Kinase 2 in T Cell Activation a Novel Role

A Novel Role for p21-Activated Protein Kinase 2 in T Cell Activation Peter C. Chu, Jun Wu, X. Charlene Liao, Jorge Pardo, Haoran Zhao, Congfen Li, Marcy K. Mendenhall, Erlina This information is current as Pali, Mary Shen, Simon Yu, Vanessa C. Taylor, Gregorio of September 25, 2021. Aversa, Susan Molineaux, Donald G. Payan and Esteban S. Masuda J Immunol 2004; 172:7324-7334; ;

doi: 10.4049/jimmunol.172.12.7324 Downloaded from http://www.jimmunol.org/content/172/12/7324

References This article cites 58 articles, 41 of which you can access for free at: http://www.jimmunol.org/content/172/12/7324.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

by guest on September 25, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2004 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

A Novel Role for p21-Activated Protein Kinase 2 in T Cell Activation

Peter C. Chu,1* Jun Wu,1,2* X. Charlene Liao,* Jorge Pardo,* Haoran Zhao,* Congfen Li,* Marcy K. Mendenhall,* Erlina Pali,* Mary Shen,* Simon Yu,* Vanessa C. Taylor,3* Gregorio Aversa,† Susan Molineaux,* Donald G. Payan,* and Esteban S. Masuda*

To identify novel components of the TCR signaling pathway, a large-scale retroviral-based functional screen was performed using CD69 expression as a marker for T cell activation. In addition to known regulators, two truncated forms of p21-activated kinase ⌬ ⌬ 2 (PAK2), PAK2 L1–224 and PAK2 S1–113, both lacking the kinase domain, were isolated in the T cell screen. The PAK2 truncation, PAK2⌬L, blocked Ag receptor-induced NFAT activation and TCR-mediated calcium ﬂux in Jurkat T cells. However, it had minimal effect on PMA/ionomycin-induced CD69 up-regulation in Jurkat cells, on anti-IgM-mediated CD69 up-regulation in B cells, or on the migratory responses of resting T cells to chemoattractants. We show that PAK2 kinase activity is increased Downloaded from in response to TCR stimulation. Furthermore, a full-length kinase-inactive form of PAK2 blocked both TCR-induced CD69 up-regulation and NFAT activity in Jurkat cells, demonstrating that kinase activity is required for PAK2 function downstream of the TCR. We also generated a GFP-fused PAK2 truncation lacking the Cdc42/Rac interactive binding region domain, GFP-

PAK283–149. We show that this construct binds directly to the kinase domain of PAK2 and inhibits anti-TCR-stimulated T cell activation. Finally, we demonstrate that, in primary T cells, dominant-negative PAK2 prevented anti-CD3/CD28-induced IL-2 http://www.jimmunol.org/ production, and TCR-induced CD40 ligand expression, both key functions of activated T cells. Taken together, these results suggest a novel role for PAK2 as a positive regulator of T cell activation. The Journal of Immunology, 2004, 172: 7324–7334.

eregulation of the acquired immune response results in determined by the set of genes activated in response to the signal- an attack directed against self-Ags and ultimately leads ing cascades that were engaged. Expression of activation markers, D to the development of autoimmune disease. Acquired including CD25, CD69, and CD40 ligand (CD40L), and the pro- immune responses are regulated in part by T cells and involve duction of cytokines, including IL-2, are well-deﬁned conse- precise control of the activation and differentiation of theses cells. quences of T cell activation.

Consequently, the regulation of T cell activation provides an at- To identify novel effectors downstream of the TCR, we de- by guest on September 25, 2021 tractive basis for the development of novel therapies for the treat- signed a retroviral-based functional genetic screen to select for ment of autoimmune disease and the prevention of transplant re- proteins that inhibit receptor-mediated T cell activation. In addi- jection. T cell activation is a complex process tightly controlled by tion to known regulators of Ag receptor signaling, we identiﬁed Ag receptors, costimulatory molecules, cytokines and chemokines. PAK2 as a regulator of TCR signaling (2). PAK2 is a member of T cell activation following engagement of the TCR involves the a family of p21-activated serine/threonine protein kinases (PAKs) coordinated activities of a diverse set of intracellular signaling composed of six mammalian isoforms (3). PAKs 1Ð3 each contain pathways, including mobilization of calcium, activation of protein a C-terminal catalytic domain and an N-terminal regulatory do- kinase C, and activation of Ras and Rho family GTPases, which main, which is inhibitory to the former (4). They share signiﬁcant result in the triggering of the mitogen-activated protein kinase homology, particularly in the kinase domain (Ͼ90%) as well as in 4 (MAPK) pathway (1). The fate of the activated T cell is ultimately a 66-aa stretch within the regulatory domain (also Ͼ90% identity). The PAKs are downstream effectors of the small GTPases, Cdc42 and Rac, members of the Rho family of GTPases, which are key *Rigel, Inc., South San Francisco, CA 94080; and †Novartis Forschungsinstitut regulators of cytoskeletal organization (5). These GTP-binding GmbH, Vienna, Austria proteins are molecular switches, which cycle between the active Received for publication April 28, 2003. Accepted for publication April 9, 2004. GTP-bound form and the inactive GDP-bound form. Binding of The costs of publication of this article were defrayed in part by the payment of page GTP-loaded Cdc42 or Rac to PAK family members through the charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Cdc42/Rac interactive binding region (CRIB) domain results in 1 P.C.C. and J.W. contributed equally to this manuscript. autophosphorylation and activation of the kinase (6, 7). In the case of PAK4, the fourth family member, which shares only 55% iden- 2 Current address: Shanghai Genomics, Inc., and Chinese National Genome Center, Zhangjiang Hi-Tech Park, Shanghai, 201203, China. tity within the kinase domain, GTPase binding does not result in 3 Address correspondence and reprint requests to Dr. Vanessa Taylor, Rigel, Inc., catalytic activation, because it lacks the regulatory kinase inhibi- 1180 Veterans Boulevard, South San Francisco, CA 94080. E-mail address: tory domain (8). PAK5 and PAK6 most closely resemble PAK4 [email protected] but show distinct tissue expression. PAK5 is highly expressed in 4 Abbreviations used in this paper: MAPK, mitogen-activated protein kinase; PAK, brain, whereas PAK6 is highly expressed in prostate and testis (3). p21-activated serine/threonine protein kinase; CRIB, Cdc42/Rac interactive binding region; PIX, PAK-interacting exchange factor; ZAP70, ␨-associated protein 70; tTA, Despite the extensive similarities between the PAK enzymes, Tet transactivator; TRE, tetracycline-responsive element; Dox, doxycycline; KI, ki- they appear to play different physiological functions and may be nase inactive; SPP, sphingosine-1-phosphate; SDF, stromal cell-derived factor; CD40L, CD40 ligand; PLC, phospholipase C; EDG, endothelial differentiation gene; differentially regulated in vivo. PAK1 plays a role in the disas- LAT, linker for activation of T cells. sembly of focal complexes, integrin-dependent sites linking the

Copyright © 2004 by The American Association of Immunologists, Inc. 0022-1767/04/$02.00 The Journal of Immunology 7325 extracellular matrix to the actin-rich cytoskeleton, and stress fibers Libraries and constructs (9) (10). This is likely mediated in part through the phosphoryla- RNA extracted from human lymph node, thymus, spleen, and bone marrow tion and inhibition of myosin light chain kinase by PAK1 (11). was used to produce two cDNA libraries; one random primed and direc- Through one of its five proline-rich elements that serve as binding tionally cloned and the second nondirectionally cloned and provided with sites for SH3 domains, PAK1 interacts with the PAK-interacting three exogenous ATG in three frames. cDNAs were cloned into the pTRA- exchange factor (PIX), allowing it to localize to the focal complex exs vector giving robust Dox-regulatable transcription of cDNAs from the TRE promoter. The total combined library complexity was 5 ϫ 107 inde- (12). Activation and autophosphorylation negatively regulate this pendent clones. ⌬ ⌬ interaction (13). Activation of PAK1 appears to promote cell mi- Dominant-negative PAK2 L1Ð224 and PAK2 S1Ð113 were cloned into gration by facilitating the breakdown of focal complexes, which is the retroviral pTRA-IRES/GFP vector (27, 28) and the mammalian expres- 6 necessary for cell migration to occur (13Ð15). PAK1 also interacts sion vector pCDEF3 (26). The full-length wild-type PAK2-V5-His was subcloned from a Research Genetics (Huntsville, AL) PAK2 clone into the with one of the three SH3 domains of Nck, which localizes it to the pCDEF3 vector. Difficulties were encountered when attempting to express membrane, where it promotes neurite outgrowth in PC12 cells the full-length wild-type kinase. PAK1 has previously been shown to be (16). In addition, PAK1 has been reported to be activated down- toxic in bacteria (9), and we found that PAK2 was also toxic to bacteria. stream of the TCR, dependent on the tyrosine kinases ␨-associated Nine of nine untagged wild-type PAK2 clones we isolated had independent protein 70 (ZAP70) and Syk (17, 18). frameshift mutations. An alternative subcloning approach to generate an untagged wild-type PAK2 yielded 125 of 126 inserts in opposite orienta- In contrast, PAK2 has been reported to be activated by cellular tion and one parental frameshift. Although the smaller PAK2-HA tag re- stresses such as hyperosmolarity and DNA damage (19, 20). In mained toxic to bacteria, wild-type PAK2-V5-His6 was successfully iso- Jurkat T cells, PAK2 is activated after cleavage by caspase 3 and lated. To generate the kinase-inactive point mutant (KI-PAK2), Lys278 was appears to mediate some of the membrane and morphological substituted with Ala using standard PCR mutagenesis methods. The iso- Downloaded from lated kinase-inactive kinase domain construct (KI-kinase-V5-His6) (resi- changes characteristic of programmed cell death (21). PAK2 has dues 162Ð524 of PAK2) was generated by a standard PCR procedure using also been reported to play an antiapoptotic role in fibroblasts, the KI-PAK2-V5-His6 as template, and expressed in the pCDEF3 plasmid. where it suppressed stress-induced cell death (22). However, more The transdominant PAK283Ð149 fragment and the isolated CRIB domain (aa recent evidence further implicates PAK2 in the induction of cy- 74Ð87) were also generated by a standard PCR procedure and expressed as tostasis in fibroblasts, which was shown to be dependent on its GFP fusions in the pCDEF3.GFP vector.

localization in the endoplasmic reticulum (23). http://www.jimmunol.org/ In this study, we demonstrate that PAK2 is a positive regulator CD69 functional screen of T cell activation. Dominant-negative forms of PAK2 block Phoenix A packaging cells were transfected using calcium phosphate with TCR-induced CD69 up-regulation, NFAT activation, and calcium a tTA regulated retroviral pTRA cDNA library. tTA Jurkat cells were ﬂux. However, dominant-negative PAK2 has minimal effect on spin-infected with the supernatant from the Phoenix A cells for 3 h at room temperature. After 4 days of cDNA expression, library-infected cells were PMA/ionomycin-stimulated T cell activation, and no effect on re- stimulated with anti-TCR (0.3 ␮g/ml C305) for 20Ð24 h, and stained with ceptor-mediated B cell activation. We also show that the effect of anti-CD69-allophycocyanin (Caltag, Burlingame, CA) for CD69 surface PAK2 is dependent on its kinase activity, which is increased in expression. Cells expressing the lowest levels of TCR-induced CD69 were response to TCR engagement. Finally, we demonstrate that dom- isolated using a MoFlo (DakoCytomation, Carpinteria, CA). Sorting was

repeated over multiple rounds with 6-day rest periods between stimulations by guest on September 25, 2021 inant-negative PAK2 can also block IL-2 production and CD40L until significant enrichment of nonresponsive clones was achieved. The expression in primary T cells. nonresponders were single-cell sorted and reanalyzed for TCR-induced CD69 expression in the presence (cDNA off) and absence (cDNA on) of Materials and Methods Dox. Clones were ranked by a ratio of C305-induced CD69 mean fluores- Cell culture cence in the presence and absence of Dox. Clones with Dox ratios of Ͼ1.5Ð2.0 were RT-PCR subcloned to pTRA-IRES-GFP for phenotype Human BJAB B cells, Jurkat T cells (clone N), and Jurkat TAg cells (ex- confirmation. pressing SV40 large T Ag) were routinely cultured in RPMI 1640 medium supplemented with 10% FCS (JRH Biosciences, Lenexa, KS), penicillin, Cell surface marker analysis and streptomycin. Phoenix A cells (24) were grown in DMEM supplemented with 10% FCS, penicillin, and streptomycin. To produce the Tet tTA-Jurkat, Jurkat-N, Jurkat-TAg, or BJAB cells were stained with an transactivator (tTA)-Jurkat cell line, Jurkat cells were infected with a ret- allophycocyanin-conjugated mouse monoclonal anti-human CD69 Ab roviral construct that constitutively expresses the tetracycline transactivator (Caltag) at 4¡C for 20 min and analyzed using a FACSCalibur instrument protein and a reporter construct that expresses Lyt2 driven by a tetracy- (BD Biosciences, Mountain View, CA) with CellQuest software. Cell sorts cline-responsive element (TRE) (25). The tTA-Jurkat cell population was were performed on a MoFlo (DakoCytomation). optimized by sorting multiple rounds for high TRE-dependent expression of Lyt2 in the absence of doxycycline (Dox) and strong repression of Lyt2 expression in the presence of Dox. The cells were also sorted for maximal CD69 up-regulation assay anti-TCR-induced expression of CD69. Dox was used at a final concen- tTA-Jurkat, Jurkat TAg, or BJAB cells were split to 2.5 ϫ 105 cells/ml 24 h tration of 10 ng/ml for at least 6 days to down-regulate expression of before stimulation. Cells were spun and resuspended at 5 ϫ 105 cells/ml in cDNAs from the TRE promoter. fresh complete RPMI 1640 medium in the presence of 1 ␮g/ml C305 (anti-TCR ␤-chain specific to Jurkat (29)), 0.3 ␮g/ml anti-IgM F(abЈ) Transfection and infection 2 (Jackson ImmunoResearch, West Grove, PA), or 50 ng/ml PMA for 20Ð26 Phoenix A packaging cells were transfected with retroviral vectors using hat37¡C, and then assayed for surface CD69 expression. calcium phosphate for 6 h (24). After 24 h, supernatant was replaced with complete RPMI 1640 medium, and virus was allowed to accumulate for an PAK1 and PAK2 TaqMan additional 24 h. Viral supernatant was collected, filtered through a 0.2-␮m filter, and mixed with BJAB/Jurkat cells at a density of 2.5 ϫ 105 cells/ml. RT-PCR was performed by real-time quantitative PCR (TaqMan) using the Cells were spun at room temperature for3hat3000 rpm, followed by following primers and probes: PAK1, forward, TTGGACTCTCATTC overnight incubation at 37¡C. Transfection and infection efficiencies were CCTTTTCC, reverse, TGGATGCGGAGACCAGGT, probe, CCCCGC monitored by GFP expression, and functional analysis was conducted 2Ð4 GCTTTCGTGAGCC; and PAK2, forward, GAATGGAAGGATCTGT days after infection. For transient transfection, various amounts of indi- TAAGCTCACT, reverse, GCCATAAGCTTTCCGTGTAACC, probe, cated plasmids, plus 10 ␮g of NFAT-Luciferase plus 2 ␮g of a second TCACCCCTGAGCAGAGCAAACGC. control TK Luc luciferase construct (Promega, Madison, WI) in the case of Human placenta total RNA (Clontech, Palo Alto, CA) was used for the the reporter assay, were electroporated into 107 BJAB or Jurkat TAg cells, standard curve and human RiP gene was used for normalization of total as previously described (26). Cells were cultured for 40 h after transfection RNA. Samples were tested in triplicate. RT-PCR was performed on an and assayed as indicated. Applied Biosystems (Foster City, CA) PRISM 7700. 7326 REGULATION OF T CELL ACTIVATION BY PAK2

NFAT Dual-Luciferase assay Results Transfected Jurkat-TAg cells (1 ϫ 105) were aliquoted into a 96-well plate An N-terminal PAK2 fragment lacking a kinase domain was (Corning, Corning, NY) 40 h after transfection and cultured in a final identified as a functional hit in a CD69 T cell screen volume of 100 ␮l of RPMI 1640 growth medium. Cells were stimulated at 37¡C in the growth medium containing either 1 ␮g/ml C305 hybridoma A functional retroviral-based T cell screen was performed to select supernatant or 50 ng/ml PMA and 1 ␮M ionomycin. After 12-h stimula- for proteins that inhibit TCR-mediated T cell activation, using tion, cells were lysed in 5ϫ lysis buffer (Promega) and assayed using the CD69 expression as the readout for T cell activation (30). tTA luciferase kit purchased from Promega. Luciferase activity was read on the Jurkat cells (25) stably transfected with pTRA cDNA library were luminometer (Dynex Technologies, Chantilly, VA) and determined in triplicate for each experimental condition. stimulated with anti-TCR, and stained for CD69 surface expression. Cells expressing the lowest levels of TCR-induced CD69 Calcium mobilization were sorted, expanded, and resorted until significant enrichment of Jurkat cells were washed in modified Tyrode’s (MT) buffer (137 mM NaCl, nonresponsive clones was achieved. Individual clones were iso-

2.7 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 20 mM HEPES (pH 7.4), 5.6 lated by single-cell sorting and were reanalyzed for TCR-induced mM glucose, and 0.1% BSA), resuspended in MT at 106 cells/ml, and CD69 expression in the presence (cDNA off) and absence (cDNA incubated with 4 ␮g/ml indo-1 (Molecular Probes, Eugene, OR) and 4 mM on) of Dox (Fig. 1A). Over 2800 single-cell clones were expanded, Probenecid (Sigma-Aldrich, St. Louis, MO) at 37¡C for 30 min. Cells were grown in the presence or absence of Dox, stimulated with anti- centrifuged and resuspended in MT. Analysis was performed on an LSR flow cytometer (BD Biosciences) at 37¡C. For stimulation, C305 was used TCR, and analyzed for CD69 surface expression. Of these, 46.8% at 300 ng/ml and ionomycin at 5 ␮M. showed a Dox-regulatable phenotype. The cDNAs were successfully identified from ϳ200 clones, representing Ͼ50 distinct genes Downloaded from Jurkat chemotaxis assay and including known regulators of the TCR signaling pathway, Jurkat cells were cultured in RPMI 1640 with 10% charcoal-stripped FBS such as ZAP70, Syk, Lck, and phospholipase C (PLC)␥1 (2). The for 24 h before the assay. Cells were washed and resuspended in serum-free 6 fourth most frequently represented cDNA was PAK2. Two kinase medium. A total of 10 cells was added to the upper chamber of a Costar domain-truncated forms of PAK2 were isolated multiple times as (Cambridge, MA) Transwell apparatus (5-␮m pore), and allowed to migrate at 37¡C for 2.5 h into the lower chamber containing either 100 nM functional hits in the CD69 screen (Fig. 1B). The phenotype of one sphingosine-1-phosphate (SPP), 100 nM stromal cell-derived factor of the PAK2 screen clones is shown (Fig. 1C). The second clone, (SDF)-1, or serum-free medium alone. The percentage of GFP-positive PAK2⌬S, which lacks the PIX binding domain, gave a similar, http://www.jimmunol.org/ cells of the migrated population (lower chamber) was analyzed by FACS. albeit less potent, phenotype (2), suggesting that the PIX binding Immunoprecipitation, in vitro kinase assay, and immunoblots domain is not completely required for the inhibitory effect. PIX directly interacts with PAK proteins and is responsible for their For in vitro kinase assays, Jurkat TAg cells were starved for2hinserum- recruitment to focal complexes (9). free RPMI 1640 to reduce basal activity. Cells were then stimulated in prewarmed PBS at 37¡C and lysed in Nonidet P-40 lysis buffer (1% Non- Although PAK1 has been implicated previously in TCR signal- idet P-40, 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 25 mM sodium ing (17, 18), this is the first report suggesting that PAK2 may play fluoride, 1 mM sodium orthovanadate, and protease inhibitors), as previ- a role in signaling downstream of the TCR. A number of lines of ously described (26). After 20 min at 4¡C, lysates were centrifuged for 10

evidence support a role for PAK2 in T cell signaling. Rudel and by guest on September 25, 2021 min at 14,000 rpm, and the supernatants were subjected to immunoprecipitation with the indicated Abs for1hat4¡C. The resulting immune Bokoch (21) showed that Jurkat T cells predominantly express complexes were washed extensively in lysis buffer. For in vitro kinase PAK2. Furthermore, Fas ligation of T cells induces caspase-me- assays, the immune complexes were resuspended in kinase buffer (20 mM diated cleavage of PAK2, but not PAK1, to generate a constitu- Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM DTT, 10 mM MgCl2,5mM tively active PAK2 fragment (21). Finally, a highly conserved HIV MnCl , 2.5 ␮M ATP, and 10 ␮Ci of [␥-32P]ATP) and incubated with 2 protein, Nef, interferes with CD3 signaling in T cells, thereby histone H4 (1 ␮g) at 30¡C for 15 min. The kinase assay samples were separated by SDS-PAGE and detected using a phosphor imager. For West- blocking up-regulation of CD69 expression, and induction of IL-2 ern blotting, the immune complexes were resolved by SDS-PAGE and and IFN-␥ mRNAs (31, 32). Interestingly, Nef specifically asso- transferred to polyvinylidene difluoride membranes. Membranes were ciates with and activates PAK2, but not PAK1, in T cells (33, 34). blocked with 5% milk in TBST (20 mM Tris (pH 7.4), 150 mM NaCl, with PAK1 and PAK2 have previously been reported to be ubiqui- 0.05% Tween 20), then incubated with the indicated Abs followed by a secondary Ab conjugated with HRP and assayed by ECL assay (ECL Plus tously expressed (4, 35). However, when we examined the expres- kit; Amersham Biosciences, Piscataway, NJ). Anti-GFP and anti-V5 Abs sion of PAK1 and PAK2 in a panel of 15 tissues by quantitative were obtained from Zymed Laboratories (San Francisco, CA) and Invitro- PCR (TaqMan), we found PAK2 to be broadly expressed, whereas gen (Carlsbad, CA), respectively. Anti-PAK1 (for Western blotting) and PAK1 was expressed predominantly in brain (Fig. 2). anti-PAK2 Abs were obtained from Cell Signaling Technology (nos. 2602 and 2608; Beverly, MA). For immunoprecipitation of endogenous PAK1, anti-PAK1 was obtained from Santa Cruz Biotechnology (SC-882; Santa PAK2⌬L specifically inhibits anti-TCR, and not PMA-induced Cruz, CA). CD69 up-regulation, in T cells Primary T cell assays To determine whether the effect of the PAK2 truncation was re- stricted to T cells, Jurkat N32H cells and BJAB cells, infected with Freshly purified PBMC (AllCells, Berkeley, CA) were incubated at 37¡C ⌬ for2htoallow the monocytes to adhere to the tissue culture flask. The the PAK2 L construct or vector alone, were stimulated with anti- cells remaining in suspension were recovered and seeded at 106 cells/ml/ TCR and anti-BCR, respectively (Fig. 3A). TCR cross-linking in- well in Yssel’s medium (Gemini Bio-Products, Woodland, CA) with pen- creased CD69 expression 5- to 8-fold in both the GFP-positive and icillin/streptomycin/L-glutamine/10% FBS and 40 U/ml rIL-2 (R&D Sys- GFP-negative vector-infected Jurkat cells. Introduction of tems, Minneapolis, MN) in a 24-well plate coated with anti-CD3 (OKT3) PAK2⌬L reduced CD69 expression to the level before stimulation. and anti-CD28 Abs (each at 5 ␮g/ml). Primary T cells were stimulated for 2Ð3 days and then infected at 0.5 ϫ 106/well with 2 ml of viral supernatant. Stimulation of vector-infected BJAB cells with anti-IgM increased Forty hours after infection, primary T cells were stimulated at 37¡Cinthe CD69 surface expression ϳ4-fold in both GFP-negative and GFP- growth medium containing either 1 ␮g/ml C305 hybridoma supernatant or positive cells. However, in BJABs, introduction of PAK2⌬L had 50 ng/ml PMA and 1 ␮M ionomycin. For IL-2 secretion, the culture me- no effect on anti-IgM-induced CD69 up-regulation. dium was harvested, and the amount of IL-2 was determined by ELISA (DuoSet ELISA; R&D Systems). For CD40L up-regulation, the cells were To gain further insight into the role played by PAK2 in TCR stained with anti-CD40L Ab (BD PharMingen, San Diego, CA) at 4¡C for signaling, we looked at whether PAK2⌬L could also block PMA- 20 min and analyzed by FACS. induced CD69 expression. The phorbol ester, PMA, bypasses early The Journal of Immunology 7327 Downloaded from http://www.jimmunol.org/

FIGURE 2. TaqMan analysis of PAK1 and PAK2 tissue expression. mRNA expression was measured by quantitative RT-PCR using either by guest on September 25, 2021 PAK1 or PAK2 primers and total RNA from a panel of 15 tissues (PBL). To normalize for mRNA input, human ribosomal protein (Hu RIP) mRNA was measured for each tissue type. Results were expressed as the ratio of PAK1 or PAK2 mRNA level to Hu RIP.

signaling events by directly activating protein kinase C and downstream signaling molecules such as ERK1/2. To obtain a brighter GFP signal and facilitate gating of the GFP-positive cells, we switched to a transient expression system, using the pCDEF3- IRES/GFP plasmid. Although PAK2⌬L blocked anti- TCR-stimulated CD69 expression in Jurkat TAg cells, it did not prevent the more potent CD69 up-regulation (Ͼ30-fold) induced by PMA (Fig. 3B). The slight effect of PAK2⌬L on PMA-induced CD69 up-regulation was minor compared with the effect on anti- TCR stimulation and was less pronounced in other experiments. Taken together, this suggests that PAK2 plays a role in early signaling events after Ag receptor stimulation in T cells, but not B FIGURE 1. CD69 cell surface marker screen in Jurkat cells. A, Screen- cells. ing strategy. tTA Jurkat cells stably transfected with pTRA cDNA library ⌬ 2ϩ were stimulated with anti-TCR, and stained for CD69 surface expression. PAK2 L inhibits TCR-mediated Ca flux in Jurkat cells Cells expressing the lowest levels of TCR-induced CD69 were sorted, ex- To determine the effect of this inhibitory PAK2 construct on cal- panded, and resorted until significant enrichment of nonresponsive clones cium mobilization, Jurkat TAg cells were again transiently trans- was achieved. The cells were single-cell sorted and were reanalyzed for fected with either pCDEF3-PAK⌬L-IRES/GFP or empty vector, TCR-induced CD69 expression in the presence (cDNA off) and absence labeled with the calcium indicator, indo-1, and stimulated with (cDNA on) of Dox. B, Two kinase domain-truncated forms of PAK2, anti-TCR, and the calcium signals of the GFP-negative and GFP- PAK2⌬L and PAK2⌬S, were isolated multiple times as functional hits in the CD69 screen. Schematic of the structure of PAK2 and the hits from the positive cells were analyzed simultaneously. In comparison with CD69 screen. C, TCR-induced CD69 expression inhibited by switching on cells transfected with vector alone, or GFP-negative cells, de- PAK2⌬L. A Jurkat TAg screen clone stably expressing PAK2⌬L cultured creased Ag-induced calcium mobilization was observed in the in the presence or absence of Dox, and stimulated with anti-TCR (1 ␮g/ml PAK2⌬L-expressing cells (Fig. 4A). However, calcium influx in- C305) for 20Ð24 h and stained for CD69 expression. duced by the ionophore, ionomycin, was unaffected, confirming 7328 REGULATION OF T CELL ACTIVATION BY PAK2 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 4. PAK2⌬L specifically blocks T cell activation, and not T cell migration. A, PAK2⌬L inhibits Ca2ϩ flux induced by TCR cross- linking in Jurkat TAg cells. Jurkat TAg cells transiently expressing either pCDEF3-PAK⌬L-IRES/GFP (lower panel) or empty vector (upper panel) were labeled with the calcium indicator, indo-1, and stimulated with anti- TCR followed by ionomycin at 37¡ConanLSRflow cytometer with UV laser. Calcium influx, measured as a ratio of bound to unbound indo-1, was monitored over a 10-min time course. The calcium signals of the GFP- negative (untransfected) (hatched line) and GFP-positive (transfected) (bold line) cells were analyzed simultaneously. In the upper panel showing the vector control, the lines representing the GFP-positive and GFP-negative cells are overlapping. B, PAK2⌬L has no effect on T cell chemotaxis. Jurkat TAg cells were transiently transfected with pCDEF3-PAK⌬L-IRES/ GFP, pCDEF3-EDG5-IRES/GFP, or empty vector, and cultured in the presence of charcoal-stripped FBS. A total of 106 cells was added to the upper chamber of a Costar Transwell apparatus (5-␮m pore), and allowed to migrate for 2.5 h into the lower chamber containing either 100 nM SPP, FIGURE 3. Kinase domain-truncated PAK2 inhibits T cell activation. 100 nM SDF, or serum-free medium alone. GFP expression of the migrated A, PAK2⌬L inhibits Ag receptor signaling in T cells, but not in B cells. population (lower chamber) was analyzed by FACS. PAK2⌬L in the pTRA-IRES/GFP vector or vector alone was infected into Jurkat N32H cells or BJAB cells. After 48 h, the Jurkat and BJAB cells were stimulated with anti-TCR (1 ␮g/ml C305) and anti-BCR (0.3 ␮g/ml Ј anti-IgM F(ab )2), respectively, for 20Ð24 h, and CD69 expression of the GFP-positive and -negative cells was compared by analytical gating. The that these cells retained the capacity to mobilize calcium. These geometrical means of allophycocyanin-CD69 fluorescence are shown for data suggest that PAK⌬L negatively regulates signaling events ⌬ GFP-negative and GFP-positive cells. B, PAK2 L blocks TCR-, but not proximal to the TCR, which lead to mobilization of calcium, and PMA-induced CD69 up-regulation. Jurkat TAg cells were transiently trans- consequent activation of NFAT. There is evidence that inhibitors fected with pCDEF3-PAK⌬L-IRES/GFP or empty vector, and after 48 h of PLC and calmodulin block activation of PAK in neutrophils, were treated with anti-TCR (1 ␮g/ml) or PMA (50 ng/ml) for 20Ð24 h. CD69 expression of the GFP-positive and -negative cells was compared by suggesting that PAK is activated downstream of PLC (36). How- analytical gating. The geometrical means of CD69-allophycocyanin fluo- ever, in mast cells, Cdc42/Rac has been placed upstream of ino- rescence of the GFP-positive cells and the GFP-negative cells are shown sitol 1,4,5-triphosphate production and calcium mobilization, pos- beneath the dot plots. sibly interacting directly with PLC␥1 (37). It would be interesting The Journal of Immunology 7329 to look at the effect of PAK2⌬L expression on PLC␥ phosphory- We wanted to test the effect of PAK2⌬L on TCR-triggered NFAT lation in T cells to determine whether PAK2 plays a role in reg- activation. To further confirm the importance of PAK2 kinase ac- ulating the activity of PLC␥ or its potential interaction with tivity in transmitting signals downstream of the TCR, we also con- Cdc42/Rac. structed a kinase dead-point mutant of PAK2. Based on homology with PAK1 and other serine threonine kinases, Lys278, a critical Jurkat cell migration toward SPP and SDF is unaffected by lysine residue in the active site of the kinase domain, was mutated ⌬ PAK2 L to alanine. The effect of the full-length wild-type and kinase-inac- Having established that PAK2 is involved in signaling downstream tive PAK2 constructs on TCR-stimulated CD69 up-regulation of the TCR, we wanted to determine whether PAK2 played a role (Fig. 5D) and NFAT activation (E) was tested. The kinase-inactive in activation-independent T cell functions. Lymphocyte trafficis point mutant inhibited both TCR-mediated CD69 up-regulation controlled by chemokines, which can be broadly classified into and NFAT activation, albeit not as potently as the PAK2 trunca- inflammatory and homeostatic. Inflammatory chemokines are re- tion, PAK2⌬L, whereas the wild-type kinase had little or no effect. sponsible for recruiting activated lymphocytes to sites of inflam- Taken together, these data suggest that stimulation of the TCR mation. In contrast, inflammation-unrelated chemokines are con- leads to enhanced PAK2 kinase activity, resulting in the activation stitutively produced and are involved in maintaining physiological of target genes. traffic, guiding precursor cells to correct sites in bone marrow and thymus to ensure proper progenitor cell development (38). SDF-1 The PAK2 fragment PAK283Ð149 has transdominant kinase and its receptor CXCR4 play a critical role in lymphopoeisis and inhibitory activity myelopoeisis. Thymocytes of all developmental stages migrate to PAKs are known to be critical effectors of Cdc42/Rac (4). Al- Downloaded from SDF-1. The lysosphingolipid SPP, thought to act both as an ex- though Cdc42/Rac binding results in PAK activation, there is a tracellular mediator and an intracellular second messenger, is also report of PAK1 mediating GTPase function independent of PAK a potent chemoattractant for resting T cells (39). SPP binds mem- kinase activity (43). Furthermore, activation of PAK kinase activ- bers of the endothelial differentiation gene (EDG) family of plasma ity independent of GTPase binding cannot be ruled out. Both dom- membrane G protein-coupled receptors. We looked at the ability of inant-negative Ras and Cdc42 inhibit TCR-induced NFAT activa-

Jurkat cells expressing PAK2⌬L to migrate toward SDF-1 and tion, but, as expected, only the latter has an effect on PAK activity http://www.jimmunol.org/ SPP. Migration of Jurkat T cells expressing PAK2⌬L to SPP and (17). Our studies so far have shown that both a kinase-domain SDF-1 was comparable with that of cells expressing control vector truncation and a kinase-inactive point mutant form of PAK2 can (Fig. 4B). In contrast, overexpression of EDG-5, a receptor for block TCR-mediated signals. However, both of these dominant SPP, blocked T cell migration toward SPP (Fig. 4B), but not negatives retain the ability to bind Cdc42/Rac and are perhaps SDF-1 (data not shown). This suggests that PAK2 specifically reg- blocking downstream of the TCR the function of Cdc42/Rac, ulates receptor-mediated activation of T cells, but plays no role in PAK1, and PAK2. Although we showed that PAK2 kinase activity other activation-independent T cell functions, such as migration of is increased in response to TCR cross-linking, this does not prove resting T cells toward SPP or SDF-1. that PAK2 is a critical link between TCR stimulation and T cell activation. To assess the relevance of PAK2 in TCR signaling, we by guest on September 25, 2021 PAK2 kinase activity is enhanced following TCR stimulation wanted to examine the effect of specifically blocking PAK2 kinase The PAK2 hits obtained from the CD69 T cell screen were all activity without affecting Cdc42/Rac. kinase domain truncations, suggesting that PAK2 kinase activity To do this, we referred to insights gained from mechanistic stud- may be critical for the signaling events mediated by PAK2 in re- ies of PAK1. N-terminal deletion of PAK1 dramatically activates sponse to TCR stimulation. We therefore cloned full-length PAK2 its kinase activity, suggesting some form of autoregulation (44). to examine the effect of TCR cross-linking on its kinase activity. Interestingly, the PAK1 fragment (70Ð149) inhibits PAK1 activa- Jurkat TAg cells transiently expressing wild-type PAK2-V5-His6 tion in vivo and in vitro and binds directly to the PAK1 kinase were stimulated with anti-TCR for 0Ð10 min. PAK2 kinase activ- domain. This inhibitory activity is not dependent on binding to ity was increased Ͼ3-fold after 5 min of anti-TCR cross-linking Cdc42/Rac. (Fig. 5A). Immunoprecipitation from Jurkat TAg cells transfected An elegant study of the crystal structure of PAK1 provided fur- with vector alone or a kinase-inactive version of PAK2 resulted in ther insights into the mechanism of PAK activation (45). PAK1 no histone H4 phosphorylation (data not shown). appears to exist as a dimer in an autoinhibited conformation. The Because activation of PAK1 downstream of the TCR has been autoregulatory fragment (70Ð149) contains a dimerization domain reported previously (17, 18), we sought to confirm that the endog- (80Ð87), an inhibitory switch domain (87Ð136), and a kinase in- enous PAK2 protein was also activated. Jurkat TAg cells were hibitory domain (137Ð149). Two PAK1 monomers associate stimulated with anti-TCR for 0Ð20 min, and PAK1 and PAK2 through the dimerization domains, and the inhibitory switch do- proteins were immunoprecipitated. Both PAK1 and PAK2 proteins main tightly associates with the C-terminal lobe of the kinase do- were activated in response to anti-TCR stimulation (Fig. 5B). T main, whereas the kinase inhibitory domain occupies the cleft of cell activation was further confirmed by measuring induction of the kinase domain and thereby stabilizes a disabled catalytic site. linker for activation of T cells (LAT) and p42/44 MAPK phos- Binding of GTP-loaded Cdc42 or Rac to the CRIB domain (75Ð90) phorylation in the cell lysates (Fig. 5C). disrupts the dimer, and unfolds the inhibitory switch domain, re- moving the kinase inhibitory domain from the cleft of the kinase PAK2 signaling downstream of the TCR is dependent on its domain. This releases the activation loop, allowing phosphoryla- kinase activity tion of Thr423, which results in activation of the kinase. Once ac- One of the outcomes of T cell activation is the production of a tivated, the kinase autophosphorylates at several Ser residues, pre- variety of cytokines such as IL-2 (40). In Jurkat cells, cross-linking venting reversion to an inactive conformation. the TCR is sufficient to activate the NFAT transcription factor, To generate a dominant-negative PAK2 that does not bind which binds and activates the IL-2 promoter. NFAT activation Cdc42/Rac, a GFP-PAK283Ð149 fusion was constructed (Fig. 6A). requires inputs from both the calcium and the Ras/MAPK path- Based on the PAK1 studies, this short PAK283Ð149 fragment ways, activated in response to Ag receptor stimulation (40Ð42). should act as a dominant negative by directly occupying the cleft 7330 REGULATION OF T CELL ACTIVATION BY PAK2 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 5. PAK2 kinase activity is required for its role in T cell signaling. A, Exogenous PAK2 kinase activity is enhanced following TCR stimulation. Jurkat TAg cells transiently expressing wild-type PAK2-V5-His6 were stimulated with anti-TCR for 0Ð10 min. The cells were lysed, and the exogenous PAK2 was immunoprecipitated through the V5 tag. The activity of the immunoprecipitated kinase at the different time points after stimulation was determined by in vitro kinase assay using histone H4 as substrate. The 32P-labeled products of the kinase reaction were separated by SDS-PAGE and visualized by phosphor imager (upper panel). The amount of PAK2-V5-His6 immunoprecipitated was determined by anti-V5 Western blot (lower panel). B, Endogenous PAK1 and PAK2 kinase activities are enhanced following TCR stimulation. Jurkat TAg cells were starved for 2 h and stimulated with anti-TCR for 0Ð20 min. The cells were lysed, and endogenous PAK1 and PAK2 were immunoprecipitated through their N terminus (SC no. 882 and CST no. 2608, respectively). The activity of the immunoprecipitated kinase at the different time points after stimulation was determined as in A. The fold of induction of PAK1 and PAK2 kinase activities was quantitated in multiple experiments and averaged (upper panel). The amount of PAK1 and PAK2 proteins immunoprecipitated was determined by anti-PAK1 (CST no. 2602) and anti-PAK2 (CST no. 2608) Western blot. C, Jurkat activation was conﬁrmed by probing Western blots of the lysates for phosphorylation of LAT and p42/44 MAPK. PAK1 and PAK2 levels are shown as loading controls. D, Full-length kinase-inactive PAK2 blocks TCR-mediated CD69 up-regulation. Jurkat TAg cells were transiently transfected with 40 ␮g of WT (wild type)-PAK2-V5-His6, KI-PAK2-V5-His6, PAK2⌬L-V5-His6, or vector alone, and cotransfected with 4 ␮g of pCDEF3.IRES/GFP. The cells were stimulated with anti-TCR (1 ␮g/ml C305) for 20Ð24 h. CD69 expression of the GFP-positive and -negative cells was compared by analytical gating. The geometrical means of CD69-allophycocyanin ﬂuorescence of the GFP-positive cells and the GFP-negative cells are shown in the respective gates. E, Full-length kinase-inactive PAK2 blocks TCR-induced NFAT activation. Jurkat TAg cells were transiently transfected with WT-PAK2-V5-His6, KI- PAK2-V5-His6, PAK2⌬L-V5-His6, or vector alone, as well as NFAT-luciferase reporter and a second control TK luciferase construct. After 48 h, the cells were stimulated with anti-TCR (1 ␮g/ml C305) for 12 h or left untreated. The cells were lysed and assayed for luciferase activity in triplicate. Fold induction of luciferase activity over the unstimulated, vector-transfected sample is shown. The data are representative of several independent experiments. The Journal of Immunology 7331 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 6. PAK283Ð149 is a transdominant inhibitor of PAK2 kinase activity. A, Schematic representation of the PAK2 domain constructs. The 6 KI-kinase domain is V5-His tagged. The isolated CRIB domain and PAK283Ð149 fragment are GFP fusions. The two histidine residues critical for 6 ء Cdc42/Rac binding are indicated ( ). B, PAK283Ð149 binds directly to the kinase domain of PAK2. The PAK2 KI-kinase-V5-His was transiently coexpressed in Jurkat cells with either GFP-PAK283Ð149 or GFP-CRIB74Ð87. The cells were lysed, and the GFP fusions were immunoprecipitated through the 6 GFP tag. Any associated PAK2 KI-kinase-V5-His was detected by Western blot using an anti-V5 Ab. C, PAK283Ð149 blocks TCR-mediated CD69 ⌬ up-regulation. Jurkat TAg cells were transiently transfected with GFP-CRIB, GFP-PAK283Ð149, pCDEF3-PAK L-IRES/GFP, or vector alone. The cells were stimulated with anti-TCR (1 ␮g/ml C305) for 20Ð24 h. CD69 expression of the GFP-positive and -negative cells was compared by analytical gating. The geometrical means of CD69-allophycocyanin ﬂuorescence of the GFP-positive cells and the GFP-negative cells are shown in the respective gates.

of the kinase domain and preventing activation of the kinase. How- Having shown that GFP-PAK283Ð149 can bind directly to the kinase ever, it should not interfere with Cdc42 or Rac binding to the CRIB domain of PAK2, we next examined whether this fragment could domain. PAK283Ð149 lacks ﬁve of eight of the core residues of the block TCR-induced CD69 up-regulation in Jurkat cells. Jurkat TAg CRIB motif, including one of the two histidines, critical for Cdc42/ cells were transiently transfected with GFP-CRIB, GFP-PAK283Ð149, Rac binding (46). A short GFP fusion containing the CRIB domain pCDEF3-PAK⌬L-IRES/GFP, or vector alone, and stimulated with

(74Ð87) was generated as a control. To determine whether the anti-TCR for 20Ð24 h (Fig. 6C). Overexpression of GFP-PAK283Ð149, 6 PAK283Ð149 fragment could bind to the kinase domain, a V5-His - but not GFP-PAK2 CRIB, inhibited TCR-induced CD69 up-regula- tagged kinase-inactive isolated PAK2 kinase domain (162Ð524) tion in Jurkat cells. Taken together, these results strongly suggest that was also constructed. Again, due to the bacterial toxicity of the GFP-PAK283Ð149 acts as a transdominant inhibitor of PAK2 kinase wild-type V5-His6-tagged PAK2 kinase domain, the kinase-inac- activity by directly interacting with the kinase domain. Direct inhibi- tive form (K278A) was used for the coimmunoprecipitation ex- tion of PAK2 kinase activity, independent of Cdc-42/Rac binding, periments. The PAK2 KI-kinase-V5-His6 was coexpressed in Ju- blocks TCR-induced T cell activation, indicating that PAK2 is a crit- rkat cells with either GFP-PAK283Ð149 or GFP-PAK2-CRIB. The ical mediator of signals originating from the TCR. kinase domain of PAK2 coimmunoprecipitated with GFP- The inhibitory effect of GFP-PAK283Ð149 is weaker than that of ⌬ PAK283Ð149, but not with GFP-PAK2 CRIB (Fig. 6B). PAK L. This could simply be due to expression levels, given that 7332 REGULATION OF T CELL ACTIVATION BY PAK2 a 10-fold excess of the inhibitory PAK1 fragment (75Ð132) was required for maximal inhibition of PAK1 kinase activity in vitro (44). In addition, PAK⌬L could block both PAK1 and PAK2 activation by potentially blocking Rac/Cdc42 function, whereas

GFP-PAK283Ð149 would block only PAK2 activation. Further- more, the presence of the Nck binding domain in PAK2⌬L may also contribute to its greater inhibitory activity. Dominant-negative Nck has previously been shown to block NFAT activation in Ju- rkat cells (17). Collectively, it is likely that the role of PAK2 in TCR signaling is mediated by its kinase activity and its multiple interactions with regulatory proteins.

PAK2⌬L inhibits T cell signaling in primary T cells Having placed PAK2 downstream of the TCR, we wanted to confirm that this novel role assigned to PAK2 is indeed physiologi- cally relevant. We therefore examined the effect of dominant-negative PAK2 on two primary T cell functions, IL-2 secretion and CD40L up-regulation. Primary T cells purified from PBMCs were infected with PAK2⌬L in the pTRA-IRES/GFP vector or vector Downloaded from alone, sorted, and stimulated with either anti-TCR/anti-CD28, or PMA/ionomycin. IL-2 secretion was determined for both GFP- positive and GFP-negative primary T cells (Fig. 7A). PAK2⌬L inhibited IL-2 secretion stimulated by anti-TCR/anti-CD28, but not in response to PMA/ionomycin. CD40L is up-regulated in response to T cell stimulation through the TCR. Expression of http://www.jimmunol.org/ PAK2⌬L in primary T cells caused a significant decrease in the total number of CD40L-expressing cells compared with primary T cells expressing GFP alone or GFP-negative cells (Fig 7B, left panel). In addition, in those PAK2⌬L-expressing cells displaying surface CD40L, a 2-fold reduction in the amount of CD40L expressed was observed (Fig. 7B, right panel). These results suggest that PAK2 plays an important role in the activation of primary T cells. by guest on September 25, 2021 Discussion In this report, we describe the use of a functional retroviral-based genetic screen to identify novel components of the TCR signaling pathway. We identified a novel role for PAK2 as a positive regu- FIGURE 7. PAK2 is a positive regulator of primary T cell activation. A, ⌬ lator of T cell activation. We have shown that a kinase domain Primary T cells were prepared from PBMCs. PAK2 L in the pCRU-IRES/ GFP vector or vector alone was infected into primary T cells. The cells truncation of PAK2 behaves as a dominant negative and blocks were sorted 48 h after infection into pure GFP-positive and GFP-negative up-regulation of cell surface expression of the activation marker populations, and stimulated for 20 h with either anti-TCR/anti-CD28 or CD69, NFAT promoter activation, and calcium flux in Jurkat cells PMA/ionomycin, or left untreated. For both GFP-positive and GFP-nega- (Figs. 4A, and 5, D and E). Simply mutating a single residue in the tive cells, IL-2 secretion was measured by ELISA. B, Primary T cells were kinase domain of PAK2, thereby rendering it inactive, mimics the infected and stimulated as above. Both the percentage of CD40L-express- effect of complete removal of the kinase domain, illustrating that ing cells in the GFP-positive and GFP-negative populations (left panel) and the function of PAK2 is dependent on its kinase activity (Fig. 5). the geometric mean fluorescence of CD40L expression (right panel) were However, dominant-negative PAK2 did not significantly block measured by FACS. phorbol ester-induced T cell activation, T cell chemotaxis, or receptor-mediated B cell activation (Figs. 3 and 4B). Given the sim- ilarity between the PAK family members, particularly PAK1 and PAK2, and given their universal ability to bind Cdc42/Rac, we PAK2, like other PAK family members, is activated upon wanted to confirm that PAK2 plays a role in linking TCR stimu- Cdc42/Rac binding to the CRIB domain within the N-terminal lation with T cell activation. Therefore, we generated a PAK2 regulatory region. Although Rho family GTPases are known to fragment (PAK283Ð149) that bound directly to the PAK2 kinase play important roles in the regulation of actin-based cytoskeletal domain, thereby inhibiting its activation (Fig. 6, A and B). Trans- organization, they also play a role in more diverse cellular func- dominant PAK283Ð149 inhibited TCR-mediated CD69 up-regula- tions, such as cell cycle progression, gene expression, and response tion, illustrating that PAK2 forms an essential link between TCR to environmental stress (5). Despite the close homology of PAK1 engagement and T cell activation (Fig. 6C). Furthermore, we dem- and PAK2, these two proteins appear to fulfill quite different func- onstrated that endogenous and transfected PAK2 kinase activity is tions. PAK1 interacts with Nck and PIX, which localize it to the enhanced following TCR cross-linking (Fig. 5, AÐC). Finally, we cell membrane, and more specifically to focal complexes in the showed that dominant-negative PAK2 had similar effects in pre- case of PIX (12, 16, 47, 48). There, in response to growth factors venting primary T cell activation (Fig. 7). Taken together, these and other stimuli, PAK1 is activated and enhances cell migration, results indicate that PAK2 is a critical regulator of the T cell at least in part by promoting the disassembly of focal complexes activation. (9, 13Ð15). No role in these processes has been attributed to PAK2. The Journal of Immunology 7333

However, both PAK1 and PAK2 can phosphorylate myosin II reg- individual Rac GTPases may help to explain how signaling spec- ulatory light chain in vitro and have been implicated in fibroblast ificity is achieved in a background of high protein homology. and endothelial cell retraction (15, 49, 50). PAK2 has primarily been attributed cytostatic properties and is References activated in response to DNA damage and different types of stress. 1. Weiss, A., and D. R. Littman. 1994. Signal transduction by lymphocyte antigen Microinjection of PAK2 into frog embryos results in arrest of cell receptors. Cell 76:263. 2. Chu, P., J. Pardo, H. Zhao, C. C. Li, E. Pali, M. M. Shen, K. Qu, S. X. Yu, cleavage (51). Furthermore, the regulatory domain of PAK2 is B. C. Huang, P. Yu, et al. 2003. Systematic identification of regulatory proteins cleaved by caspase-3 in Jurkat cells undergoing apoptosis (21). critical for T-cell activation. J. Biol. 2:21. 3. Jaffer, Z. M., and J. Chernoff. 2002. p21-activated kinases: three more join the This produces a constitutively active fragment that regulates some Pak. Int. J. Biochem. Cell Biol. 34:713. of the morphological changes seen in apoptotic cells, which in- 4. Bagrodia, S., and R. A. Cerione. 1999. Pak to the future. Trends Cell Biol. 9:350. clude the formation of apoptotic bodies and the exposure of phos- 5. Bar-Sagi, D., and A. Hall. 2000. Ras and Rho GTPases: a family reunion. Cell 103:227. phatidylserine. This caspase-mediated activation has been reported 6. Chong, C., L. Tan, L. Lim, and E. Manser. 2001. The mechanism of PAK acti- only for PAK2. In addition, PAK2, but not PAK1, is activated and vation: autophosphorylation events in both regulatory and kinase domains control activity. J. Biol. Chem. 276:17347. translocates to the particulate fraction in response to hyperosmotic 7. Buchwald, G., E. Hostinova, M. G. Rudolph, A. Kraemer, A. Sickmann, stress (19). However, PAK2 can phosphorylate the proapoptotic H. E. Meyer, K. Scheffzek, and A. Wittinghofer. 2001. Conformational switch family member Bad, and protect against cell death in BALB3T3 and role of phosphorylation in PAK activation. Mol. Cell. Biol. 21:5179. 8. Abo, A., J. Qu, M. S. Cammarano, C. Dan, A. Fritsch, V. Baud, B. Belisle, and fibroblasts (22, 52). This antiapoptotic role appears to be a function A. Minden. 1998. PAK4, a novel effector for Cdc42Hs, is implicated in the of the full-length kinase rather than the caspase-cleaved form. reorganization of the actin cytoskeleton and in the formation of filopodia. EMBO

J. 17:6527. Downloaded from In this paper, we have demonstrated a novel function of PAK2 9. Manser, E., H. Y. Huang, T. H. Loo, X. Q. Chen, J. M. Dong, T. Leung, and in the activation of T cells. To date, only PAK1 has been attributed L. Lim. 1997. Expression of constitutively active ␣-PAK reveals effects of the a role in TCR signaling. Yablonski et al. (17) demonstrated a role kinase on actin and focal complexes. Mol. Cell. Biol. 17:1129. 10. Zhao, Z. S., E. Manser, X. Q. Chen, C. Chong, T. Leung, and L. Lim. 1998. A for PAK1 downstream of Vav and Cdc42 in the TCR-induced conserved negative regulatory region in ␣PAK: inhibition of PAK kinases reveals activation of ERK2 and NFAT. A requirement for a complex be- their morphological roles downstream of Cdc42 and Rac1. Mol. Cell. Biol. 18:2153. tween SLP-76, Nck, and Vav was also demonstrated (53). Subse- 11. Sanders, L. C., F. Matsumura, G. M. Bokoch, and P. de Lanerolle. 1999. Inhi- quently, however, TCR engagement was shown to stimulate acti- bition of myosin light chain kinase by p21-activated kinase. Science 283:2083. http://www.jimmunol.org/ vation of PAK1 only if in a complex with PIX and p95PKL, in a 12. Manser, E., T. H. Loo, C. G. Koh, Z. S. Zhao, X. Q. Chen, L. Tan, I. Tan, T. Leung, and L. Lim. 1998. PAK kinases are directly coupled to the PIX family ZAP70-, but not LAT-, SLP-76-, or Nck-dependent manner (18). of nucleotide exchange factors. Mol. Cell 1:183. Furthermore, two reports suggest that stimulation of CD28 alone is 13. Zhao, Z. S., E. Manser, T. H. Loo, and L. Lim. 2000. Coupling of PAK-interacting exchange factor PIX to GIT1 promotes focal complex disassembly. Mol. sufficient to activate PAK1 and that costimulation of CD28 with Cell. Biol. 20:6354. TCR enhances PAK1 activity even further than stimulation of ei- 14. Kiosses, W. B., R. H. Daniels, C. Otey, G. M. Bokoch, and M. A. Schwartz. ther alone (54, 55). In contrast, Yablonski et al. (17) did not detect 1999. A role for p21-activated kinase in endothelial cell migration. J. Cell Biol. 147:831. any effect of CD28 stimulation, either alone or in conjunction with 15. Sells, M. A., J. T. Boyd, and J. Chernoff. 1999. p21-activated kinase 1 (Pak1) regulates cell motility in mammalian fibroblasts. J. Cell Biol. 145:837. anti-TCR, on the activity of PAK1. by guest on September 25, 2021 16. Daniels, R. H., P. S. Hall, and G. M. Bokoch. 1998. Membrane targeting of We identified PAK2 in a screen designed to discover novel pro- p21-activated kinase 1 (PAK1) induces neurite outgrowth from PC12 cells. teins involved in TCR-induced T cell activation. PAK2 has been EMBO J. 17:754. reported to be activated as a result of caspase-mediated cleavage in 17. Yablonski, D., L. P. Kane, D. Qian, and A. Weiss. 1998. A Nck-Pak1 signaling module is required for T-cell receptor-mediated activation of NFAT, but not of T cells undergoing Fas-induced cell death (21). This is the first JNK. EMBO J. 17:5647. report of a requirement for PAK2 in the activation of T cells. We 18. Ku, G. M., D. Yablonski, E. Manser, L. Lim, and A. Weiss. 2001. A PAK1- PIX-PKL complex is activated by the T-cell receptor independent of Nck, Slp-76 propose that PAK2 is a critical player in T cell activation. Several and LAT. EMBO J. 20:457. lines of evidence support this idea. Jurkat T cells predominantly 19. Roig, J., Z. Huang, C. Lytle, and J. A. Traugh. 2000. p21-activated protein kinase express PAK2 (21). Furthermore, the highly conserved HIV pro- ␥-PAK is translocated and activated in response to hyperosmolarity: implication of Cdc42 and phosphoinositide 3-kinase in a two-step mechanism for ␥-PAK tein, Nef, specifically associates with and activates PAK2, but not activation. J. Biol. Chem. 275:16933. PAK1, in T cells (33, 34, 56). Nef plays an essential role in the 20. Tang, T. K., W. C. Chang, W. H. Chan, S. D. Yang, M. H. Ni, and J. S. Yu. 1998. pathogenesis of HIV and subsequent development of AIDS, by Proteolytic cleavage and activation of PAK2 during UV irradiation-induced apoptosis in A431 cells. J. Cell. Biochem. 70:442. promoting efficient viral budding and enhancing viral particle in- 21. Rudel, T., and G. M. Bokoch. 1997. Membrane and morphological changes in fectivity. Nef down-regulates CD4 and MHC class I surface ex- apoptotic cells regulated by caspase-mediated activation of PAK2. Science 276:1571. pression, and interferes with signaling downstream of the TCR. 22. Jakobi, R., E. Moertl, and M. A. Koeppel. 2001. p21-activated protein kinase More recently, Nef was shown to prevent T cell apoptosis through ␥-PAK suppresses programmed cell death of BALB3T3 fibroblasts. J. Biol. the activation of PAK2 and phosphorylation and inactivation of the Chem. 276:16624. 23. Huang, Z., J. Ling, and J. A. Traugh. 2003. Localization of p21-activated protein proapoptotic Bcl-2 family member, Bad (57). Taken together, kinase ␥-PAK/Pak2 in the endoplasmic reticulum is required for induction of these data implicate PAK2 as a critical regulator of T cell function. cytostasis. J. Biol. Chem. 278:13101. 24. Swift, S. E., J. B. Lorens, P. Achacoso, and G. P. Nolan. 2000. Rapid production In a recent report, the role of the hemopoietic-specific GTPase of retroviruses for efficient gene delivery to mammalian cells using 293T cell- Rac2 in T cell activation was studied using Rac2 knockout mice based systems. In Current Protocols in Immunology, Vol. 10.17C. J. E. Coligan, (58). The effects of TCR engagement, including proliferation, ac- A. M. Kruisbeek, D. H. Margulies, E. M. Shevach, and W. Strober, eds. Wiley, New York, p. 1Ð17. tivation of MAPKs, calcium mobilization, and actin polymeriza- 25. Gossen, M., and H. Bujard. 1992. Tight control of gene expression in mammalian tion were reduced in RacϪ/Ϫ T cells. Rac2 acts, most likely down- cells by tetracycline-responsive promoters. Proc. Natl. Acad. Sci. USA 89:5547. 26. Wu, J., S. Katzav, and A. Weiss. 1995. A functional T-cell receptor signaling stream of Vav, as an important mediator of both transcriptional and pathway is required for p95vav activity. Mol. Cell. Biol. 15:4337. cytoskeletal changes during T cell activation. Having demon- 27. Xu, X., C. Leo, Y. Jang, E. Chan, D. Padilla, B. C. Huang, T. Lin, T. Gururaja, strated the key role played by PAK2 in T cell activation, combined Y. Hitoshi, J. B. Lorens, et al. 2001. Dominant effector genetics in mammalian cells. Nat. Genet. 27:23. with the evidence that PAK2 is activated by Cdc42/Rac down- 28. Lorens, J. B., M. K. Bennett, D. M. Pearsall, W. R. Throndset, A. B. Rossi, stream of Vav, it is tempting to speculate that Rac2 connects sig- R. J. Armstrong, B. P. Fox, E. H. Chan, Y. Luo, E. Masuda, et al. 2000. Retroviral delivery of peptide modulators of cellular functions. Mol. Ther. 1:438. nals from the TCR with cytoskeletal changes through the specific 29. Weiss, A., and J. D. Stobo. 1984. Requirement for the coexpression of T3 and the activation of PAK2. Matching different PAK effector kinases with T cell antigen receptor on a malignant human T cell line. J. Exp. Med. 160:1284. 7334 REGULATION OF T CELL ACTIVATION BY PAK2

30. Holland, S. J., X. C. Liao, M. K. Mendenhall, X. Zhou, J. Pardo, P. Chu, 45. Lei, M., W. Lu, W. Meng, M. C. Parrini, M. J. Eck, B. J. Mayer, and C. Spencer, A. Fu, N. Sheng, P. Yu, et al. 2001. Functional cloning of Src-like S. C. Harrison. 2000. Structure of PAK1 in an autoinhibited conformation reveals adapter protein-2 (SLAP-2), a novel inhibitor of antigen receptor signaling. a multistage activation switch. Cell 102:387. J. Exp. Med. 194:1263. 46. Burbelo, P. D., D. Drechsel, and A. Hall. 1995. A conserved binding motif de- 31. Luria, S., I. Chambers, and P. Berg. 1991. Expression of the type 1 human fines numerous candidate target proteins for both Cdc42 and Rac GTPases. immunodeficiency virus Nef protein in T cells prevents antigen receptor-medi- J. Biol. Chem. 270:29071. ated induction of interleukin 2 mRNA. Proc. Natl. Acad. Sci. USA 88:5326. 47. Galisteo, M. L., J. Chernoff, Y. C. Su, E. Y. Skolnik, and J. Schlessinger. 1996. 32. Iafrate, A. J., S. Bronson, and J. Skowronski. 1997. Separable functions of Nef The adaptor protein Nck links receptor tyrosine kinases with the serine-threonine disrupt two aspects of T cell receptor machinery: CD4 expression and CD3 sig- kinase Pak1. J. Biol. Chem. 271:20997. naling. EMBO J. 16:673. 48. Zhao, Z. S., E. Manser, and L. Lim. 2000. Interaction between PAK and nck: a 33. Renkema, G. H., A. Manninen, D. A. Mann, M. Harris, and K. Saksela. 1999. template for Nck targets and role of PAK autophosphorylation. Mol. Cell. Biol. Identification of the Nef-associated kinase as p21-activated kinase 2. Curr. Biol. 20:3906. 9:1407. 49. Goeckeler, Z. M., R. A. Masaracchia, Q. Zeng, T. L. Chew, P. Gallagher, and 34. Renkema, G. H., A. Manninen, and K. Saksela. 2001. Human immunodeficiency R. B. Wysolmerski. 2000. Phosphorylation of myosin light chain kinase by p21- virus type 1 Nef selectively associates with a catalytically active subpopulation of activated kinase PAK2. J. Biol. Chem. 275:18366. ␤ p21-activated kinase 2 (PAK2) independently of PAK2 binding to Nck or -PIX. 50. Zeng, Q., D. Lagunoff, R. Masaracchia, Z. Goeckeler, G. Cote, and J. Virol. 75:2154. R. Wysolmerski. 2000. Endothelial cell retraction is induced by PAK2 mono- 35. Daniels, R. H., and G. M. Bokoch. 1999. p21-activated protein kinase: a crucial phosphorylation of myosin II. J. Cell Sci. 113:471. component of morphological signaling? Trends Biochem. Sci. 24:350. 51. Rooney, R. D., P. T. Tuazon, W. E. Meek, E. J. Carroll, J. J. Hagen, E. L. Gump, 36. Lian, J. P., L. Crossley, Q. Zhan, R. Huang, P. Coffer, A. Toker, D. Robinson, and C. A. Monnig, T. Lugo, and J. A. Traugh. 1996. Cleavage arrest of early frog J. A. Badwey. 2001. Antagonists of calcium fluxes and calmodulin block acti- embryos by the G protein-activated protein kinase PAK I. J. Biol. Chem. vation of the p21-activated protein kinases in neutrophils. J. Immunol. 166:2643. 271:21498. 37. Hong-Geller, E., and R. A. Cerione. 2000. Cdc42 and Rac stimulate exocytosis 52. Schurmann, A., A. F. Mooney, L. C. Sanders, M. A. Sells, H. G. Wang, of secretory granules by activating the IP /calcium pathway in RBL-2H3 mast 3 J. C. Reed, and G. M. Bokoch. 2000. p21-activated kinase 1 phosphorylates the cells. J. Cell Biol. 148:481.

death agonist bad and protects cells from apoptosis. Mol. Cell. Biol. 20:453. Downloaded from 38. Moser, B., and P. Loetscher. 2001. Lymphocyte traffic control by chemokines. Nat. Immunol. 2:123. 53. Bubeck Wardenburg, J., R. Pappu, J. Y. Bu, B. Mayer, J. Chernoff, D. Straus, and 39. Pyne, S., and N. J. Pyne. 2000. Sphingosine 1-phosphate signalling in mamma- A. C. Chan. 1998. Regulation of PAK activation and the T cell cytoskeleton by lian cells. Biochem. J. 349:385. the linker protein SLP-76. Immunity 9:607. 40. Durand, D. B., J. P. Shaw, M. R. Bush, R. E. Replogle, R. Belagaje, and 54. Kaga, S., S. Ragg, K. A. Rogers, and A. Ochi. 1998. Activation of p21-CDC42/ G. R. Crabtree. 1988. Characterization of antigen receptor response elements Rac-activated kinases by CD28 signaling: p21-activated kinase (PAK) and MEK within the interleukin-2 enhancer. Mol. Cell. Biol. 8:1715. kinase 1 (MEKK1) may mediate the interplay between CD3 and CD28 signals. 41. Emmel, E. A., C. L. Verweij, D. B. Durand, K. M. Higgins, E. Lacy, and J. Immunol. 160:4182. 55. Salojin, K. V., J. Zhang, and T. L. Delovitch. 1999. TCR and CD28 are coupled G. R. Crabtree. 1989. Cyclosporin A specifically inhibits function of nuclear http://www.jimmunol.org/ proteins involved in T cell activation. Science 246:1617. via ZAP-70 to the activation of the Vav/Rac-1- /PAK-1/p38 MAPK signaling 42. Woodrow, M., N. A. Clipstone, and D. Cantrell. 1993. p21ras and calcineurin pathway. J. Immunol. 163:844. synergize to regulate the nuclear factor of activated T cells. J. Exp. Med. 56. Arora, V. K., R. P. Molina, J. L. Foster, J. L. Blakemore, J. Chernoff, 178:1517. B. L. Fredericksen, and J. V. Garcia. 2000. Lentivirus Nef specifically activates 43. Hullinger, T. G., R. L. Panek, X. Xu, and S. K. Karathanasis. 2001. p21-activated Pak2. J. Virol. 74:11081. kinase-1 (PAK1) inhibition of the human scavenger receptor class B, type I pro- 57. Wolf, D., V. Witte, B. Laffert, K. Blume, E. Stromer, S. Trapp, P. d’Aloja, moter in macrophages is independent of PAK1 kinase activity, but requires the A. Schurmann, and A. S. Baur. 2001. HIV-1 Nef associated PAK and PI3-kinases GTPase-binding domain. J. Biol. Chem. 276:46807. stimulate Akt-independent Bad-phosphorylation to induce anti-apoptotic signals. 44. Frost, J. A., A. Khokhlatchev, S. Stippec, M. A. White, and M. H. Cobb. 1998. Nat. Med. 7:1217. Differential effects of PAK1-activating mutations reveal activity-dependent and 58. Yu, H., D. Leitenberg, B. Li, and R. A. Flavell. 2001. Deficiency of small GTPase -independent effects on cytoskeletal regulation. J. Biol. Chem. 273:28191. Rac2 affects T cell activation. J. Exp. Med. 194:915. by guest on September 25, 2021