PIPKIγ90 Negatively Regulates LFA-1–Mediated Adhesion and Activation in Antigen-Induced CD4⁺ T Cells

T cells are highly motile cells that can rapidly transition to form stable, long-lasting contacts with APCs bearing a peptide fragment in the context of the MHC. Contact sites between T cells and APCs, known as immune synapses, are characterized by the polarization of signaling and adhesive molecules at the contact interfaces, including the integrin LFA-1 (1). Recent advances in imaging have shown that T cells can maintain contact with Ag-bearing dendritic cells for up to 24 h (2) and that the duration of contact between T cells and APCs correlates with the degree of proliferation and cytokine production both in vitro and in vivo (3–5).

Formation of a stable immune synapse and stable T cell–APC interactions require the polarization of the integrin LFA-1 (CD11aCD18). LFA-1-deficient T cells and those treated with LFA-1 blocking Abs fail to conjugate (1, 6). Additionally, CD18-deficient mice show defects in CD4⁺ T cell proliferation and Ab generation in response to Ag challenge (7). LFA-1 activity is regulated both by an upregulation of affinity for its ligand, ICAM-1, and by clustering at the immune synapse following TCR stimulation (reviewed in Ref. 8). The cytoskeletal regulatory protein talin is an important positive regulator of both LFA-1 affinity and clustering in T cells. Binding of the talin FERM (protein 4.1, ezrin, radixin, moesin) domain to the cytoplasmic tail of CD18 can increase integrin ligand affinity, and the talin rod domain can promote integrin clustering (9). Talin-deficient T cells fail to conjugate normally due to impaired LFA-1 clustering at the immune synapse (9, 10). Other identified positive regulators of LFA-1 activity in T cells include adhesion- and degranulation- promoting adaptor protein (11), Rap1 (12), Rap1 ligand (12, 13), Src-kinase associated protein of 55 kDa (14), and mammalian sterile twenty-like-1 (15) (reviewed in Ref. 16). However, little is known about negative regulators of integrin activity in T cells.

Previous work in nonhematopoietic cells has shown that an extended isoform of phosphatidylinositol phosphate kinase type I γ (PIPKIγ90) is an important modifier of talin–integrin interactions (17). PIPKIγ90 phosphorylates phosphatidylinositol 4-phosphate to generate phosphatidylinositol 4,5-biphosphate [PI(4,5)P2], which promotes talin–integrin interactions (18), actin dynamics, and endocytosis, and can be modified further to form the signaling intermediates phosphatidylinositol 3,4,5-triphosphate, inositol 1,4,5-triphosphate, and diacylglycerol (reviewed in Ref. 19). Several splice variants of PIPKIγ proteins, which all generate PI(4,5)P2 but differ in their subcellular localizations, have been identified, including 635-aa (87 kDa) and 661-aa (90 kDa) isoforms (19). The 90-kDa isoform of PIPKIγ90 differs from other isoforms of PIPKIγ by a C-terminal extension capable of binding talin at the same site that talin uses to bind the cytoplasmic tails of β integrins (17, 20). Given the importance of the 90-kDa isoform in regulating integrin activity during focal adhesion formation in fibroblasts (21), (22), we were interested in determining if PIPKIγ90 modulates LFA-1 activity in T cells.

To our knowledge, no previous studies have investigated the role of PIPKIγ90 in the context of T cell adhesion and activation. Recent studies investigating the function of PIPKIγ in NK cells (23), mast cells (24), and T cells (25) have depleted both PIPKIγ isoforms. This has clouded investigations into the relative contributions of the 87- and 90-kDa isoforms. Previous work has shown that the 87-kDa isoform is the primary PIPKIγ isoform responsible for generating PI(4,5)P2 that mediates calcium signaling downstream of G protein-coupled receptors (26). Thus, recent findings showing that PIPKIγ depletion in T cells decreases chemokine-mediated adhesion to ICAM-1 are likely due to the depletion of the 87-kDa isoform but not the 90-kDa isoform (25). Supporting this idea is a recent report showing an essential role for PIPKIγ isoforms in target cell lysis by NK cells; however, they also demonstrate that loss of PIPKIγ results in a significant decrease in inositol 1,4,5-triphosphate generation (23).

We previously reported that primary T cells express PIPKIγ90 and that it localizes to the T cell uropod (27). In this study, we investigate the role that PIPKIγ90 plays in CD4⁺ T cell LFA-1–mediated adhesion and proliferation. Using mice that are specifically deficient in the PIPKIγ 90-kDa isoform, we find that CD4⁺ T cell development is normal. However, cells from these mice have increased adhesion to ICAM-1 and increased T cell–APC conjugate formation. Immune synapses from PIPK90^−/− CD4⁺ T cells show increased LFA-1 polarization, suggesting that PIPKIγ90 is a negative regulator of T cell adhesion and LFA-1 clustering. These increased rates of adhesion correspond to increased rates of proliferation and Th1 cytokine production in PIPKIγ90^−/− cells compared with those of wild-type cells. Together, these findings suggest the PIPKIγ90 is a novel negative regulator of T cell activation, adhesion, and proliferation.

PIPKIγ90^−/− mice were generated by deleting the exon encoding the talin-binding sequence (exon 17) of PIPKIγ90 and backcrossed onto a C57BL/6 background for six generations (K. Legate and R. Fassler, manuscript in preparation). Knockout mice were crossed with mice expressing the OTII TCR transgene that recognizes OVA peptide 223–230. All of the experiments used wild-type and knockout siblings confirmed by DNA genotyping of tail clips with the primers 5′-TACTAACTGCTTCCCGCTGCTGC-3′ (forward) and 5′-TTCCTGGGTTTTCTGTGTCTTGTCG-3′ (reverse) at an annealing temperature of 60.2°C. OTII TCR transgenic and C57BL/6 mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All of the experimental protocols involving the use of mice were approved by the Institutional Animal Care and Use Committee at the University of Wisconsin.

For flow cytometry, PE anti-CD4, allophycocyanin anti-CD4, PE anti-CD3, FITC anti-CD8, PE anti-CD62L, FITC anti-CD44, allophycocyanin anti-CD69, PE anti-β₂, PE anti-β₁, FITC anti-α₄, PE anti-α_L, PE IFN-γ, FITC IL-2, PE IL-10, FITC IL-4, allophycocyanin B220, FITC CD11b, and PE CD11c were all from eBioscience (San Diego, CA).

For immunoblotting, talin (8d4) and actin were from Sigma-Aldrich (St. Louis, MO). A PIPKIγ-specific Ab was obtained by immunizing rabbits against a keyhold limpet hemocyanin-conjugated peptide corresponding to amino acids 637–656 of the mouse PIPKIγ90 isoform (TDERSWVYSPLHYSARPASD).

Single-cell suspensions of primary mouse T cells were made from lymph nodes and spleen from mice that were between 12 and 20 wk of age. After RBC lysis, mixed lymphocyte populations were resuspended in complete RPMI 1640 supplemented with 25 U/ml IL-2 (Chiron, Emeryville, CA) and stimulated with OVA_223–230 (Fremont, CA). OVA peptide-expanded cells were used on days 7–10 after isolation for Ag-dependent in vitro assays. Alternatively, CD4⁺ T cells were isolated from the cell suspension by negative selection and autoMACS sorting (Miltenyi Biotec, Auburn, CA). Isolated CD4⁺ cells were stimulated 1:1 with anti-CD3/CD28–coated beads according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA) and maintained in RPMI 1640 supplemented with IL-2 (Chiron). Anti-CD3/CD28 bead-activated cells were used on days 7–10 after isolation for in vitro assays.

LB27.4 B cells were purchased from American Type Culture Collection (Manassas, VA) and maintained in RPMI 1640 complete medium. Mouse D10 T cells and CH12 cells were maintained in culture as previously described (28). D10 cells were retrovirally transduced with PMX-GFP or PMX-GFP-PIPKIγ90 as described (27) and sorted for GFP expression using FACS.

RNA was isolated from anti-CD3/CD28 bead-activated CD4⁺ T cells using a RNeasy Protect Kit (Qiagen, Valencia, CA). PIPKIγ90 transcript was amplified using a OneStep RT-PCR Kit with the primers 5′-GTGCACAACATCGATCAGCAGGA-3′ (forward) and 5′-CTATAGTGAAGCGGGGAGTACAC-3′ (reverse). PIPKIγ87 used the same forward primer as for PIPKIγ90 and the reverse primer 5′-GCTGCTCCGATGTATCTGAAGG-3′. GAPDH was amplified with the primers 5′-GAGTCAACGGATTTGGTCGTAT-3′ (forward) and 5′-AGTCTTCTGGGTGGCAGTGAT-3′ (reverse).

For immunoblotting, anti-CD3/CD28 bead-activated CD4⁺ T cells were lysed in 50 mM Tris (pH 7.6), 0.15 M NaCl, 0.1% SDS, 0.5% deoxycholate, and 1% Nonidet P-40 containing 0.2 mM PMSF, 1 μg/ml pepstatin, 1 μg/ml apoprotinin, 1 μg/ml leupeptin, and 1 mM sodium orthovanadate on ice and cleared by centrifugation. Protein concentration was determined by bicinchoninic acid protein assay kit (ThermoScientific, Waltham, MA), and equal concentrations of protein were added to SDS sample buffer, boiled, and run on a 6–20% acrylamide gradient gel. Proteins were transferred to a nitrocellulose membrane and stained. Blots were imaged with an Odyssey infrared imaging system (Licor Biosciences, Lincoln, NE).

Single-cell suspensions were made from blood, two inguinal and cervical lymph nodes, three Peyer’s patch lymph nodes, and/or spleen from wild-type and PIPKIγ90^−/− mice. Cells were counted using trypan blue exclusion, stained with Abs as described, and analyzed by flow cytometry to determine total cell and subset numbers.

Adhesion assays were performed as previously described with minor modifications (9). Briefly, 96-well plates were coated with 3 μg/ml rmICAM-1 (R&D Systems, Minneapolis, MN) and blocked with 1% BSA. Anti-CD3/CD28 bead-activated CD4⁺ T cells were stained with 0.5 μg/ml calcein AM (Invitrogen) and left untreated or stimulated with 0.5 μg/ml biotinylated anti-CD3 (eBioscience) and streptavidin (The Jackson Laboratory), 20 ng/ml PMA (Sigma-Aldrich), or 10 μM MnCl₂ (Sigma-Aldrich). Cells were allowed to adhere to plates for 25 min at 37°C, and prewash fluorescence emission was measured on a VictorV3 plate reader (PerkinElmer, Waltham, MA). The plates were washed by pipetting, and postwash fluorescence emission was measured. Percentage adhesion was determined by (fluorescence_initial − fluorescence_final)/fluorescence_initial.

LB27.4 B cells were stained with 2.5 μM PKH-26 (Sigma-Aldrich) in 5% dextrose according to the manufacturer’s instructions and left untreated or loaded with 2.5 μg/ml OVA peptide for 30 min at 37°C. OVA peptide-expanded CD4⁺ T cells were stained with 0.5 μg/ml calcein AM (Invitrogen). Both cell types were resuspended in HBSS (Mediatech, Manassas, VA) supplemented with 2 mg/ml BSA (Sigma-Aldrich) and 1 mM HEPES (Mediatech). Equal numbers were combined on ice and centrifuged at 0.6 relative centrifugal force for 5 min. D10 cells overexpressing GFP or PIPKIγ90 were coincubated with CH12 cells cultured for 18 h with or without 250 μg/ml conalbumin (Sigma-Aldrich) and stained as above. Pellets were incubated at 37°C for indicated times prior to vortexing to dissociate nonspecific conjugates. Conjugates were analyzed using a FACSCalibur (BD Biosciences, San Jose, CA), and the percentage conjugate formation was determined by the percentage of double-positive cells divided by the sum of the percentage of double-positive and single-positive cells.

LB27.4 B cells were stained with 1 μM 7-amino-4-chloromethylcoumarin (Invitrogen) according to the manufacturer’s directions and pulsed with 2.5 μg/ml OVA peptide for 30 min at 37°C. CH12 cells were loaded for 18 h with 250 μg/ml conalbumin and stained with 1 μM 7-amino-4-chloromethylcoumarin. Equal numbers of T cells and B cells in RPMI 1640 were combined, centrifuged, and incubated at 37°C for 30 min prior to resuspension in PBS and pulse vortexing. Cells were allowed to adhere to poly-l-lysine–coated (Sigma-Aldrich) coverslips for 5 min prior to fixing with 3% paraformaldehyde (Electron Microscopy Services, Hatfield, PA) for 15 min. Cells were permeabilized with 0.2% Triton X-100 and blocked in goat serum. Cells were stained with anti-protein kinase C-θ (PKC-θ; Santa Cruz Biotechnology, Santa Cruz, CA), anti-LFA-1 (M17/4) (eBioscience), and rhodamine phalloidin (Invitrogen) along with FITC/tetramethylrhodamine isothiocyanate-conjugated anti-rat or FITC-conjugated anti-rabbit secondary Abs (The Jackson Laboratory). Images were acquired on a laser scanning confocal microscope (Olympus, Center Valley, PA) using a 60× Plan Apo/1.45 oil immersion objective with a 1× or 10× zoom factor and captured into Fluoview software (FV10-ASW version 01.07; Olympus).

Calcium flux was determined essentially as described previously in Ref. 29. Briefly, CD4⁺ T cells were stained with 10 μM Indo-1 in PBS for 30 min and washed in complete RPMI 1640. Cells were coated with biotinylated anti-CD3 (2C11) (eBioscience) for 10 min on ice, and a baseline reading of fluorescence at 495 and 450 nm was acquired on the LSRII flow cytometer (BD Biosciences) for 45 s. Streptavidin (The Jackson Laboratory) was added, and fluorescence was measured for an additional 5 min. The ratio of fluorescence at 405 and 495 nm was analyzed by FlowJo (Tree Star, Ashland, OR).

Live imaging of T cell–APC interactions was done using methods adapted from previous studies (30). CH12 APCs were loaded overnight with 250 μg/ml conalbumin. On the following day, CH12 cells were labeled with 2.5 μM PKH-26 and allowed to adhere to the bottom of a poly-l-lysine–coated glass-bottom plate. D10 T cells resuspended in HBSS (Mediatech), supplemented with 1 mM HEPES (Mediatech), 10% FBS (HyClone, Waltham, MA), and 0.25% low melt agarose (Fisher, Pittsburgh, PA), were plated on APCs and overlaid with 1 ml mineral oil (Sigma-Aldrich). Cells were maintained at 37°C for the duration of acquisition. Confocal images of GFP localization were obtained using a laser scanning confocal microscope (Olympus) using a 60× Plan Apo/1.45 oil immersion objective with one image every 2 min. Long-term conjugation was measured using an epifluorescent microscope (Nikon, Melville, NY) and a Coolsnap ES2 camera (Photometrics, Tuscon, AZ). One bright-field image was acquired every 2 min, and a fluorescent image was acquired every 8 min for 5 h. Images were acquired using MetaMorph imaging software (MDS Analytical Technologies, Downingtown, PA). Duration of conjugation was calculated as the time from initial T cell–APC contact to reformation of a T cell uropod.

OVA peptide-expanded CD4⁺ T cells were stained with 0.25 μM CFSE (Invitrogen) according to the manufacturer’s directions. Cells were left unstimulated or stimulated with one anti-CD3/CD28 coated bead (Invitrogen) per cell, 5 ng/ml PMA and 0.5 μg/ml ionomycin (Sigma-Aldrich). Additionally, CD4⁺ T cells were stimulated with irradiated splenocytes (3000 Gy) loaded with 0, 0.1, or 1 μg/ml OVA peptide. Seventy-two hours after activation, cells were stained with anti-CD4, and CFSE dye dilution in CD4⁺ T cells was analyzed using a FACSCalibur (BD Biosciences). Percentage of cells per division and proliferative index were determined using the ModFit 3.2.1 (Verity, Topsham, ME) analysis program.

In vivo proliferation was performed essentially as described in Ref. 7. Briefly, CD4⁺ cells from wild-type and knockout mice were isolated by CD4⁺ negative selection and stained with 2.5 μM CFSE (Invitrogen). Five million cells were injected i.v. into age- and sex-matched recipient mice. Eighteen hours later, 25 μg LPS (Sigma-Aldrich) or 25 μg LPS and 50 μg OVA (Sigma-Aldrich) was injected i.p. Seventy-two hours later, mice were sacrificed, and splenocytes were isolated and stained for CD4. The degree of CFSE dye dilution was determined for CD4⁺ T cells on a FACSCalibur (BD Biosciences).

A total of 1.5 × 10⁶ OVA peptide-expanded T cells were restimulated on days 7–10 after isolation on 24-well plates coated with 1 μg/ml anti-CD3 (2C11) (Biolegend, San Diego, CA) along with 2 μg/ml soluble anti-CD28 (eBioscience) in the presence of brefeldin A (eBioscience). Four hours after restimulation, cells were stained with anti-CD4 and fixed with 4% paraformaldehyde (Electron Microscopy Services). Intracellular staining was performed essentially as described in Ref. 31. Briefly, cells were permeabilized in PBS supplemented with 0.1% saponin (Alfa Aesar, Ward Hill, MA), 2% FBS (HyClone), and 0.01% sodium azide (Sigma-Aldrich), blocked with CD32/CD16 (eBioscience), and stained with PE IFN-γ, FITC IL-2, PE IL-10, and FITC IL-4 (eBioscience). The percentage of CD4⁺ T cells producing cytokines was determined by flow cytometry using a FACSCalibur (BD Biosciences) and analyzed using FlowJo (Tree Star).

Statistical analyses were performed using Prism 4 software (GraphPad, La Jolla, CA). Two-tailed paired t test or one-way ANOVA with Tukey posttest was used with p < 0.05 considered statistically significant.

We previously identified PIPKIγ90 as a novel component of the T cell uropod (27). To determine if PIPKIγ90 regulates T cell function, we characterized CD4⁺ T cells from PIPKIγ90-deficient mice. PIPKIγ90^−/− mice were generated by deleting the PIPKIγ exon encoding the talin-binding sequence (exon 17) (K. Legate and R. Fassler, manuscript in preparation). These mice were crossed onto the OTII TCR transgenic line, which recognizes OVA_223–230. We found that PIPKIγ90^−/− mice and wild-type mice had similar numbers of T cells, B cells, dendritic cells, and myeloid cells in the blood, spleen, and Peyer’s patches (Supplemental Table I). Although there was a trend toward fewer cells in the peripheral lymph nodes of knockout mice compared with those of control mice, this difference was due to fewer B cells and dendritic cells, not T cells, suggesting that T cell development and distribution remain intact in PIPKIγ90^−/− mice (Supplemental Table I).

CD4⁺ T cells from PIPKIγ90^−/− mice, as compared with T cells from wild-type littermate control mice, specifically lacked expression of the 90-kDa isoform of PIPKIγ by both RT-PCR and immunoblotting using isoform-specific Abs (Fig. 1). T cells from PIPKIγ90-deficient mice had similar surface expression of CD3, CD4, and CD8 on cells from the thymus, spleen, and peripheral lymph nodes (Fig. 2A). Further analysis of CD4⁺ subsets also showed similar expression of CD62L and CD44 (Fig. 2B), indicating that T cells from PIPKIγ90-deficient mice develop normally. We observed an increase in CD69 expression, but not CD25 expression, in CD4⁺ T cells from PIPKIγ90-deficient mice compared with those from wild-type mice, suggesting an increase in T cell activation in both lymph node and splenic populations (Fig. 2C, 2D). Importantly, although loss of the 87-kDa isoform has been reported to result in calcium signaling defects in fibroblasts (26), T cells lacking the 90-kDa isoform demonstrated no defects in calcium signaling in response to TCR cross-linking (Fig. 2E). Taken together, the findings suggest that PIPKIγ90 is not required for T cell development in vivo.

Previous studies have shown that PIPKIγ90 can bind to the talin FERM (protein 4.1, ezrin, radixin, moesin) domain at a position overlapping with talin’s primary integrin binding site and can impair integrin activation (17, 20). In addition, it was reported recently that T cells deficient in both 87- and 90-kDa isoforms of PIPKIγ have impaired chemokine-mediated adhesion to ICAM-1 (25). To determine how the PIPKIγ90 isoform specifically regulates LFA-1–mediated adhesion, we assayed adhesion of PIPKIγ90-deficient T cells on plate-bound ICAM-1. Although there were no differences in basal levels of adhesion to ICAM-1, we observed a significant increase in TCR-induced adhesion of PIPKIγ90^−/− CD4⁺ T cells to ICAM-1 compared with that of control cells (Fig. 3A), suggesting that PIPKIγ90 is a negative regulator of T cell LFA-1–mediated adhesion induced by TCR stimulation. Furthermore, we found that the defect in adhesion was not present when T cells were stimulated with either MnCl₂ or PMA, suggesting that PIPKIγ90 specifically regulates LFA-1 adhesion in response to TCR ligation. To determine if Ag-induced T cell–APC conjugation also was altered in PIPKIγ90-deficient T cells, we examined conjugate formation to OVA_223–230-loaded LB27.4 B cells. Using a flow cytometry-based conjugation assay, we found that depletion of PIPKIγ90 resulted in a 20% increase in T cell–APC conjugate formation after 2, 5, and 10 min of contact (Fig. 3B). Taken together, these findings suggest that PIPKIγ90 negatively regulates both TCR-induced ICAM-1 adhesion and T cell–APC interactions.

Previous work has shown that depletion of talin results in impaired LFA-1 polarization, but not actin polarization, to the immune synapse, suggesting that talin is required for LFA-1 clustering at the immune synapse (9). To determine if LFA-1 polarization is altered in PIPKIγ90-deficient T cells, we examined polarization of PKC-θ, LFA-1, and actin in T cell–APC conjugates in CD4⁺ T cells from wild-type and PIPKIγ90^−/− mice. We found that PKC-θ, LFA-1, and actin polarize to the immune synapse in both wild-type and knockout cells (Fig. 4A, 4B). Interestingly, we observed an increase in LFA-1 polarization at the T cell–APC contact site in PIPKIγ90-deficient T cells as compared with that in control cells (Fig. 4C). Using line scans of fluorescent intensity through the immune synapse of fixed conjugates, we found no difference in polarized actin fluorescence intensity between wild-type and PIPKIγ90^−/− conjugates (Fig. 4C, 4E). However, there was a 25% increase in peak LFA-1 fluorescence intensity at the immune synapse in PIPKIγ90^−/− conjugates compared with that in control conjugates (Fig. 4C, 4D). This increase in LFA-1 fluorescence intensity was not caused by an increase in LFA-1 surface expression, because flow cytometry analysis of surface integrins indicated no difference in integrin expression between PIPKIγ90^−/− and wild-type CD4⁺ T cells (Fig. 4F). Together, these data suggest that PIPKIγ90 negatively regulates LFA-1 clustering at the immune synapse and its absence potentially contributes to the observed increase in T cell–APC conjugate formation.

To characterize further how PIPKIγ90 expression affects T cell–APC conjugate formation, we ectopically expressed wild-type GFP-PIPKIγ90 in D10 T cells and examined PIPKIγ90 localization and conjugation with CH12 APCs loaded with conalbumin. The advantage of this system is that it is amenable to live imaging of GFP-PIPKIγ90 dynamics in T cells contacting APCs. In accordance with our findings indicating that PIPKIγ90 negatively regulates T cell–APC interactions, we found that ectopic expression of PIPKIγ90 impaired T cell–APC conjugate formation at early time points by flow cytometry (Fig. 5C).

Previous studies have demonstrated that talin concentrates at the immune synapse (1, 9). To determine if PIPKIγ90 also localizes to the immune synapse, we examined localization of GFP-PIPKIγ90 in T cell–APC conjugates by immunofluorescence. Although LFA-1 localized to the T cell–APC contact site, GFP-PIPKIγ90 showed a diffuse localization at the membrane without specific concentration at the immune synapse (Fig. 5A, 5B). In contrast, we found that GFP-PIPKIγ87 concentrated at the T cell–APC contact site and colocalized with LFA-1 (Fig. 5A, 5B). These findings suggest that PIPKIγ90 and PIPKIγ87 have distinct intracellular distributions in T cell–APC conjugates.

To characterize further the dynamics of GFP-PIPKIγ90 in T cell–APC conjugates, live time-lapse fluorescent imaging was performed. Using live cell imaging, we found that GFP-PIPKIγ90 localized to the uropod of motile T cells not in contact with APCs (27). Interestingly, upon contact with Ag-loaded CH12 cells, PIPKIγ90’s uropod localization was lost, and GFP-PIPKIγ90 became localized diffusely at the cell membrane (Fig. 5A, Supplemental Video 1). Quantification of T cell–APC contact times showed a reduction in conjugation duration from ∼4 h in control cells expressing GFP alone to <1.5 h in cells expressing GFP-PIPKIγ90 (Fig. 5B), providing further evidence that PIPKIγ90 negatively regulates T cell–APC interactions. Further analysis of the live cell imaging revealed that PIPKIγ90 reestablishes uropod localization prior to T cell migration away from the APC, suggesting that PIPKIγ90 may play a role in release of T cell–APC interactions (Fig. 5A, 5C).

Previous studies have reported that increased duration of T cell–APC interactions can result in increased rates of proliferation both in vitro and in vivo (3–5). To determine if altered conjugation duration is associated with increased proliferation of PIPKIγ90-deficient T cells, T cell proliferation was characterized both in vitro and in vivo. After CFSE labeling of CD4⁺ T cells from wild-type and PIPKIγ90^−/− mice, we stimulated cells with anti-CD3/CD28–coated beads or PMA and ionomycin (Fig. 6A). Each cell division corresponds to a 50% decrease in CFSE intensity, and the percentage of CD4⁺ T cells that had undergone anywhere from zero to six cell divisions was calculated. We found that PIPKIγ90^−/− CD4⁺ T cells proliferated more than wild-type cells in response to both anti-CD3/CD28–coated beads and PMA/ionomycin stimulation. Indeed, the proliferative index of PIPKIγ90^−/− cells was 50% greater than that for wild-type cells. To determine if altered T cell proliferation also was observed with a specific Ag, CD4⁺ T cells from OTII⁺ wild-type and PIPKIγ90^−/− mice were stimulated with irradiated splenocytes and varying concentrations of OVA peptide (0, 0.1, and 1 μg/ml). Similar to our bead and PMA/ionomycin stimulation, we found that CD4⁺ T cells from PIPKIγ90^−/− mice proliferate more than wild-type cells (Supplemental Fig. 1).

Finally, to determine if proliferation was also altered in vivo, we injected mice with CFSE-labeled CD4⁺ wild-type and PIPKIγ90^−/− T cells. Eighteen hours later, mice were injected i.p. with LPS alone or LPS and OVA. Spleens were isolated 72 h after injection, and CD4⁺ T cells were analyzed for CFSE dilution by flow cytometry. Consistent with our in vitro studies, we observed an approximately 2-fold increase in proliferation of PIPKIγ90^−/− cells compared with that of wild-type cells in vivo (Fig. 7). Taken together, these findings indicate that PIPKIγ90 negatively regulates T cell activation both in vitro and in vivo.

Because we observed increased CD4⁺ T cell proliferation in PIPKIγ90-deficient mice, we wanted to determine if PIPKIγ90 also negatively regulated cytokine production in vitro. To investigate this, control and PIPKIγ90-deficient CD4⁺ T cells were activated in vitro with OVA peptide. Seven days after activation, cells were restimulated for 4 h with anti-CD3 and anti-CD28 in the presence of brefeldin A, and cytokine production was assessed by intracellular staining and flow cytometry. Although unstimulated CD4⁺ cells did not have a measureable cytokine response (data not shown), we found that CD4⁺ T cells from both wild-type and knockout mice produced Th1 cytokines (Fig. 8A, 8B). Interestingly, there were nearly twice as many IFN-γ- and 50% more IL-2–producing PIPKIγ90^−/− CD4+ cells compared with control cells, suggesting that PIPKIγ90 negatively regulates Th1 cytokine production.

Appropriate activation of the integrin LFA-1 is critical for establishing and maintaining T cell–APC interactions and generating T cell immune responses. Previous studies have demonstrated that altered LFA-1–mediated adhesion is associated with impaired immune responses and has significant pathological consequences (7, 32). Several positive regulators of LFA-1 activation downstream of TCR signaling have been identified, including Rap1, Rap1 ligand, Src-kinase associated protein of 55 kDa, mammalian sterile twenty-like-1, and talin. Depletion of these signaling components results in defects in LFA-1 polarization toward the immune synapse (reviewed in Ref. 16). Although activation of LFA-1 is required for generation of immune responses, its deactivation is also necessary (33). Factors involved in the negative regulation of LFA-1–mediated adhesion in T cells are largely unknown. Most negative regulators of T cell activation, including the surface receptors CTLA-4, programmed death-1, and CD5, influence proximal TCR signaling pathways and interfere with TCR signal transduction (reviewed in Ref. 34). This is the first study to identify PIPKIγ90 as a novel negative regulator of T cell activation that affects LFA-1 polarization and adhesion induced by TCR ligation.

Previous studies have demonstrated that the PIPKIγ87 isoform is necessary for LFA-1–mediated adhesion induced by chemokines (25). In contrast, we show that PIPKIγ90 impairs TCR-induced LFA-1 adhesion, suggesting that the two isoforms may have opposing functions in regulating T cell adhesion. We show that depletion of PIPKIγ90 enhances T cell LFA-1 adhesion to ICAM-1 and T cell–APC interactions, whereas overexpression of PIPKIγ90 impairs the duration of T cell–APC interactions. The opposing roles of the two PIPKIγ isoforms is especially intriguing, because the two isoforms differ by only 26 aa in the C terminus. It is possible that these opposing functions are caused by the distinct intracellular distributions of the two isoforms. PIPKIγ90 localizes to the T cell uropod and is not polarized at the immune synapse. In contrast, PIPKIγ87 is not a uropod component in motile T cells (27) but colocalizes with LFA-1 at the immune synapse (Fig. 5).

There have been several interacting proteins identified that specifically bind to the C-terminal 26 aa of PIPKIγ90, including talin. It is possible that the association of PIPKIγ90 with talin contributes to its negative regulation of LFA-1 adhesion and polarization to the synapse. An attractive hypothesis is that PIPKIγ90 binds talin and sequesters it away from the synapse, thereby negatively regulating LFA-1 clustering at the immune synapse. Unfortunately, poor Ab staining and the large size of talin prohibited us from testing this directly. However, we observed increased LFA-1 polarization at the T cell–APC contact site in PIPKIγ90-deficient T cells. Alternatively, the C-terminal tail of PIPKIγ90 also interacts with other binding partners that may affect T cell–APC interactions and LFA-1 function. For example, previous work has supported a role for PIPKIγ90 as a key regulator of clathrin-mediated endocytosis through both its generation of PI(4,5)P2 and its association with endocytic protein adaptor protein 2 (35–38). Although it is possible that PIPKIγ90 influences T cell conjugation and proliferation by affecting endocytosis and internalization of the TCR or LFA-1, the finding that hyperproliferation is observed with PMA and ionomycin stimulation argues against the possibility that PI(4,5)P2 generation by PIPKIγ90 is solely responsible for the adhesion and proliferation defects observed with PIPKIγ90-deficient T cells.

Live imaging of GFP-PIPKIγ90 dynamics in motile T cells supports the hypothesis that PIPKIγ90 localization to the uropod may sequester key components that mediate T cell–APC contact. We found that PIPKIγ90 was predominantly at the uropod in motile T cells and that with T cell–APC contact became diffusely localized to the membrane. Prior to cessation of T cell–APC contact, we noted an early relocalization of PIPKIγ90 back to the uropod and the subsequent migration of T cells away from the APC. These findings support the intriguing possibility that PIPKIγ90 may negatively regulate T cell–APC contacts and T cell proliferation by sequestering interacting proteins, such as talin, away from the synapse, thereby promoting release of T cell–APC contact. Interestingly, other T cell uropod components, such as CD43, when depleted, also are associated with increased T cell adhesion and proliferation, suggesting that the uropod may in part function to negatively regulate T cell–APC contact and activation, although the mechanisms by which this occurs may be distinct (39–41). Taken together, these findings suggest that uropod components may potentially contribute to release of T cell–APC interactions and serve as negative regulators of T cell activation.

Maintenance of T cell–APC interactions is critical for subsequent T cell proliferation with increased duration of conjugation directly correlated with increased proliferation (4, 5). Previous reports have shown that continuous calcium flux and phosphatidylinositol 3,4,5-triphosphate signaling are necessary for the maintenance of T cell–APC interactions and T cell proliferation (4, 42). In this study, we found that, in contrast to reports of PIPKIγ87 (26), PIPKIγ90 has no effect on TCR-induced calcium flux. We also found that PIPKIγ90 negatively regulates T cell proliferation, because T cell proliferation was increased significantly in PIPKIγ90-deficient T cells both in vitro and in vivo. Although we do not know how PIPKIγ87 affects T cell proliferation, its reported role in calcium flux suggests that it may be required for proper T cell activation by Ag.

In summary, our findings indicate that PIPKIγ90 is working downstream of proximal TCR signaling to negatively regulate T cell LFA-1 adhesion and proliferation. Our findings raise the possibility that PIPKIγ90 may negatively regulate T cell activation by affecting T cell uropod formation and the localization or activation of interacting partners, including talin. A challenge for future investigations will be to identify the key binding partners that mediate PIPKIγ90 localization at the uropod and play a role in its inhibitory function during T cell activation. It will also be interesting to determine if the expression of PIPKIγ90 alters susceptibility to the development of autoimmunity in animal models. It is intriguing to speculate that PIPKIγ90 may be an important downstream molecule responsible for downregulating T cell activation in response to self-Ags. Although there is no evidence of systemic autoimmunity in PIPKIγ90^−/− mice to date, several known negative regulators of TCR signaling also show no signs of systemic autoimmunity (reviewed in Ref. 34). Future investigations will be focused on challenging PIPKIγ90^−/− mice to determine how PIPKIγ90 expression affects susceptibility to autoimmunity in animals, including models of diabetes and experimental autoimmune encephalomyelitis.

We thank Lisa Fox, Subbe Hegde, and the University of Wisconsin Paul P. Carbone Comprehensive Cancer Center Flow Cytometry Facility for technical assistance.

Disclosures The authors have no financial conflicts of interest.