Abl family kinases modulate T cell-mediated inflammation and chemokine-induced migration through the adaptor HEF1 and the GTPase Rap1

Abstract

Chemokine signaling is critical for T cell function during homeostasis and inflammation and directs T cell polarity and migration through the activation of specific intracellular pathways. Here, we uncovered a previously uncharacterized role for the Abl family tyrosine kinases Abl and Arg in the regulation of T cell-dependent inflammatory responses and showed that the Abl family kinases were required for chemokine-induced T cell polarization and migration. Our data demonstrated that Abl and Arg were activated downstream of chemokine receptors and mediated the chemokine-induced tyrosine phosphorylation of human enhancer of filamentation 1 (HEF1), an adaptor protein that is required for the activity of the guanosine triphosphatase Rap1, which mediates cell adhesion and migration. Phosphorylation of HEF1 by Abl family kinases and activation of Rap1 were required for chemokine-induced T cell migration. Mouse T cells that lacked Abl and Arg exhibited defective homing to lymph nodes and impaired migration to sites of inflammation. These findings suggest that Abl family kinases are potential therapeutic targets for the treatment of T cell-dependent immune disorders that are characterized by chemokine-mediated inflammation.

Fig. 1

Abl family kinases are required for chemokine-induced T cell migration and are activated by chemokines. (A) Primary wild-type (WT) and Abl/Arg null T cells were tested for their ability to migrate to either SDF-1α or CCL21 in transwell chambers. Data are means ± SEM (n = 5 experiments). (B) WT, Abl-deficient, Arg-deficient, and Abl and Arg doubly deficient (Abl/Arg null) T cells were tested for migration to CCL21. Data are means ± SEM (WT, n=4; Abl^−/−, n=3; Arg^−/−, n=4; Abl/Arg^−/−, n=3 experiments). (C) Mouse primary T cells were stimulated with SDF-1α or CCL21 for 30 s, and activation of Abl family kinases was detected by Western blotting analysis of the phosphorylation of CrkL (pCrkL, Y207). Blots were stripped and then analyzed for total CrkL protein. (D) WT or Abl/Arg null T cells were stimulated with SDF-1α and analyzed by Western blotting for the presence of pCrkL as described for (C). Total cell lysates were analyzed by Western blotting for Abl and Arg. Quantification of the amount of pCrkL normalized to that of total CrkL is shown in the bar graph below (n = 4 experiments). (E) Human H9 T cells that were untreated or were pretreated with STI571 were plated onto transwell inserts precoated with fibronectin. Migration of the cells to increasing amounts of SDF-1α is shown. Data are means ± SD of triplicate values. (F) H9 T cells that were untreated or were pretreated with STI571 were stimulated with SDF-1α at the indicated concentrations for 30 s, and lysates were analyzed by Western blotting for pCrkL as described for (C). Blots were then analyzed for total CrkL protein. Data are representative of three (C and E) or two (F) independent experiments. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 2

Abl/Arg null T cells exhibit decreased migration velocity and impaired chemokinesis. Primary T cells isolated from either WT or Abl/Arg null mice were differentially fluorescently labeled, mixed, and plated onto microwell dishes that were precoated with BSA as a control or with ICAM-1 and CCL21. Cell movement was tracked by time-lapse video microscopy. (A) Cell tracking of WT (top) and Ab/Arg null (bottom) T cells on ICAM-1- and CCL21-coated surfaces. (B) Migration velocities of WT T cells plated on BSA, or of WT and Abl/Arg null T cells plated on ICAM-1- and CCL21-coated wells were calculated with Metamorph software (n = 30 cells). Data are representative of three independent experiments. (C) T cells isolated from either WT or Abl/Arg null mice were tested in a transwell migration assay with the indicated amounts of CCL21 present in both the upper and bottom chambers. Data are representative of four independent experiments and are presented as the mean ± SD of triplicate values.

Fig. 3

Abl family kinases are required for chemokine-induced T cell polarization and F-actin polymerization. (A) H9 T cells that were untreated or were pretreated with STI571 were stimulated with SDF-1α. Polarization was analyzed by staining the uropod with antibody against ICAM-3 for human T cells. Representative immunofluorescence images are shown. (B) Quantification of polarization for control or STI571-treated H9 T cells in the presence or absence of SDF-1α. Data are presented as mean ± SEM (n=3 experiments). (C) WT mouse primary T cells that were untreated or were pretreated with STI571 and Abl/Arg null T cells were stimulated with CCL21 at the indicated concentrations. Polarization was analyzed by staining the uropod with antibody against CD44 for mouse T cells. Data are presented as mean ± SD. **P < 0.01, ***P < 0.001. (D) Molt-4 T cells that were untreated or were pretreated with STI571 were stimulated with SDF-1α for the indicated times. Cells stained with phalloidin (to detect F-actin) were analyzed by flow cytometry. The y-axis indicates mean fluorescence intensity (MFI). (E) WT or Abl/Arg null T cells were stimulated with CCL21, and actin polymerization was assayed as described for (D). Data are means ± SD of triplicate values. (F) WT primary T cells that were untreated or were pretreated with STI571 and Abl/Arg null T cells were stimulated with CCL21, and GTP-bound Rac1 was detected by pull-down assay with GST-PBD and analyzed by Western blotting with antibody against Rac1. Total cellular lysates were analyzed by Western blotting for total Rac1 and Abl proteins. Quantification of the amount of GTP-Rac1 normalized to that of total Rac1 l is shown in the bar graph below. Data are representative of three (C and E) or two (D and F) independent experiments.

Fig. 4

Abl-mediated activation of Rap1 in response to chemokine is required for T cell polarization and migration. (A) H9 T cells that were untreated or were pretreated with STI571 were stimulated with SDF-1α. GTP-bound Rap1 was pulled down with GST-Ral-GDS and detected by Western blotting analysis with an antibody against Rap1. Data are representative of three independent experiments. (B) WT primary T cells that were untreated or were pretreated with STI571 and Abl/Arg null T cells were stimulated with SDF-1α, and GTP-Rap1 amounts were measured as described for (A). Quantification of GTP-Rap1 abundance normalized to that of total Rap1 is shown in the bar graph below (n = 3 experiments). (C) H9 T cells were transduced with lentiviruses encoding the pCSCGW-GFP vector (V) or constitutively active Rap1a-63E (63E). Active GTP-bound Rap1 was detected as described for (A), and total cell lysates were analyzed by Western blotting for the presence of Rap1 and β-tubulin. (D and E) H9 T cells transduced with empty plasmid or with plasmid encoding Rap1a-63E were left untreated or were pretreated with STI571 and then were analyzed for (D) polarization in the presence of SDF-1α and (E) migration in the presence or absence of SDF-1α. Data are means ± SD of triplicate values and are representative of two independent experiments. ***P < 0.001.

Fig. 5

Abl kinases are required for chemokine-induced tyrosine phosphorylation of HEF1. (A) Human Molt-4 T cells were transduced with lentiviruses encoding either the pLKO vector (V) or HEF1-specific shRNA (sh-HEF). Total cells lysates were analyzed by Western blotting for HEF1 and actin (right panel). Transduced cells were untreated or were stimulated with SDF-1α, and active GTP-Rap1 was detected as described for Fig. 4A. Total cells lysates were analyzed by Western blotting for Rap1 (left panel). Quantification of GTP-Rap1 abundance normalized to that of total Rap1 is shown in the bar graph below. (B) Human H9 T cells that were untreated or were pretreated with STI571 were stimulated with SDF-1α at the indicated concentrations for 30 s. HEF1 protein was immunoprecipitated (IP) and samples were analyzed by Western blotting with phospho-tyrosine-specific antibody (pTyr). Total cellular lysates were analyzed by Western blotting for HEF1. (C) H9 T cells were transduced with lentiviruses encoding either scrambled miRNA (Scr) or miRNAs specific for Abl and Arg. GFP-sorted cells were left untreated or were stimulated with SDF-1α, and tyrosine phosphorylation of HEF1 was detected as described for (B). GTP-bound Rap1 was analyzed as described for (A). Total cell lysates were analyzed by Western blotting for HEF1 and Rap1 (left panel). Depletion of Abl and Arg in the knockdown cells was verified by Western blotting analysis (right panel). Quantification of the abundance of tyrosine-phosphorylated HEF1 normalized to that of total HEF1 protein and of the amount of GTP-Rap1 normalized to that of total Rap1 is shown in bar graphs below (n = 3 experiments). (D) WT primary T cells that were untreated or were pretreated with STI 571 and Abl/Arg null T cells were stimulated with SDF-1α. Tyrosine phosphorylation of HEF1 was detected as described for (B). Total cellular lysates were analyzed by Western blotting for the indicated proteins. Quantification of the abundance of HEF1-pTyr normalized to that of total HEF1 protein is shown in the bar graph below (n = 4 experiments). Data in (A) to (C) are representative of at least two independent experiments.

Fig. 6

Abl-dependent tyrosine phosphorylation of HEF1 is required for chemokine-induced T cell migration. (A) HEK 293T cells were cotransfected with plasmids encoding HA-tagged HEF1 and the indicated Abl constructs. Phosphorylated HEF1 was detected by immunoprecipitation with an antibody against HA and Western blotting with antibody against pTyr. HEF1 protein was detected by incubating stripped blots with antibody against HA. Total cell lysates were analyzed by Western blotting for Abl and HEF1. (B) HEK 293T cells were transfected with plasmids encoding GFP-tagged WT HEF1 or the indicated (Y-F) HEF1 mutants, and cotransfected with plasmids encoding the inactive kinase mutants Abl-KM and Arg-KR or the constitutively active kinases Abl-PP and Arg-PP. Immunoprecipitation of GFP-HEF1 with antibody against GFP was followed by Western blotting analysis with antibody against pTyr. Total cellular lysates were analyzed by Western blotting for the indicated proteins. (C) Human Molt-4 T cells were transduced with lentiviruses encoding either the pGZ vector with scrambled shRNA (sh-Scr) or HEF1-specific shRNA (sh-HEF1). These cells were then transfected with plasmids encoding HEF1-WT or HEF1-7YF constructs. Total cellular lysates were analyzed by Western blotting for HEF1 and actin. (D) Migration of control and HEF1-depleted T cells in the presence or absence of SDF-1α. (E) T cells transduced with lentiviruses expressing the indicated shRNAs and then transfected with plasmids encoding HEF1-WT or HEF1-7YF proteins were analyzed for their ability to migrate in the presence of SDF-1α. Data are means ± SD of triplicate values. Data are representative of three (A and B) or two (C to E) independent experiments. ***P < 0.001.

Fig. 7

Abl/Arg null T cells exhibit defective homing to lymph nodes and impaired migration to sites of inflammation in mice. (A) WT or Abl/Arg null T cells were differentially fluorescently labeled and injected into recipient mice. Inguinal lymph nodes (Ing-LN), mesenteric lymph nodes (Mes-LN), spleen, and blood were collected after 2 hours and analyzed by flow cytometry. Input indicates the initial 1:1 ratio of mutant to WT T cells injected. Data are means ± SEM (n = 3 experiments). (B) In the adoptive transfer CHS model, extravasation of adoptively transferred CMTPX-labeled T cells (red) in each ear was visualized by confocal microscopy. Representative images are shown. (C) Infiltrated and labeled T cells were quantified with Metamorph software. Data are means ± SEM (WT: n=4 mice; Abl/Arg null: n=5 mice). The numbers of infiltrating cells were normalized to the total numbers of cells injected. (D) Draining lymph node cells were isolated from recipient mice 20 hours after challenge with hapten. Cells were stained with FITC-conjugated antibodies against CD4 or CD8. Double positive T cells for adoptively transferred CD4-CMTPX or CD8-CMTPX were plotted as percentages of the total numbers of CD4+ or CD8+ T cells. Data are means ± SEM (WT: n=4 mice; Abl/Arg null: n=5 mice). (E) WT and Abl/Arg null mice were immunized with hapten, and draining lymph node cells were isolated and stimulated in vitro with antibody against CD3. Culture media were assayed for the presence of IFN-γ by ELISA. Control WT mice were not exposed to hapten. Data are representative of two independent experiments and are presented as means ± SD of triplicates. *P < 0.05, ** P < 0.01, and *** P < 0.001. (F) Proposed model for Abl-mediated signaling during chemokine-induced polarization and migration.