CRK proteins selectively regulate T cell migration into inflamed tissues

Abstract

Effector T cell migration into inflamed sites greatly exacerbates tissue destruction and disease severity in inflammatory diseases, including graft-versus-host disease (GVHD). T cell migration into such sites depends heavily on regulated adhesion and migration, but the signaling pathways that coordinate these functions downstream of chemokine receptors are largely unknown. Using conditional knockout mice, we found that T cells lacking the adaptor proteins CRK and CRK-like (CRKL) exhibit reduced integrin-dependent adhesion, chemotaxis, and diapedesis. Moreover, these two closely related proteins exhibited substantial functional redundancy, as ectopic expression of either protein rescued defects in T cells lacking both CRK and CRKL. We determined that CRK proteins coordinate with the RAP guanine nucleotide exchange factor C3G and the adhesion docking molecule CASL to activate the integrin regulatory GTPase RAP1. CRK proteins were required for effector T cell trafficking into sites of inflammation, but not for migration to lymphoid organs. In a murine bone marrow transplantation model, the differential migration of CRK/CRKL-deficient T cells resulted in efficient graft-versus-leukemia responses with minimal GVHD. Together, the results from our studies show that CRK family proteins selectively regulate T cell adhesion and migration at effector sites and suggest that these proteins have potential as therapeutic targets for preventing GVHD.

Figure 8. CRK/CRKL Dko donor T cells can carry out GVL with minimal GVHD.

(A–C) T cell–depleted bone marrow cells alone (BM, H-2b), mixed with purified WT CD4⁺ T cells (BM + WT CD4 T cells, H-2b), or mixed with purified CRK/CRKL Dko CD4⁺ T cells (BM + Dko CD4 T cells, H-2b) were injected into lethally irradiated BALB/c hosts (H-2d). Host GVHD clinical scores (A), body weight change (B), and survival (C) are summarized. Combined data from 2 different experiments are shown, with a total of 5–10 mice for each experimental group. (D–G) Lethally irradiated BALB/c hosts were injected with T cell–depleted bone marrow cells alone (BM control), or together with luciferase-transduced A20 cells (BM + A20luc), purified WT CD8⁺ T cells (BM + A20luc + WT CD8 T), or CRK/CRKL Dko CD8⁺ T cells (BM + A20luc + Dko CD8 T). Host body weight change (D), survival (E), and tumor burden (reflected by the luciferase signal) (F and G) are summarized. E shows combined data from 2 different experiments, with a total of 8–10 mice per experimental group. D and F show data from 1 of 2 independent experiments, with 5 mice for each experimental group. (H) To assess T cell migration during GVHD, congenically marked CD8⁺ T cells from WT or CRK/CRKL Dko mice were mixed 1:1 with competitor SJL T cells (all H-2b) and injected together with T cell–depleted bone marrow cells into lethally irradiated BALB/c host mice (H-2d). Spleen and liver were harvested after 10–12 days, and the ratio of experimental (WT or Dko) to competitor adoptively transferred cells was determined. Paired data (combined from 2 experiments) are shown, with a total of 5 (WT) or 6 (Dko) recipient mice. Error bars represent mean ± SD. Statistical analysis in C, E, and H was performed using a 1-way ANOVA.

Figure 7. CRK/CRKL-deficient T cells have defects in migration to inflamed skin.

(A) Resting WT or Dko CD4⁺ T lymphoblasts were labeled with CFSE or CMTMR, mixed, and injected into recipient mice. One hour after injection, cells were collected from blood, peripheral lymph nodes (PLN), mesenteric lymph nodes (MLN), and spleen, and the ratio of Dko to WT T cells was determined. Pooled data from 3 experiments is shown (n = 9 total mice). (B) PLN lymphocytes from WT mice (CD45.2⁺) were labeled with CFSE, mixed with PLN lymphocytes from B6.SJL mice (CD45.1⁺), and injected i.v. into recipient mice. In parallel, PLN lymphocytes from CRK/CRKL Dko Rosa-YFP mice (CD45.2⁺) were mixed with PLN lymphocytes from B6.SJL mice (CD45.1⁺) and injected into recipient mice (CD45.2⁺). Cells were harvested from blood, PLN, MLN, and spleen 1 hour after transfer, and the ratio of naive (CD62L^hi, CD44^lo) experimental (CFSE⁺ WT or EYFP⁺ Dko) to competitor (CD45.1⁺) CD4⁺ T cells in the recovered populations was determined and normalized to the input ratio. Data represent individual mice from 1 experiment (n = 7 total mice). (C) CD4⁺ T cells were purified from WT and Dko mice, and Th1 cells were cultured in vitro. Percentages of WT and Dko Th1 cells were analyzed by intracellular flow. (D) Skin inflammation was induced in recipient mice using DNFB as described in Methods. WT and Dko Th1 cells cultured as in C were stained with CFSE or CMTMR, mixed 1:1, and injected intravenously into recipient mice. Twenty-four hours after injection, blood, draining lymph nodes (DLN), PLN, MLN, spleen, control skin, and inflamed skin were collected, and adoptively transferred cells were analyzed by flow cytometry. Combined data from 3 experiments (n = 10 total mice) are shown. **P < 0.01. Statistical analysis was performed using a paired 2-tailed Student’s t test.

Figure 6. CRK and CRKL interact with C3G and CASL to regulate T cell diapedesis.

(A) Schematic drawing showing a proposed CHAT-H–CASL–CRK/CRKL–C3G module that could regulate RAP1 activation in response to chemokine stimulation. CASL binds to the SH2 domains of CRK proteins, and mutation of the SH2 domain (CRKII R38K) interrupts this association. Both C3G and c-ABL bind to the nSH3 domains of CRK and CRKL, and mutation in the SH3 domain (CRKII R169K) interrupts these interactions. c-ABL phosphorylates CASL, CRKII at Y221, and CRKL at Y207. CHAT-H constitutively associates with CASL and recruits CASL to the plasma membrane. Chemokine stimulation induces CASL tyrosine phosphorylation by ABL family kinases and provides binding sites for the SH2 domains of CRK proteins. The association between CRK and CASL brings C3G to the membrane and activates RAP1. (B and C) Preactivated CD4⁺ T cells were stimulated with CXCL10 for the indicated times, lysed, and incubated with recombinant CRKL-SH3 protein to pull down C3G or immunoprecipitated with anti-CASL. The precipitants were blotted with the indicated antibodies. (D) CRK/CRKL Dko CD4⁺ T cells were reconstituted with the indicated CRKII mutants, and diapedesis was assessed. Data represent mean ± SD of replicate samples from 1 experiment, representative of 4 independent experiments. *P < 0.05, **P < 0.01. Statistical analysis was performed using a paired 2-tailed Student’s t test.

Figure 5. RAC1 activation is intact, but RAP1 and CDC42 activation is defective in CRK/CRKL-deficient T cells.

(A) Preactivated WT and CRK/CRKL Dko CD4⁺ T cells were stimulated with 10 nM CCL21 for the indicated times. Whole cell lysates (WCL) were subjected to pull-down with GST-RalGDS (GST-Ral PD) or GST-PakRBD (GST-PAK PD). Bound proteins were analyzed by SDS-PAGE and immunoblotted with the indicated antibodies. (B–D) Signals obtained for GTP-RAP1, GTP-CDC42, and GTP-RAC1 as shown in A were quantified using a fluorescence-based detection system. (E) WT and CRK/CRKL Dko CD4⁺ T cells were cultured under Th1-skewing conditions and stimulated with 10 nM CXCL10. Activation of RAP1 was assessed as in A and quantified in F. Representative data from at least 3 experiments are shown.

Figure 4. CRK/CRKL-deficient T cells show defects in diapedesis.

(A and B) 3B-11 endothelial cell monolayers were treated with TNF-α overnight, and preactivated WT or CRK/CRKL Dko CD4⁺ T cells were added on top. Images were collected every 30 seconds for 2 hours. Selected time-lapse images of WT T cells (left, arrows) and CRK/CRKL Dko T cells (right, arrows) acquired at the indicated time points are shown (A). T cells that transmigrated across the endothelial monolayer were scored, and average ± SD values from 4 different experiments are shown. A total of 80 cells per genotype were scored (B). (C) 3B-11 cells were grown as a monolayer on top of Transwell inserts and treated with TNF-α. CRK/CRKL Dko CD4⁺ T cells were reconstituted with the indicated CRK isoforms and applied to the Transwell inserts. T cells that underwent diapedesis were collected and analyzed by flow cytometry. (D) Cells used in C were blotted with the indicated antibodies. Data represent mean ± SD for triplicate samples from 1 experiment, representative of 4 independent experiments. (E) Preactivated CD4⁺ T cells prepared from mice that were singly deficient (Sko) for either CRK or CRKL were applied to 3B-11 monolayers on Transwell inserts, and T cell diapedesis was assessed as in C. Data represent mean ± SD from 3 independent experiments. *P < 0.05, **P < 0.01, ***P < 0.005. Statistical analysis was performed using unpaired (B) or paired (C and E) 2-tailed Student’s t tests.

Figure 3. CRK/CRKL-deficient T cells show defects in chemotaxis.

Preactivated WT and CRK/CRKL Dko CD4⁺ T cells were prepared, and migration in response to CCL21 (A) or CXCL10 (B) was tested by using a 3-μm-pore-size Transwell chamber. Average ± SD values from quadruplicate wells from 1 experiment, representative of 5 separate experiments, are shown. *P < 0.05, **P < 0.01. Statistical analysis was performed using paired 2-tailed Student’s t tests. (C) Flow cytometric analysis of chemokine receptor expression. Left: Preactivated WT and CRK/CRKL Dko CD4⁺ T cells were incubated with recombinant mouse CCL19-Fc, followed by labeling with anti–human Fcγ–biotin, then streptavidin-PE, to detect CCR7 surface expression. Right: Preactivated WT and CRK/CRKL Dko CD4⁺ T cells were stained with anti-CXCR3–PE. (D) Preactivated WT or CRK/CRKL Dko CD4⁺ T cells were placed in a 5-μm-pore collagen gel in the presence of a CCL21 gradient, and cell migration was imaged for 4 hours at 37°C. Tracks of individual cells are presented with the same point of origin. Data are representative of 3 experiments. Quantitative analysis of cell movements is shown in Table 1. Tracks in D are from experiment 3.

Figure 2. CRK/CRKL-deficient T cells show impaired integrin-dependent adhesion.

Ninety-six-well plates were coated with 1 μg/ml recombinant mouse ICAM-1 (A) or 3 μg/ml fibronectin (FN) (B). Preactivated WT and CRK/CRKL Dko CD4⁺ T cells were stained with Calcein-AM, applied to ICAM-1– or fibronectin-coated plates, and allowed to warm to 37°C for 10 minutes without stimulus (No) or in the presence of 10 nM CXCL12, 10 nM CCL21, 1 μg/ml anti-CD3, or 10 ng/ml PMA. Unbound cells were washed off, and adherent cells were quantified using a fluorescence microplate reader. Data represent averages ± SD of triplicate samples from 1 experiment, representative of 5 separate experiments. *P < 0.05, **P < 0.01. (C) Preactivated WT and CRK/CRKL Dko CD4⁺ T cells were stained with anti-CD11a–FITC or anti-CD29–PE and analyzed by flow cytometry. Representative data from 3 experiments are shown. (D and E) Preactivated WT and CRK/CRKL Dko CD4⁺ T cells were applied to fibronectin-coated coverslips. After incubation at 37°C for 30 minutes, cells were fixed and analyzed by differential interference contrast (DIC) microscopy. Arrows indicate polarized T cells with a clear uropod. (E) Cells prepared as in D were quantified. Data are average ± SD values from 3 experiments, totaling 433 cells for WT and 387 for CRK/CRKL Dko. *P < 0.05. Scale bar: 20 μm. Statistical analysis was performed using paired (A and B) or unpaired (E) 2-tailed Student’s t tests.

Figure 1. Crk and Crkl are deleted in T cells of CRK/CRKL Dko mice.

(A) CD4⁺ T cells were purified from lymph nodes of CRK/CRKL Dko and WT mice. Preactivated T cells were prepared by stimulating with plate-bound anti-CD3 and anti-CD28 for 2 days and culturing without stimuli for an additional 5 days. Whole cell lysates were analyzed by SDS-PAGE and immunoblotted with anti-CRKL, anti-CRK, and anti-ZAP70. (B) (C) Preactivated CD4⁺ T cells made as in A were fixed and permeabilized, and the intracellular staining of CRK (B) and CRKL (C) was determined by flow cytometry. (D) Thymus, spleens, and lymph nodes (LNs) were isolated from WT and CRK/CRKL Dko mice. Single-cell suspensions were analyzed by flow cytometry using the indicated antibodies. (E) Gated CD4⁺ and CD8⁺ T cells from lymph nodes (mixed peripheral and mesenteric lymph nodes) were analyzed for the indicated surface markers.