of cellular proteins are involved in this phenomena was explored by monitoring the adhesion of T cells to immobilized ECM proteins. A short exposure of “Cr-labeled human CD4+ T cells to phorbol estersin vitro induced a rapid pl-integrin-mediated adhesion to both fibronectin and laminin, as determined by inhibition with anti-integrin antibodies. Adhesion was reversible; detachment from the immobilized ECM ligands occurred between 20 and 120 min without further intervention. This T cell adhesion was regulated by the activation of protein kinase C because(a) staurosporine and H-7 inhibitors of protein kinase C suppressedT cell adhesion, and (b) PMAinduced down-regulation of intracellular levels of protein kinase C was associatedwith the abrogation of the T cell adhesivenessto fibronectin and laminin. Furthermore, inhibition of protein phosphatasesactivity by okadaic acid delayed the detachment of the T cells from fibronectin or laminin. Thus, we suggestthat T cell-ECM interactions such as adhesion and detachment are regulated, respectively, by protein kinase C and protein phosphatases. o 1992 Academic PW., IX.
INTRODUCTION CD4+ T lymphocytes circulate within the body in search of peptide antigens presented on accessorycells in the context of the MHC class II molecules. Additionally, T cells interact in a precisely regulated fashion with normal or inflamed endothelium and thereby extravasate (1, 2). Simultaneously, enzymes are activated to enable the penetration of the ECM toward inflamed sites (3). Therefore, T cell adhesion to and dissociation from extracellular matrix (ECM)’ proteins are pivotal steps in T cell migration, a basis of immune surveillance (4, 5). ’ O.L. is the recipient of the Alon scholarshipand incumbent of the Weizmann LeagueCareer Development Chair in Children’s Diseases.Address correspondence to Dr. 0. Lider, The Department of Cell Biology, The Weizmann Institute of Science, P.O. Box 26, Rehovot, 76 100, Israel. * Abbreviations used: ECM, extracellular matrix; FN, fibronectin; LN, laminin; PKC, protein kinase C; PMA, phorbol- 12-myristate-13-acetate;PrP, serine/threonine protein phosphatases;TPA, I2-O-tetradecanoylphorbol- 13-acetate;mAb, monoclonal antibody. 182 0008-8749/92 $5.00 Copyright 0 1992 by Academic Press,Inc. All rights of reproduction in any form reserved
REGULATION
OF T CELL-EXTRACELLULAR
MATRIX
INTERACTIONS
183
T cells recognize protein components of the basement membrane and ECM, such as fibronectin (FN) or laminin (LN), through /31-integrin receptors (CD-29) designated very late antigens (4). T cell-ECM adhesion is rapidly augmented upon cell activation, without changesin the level of expression of the VLA4, VLA-5, and VLA-6 receptors (4, 5). The molecular basis for this T cell adhesion following activation is unknown. Nevertheless, it has been suggestedthat this processis regulated by intracellular signals which induce conformational change(s)of the integrin receptor, resulting in increased avidity of the T cell integrin to its ECM ligand (5). Indeed, the possibility that intracellular signals might regulate integrin affinity was recently examined by us. We have shown that inhibition of tyrosine protein kinase (6) resulted in partial abrogation of activated CD4+ T cell adhesion to ECM ligands (7). Upon antigen recognition by the T cell receptor-CD3 complex, T cells interact with accessory cells in a dynamically regulated fashion. A cascadeof protein phosphorylation and/or production of second messengersleads to cell-cell adhesion via, for example, LFA- 1 and ICAM- 1 receptors, followed by a decreasein receptor avidity and subsequent detachment (1, 2). We asked whether these types of transient cellcell and cell-ECM interactions might be modulated and reversed by protein phosphatases. It has been demonstrated recently that the pattern of T cell-ECM transient adhesiveness was identical whether the cells were activated by phorbol-12-myristate-13acetate (PMA), by 12-O-tetradecanoylphorbol- 13-acetate (TPA), or via the antigenspecific T cell receptor-CD3 pathways (8). In the present study we examined the regulatory role of protein kinase C (PKC) in PMA-induced T cell adhesion to FN and LN, since this kinase is also involved in the initial steps of T cell activation (9) and in cell-cell interactions and adhesion ( lo- 12). We demonstrate that PKC is involved in CD4+ T cell-IN and -LN interactions and suggestthat the rapid progression from adhesion to detachment states is due to dephosphorylation induced by protein phosphatases(PrP). MATERIALS AND METHODS T cells. CD4+ T cells were obtained from peripheral blood mononuclear leukocytes obtained from healthy human donors as follows. The mononuclear cells were isolated on a Ficoll gradient, washed, and incubated at 37°C in a 10% CO;! humidified atmosphere. After 2 hr, the nonadherent cells were removed and incubated on nylonwool columns (Fenwall, IL) for 1 hr at 37°C in a 10% CO* humidified atmosphere. Nonadherent cells were eluted and washed. CD4+ T cells were negatively selected by exposure of the eluted cells to a mixture of the following monoclonal antibodies (mAb): anti-CDs, CD1 9, and CD 14 conjugated to magnetic beads (Advanced Magnetics, MA). The remaining cells were recovered and exposed to a second round of negative selection. The resulting cell population consistedof over 90% CD3+CD4+ asdetermined by FACScan analysis. Tcell adhesion assay. T cell adhesion to the ECM protein components was designed and performed as previously described (4, 8). FN or LN (Sigma, St. Louis, MO), 40kDa heparin-binding site, and 120-kDa cell attachment site of FN (Telios Pharmaceuticals, Inc., San Diego, CA) were used as adhesion substratesat a concentration of 1 pug/wellin 100 yl PBS which was washed away after 16 hr of incubation. To activate the T cells, different concentrations of PMA or TPA (Sigma) in DMSO (final concen-
184
MIRON
ET AL.
tration 0.0 1%)were added to the microtiter wells containing 5’Cr labeled cells. Results are expressedasthe mean percentageof T cells binding from triplicate or quadriplicate wells per experimental group *SD. Modulation of T cell adhesion. 51Cr-labeled purified human CD4+ T cells were incubated with mAb against integrin receptors, PKC inhibitors, or okadaic acid either prior to or upon initiation of the adhesion assayto FN- or LN-coated wells. Labeled T cells were incubated with either Gly-Arg-Gly-Asp-Ser-Pro-Lys (GRGDSPK) or Gly-Arg-Gly-Glu-Ser-Pro (GRGESP) peptides (50 pg/ml; Sigma) for 30 min at 37“C before the start of the assay of adhesion. Similarly labeled T cells were preincubated with anti-VLA-5 (Telios), anti-VLA-6 (CDw49f, GoH3) mAb (CLB, Amsterdam, Holland), or anti-CD-29 (Serotec, GB) mAb to inhibit T cell adhesion to FN and/or LN. Monoclonal antibodies (clone 3E3 diluted l/400) to the cell attachment site of FN (Boehringer-Mannheim) were used to pretreat the coated wells for 1 hr at 37°C at 10% CO2 humidified atmosphere before the start of the adhesion assay. To activate PKC, labeled T cells were exposed to PMA. To inhibit PKC, labeled CD4+ T cells were treated (15 min at 37°C) with staurosporine (0.0 l-O.5 piW; Sigma) or with H-7 (lo-100 PM; Siekagaku America, Inc., Rockville, MD). Freshly isolated CD4+ T cells were incubated for 18 hr in a humidified incubator in culture medium, washed, and labeled. These cells were then treated with staurosporine, transferred into the culture wells, and activated with PMA, and their adhesion to FN or LN was examined. To down-regulate PKC, labeled T cells were incubated with PMA (0- 100 rig/ml) for 18 hr ( 10) prior to washing and initiation of the adhesion assayin the presence or absenceof the PKC activator PMA. To inhibit PrP T cells were treated with 1 PALM okadaic acid. The okadaic acid was added simultaneously with or following PMA (10 rig/ml) addition, which was added at the initiation of the assay. RESULTS AND DISCUSSION Eficts of PMA on CD4’ T cell adhesion to FN and LN. To study the kinetics of binding and detachment of CD4+ T cells to and from ECM proteins, the T cells were radioactively labeled and seededin wells coated with either FN or LN and then activated with PMA. Nonactivated T cells did not adhere to FN nor to LN. In contrast, PMAactivated CD4+ T cell adhesion to both FN (Fig. 1A) and LN (Fig. 1B) and maximal cell adhesion were achieved by 15 min of activation, irrespective of the PMA concentration (10, 50, or 100 rig/ml). T cell adhesion to FN and LN induced by the lower PMA concentrations (lo-50 rig/ml) was higher than that induced by a higher dose (100 rig/ml). By 60 min of PMA activation, T cell adhesion decreased. Similarly, activation of the T cells with 80 nM TPA, a co-mitogen which induces cell adhesion by PKC activation (l), could induce T cell adhesion in a time-dependent fashion (Figs. 1C and 1D). Hence, the adhesion state is transient and is followed by detachment of the cells from the ECM protein components. Regulation by integrin receptors of PMA-induced CD4’ human T cell adhesion to FN and LN. T cell adhesion to wells coated with FN or the 120-kDa cell attachment site of FN, but not to LN or to the heparin-binding site of FN (40-kDa fragment; data not shown), was inhibited by anti-FN antibodies and by anti-CD-29 mAb, which is specific for the pl chain of the integrin receptor (Table 1). Anti-VLA-5 (a5/31) mAb
REGULATION
OF T CELL-EXTRACELLULAR 80
60
MATRIX
185
INTERACTIONS
c
1
50 40 30 20 10 0 i
0
20
40
60
80
100
120
0
20
40
60
80
100
120
60
80
100
120
0
20
40
60
80
100
120
e
0
20
40
Time
(mln)
FIG. 1. Effect of PMA or TPA induced activation on human CD4+ T cell adhesion to FN and LN. Purified CD4+ T cells were seededon FN (A and C) or LN (B and D) coated microtiter wells. The cells were then either left intact (a) or activated with different concentrations of PMA (A and B): IO (0) 50 (0) or 100 rig/ml PMA (H), or with TPA (W)(C and D). Results are expressedas percentage of T cell (&SD) adhesion to the ECM proteins. The results shown here represent five experiments that produced essentially similar results.
inhibited T cell adhesion to FN and its 120-kD cell-attachment site but did not affect cell adhesion to LN, whereas anti-VLA-6 (a6/31) mAb inhibited T cell adhesion to LN but not to FN. Thus, the adhesion of activated CD4+ T cells to FN and LN is primarily regulated by way of specific /31-integrins. Lymphoid cells recognize and attach to immobilized FN via the VLA-5 receptor which recognizesthe tripeptide Arg-Gly-Asp (RGD) sequencepresent in the cell-binding domain of FN ( 13, 14). The binding of human CD4+ T cells to FN was specifically inhibited by the soluble RGD-containing peptide GRGDSPK but not by the nonRGD containing peptide GRGESP (Table 1). In contrast, the RGD-containing peptide did not interfere with T cell adhesivenessto immobilized LN. Indeed, it has been demonstrated that none of the integrins bind LN in an RGD-dependent manner (1315). Thus, PMA-induced CD4+ T cell adhesion to FN and LN appearsto be regulated by FN- or LN-specific VLA receptors. Involvement of PKC activation in T cell adhesion to FN or LN. It has been demonstrated that PKC, activated directly by PMA (16) is inhibited by the alkaloid compound staurosporine and by H-7 ( 1, 17, 18). Moreover, prolonged exposure of human
186
MIRON ET AL. TABLE 1 Effects of FN Derivatives and mAb against Integrin Receptors on PMA-Induced CD4+ T Cell Adhesion to FN or LN Percentageof T cell adhesion (% inhibition)
Inhibition of T cell adhesion mAb None FN CD-29 VLA-5 VLA-6 Peptides GRGDSPK GRGESP
FN 60 + 12 f 8* 13 f 57 t
6 3 (80) 3 (87) 4 (79) 4 (0)
10 * 4 59 k 6 (0)
FN: 120 kDa 65 + 8+ 6+ 8+ 62 f
5 2 (88) 3 (91) 5 (88) 2 (0)
623 62 + 3 (0)
LN
50 + 54 f 7f 55 f 5+
5 2 (0) 4 (86) 3 (0) 3 (90)
55 k 6 (0) 48 f 4 (0)
Note. Purified human CD4+ T cells were labeled with “Cr and pretreated with anti-CD-29, anti-VLA-5, and VLA-6 mAb, or with the RGD- or RGE-containing peptides. Where indicated, the coated wells were pretreated with mAb specific for the cell attachment site of FN. The T cells were then transferred into the coated microtiter wells, allowed to settle down, and activated with 10 rig/ml PMA. T cell adhesion to the ECM substrateswas then examined. T cell adhesion to BSA or to the heparin binding site of FN (40 kDa) was lower than 5 f 3%, data not shown. The results shown here represent three experiments which yielded essentially the same results.
T cells to PMA down-regulates cellular PKC levels and PKC activity (18, 19). To determine whether PMA-induced T cell adhesion to FN or LN is indeed mediated by the activation of PKC, the PKC activity of human CD4+ T cells was either inhibited by staurosporine or H-7 or down-regulated by prior PMA activation. Low concentrations of staurosporine or H-7 did not suppressT cell adhesion to FN or LN, whereas higher concentrations (0.5 pcM) (20) of staurosporine and 50 and 100 pF1M of H-7 did. If T cells were allowed to recover from staurosporine and H-7 treatment (washing followed by culture for 24 hr), their adhesion to FN or LN was not affected, indicating that staurosporine and H-7, at the concentrations used, were not toxic (data not shown). Thus, PKC activation is required for T cell adhesion to the FN and LN components of the ECM. Further support for this hypothesis comes from the results of the next experiment in which the intracellular levels of PKC were down-regulated by long-term exposure of the T cells to 50- 100 rig/ml PMA. Such treatment of the T cells completely inhibited their adhesion to both FN and LN (Table 2). Thus, abrogation of T cell adhesiveness to ECM protein components is associated with a decreasein cellular levels of PKC induced by PMA (18). Efect of a PrP inhibitor okadaic acid on the detachment of activated T cells from ECM Zigands. Intracellular events, such as PKC-induced Ser/Thr phosphorylation, probably induce conformational changesin the integrin-extracellular binding domain that affect receptor avidity to its ligand. However, since T cell adhesion to FN or LN is a transient event (Fig. 1) (4-5), the PKC-induced changesmust be reversible. What is the mdecular mechanism responsible for this integrin avidity cycle? Protein phosphatases1 and 2A (PrP 1, PrP2A) are cytosolic and plasma membraneassociated Ser/Thr phosphatasesthat are thought to reverse PKC activity by dephos-
REGULATION
OF T CELL-EXTRACELLULAR
MATRIX
187
INTERACTIONS
TABLE 2 Effects of Modulation of PKC Activity on T Cell Adhesion to FN and LN Second treatment % T cell adhesion (W inhibition) First treatment PMA (&ml)
PMA ( 10 wh-4
PKC inhibitor (ww
FN
LN
40 * 4
32 k 5
0
0
44 f 5 (0) 36 it- 2 (10) 14 f 3 (69) 7 + I (85)
34 t 35 f 15 t 7+
0 0 0
Staurosporine 0.01 0.1 0.3 0.5 H-7 10 50 100
44 f 3 (0) 30 f 2 (25) 20 * 4 (50)
33 * 4 (0) 22 k 4 (32) 12 + 1 (62)
0 0 0 0
30 + 3 (70) 7 f 2 (83) 5 f I (88) 4k3
10 + 3 (68) 5 + 2 (85) 4 f 2 (88) 5f2
10 50 100 100
+ + + -
4 (0) 3 (0) 2 (66) 2 (80)
Note. Human CD4+ T cells were either pretreated for 18 hr with different concentrations of PMA or left intact (first treatment). After 18 hr, the untreated cells were treated with different concentrations of staurosporine or H-7 and their adhesion following PMA activation was examined (secondtreatment). The PMAtreated cells were washed and seededinto the precoated wells and exposed to a second treatment with PMA and T cell adhesion was examined. The results shown here represent one of four experiments.
phorylating substrate proteins modified by PKC. We therefore asked whether PrPl and PrP2A are involved in regulation of integrin avidity. Thus, we chose okadaic acid, a specific cell-permeable inhibitor of PrPl and PrP2A (1 l), which does not affect any of the protein kinases, including PKC, tested to date (2 1). Okadaic acid was added to the cell cultures during or after the addition of PMA. The results are shown in Fig. 2. The adhesion of PMA-activated CD4+ T cells to FN or LN in the presence of 1 puM okadaic acid delayed (P < 0.05) T cell detachment when okadaic acid was administrated during the early stagesof T cell activation. Treatment of T cells with okadaic acid alone did not induce T cell adhesion, nor was the acid toxic since it did not alter T cell proliferative responses(data not shown). Thus, detachment of T cells from ECM proteins appears to be regulated by Ser/Thr dephosphorylation. It has been demonstrated that T cell-ECM transient adhesivenessis primarily regulated by cell activation, whether the lymphocytes were stimulated by phorbol esters (PMA or TPA), mitogens, anti-CD3 mAb, or specific antigens (4, 7, 8, 13). It is reasonable to speculate, therefore, that one would obtain similar patterns of inhibition of T cell adhesion (by staurosporine and H-7) to and detachment from (okadaic acid) immobilized LN or FN using antigen-dependent inducers of cell activation. Nevertheless, additional studies should be conducted to delineate this prospect. To ensure rapid lymphocyte extravasation, adhesion of T cells to components of the ECM must be induced rapidly and in a reversible fashion. The results herein suggestthat PKC is involved in the induction of T cell-ECM interactions induced
188
MIRON ET AL.
60 Y
0
40
20 Time
from
PMA
60 actlvation
80
100
(min)
FIG. 2. Effect of okadaic acid (I PM), an inhibitor of protein phosphatase on the detachment of human CD4+ T cells from FN (A) and LN (B). Purified human CD4+ T cells were seededon FN- or LN-coated microtiter wells in the presence of PMA (10 rig/ml). The T cells then either were treated with protein phosphataseinhibitor okadaic acid (open symbols) given at various time points (illustrated by black arrows) or were controls exposedonly to PMA (0). Okadaic acid was administrated to the culture wells simultaneously with (a), 20 (Cl), or 60 min (0) after initiation of the adhesion assay of T cells to FN or LN. The results obtained by treatment of the CD4+ T cells by okadaic acid administrated into the culture wells immediately or after 20 min from PMA activation differ markedly from those of the controls (P < 0.05). The results shown here represent the data obtained in one of five experiments.
either by direct conformational changes in the pl-integrin receptors or by a phosphorylation of a cytoskeletal protein linked to the cytoplasmic domain of these receptors. The ensuing reversed conformational modifications, induced by protein phosphatases,appear to change the T cells from an adhesive to a nonadhesive state. Our findings are in agreement with previous studies demonstrating that the p 1-integrin molecule associated with cytoskeletal components is capable of undergoing phosphorylation (22,23). Under physiological conditions, T cell adhesion and detachment from ECM proteins are probably dynamically regulated by “inside-out” signaling (l), the molecular details of which remain to be elucidated. REFERENCES I. Dustin, M. L., and Springer, T. A., Nature 341,6 19, 1989. 2. Springer, T. A., Nature346, 425, 1990.
REGULATION
OF T CELL-EXTRACELLULAR
MATRIX
INTERACTIONS
189
3. Fridman, R., Lider, O., Naparstek, Y., Fuks, Z., Vlodavsky, I., and Cohen, I. R., J. Ceil. P&id/. 130, 85, 1986. 4. Shimizu, Y., van Seventer, G. A., Horgan, K. J., and Shaw, S., Nature 345,250, 1990. 5. van Seventer, G. A., Shimizu, Y., and Shaw, S., Curr. Opinion Immunol. 3, 294, 1991. 6. Mustelin, T., Coggeshal, K. M., Isacov, N., and Altman, A., Science 247, 1584, 1990. 7. Hershkoviz, R., Miron, S., Cohen, I. R., Miller, A., and Lider, O., Eur. J. Immunol. 22, 7, 1992. 8. Chan, B. M. C., Wang, J. G. P., Rao, A., and Hemler, M. E., J. Immunol. 147, 398, 1991. 9. Alexander, D. R., and Cantrell, D. A., Immunol. Today 10, 200, 1989. 10. Kraft, A., Anderson, W., Cooper, L., and Sando, J. J., J. Biol. Chem. 257, 131, 1982. I 1. Hardie, D. G., Haystead, T. J., and Sim, A. T. R., Methods Enzymol. 201, 469, 1991. 12. Chatila, T. A., and Geha, R. S., J. Immunol. 140, 4308, 1988. 13. Shimizu, Y., van Seventer, G. A., Horgan, K. J., and Shaw, S., Immunol. Rev. 114, 109, 1990. 14. Ruoslahti, E., and Pierschbacher, M. D., Science 238, 49 1, 1987. 15. Mecham, R. P., FASEB J. 5,2538, 1991. 16. Tamaoki, T., Nomoto, H., Takahashi, I., Kato, Y., Morimoto, M., and Tomota, F., Biochem. Biophys. Rex Commun. 135, 397, 1986. 17. Weeks, B. S., Klotman, M. E., Dhawan, S., Kibbey, M., Rappaport, J., Klienman, H., Yamada, K. M., and Klotman, P. E., J. Cell Biol. 114, 847, 1991. 18. Larsen, C. S., Christiansen, C. O., and Esmann, V., Biochem. Biophys. Acta 969, 28 1, 1988. 19. Nel, A. E., Hanekom, C., Rheeder, A., Williams, K., Pollack, S., Katz, R., and Landreth, G. E., J. Immunol. 144,2683, 1990. 20. Merrill, J. T., Slade, S. G., Weissmann, G., Winchester, R., and Buyon, J. P., J. Immunol. 145, 2608, 1990. 21. Thevenin, C., Kim, S-J., and Kehrl, J. H., J. Biol. Chem. 266,9363, 1991. 22. Otey, C. A., Pavalko, F. M., and Burridge, K., J. Cell Biol. 111, 7217, 1990. 23. Hirst, R., Horwitz, A., and Rohrschneider, L., Proc. Natl. Acad. Ski. USA 83, 6470, 1986.
IMMUNOLOGY
144, 182- 189 ( 1992)
Involvement of a Protein Kinase C and Protein Phosphatases in Adhesion of CD4+ T Cells to and Detachment from Extracellular Matrix Proteins SHMUELMIRON,* RAMI HERSHKOVIZ,*IRIS TIROSH,* YORAM SHECHTER,t AVNER YAYUN,$ AND OFER LIDER*,’ Departments of *Cell Biology, $Chemical Immunology, and -fHormone Research, The Weizmann Institute of Science,Rehovot 76100, Israel Received April 16, 1992; acceptedJune 13, 1992 For immune surveillance and function to be effective, T lymphocytes constantly recirculate via lymph and blood between lymphoid organsand body tissues.To enable efficient cell movement and migration, cell adhesion to components of the basement membrane and the extracellular matrix (ECM) must be a rapid and transitory process. Whether phosphorylation and dephosphorylation of cellular proteins are involved in this phenomena was explored by monitoring the adhesion of T cells to immobilized ECM proteins. A short exposure of “Cr-labeled human CD4+ T cells to phorbol estersin vitro induced a rapid pl-integrin-mediated adhesion to both fibronectin and laminin, as determined by inhibition with anti-integrin antibodies. Adhesion was reversible; detachment from the immobilized ECM ligands occurred between 20 and 120 min without further intervention. This T cell adhesion was regulated by the activation of protein kinase C because(a) staurosporine and H-7 inhibitors of protein kinase C suppressedT cell adhesion, and (b) PMAinduced down-regulation of intracellular levels of protein kinase C was associatedwith the abrogation of the T cell adhesivenessto fibronectin and laminin. Furthermore, inhibition of protein phosphatasesactivity by okadaic acid delayed the detachment of the T cells from fibronectin or laminin. Thus, we suggestthat T cell-ECM interactions such as adhesion and detachment are regulated, respectively, by protein kinase C and protein phosphatases. o 1992 Academic PW., IX.
INTRODUCTION CD4+ T lymphocytes circulate within the body in search of peptide antigens presented on accessorycells in the context of the MHC class II molecules. Additionally, T cells interact in a precisely regulated fashion with normal or inflamed endothelium and thereby extravasate (1, 2). Simultaneously, enzymes are activated to enable the penetration of the ECM toward inflamed sites (3). Therefore, T cell adhesion to and dissociation from extracellular matrix (ECM)’ proteins are pivotal steps in T cell migration, a basis of immune surveillance (4, 5). ’ O.L. is the recipient of the Alon scholarshipand incumbent of the Weizmann LeagueCareer Development Chair in Children’s Diseases.Address correspondence to Dr. 0. Lider, The Department of Cell Biology, The Weizmann Institute of Science, P.O. Box 26, Rehovot, 76 100, Israel. * Abbreviations used: ECM, extracellular matrix; FN, fibronectin; LN, laminin; PKC, protein kinase C; PMA, phorbol- 12-myristate-13-acetate;PrP, serine/threonine protein phosphatases;TPA, I2-O-tetradecanoylphorbol- 13-acetate;mAb, monoclonal antibody. 182 0008-8749/92 $5.00 Copyright 0 1992 by Academic Press,Inc. All rights of reproduction in any form reserved
REGULATION
OF T CELL-EXTRACELLULAR
MATRIX
INTERACTIONS
183
T cells recognize protein components of the basement membrane and ECM, such as fibronectin (FN) or laminin (LN), through /31-integrin receptors (CD-29) designated very late antigens (4). T cell-ECM adhesion is rapidly augmented upon cell activation, without changesin the level of expression of the VLA4, VLA-5, and VLA-6 receptors (4, 5). The molecular basis for this T cell adhesion following activation is unknown. Nevertheless, it has been suggestedthat this processis regulated by intracellular signals which induce conformational change(s)of the integrin receptor, resulting in increased avidity of the T cell integrin to its ECM ligand (5). Indeed, the possibility that intracellular signals might regulate integrin affinity was recently examined by us. We have shown that inhibition of tyrosine protein kinase (6) resulted in partial abrogation of activated CD4+ T cell adhesion to ECM ligands (7). Upon antigen recognition by the T cell receptor-CD3 complex, T cells interact with accessory cells in a dynamically regulated fashion. A cascadeof protein phosphorylation and/or production of second messengersleads to cell-cell adhesion via, for example, LFA- 1 and ICAM- 1 receptors, followed by a decreasein receptor avidity and subsequent detachment (1, 2). We asked whether these types of transient cellcell and cell-ECM interactions might be modulated and reversed by protein phosphatases. It has been demonstrated recently that the pattern of T cell-ECM transient adhesiveness was identical whether the cells were activated by phorbol-12-myristate-13acetate (PMA), by 12-O-tetradecanoylphorbol- 13-acetate (TPA), or via the antigenspecific T cell receptor-CD3 pathways (8). In the present study we examined the regulatory role of protein kinase C (PKC) in PMA-induced T cell adhesion to FN and LN, since this kinase is also involved in the initial steps of T cell activation (9) and in cell-cell interactions and adhesion ( lo- 12). We demonstrate that PKC is involved in CD4+ T cell-IN and -LN interactions and suggestthat the rapid progression from adhesion to detachment states is due to dephosphorylation induced by protein phosphatases(PrP). MATERIALS AND METHODS T cells. CD4+ T cells were obtained from peripheral blood mononuclear leukocytes obtained from healthy human donors as follows. The mononuclear cells were isolated on a Ficoll gradient, washed, and incubated at 37°C in a 10% CO;! humidified atmosphere. After 2 hr, the nonadherent cells were removed and incubated on nylonwool columns (Fenwall, IL) for 1 hr at 37°C in a 10% CO* humidified atmosphere. Nonadherent cells were eluted and washed. CD4+ T cells were negatively selected by exposure of the eluted cells to a mixture of the following monoclonal antibodies (mAb): anti-CDs, CD1 9, and CD 14 conjugated to magnetic beads (Advanced Magnetics, MA). The remaining cells were recovered and exposed to a second round of negative selection. The resulting cell population consistedof over 90% CD3+CD4+ asdetermined by FACScan analysis. Tcell adhesion assay. T cell adhesion to the ECM protein components was designed and performed as previously described (4, 8). FN or LN (Sigma, St. Louis, MO), 40kDa heparin-binding site, and 120-kDa cell attachment site of FN (Telios Pharmaceuticals, Inc., San Diego, CA) were used as adhesion substratesat a concentration of 1 pug/wellin 100 yl PBS which was washed away after 16 hr of incubation. To activate the T cells, different concentrations of PMA or TPA (Sigma) in DMSO (final concen-
184
MIRON
ET AL.
tration 0.0 1%)were added to the microtiter wells containing 5’Cr labeled cells. Results are expressedasthe mean percentageof T cells binding from triplicate or quadriplicate wells per experimental group *SD. Modulation of T cell adhesion. 51Cr-labeled purified human CD4+ T cells were incubated with mAb against integrin receptors, PKC inhibitors, or okadaic acid either prior to or upon initiation of the adhesion assayto FN- or LN-coated wells. Labeled T cells were incubated with either Gly-Arg-Gly-Asp-Ser-Pro-Lys (GRGDSPK) or Gly-Arg-Gly-Glu-Ser-Pro (GRGESP) peptides (50 pg/ml; Sigma) for 30 min at 37“C before the start of the assay of adhesion. Similarly labeled T cells were preincubated with anti-VLA-5 (Telios), anti-VLA-6 (CDw49f, GoH3) mAb (CLB, Amsterdam, Holland), or anti-CD-29 (Serotec, GB) mAb to inhibit T cell adhesion to FN and/or LN. Monoclonal antibodies (clone 3E3 diluted l/400) to the cell attachment site of FN (Boehringer-Mannheim) were used to pretreat the coated wells for 1 hr at 37°C at 10% CO2 humidified atmosphere before the start of the adhesion assay. To activate PKC, labeled T cells were exposed to PMA. To inhibit PKC, labeled CD4+ T cells were treated (15 min at 37°C) with staurosporine (0.0 l-O.5 piW; Sigma) or with H-7 (lo-100 PM; Siekagaku America, Inc., Rockville, MD). Freshly isolated CD4+ T cells were incubated for 18 hr in a humidified incubator in culture medium, washed, and labeled. These cells were then treated with staurosporine, transferred into the culture wells, and activated with PMA, and their adhesion to FN or LN was examined. To down-regulate PKC, labeled T cells were incubated with PMA (0- 100 rig/ml) for 18 hr ( 10) prior to washing and initiation of the adhesion assayin the presence or absenceof the PKC activator PMA. To inhibit PrP T cells were treated with 1 PALM okadaic acid. The okadaic acid was added simultaneously with or following PMA (10 rig/ml) addition, which was added at the initiation of the assay. RESULTS AND DISCUSSION Eficts of PMA on CD4’ T cell adhesion to FN and LN. To study the kinetics of binding and detachment of CD4+ T cells to and from ECM proteins, the T cells were radioactively labeled and seededin wells coated with either FN or LN and then activated with PMA. Nonactivated T cells did not adhere to FN nor to LN. In contrast, PMAactivated CD4+ T cell adhesion to both FN (Fig. 1A) and LN (Fig. 1B) and maximal cell adhesion were achieved by 15 min of activation, irrespective of the PMA concentration (10, 50, or 100 rig/ml). T cell adhesion to FN and LN induced by the lower PMA concentrations (lo-50 rig/ml) was higher than that induced by a higher dose (100 rig/ml). By 60 min of PMA activation, T cell adhesion decreased. Similarly, activation of the T cells with 80 nM TPA, a co-mitogen which induces cell adhesion by PKC activation (l), could induce T cell adhesion in a time-dependent fashion (Figs. 1C and 1D). Hence, the adhesion state is transient and is followed by detachment of the cells from the ECM protein components. Regulation by integrin receptors of PMA-induced CD4’ human T cell adhesion to FN and LN. T cell adhesion to wells coated with FN or the 120-kDa cell attachment site of FN, but not to LN or to the heparin-binding site of FN (40-kDa fragment; data not shown), was inhibited by anti-FN antibodies and by anti-CD-29 mAb, which is specific for the pl chain of the integrin receptor (Table 1). Anti-VLA-5 (a5/31) mAb
REGULATION
OF T CELL-EXTRACELLULAR 80
60
MATRIX
185
INTERACTIONS
c
1
50 40 30 20 10 0 i
0
20
40
60
80
100
120
0
20
40
60
80
100
120
60
80
100
120
0
20
40
60
80
100
120
e
0
20
40
Time
(mln)
FIG. 1. Effect of PMA or TPA induced activation on human CD4+ T cell adhesion to FN and LN. Purified CD4+ T cells were seededon FN (A and C) or LN (B and D) coated microtiter wells. The cells were then either left intact (a) or activated with different concentrations of PMA (A and B): IO (0) 50 (0) or 100 rig/ml PMA (H), or with TPA (W)(C and D). Results are expressedas percentage of T cell (&SD) adhesion to the ECM proteins. The results shown here represent five experiments that produced essentially similar results.
inhibited T cell adhesion to FN and its 120-kD cell-attachment site but did not affect cell adhesion to LN, whereas anti-VLA-6 (a6/31) mAb inhibited T cell adhesion to LN but not to FN. Thus, the adhesion of activated CD4+ T cells to FN and LN is primarily regulated by way of specific /31-integrins. Lymphoid cells recognize and attach to immobilized FN via the VLA-5 receptor which recognizesthe tripeptide Arg-Gly-Asp (RGD) sequencepresent in the cell-binding domain of FN ( 13, 14). The binding of human CD4+ T cells to FN was specifically inhibited by the soluble RGD-containing peptide GRGDSPK but not by the nonRGD containing peptide GRGESP (Table 1). In contrast, the RGD-containing peptide did not interfere with T cell adhesivenessto immobilized LN. Indeed, it has been demonstrated that none of the integrins bind LN in an RGD-dependent manner (1315). Thus, PMA-induced CD4+ T cell adhesion to FN and LN appearsto be regulated by FN- or LN-specific VLA receptors. Involvement of PKC activation in T cell adhesion to FN or LN. It has been demonstrated that PKC, activated directly by PMA (16) is inhibited by the alkaloid compound staurosporine and by H-7 ( 1, 17, 18). Moreover, prolonged exposure of human
186
MIRON ET AL. TABLE 1 Effects of FN Derivatives and mAb against Integrin Receptors on PMA-Induced CD4+ T Cell Adhesion to FN or LN Percentageof T cell adhesion (% inhibition)
Inhibition of T cell adhesion mAb None FN CD-29 VLA-5 VLA-6 Peptides GRGDSPK GRGESP
FN 60 + 12 f 8* 13 f 57 t
6 3 (80) 3 (87) 4 (79) 4 (0)
10 * 4 59 k 6 (0)
FN: 120 kDa 65 + 8+ 6+ 8+ 62 f
5 2 (88) 3 (91) 5 (88) 2 (0)
623 62 + 3 (0)
LN
50 + 54 f 7f 55 f 5+
5 2 (0) 4 (86) 3 (0) 3 (90)
55 k 6 (0) 48 f 4 (0)
Note. Purified human CD4+ T cells were labeled with “Cr and pretreated with anti-CD-29, anti-VLA-5, and VLA-6 mAb, or with the RGD- or RGE-containing peptides. Where indicated, the coated wells were pretreated with mAb specific for the cell attachment site of FN. The T cells were then transferred into the coated microtiter wells, allowed to settle down, and activated with 10 rig/ml PMA. T cell adhesion to the ECM substrateswas then examined. T cell adhesion to BSA or to the heparin binding site of FN (40 kDa) was lower than 5 f 3%, data not shown. The results shown here represent three experiments which yielded essentially the same results.
T cells to PMA down-regulates cellular PKC levels and PKC activity (18, 19). To determine whether PMA-induced T cell adhesion to FN or LN is indeed mediated by the activation of PKC, the PKC activity of human CD4+ T cells was either inhibited by staurosporine or H-7 or down-regulated by prior PMA activation. Low concentrations of staurosporine or H-7 did not suppressT cell adhesion to FN or LN, whereas higher concentrations (0.5 pcM) (20) of staurosporine and 50 and 100 pF1M of H-7 did. If T cells were allowed to recover from staurosporine and H-7 treatment (washing followed by culture for 24 hr), their adhesion to FN or LN was not affected, indicating that staurosporine and H-7, at the concentrations used, were not toxic (data not shown). Thus, PKC activation is required for T cell adhesion to the FN and LN components of the ECM. Further support for this hypothesis comes from the results of the next experiment in which the intracellular levels of PKC were down-regulated by long-term exposure of the T cells to 50- 100 rig/ml PMA. Such treatment of the T cells completely inhibited their adhesion to both FN and LN (Table 2). Thus, abrogation of T cell adhesiveness to ECM protein components is associated with a decreasein cellular levels of PKC induced by PMA (18). Efect of a PrP inhibitor okadaic acid on the detachment of activated T cells from ECM Zigands. Intracellular events, such as PKC-induced Ser/Thr phosphorylation, probably induce conformational changesin the integrin-extracellular binding domain that affect receptor avidity to its ligand. However, since T cell adhesion to FN or LN is a transient event (Fig. 1) (4-5), the PKC-induced changesmust be reversible. What is the mdecular mechanism responsible for this integrin avidity cycle? Protein phosphatases1 and 2A (PrP 1, PrP2A) are cytosolic and plasma membraneassociated Ser/Thr phosphatasesthat are thought to reverse PKC activity by dephos-
REGULATION
OF T CELL-EXTRACELLULAR
MATRIX
187
INTERACTIONS
TABLE 2 Effects of Modulation of PKC Activity on T Cell Adhesion to FN and LN Second treatment % T cell adhesion (W inhibition) First treatment PMA (&ml)
PMA ( 10 wh-4
PKC inhibitor (ww
FN
LN
40 * 4
32 k 5
0
0
44 f 5 (0) 36 it- 2 (10) 14 f 3 (69) 7 + I (85)
34 t 35 f 15 t 7+
0 0 0
Staurosporine 0.01 0.1 0.3 0.5 H-7 10 50 100
44 f 3 (0) 30 f 2 (25) 20 * 4 (50)
33 * 4 (0) 22 k 4 (32) 12 + 1 (62)
0 0 0 0
30 + 3 (70) 7 f 2 (83) 5 f I (88) 4k3
10 + 3 (68) 5 + 2 (85) 4 f 2 (88) 5f2
10 50 100 100
+ + + -
4 (0) 3 (0) 2 (66) 2 (80)
Note. Human CD4+ T cells were either pretreated for 18 hr with different concentrations of PMA or left intact (first treatment). After 18 hr, the untreated cells were treated with different concentrations of staurosporine or H-7 and their adhesion following PMA activation was examined (secondtreatment). The PMAtreated cells were washed and seededinto the precoated wells and exposed to a second treatment with PMA and T cell adhesion was examined. The results shown here represent one of four experiments.
phorylating substrate proteins modified by PKC. We therefore asked whether PrPl and PrP2A are involved in regulation of integrin avidity. Thus, we chose okadaic acid, a specific cell-permeable inhibitor of PrPl and PrP2A (1 l), which does not affect any of the protein kinases, including PKC, tested to date (2 1). Okadaic acid was added to the cell cultures during or after the addition of PMA. The results are shown in Fig. 2. The adhesion of PMA-activated CD4+ T cells to FN or LN in the presence of 1 puM okadaic acid delayed (P < 0.05) T cell detachment when okadaic acid was administrated during the early stagesof T cell activation. Treatment of T cells with okadaic acid alone did not induce T cell adhesion, nor was the acid toxic since it did not alter T cell proliferative responses(data not shown). Thus, detachment of T cells from ECM proteins appears to be regulated by Ser/Thr dephosphorylation. It has been demonstrated that T cell-ECM transient adhesivenessis primarily regulated by cell activation, whether the lymphocytes were stimulated by phorbol esters (PMA or TPA), mitogens, anti-CD3 mAb, or specific antigens (4, 7, 8, 13). It is reasonable to speculate, therefore, that one would obtain similar patterns of inhibition of T cell adhesion (by staurosporine and H-7) to and detachment from (okadaic acid) immobilized LN or FN using antigen-dependent inducers of cell activation. Nevertheless, additional studies should be conducted to delineate this prospect. To ensure rapid lymphocyte extravasation, adhesion of T cells to components of the ECM must be induced rapidly and in a reversible fashion. The results herein suggestthat PKC is involved in the induction of T cell-ECM interactions induced
188
MIRON ET AL.
60 Y
0
40
20 Time
from
PMA
60 actlvation
80
100
(min)
FIG. 2. Effect of okadaic acid (I PM), an inhibitor of protein phosphatase on the detachment of human CD4+ T cells from FN (A) and LN (B). Purified human CD4+ T cells were seededon FN- or LN-coated microtiter wells in the presence of PMA (10 rig/ml). The T cells then either were treated with protein phosphataseinhibitor okadaic acid (open symbols) given at various time points (illustrated by black arrows) or were controls exposedonly to PMA (0). Okadaic acid was administrated to the culture wells simultaneously with (a), 20 (Cl), or 60 min (0) after initiation of the adhesion assay of T cells to FN or LN. The results obtained by treatment of the CD4+ T cells by okadaic acid administrated into the culture wells immediately or after 20 min from PMA activation differ markedly from those of the controls (P < 0.05). The results shown here represent the data obtained in one of five experiments.
either by direct conformational changes in the pl-integrin receptors or by a phosphorylation of a cytoskeletal protein linked to the cytoplasmic domain of these receptors. The ensuing reversed conformational modifications, induced by protein phosphatases,appear to change the T cells from an adhesive to a nonadhesive state. Our findings are in agreement with previous studies demonstrating that the p 1-integrin molecule associated with cytoskeletal components is capable of undergoing phosphorylation (22,23). Under physiological conditions, T cell adhesion and detachment from ECM proteins are probably dynamically regulated by “inside-out” signaling (l), the molecular details of which remain to be elucidated. REFERENCES I. Dustin, M. L., and Springer, T. A., Nature 341,6 19, 1989. 2. Springer, T. A., Nature346, 425, 1990.
REGULATION
OF T CELL-EXTRACELLULAR
MATRIX
INTERACTIONS
189
3. Fridman, R., Lider, O., Naparstek, Y., Fuks, Z., Vlodavsky, I., and Cohen, I. R., J. Ceil. P&id/. 130, 85, 1986. 4. Shimizu, Y., van Seventer, G. A., Horgan, K. J., and Shaw, S., Nature 345,250, 1990. 5. van Seventer, G. A., Shimizu, Y., and Shaw, S., Curr. Opinion Immunol. 3, 294, 1991. 6. Mustelin, T., Coggeshal, K. M., Isacov, N., and Altman, A., Science 247, 1584, 1990. 7. Hershkoviz, R., Miron, S., Cohen, I. R., Miller, A., and Lider, O., Eur. J. Immunol. 22, 7, 1992. 8. Chan, B. M. C., Wang, J. G. P., Rao, A., and Hemler, M. E., J. Immunol. 147, 398, 1991. 9. Alexander, D. R., and Cantrell, D. A., Immunol. Today 10, 200, 1989. 10. Kraft, A., Anderson, W., Cooper, L., and Sando, J. J., J. Biol. Chem. 257, 131, 1982. I 1. Hardie, D. G., Haystead, T. J., and Sim, A. T. R., Methods Enzymol. 201, 469, 1991. 12. Chatila, T. A., and Geha, R. S., J. Immunol. 140, 4308, 1988. 13. Shimizu, Y., van Seventer, G. A., Horgan, K. J., and Shaw, S., Immunol. Rev. 114, 109, 1990. 14. Ruoslahti, E., and Pierschbacher, M. D., Science 238, 49 1, 1987. 15. Mecham, R. P., FASEB J. 5,2538, 1991. 16. Tamaoki, T., Nomoto, H., Takahashi, I., Kato, Y., Morimoto, M., and Tomota, F., Biochem. Biophys. Rex Commun. 135, 397, 1986. 17. Weeks, B. S., Klotman, M. E., Dhawan, S., Kibbey, M., Rappaport, J., Klienman, H., Yamada, K. M., and Klotman, P. E., J. Cell Biol. 114, 847, 1991. 18. Larsen, C. S., Christiansen, C. O., and Esmann, V., Biochem. Biophys. Acta 969, 28 1, 1988. 19. Nel, A. E., Hanekom, C., Rheeder, A., Williams, K., Pollack, S., Katz, R., and Landreth, G. E., J. Immunol. 144,2683, 1990. 20. Merrill, J. T., Slade, S. G., Weissmann, G., Winchester, R., and Buyon, J. P., J. Immunol. 145, 2608, 1990. 21. Thevenin, C., Kim, S-J., and Kehrl, J. H., J. Biol. Chem. 266,9363, 1991. 22. Otey, C. A., Pavalko, F. M., and Burridge, K., J. Cell Biol. 111, 7217, 1990. 23. Hirst, R., Horwitz, A., and Rohrschneider, L., Proc. Natl. Acad. Ski. USA 83, 6470, 1986.
