Defective T Cell Development and Function in the Absence of Abelson Kinases¹

Successful T cell development in the thymus is critical for both cellular and humoral immunity. Bone marrow-derived prothymocytes undergo a tightly controlled developmental program that is regulated by signals downstream of the pre-TCR, and subsequently the TCR (1, 2). Immature thymocytes lack expression of the CD4 and CD8 cell surface receptors, and proceed from the CD4⁻CD8⁻ double-negative (DN)³ stage through a CD4⁺CD8⁺ double-positive (DP) stage to become mature CD4⁺ or CD8⁺ single-positive (SP) T cells. The transition from DN into DP cells is regulated by signaling through the pre-TCR, whereas differentiation from DP cells to mature T cells requires a functional TCRαβ (2, 3). A subset of protein tyrosine kinases has been shown to play pivotal roles as proximal signaling molecules downstream of both the pre-TCR and the TCRαβ. Mice deficient for Lck, a member of the Src family of tyrosine kinases, show a partial block in thymocyte development at the DN stage and lack mature T cells (4). Mice deficient for both the Syk and ZAP70 tyrosine kinases fail to initiate clonal expansion of DN cells, which fail to differentiate into DP cells (5). In addition to these tyrosine kinases, adaptor proteins such as linker for activation of T cells (LAT), SH2-domain-containing leukocyte protein of 76 kDa, and Shc have been shown to contribute to pre-TCR signaling and regulate T cell development (6, 7, 8, 9).

The Abelson (Abl) kinases, Abl (Abl1) and Arg (Abl2), are highly regulated nonreceptor tyrosine kinases that transduce signals downstream of receptor tyrosine kinases and other receptors (10, 11). Abl kinases have been shown to regulate cell proliferation, migration, and survival (12), and altered forms of the Abl kinases produced from chromosomal translocation events, such as Bcr-Abl and Tel-Arg, are implicated in the development of some human leukemias (13). We have previously demonstrated that endogenous Abl tyrosine kinase activity is elevated following activation of the TCR, and that inhibition of Abl kinase activity or partial loss of Abl and Arg proteins results in down-regulation of mature TCR-induced signaling in vitro (14). Whether the Abl family kinases regulate pre-TCR and TCR signaling in vivo, thereby affecting T cell development and function, remained to be determined.

Both Abl and Arg are expressed throughout embryonic development and in the adult (15). Expression of Abl is higher in the thymus, spleen, and testes compared with other tissues in postnatal mice (16), and Arg expression levels are highest in the brain, followed by thymus and spleen (17). Single deletion of Abl (Abl1) in mice results in severe immune system dysfunction, including splenic and thymic atrophy, lymphopenia, and increased susceptibility to infection (18, 19). However, the cellular and molecular basis for the immune phenotypes observed in the Abl1 single-knockout mice have remained elusive, and have been compounded by the pleiotropic roles of the Abl kinases in the regulation of multiple cell types, including B cells and stromal cells (11, 12, 18, 19). In contrast to the Abl1 knockout mice, Arg-deficient mice do not display obvious abnormalities in the thymus and spleen (17). Notably, whereas mice lacking either abl1 or abl2 can survive to adulthood, mice knockout for both abl1 and abl2 genes die during embryogenesis, thereby suggesting that these kinases share overlapping functions during embryonic development (17).

To study whether Abl kinases are involved in T cell development and dissect the role of the Abl kinases in T cell function in vivo, we have conditionally inactivated the Abl kinases in T cells by crossing the Lck-Cre transgenic mice with mice carrying loxP-flanked abl1 sequences in the abl2^−/− background. In this study, we report that Abl/Arg-conditional knockout mice have impaired thymocyte development and show for the first time that Abl kinase activity is required for pre-TCR signaling. We have found that specific deletion of Abl family kinases in mature T cells results in impaired proliferation in response to TCR stimulation, as well as reduced IL-2 and IFN-γ production. Significantly, Abl/Arg-conditional knockout mice are compromised to mount a humoral immune response to T cell-dependent Ag in vivo, and are severely impaired in the production of CTL during Listeria monocytogenes infection. Together these findings reveal an important role for Abl family kinases in T cell development and T cell function in vitro and in vivo.

The Abl^flox mice were generated, as previously described (20), and were crossed into the arg^−/− background to generate abl flox/flox arg^−/− mice. These mice were subsequently crossed with the Lck-Cre transgenic mice (Taconic Farms) to generate conditional loss of Abl kinases in the T cell lineage. Genotypes were confirmed by PCR. All mice were in the C57BL/6 genetic background. Mice were housed under specific pathogen-free conditions in the Duke University Cancer Center Isolation Facility. All studies using mice have followed the protocols reviewed and approved by Duke Institutional Animal Care & Use Committee.

Primary T cells purified from mouse spleen were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 55 μM 2-ME, and 100 μg/ml each penicillin and streptomycin. The pre-TCR SL-12β12 cell line (gift from D. Wiest, Fox Chase Cancer Center, Philadelphia, PA) was maintained in DMEM supplemented with 10% heat-inactivated FBS, 10 mM HEPES, 1 mM sodium pyruvate, 55 μM 2-ME, 0.1% gentamicin, and 0.5 mg/ml geneticin. Anti-Abl Ab specific for the Abl kinase domain (clone 8E9) was obtained from BD Pharmingen, and anti-Abl Ab specific for the Abl C-terminal region (Ab-3) was obtained from Calbiochem.

Thymocytes or splenocytes in single-cell suspension were lysed of RBC and washed in FACS buffer (PBS containing 2% heat-inactivated FBS, 0.05% sodium azide). For surface staining, ∼1 × 10⁶ cells were first incubated 5 min with anti-CD16/CD32 mouse Fc block (BD Pharmingen) and followed by staining with FITC-, PE-, PE-CY5-labeled anti-CD4, CD8, B220, TCRβ, CD44, and CD25 (BD Pharmingen; eBioscience) on ice for 30 min. Cells were washed twice, collected on a FACScan flow cytometer, and analyzed using CellQuest software (BD Biosciences). For BrdU staining, 50 μM BrdU was added at the end of T cell activation for 30 min. After fixation in EtOH, cells were then treated with 0.08% pepsin in 0.1 N HCl to isolate nuclei, followed by denaturing DNA with 2 N HCl and neutralizing with 0.1 M sodium borate (pH 8.5). After wash, cells were stained with FITC anti-BrdU Ab (BD Pharmingen) and propidium iodide (PI; Sigma-Aldrich) (10 μg/ml) in the presence of RNase A (250 μg/ml) in staining buffer (10 mM HEPES (pH 7.5), 150 mM NaCl, 4% FBS, 0.5% Tween 20, and 0.1% sodium azide).

T cells were purified from total splenocytes using Pan T Cell Isolation Kit with LS columns (Miltenyi Biotec). Approximately 2 × 10⁵ T cells in 100 μl were plated in triplicates into 96-well plate precoated with anti-CD3 (clone145-2C11; BD Pharmingen) or anti-CD3 plus anti-CD28 Abs (clone 37.51; BD Pharmingen), incubated in a 37°C 5% CO₂ incubator for 48 h. [³H]Thymidine (PerkinElmer Life & Analytical Sciences) (1 μCi/well) was added during the last 16–20 h, and cells were harvested using a scintillation counter (PerkinElmer Life & Analytical Sciences). For MLR, splenocytes from DBA mice (H-2^d) were first treated with 50 μg/ml mitomycin C (Sigma-Aldrich) and mixed with splenocytes isolated from either wild-type or Abl/Arg-conditional null mice at E:T = 1:2.5 for 72 h; [³H]thymidine was added during the last 16 h. For cytokine production, T cells were stimulated as for the proliferation assay, and culture supernatant was collected at 24 h. Levels of IL-2, IL-4, and IFN-γ were measured using DuoSet ELISA kits for mouse IL-2, IL-4, and IFN-γ (R&D Systems).

Splenocytes isolated from either Abl^flox/flox/Arg^+/+ or Abl^flox/flox/Arg^−/−, lck-Cre-negative mice were stimulated with anti-CD3 Ab (5 μg/ml) for 24 h. Cells were then retrovirally transduced with either vector-encoding GFP alone or Cre-GFP-containing vector in the presence of mouse rIL-2 (10 ng/ml; R&D Systems) and 8 μg/ml polybrene (Sigma-Aldrich). GFP-positive cells were sorted 3 days after infection using FACS and were used for experiments. For proliferation, 5 × 10⁴ cells were cultured in the presence of mouse rIL-2 for 24 h, and [³H]thymidine (1 μCi/well) was added during the last 12 h. For IFN-γ secretion, cells were restimulated with anti-CD3, and culture supernatant was collected between 3 and 8 h.

For analyzing TCR-induced apoptosis, purified T cells were stimulated with plate-bound anti-CD3 for 48 h, and stained with annexin V and 7-aminoactinomycin D using Annexin V-PE Apoptosis Detection Kit (BD Pharmingen). For caspase 3 activation, purified T cells were plated onto 96-well plate precoated with anti-CD3 (5 μg/ml) Ab for the indicated times. Cell lysates were then analyzed using both anti-caspase 3 and anti-cleaved caspase 3 Abs (Cell Signaling Technology).

For pre-TCR activation, SL-12β12 cells in DMEM plus 0.1% BSA were incubated with or without imatinib (5 μM) for 1 h at 37°C. Cells were then incubated with biotinylated anti-TCRβ Ab (10 μg/ml; BD Pharmingen) on ice for 30 min. After wash, streptavidin (25 μg/ml; Sigma-Aldrich) was added to each sample, incubated at 37°C for the indicated time, and stopped by adding 1 ml of cold PBS. Cells were lysed in lysis buffer containing 50 mM Tris-HCl (pH 7.4), 0.5% Triton X-100, 150 mM NaCl, 1 mM EDTA, 1 mM sodium orthovanadate (Na₃VO₄), and 1× proteinase and phosphatase inhibitors (Sigma-Aldrich). Purified splenic T cells were plated onto 24-well plate precoated with anti-CD3 (5 μg/ml) Ab at 37°C for the indicated time and stopped by adding 5× radioimmunoprecipitation assay buffer (1× buffer: 50 mM HEPES (pH 7.0), 150 mM NaCl, 2 mM EGTA, 1% Triton X-100, 0.25% sodium deoxycholate, 1 mM Na₃VO₄, and proteinase and phosphatase inhibitors). For analyzing IκBα protein, T cells were pretreated with proteosome inhibitor MG132 (5 μM) at 37°C for 1 h and then stimulated with either anti-CD3 (5 μg/ml) or anti-CD3 (2.5 μg/ml) plus anti-CD28 (5 μg/ml) for 20 min. Protein was quantitated by Bio-Rad Protein Assay (Bio-Rad). Equal amount of protein was separated on SDS-PAGE. Western blots were performed per manufacturer’s instruction with the following phospho-Abs: anti-ZAP70 (Tyr³¹⁹/Syk³⁵²), anti-phospholipase C (PLC) γ1 (Tyr⁷⁸³), anti-Shc (Tyr^239/240), anti-ERK (Thr²⁰²/Tyr²⁰⁴), anti-SAPK/JNK (Thr¹⁸³/Tyr¹⁸⁵), anti-p38 (Thr¹⁸³/Tyr¹⁸⁵), and anti-IκBα (Ser³²) from Cell Signaling Technology, and anti-LAT (Tyr¹³²) from BioSource International. Phospho-specific Ab against the autophosphorylation site of Lck (Y394) was a gift from A. Shaw (Washington University School of Medicine, St. Louis, MO). Membranes were washed with stripping buffer (62.5 mM Tris-HCl (pH 6.8), 2% SDS, 0.7% 2-ΜΕ) and reprobed with the following Abs for total protein: anti-ZAP70, anti-LAT, and anti-p38 (Cell Signaling Technology); anti-PLCγ, anti-ERK1, anti-JNK1, and anti-IκBα from Santa Cruz Biotechnology; anti-Shc from BD Transduction Laboratories; and anti-p56^lck from BioSource International.

Wild-type and Abl/Arg-conditional null mice were immunized i.p. at day 0 with 50 μg of alum-precipitated nitrophenylacetyl (NP)-chicken γ-globulin (T cell-dependent Ag) or NP-LPS (T cell-independent Ag) (Biosearch Technologies). For T cell-dependent Ag response, mice were challenged with same dose at day 21. Serum was collected from tail vein at indicated times, and NP-specific IgM and IgG titers were determined by ELISA on 96-well plates coated with NP-BSA (Biosearch Technologies), followed by alkaline phosphatase-conjugated goat anti-mouse IgM or IgG, and developed in pNPP substrate (Southern Biotechnology Associates). NP-specific mouse IgM and IgG (gifts from W. Zhang, Duke University Medical Center, Durham, NC) were used as standards.

A recombinant L. monocytogenes strain engineered to secrete chicken OVA and pMHC/peptide tetramers was originally provided by M.J. Bevan (University of Washington, Seattle, WA). Wild-type and Abl/Arg-conditional null mice were infected i.v. with 5 × 10³ CFU of bacteria. Splenocytes were prepared at day 7 after infection and were then challenged with or without 0.2 μM of either OVA (257–264) peptide for CD8 or listeriolysin O (190–201) peptide for CD4 in the presence of monensin (3 μM) at 37°C for 5.5 h. Cells were surface stained with PE-conjugated anti-CD8 or anti-CD4 Abs, fixed in 2% paraformaldehyde, and permeabilized with 0.1% saponin, followed by staining with FITC anti-IFN-γ (BD Pharmingen) Ab. Activation of Ag-specific CD4 cells was also analyzed by staining for both CD4 and CD154. For H-2K^b-OVA-binding experiment, DimerX I (BD Biosciences) was loaded with OVA peptides (257–264) overnight at 37°C and preincubated with PE-conjugated anti-mouse IgG1 for 4 h before incubation with splenocytes harvested 7 days after L. monocytogenes infection. The number of cells binding to the H-2K^b peptide was analyzed by gating on the CD8⁺ cells. A total of 200,000 cells was collected on FACScan and analyzed by CellQuest.

All statistics were performed using Student’s unpaired, two-tailed t test.

The Abl^flox mice were generated, as previously described (20), and were crossed into the arg^−/− background to generate abl flox/flox arg^−/− mice. These mice were subsequently crossed with the Lck-Cre transgenic mice to generate conditional loss of Abl kinases in the T cell lineage. Protein analysis of Abl/Arg-double-null thymocytes and purified splenic T cells demonstrated the absence of full-length Abl proteins by immunoblotting with an Ab specific for the kinase domain of Abl (Fig. 1, A, upper panel, and B, lane 2). In contrast, Abl protein was present in cells other than T cells, indicating specific inactivation of the abl1 gene in T cells (Fig. 1,B). Consistent with previous findings (20), a small amount of a truncated Abl protein was detected in Abl/Arg-null thymocytes by blotting with Ab specific for the Abl C-terminal region (Fig. 1 A, bottom panel). This truncated version of the Abl protein was previously demonstrated to completely lack Abl kinase activity (20).

We observed that Abl/Arg-conditional knockout mice consistently exhibited greater than 50% reduction in the total number of thymocytes compared with wild-type controls (Fig. 2,A). Thymocytes isolated from the Abl/Arg-conditional knockout mice and their wild-type littermate controls (4–6 wk old) were surface stained with anti-CD4 and anti-CD8 Abs and analyzed by flow cytometry. Although the percentage of CD4⁺CD8⁺ DP and CD4⁺ or CD8⁺ SP cells was not significantly different between wild-type and mutant mice, the percentage of the CD4⁻CD8⁻ DN population was consistently increased by 2- to 4-fold in the Abl/Arg-null mice (Fig. 2,B). Comparison of the absolute cell numbers for each subpopulation revealed that the reduction in the total number of thymocytes in the Abl/Arg mutant mice compared with wild-type mice was primarily due to a ∼50% decrease in DP cells as well as similar decreases in CD4-SP and CD8-SP cells, with no significant difference in the total number of DN cells (Fig. 2 C). These results suggested potential developmental defects during the DN to DP transition in the absence of Abl and Arg kinases.

The DN thymocytes can be subdivided into four developmental stages (DN1, DN2, DN3, and DN4) based on the expression of the CD25 and CD44 cell surface molecules (2, 3). These stages progress in the following order: CD25⁻CD44⁺ (DN1), CD25⁺CD44⁺ (DN2), CD25⁺CD44⁻ (DN3), and CD25⁻CD44⁻ (DN4). TCRβ gene rearrangements are initiated at the DN2 stage and continue during the DN3 stage. The rearranged TCRβ chain heterodimerizes with the pre-Tα chain to form the pre-TCR at the cell surface. Signals derived from the pre-TCR at the DN3 stage are important for maturation to the DN4 stage and are implicated in proliferative expansion, survival, allelic exclusion of the TCRβ locus, and induction of TCRα rearrangement (2). To define the defect in thymocyte development in the Abl/Arg-conditional knockout mice, we analyzed the expression of CD25 and CD44 among the DN populations. Thymocytes from Abl/Arg-conditional knockout mice consistently presented a higher percentage of CD25⁺CD44⁻ (DN3) cells with a corresponding decrease in DN4 stage cells compared with wild-type mice (Fig. 3 A), which suggests that loss of Abl kinases may affect the DN3 to DN4 transition during thymocyte development. Successful TCRβ gene rearrangement and expression of the pre-TCR occur at DN3 stage, and functional signaling through pre-TCR is critical for thymocyte development (2). Therefore, we analyzed the expression of TCRβ on subsets of thymocytes by FACS analysis. There was no significant difference in surface expression of TCRβ in all four thymocyte subpopulations between wild-type and Abl/Arg mutant mice (data not shown). Thus, Abl kinases are not involved in TCRβ chain rearrangement and surface expression. Similarly, analysis of positive selection by staining thymocytes for DP expression of TCRβ and CD69 (21) demonstrated that there was no significant difference between Abl/Arg-null thymocytes and their littermate controls (data not shown).

To address whether the defect in the transition from DN3 to DN4 elicited by the absence of Abl kinases could result from impaired signaling downstream of the pre-TCR, we performed in vivo injections with anti-CD3ε Ab, which activates the pre-TCR and accelerates thymocyte maturation (22, 23). Mice were injected i.p. with anti-CD3ε Ab and thymocytes were analyzed 4 days later. As expected, wild-type mice exhibited accelerated maturation through the DN3 stage with few cells remaining at the DN2 and DN3 stages (Fig. 3,B). In contrast, a high number of DN3 cells accumulated in the Abl/Arg-double-null thymocytes (Fig. 3 B). Together our data support a role for Abl kinases in thymocyte development and pre-TCR signaling in vivo.

To directly examine the role of Abl kinases in pre-TCR signaling, we used the pre-T cell line SL-12β12, derived from a spontaneous SCID mouse thymoma that stably expresses functionally rearranged TCRβ with endogenous pre-TCRα chains on the cell surface (24). To assess whether loss of Abl kinase activity affected pre-TCR signaling, SL-12β12 cells were pretreated with the Abl kinase inhibitor imatinib mesylate (5 μM for 1 h), followed by stimulation with biotinylated anti-TCRβ Ab and cross-linking with streptavidin for the indicated times (Fig. 4). We observed enhanced phosphorylation of ZAP70/Syk, LAT, and PLCγ1, as well as the adaptor protein Shc following pre-TCR activation in control cells. In contrast, the tyrosine phosphorylation of these molecules was greatly diminished in cells treated with the Abl kinase inhibitor (Fig. 4, A–D). Phosphorylation of ERK was also strongly enhanced after pre-TCR cross-linking in control cells, and it was only slightly reduced at the earliest time point after pre-TCR stimulation in the imatinib-treated cells (Fig. 4 E). Together these results suggest that Abl kinases are involved in pre-TCR signaling.

Splenocytes were isolated from Abl/Arg-conditional knockout mice and their littermate controls (5–9 wk), and stained with cell surface markers for T and B cells, including TCRβ, CD4, CD8, and B220. We observed a ∼40–50% decrease in the percentage of total T cells, as well as CD4⁺ and CD8⁺ cell populations, and a relative small increase in the percentage of B cells in the Abl/Arg-conditional null compared with wild-type mice (Fig. 5,A). B cell proliferation was not affected in these mice because there was no significant difference in proliferation following stimulation of total splenocytes with the B cell mitogens anti-IgM or LPS (data not shown). Although the absolute number of total splenocytes was not significantly different between mutant and wild-type mice, the total T cell numbers were significantly reduced in the absence of Abl family kinases, with significant reductions in both CD4⁺ and CD8⁺ subsets in the Abl/Arg-conditional knockout mice (Fig. 5,B). The reduction in splenic T cells in the mutant mice suggested that Abl kinases might be involved in T cell proliferation or survival. Splenic T cells from wild-type, Abl-null, Arg-null, and Abl/Arg-double-null mice were stimulated with either anti-CD3 alone or anti-CD3 plus anti-CD28 for 48 h, and proliferation was measured by [³H]thymidine incorporation (Fig. 6,A). Resting T cells did not show significant [³H]thymidine incorporation in both wild-type and mutant cells (data not shown). Abl- and Arg-single-null T cells showed a significant reduction in proliferation upon TCR stimulation with low amounts (1 μg/ml) of anti-CD3 (Fig. 6,A). At higher doses of anti-CD3 (5 μg/ml) alone or anti-CD3 plus higher dose of anti-CD28 (5 μg/ml), Arg-null T cells proliferated similar to wild-type controls. Notably, loss of both Abl and Arg resulted in a more profound defect in T cell proliferation compared with the loss of either kinase alone at both low and high doses of anti-CD3 in the absence or presence of CD28 costimulation (Fig. 6,A). Addition of IL-2 failed to reverse the proliferation defect in the Abl/Arg-double-null T cells (data not show). We did not observe significant differences in proliferation between wild-type and Abl/Arg-double-null T cells in response to PMA and ionomycin stimulation (data not shown). We also observed significantly less proliferation in lymphocytes from Abl/Arg-conditional null mice compared with wild-type mice in response to allogeneic stimulation in a MLR (Fig. 6 B). Thus, loss of Abl/Arg kinases does not result in a general defect in T cell proliferation, but specifically inhibits TCR-induced cell proliferation, and suggests that Abl kinases function to transduce proximal signals downstream of the TCR.

To study whether T cell function is impaired in the absence of Abl kinases, wild-type or Abl/Arg-double-null T cells were stimulated with anti-CD3 in the absence or presence of anti-CD28 Abs for 24 h, and the levels of IL- 2, IL-4, and IFN-γ secreted into culture supernatant were measured by ELISA (Fig. 6,C). Both IL-2 and IFN-γ levels were markedly reduced by ∼70% in Abl/Arg-double-null T cells compared with wild type (Fig. 6,C, top and bottom panels), whereas a smaller reduction of IL-4 (∼30%) was observed in the Abl/Arg mutant mice (Fig. 6 C, middle panel). Thus, Abl kinases regulate Th function, and specifically TCR-mediated IL-2 and IFN-γ secretion.

To determine whether the defect in mature T lymphocyte proliferation and cytokine secretion in the absence of Abl kinases is due to defective T cell development or to an intrinsic cell-autonomous defect, we have deleted the abl gene in vitro following transduction with a retroviral Cre vector. T cells were stimulated with anti-CD3 before retroviral infection, and deletion of the abl gene was confirmed by the absence of Abl protein (Fig. 6,D, left). Proliferation of the CD3-stimulated T cells was measured in the presence of IL-2 for 24 h, and IFN-γ secretion was measured 3–8 h after restimulation with anti-CD3 Ab. We observed a significant reduction in both proliferation and IFN-γ secretion in the Abl/Arg-double-null T cells compared with wild-type T cells (Fig. 6 D, middle and right panels). Thus, Abl kinases have a T cell-autonomous role and are required for maximal proliferation and cytokine secretion in mature T cells. Prolonged restimulation of these activated T cells with anti-CD3 Ab resulted in apoptosis of both wild-type and Abl/Arg-null cells (data not shown).

To further investigate the possible mechanism(s) responsible for the reduced T cell proliferation induced by loss of the Abl kinases, cells were pulse labeled for a short time with BrdU at the end of anti-CD3 stimulation, and the cell cycle was analyzed by the DNA profile. Abl/Arg-double-null T cells or wild-type T cells treated with the Abl kinase inhibitor imatinib showed a consistently lower percentage of cells entering into S phase compared with wild-type T cells (Fig. 7,A). There was no significant difference in the percentage of sub-G₁ cells, suggesting no increased apoptosis in Abl/Arg-double-null T cells. Moreover, analysis of T cell survival upon TCR stimulation using annexin V and PI staining (Fig. 7,B), as well as Western blot analysis of caspase 3 activation (Fig. 7 C) did not show differences in apoptosis between wild-type and Abl/Arg-double-null T cells. These results indicate that the reduced T cell proliferation in the absence of Abl kinases is primarily due to decreased cell cycle entry.

To define the possible mechanism(s) responsible for impaired T cell proliferation, we examined expression of early T cell surface markers following TCR activation. Wild-type and Abl/Arg-double-null T cells were stimulated with anti-CD3 Ab for 18 h and stained with either anti-CD25α (IL-2Rα chain) or the CD69 early T cell activation marker. There was no significant difference in the induction of either CD25α or CD69 expression between wild-type and mutant T cells (data not shown), indicating that the proliferation defect in the Abl/Arg-double-null T cells is not due to decreased IL-2R on the cell surface. These findings suggested that inhibition of cell proliferation in T cells lacking Abl and Arg kinases may result from impaired TCR-mediated signaling. Indeed, we found that TCR-induced phosphorylation of ZAP70, PLCγ1, and LAT was greatly reduced in Abl/Arg-double-null T cells compared with wild-type cells (Fig. 8, A–C). These results are consistent with our previous findings of reduced phosphorylation of these signaling proteins in imatinib-treated primary T cells and splenic T cells derived from abl1^+/−abl2^−/− mice (14). Notably, TCR-induced Lck activation measured by tyrosine 394 phosphorylation was unchanged in T cells lacking Abl family kinases compared with wild-type controls (data not shown). We also analyzed the phosphorylation of the Shc adaptor protein following TCR stimulation of purified T cells derived from control and Abl/Arg-conditional knockout mice, and show for the first time that Abl/Arg-null T cells exhibit a dramatic decrease in Shc tyrosine phosphorylation following TCR stimulation (Fig. 8,D). Activation of the MAPK pathways following TCR stimulation was also impaired in Abl/Arg-double-null T cells, with loss of the Abl kinases leading to a more severe reduction in JNK activation than that observed for ERK activation (Fig. 8, E and F). In contrast, we found that phosphorylation of the p38 MAPK was unchanged in Abl/Arg-double-null T cells compared with wild-type T cells (Fig. 8 G).

We have shown in this study that IL-2 production was markedly reduced in Abl/Arg-null T cells, and have previously observed that inhibition of Abl kinase activity reduced activation of the CD28RE/AP response element of the IL-2 promoter following TCR activation in Jurkat T cells (14). The NF-κB transcription factor is known to bind to the CD28RE region and activate IL-2 gene transcription (25), and is also required for chromatin remodeling of the IL-2 promoter upon TCR stimulation (26). Thus, we examined whether NF-κB activation was impaired in T cells lacking the Abl kinases. Inactive NF-κB is sequestered by binding to IκB in the cytoplasm of resting T cells. Upon TCR activation, IκBα is phosphorylated, which triggers the ubiquitination and degradation of IκBα, thereby releasing NF-κB and promoting its translocation to nucleus to regulate transcription (27, 28). As shown in Fig. 8 H, phosphorylation of IκBα was markedly decreased in Abl/Arg-double-null T cells compared with wild-type controls following stimulation with either anti-CD3 alone or anti-CD3 plus anti-CD28. Thus, decreased IL-2 production in the absence of Abl kinases may be mediated in part by decreased NF-κB activation.

To determine whether Abl kinases play a role in T cell function in vivo, mice were injected with T cell-dependent or T cell-independent Ags, and Ab production was analyzed in Abl/Arg-conditional knockout mice and control littermates. As shown in Fig. 9,A, wild-type mice produced high levels of IgM in response to initial injection of T cell-dependent Ag, and high levels of IgG upon secondary boost with same Ag. In contrast, Abl/Arg-conditional knockout mice were markedly impaired in both early IgM secretion and secondary isotype switch to IgG secretion, suggesting that Th function is severely impaired in these mice. As expected, wild-type and mutant mice secreted equal amounts of IgM in response to a T cell-independent Ag (Fig. 9 B), indicating that B cell function is not affected in the Abl/Arg T cell-conditional knockout mice.

To investigate whether deficiency in the expression of Abl family kinases results in altered susceptibility to infection, Abl/Arg-conditional knockout mice and littermate controls were infected with a low dose of L. monocytogenes, and the mice were analyzed for the development of Ag-specific CD8⁺ and CD4⁺ T cells. We found that Abl/Arg mutant mice developed very few Ag-induced IFN-γ-producing CD8⁺ effector T cells compared with wild-type mice (Fig. 10,A). Notably, the total number of IFN-γ-positive CD8⁺ T cells in the Abl/Arg-conditional knockout mice was reduced by greater than 80%, and it was only 18% compared with the wild-type controls (Fig. 10,B). The number of specific CD8⁺ T cells binding to the H-2K^b peptide was also significantly reduced in the Abl/Arg-null mice (73% reduction compared with the wild type; Fig. 10,C). Thus, expansion of specific CD8⁺ T cells upon Listeria infection was impaired in Abl/Arg-null mice. In contrast, the percentage of IFN-γ-positive CD4⁺ T cells was similar between wild-type and mutant mice (Fig. 10,D). However, the total number of IFN-γ-positive CD4⁺ T cells was ∼50% of the wild type (Fig. 10,E), which most likely reflects the reduced number of splenocytes in the mutant mice after infection (52.6% of the wild-type mice). The total number of CD4⁺/CD154⁺ DP cells in the mutant mice was also ∼50% of the wild-type mice (Fig. 10 F). Together, these results reveal that Abl kinases are principally required for the development of effector CD8⁺ T cells in vivo in response to Listeria infection.

Our work has uncovered cell-autonomous roles for Abl family kinases in the regulation of both thymocyte development and mature T cell function, and showed that Abl kinases signal downstream of both the pre-TCR and TCR. The generation of T cell-specific Abl/Arg-null mice permitted us to bypass the embryonic lethality associated with complete loss of Abl and Arg, and revealed a previously unappreciated role for these kinases in thymocyte development. We found that Abl/Arg-double-null thymocytes were partially blocked at the DN3 stage, a developmental stage that requires signaling through the pre-TCR to progress to the DN4 stage (2). Abl/Arg-double-null thymocytes expressed wild-type levels of TCRβ on the DN populations, which suggests that the Abl/Arg-double-null thymocytes have normal amounts of pre-TCR surface expression. Rather, we found that signaling through the pre-TCR was impaired in the absence of Abl kinases. In this regard, Abl/Arg-double-null thymocytes failed to progress through the DN3 stage induced by in vivo injection of anti-CD3ε Ab, which results in complete mobilization of wild-type thymocytes to the DN4 stage with few remaining DN3 stage cells. Moreover, loss of Abl kinase activity produced a marked decrease in the activation of several signaling molecules in pre-TCR cells.

Knockout mouse studies have identified a critical role for the Lck and ZAP70/Syk tyrosine kinases in thymocyte development (4, 5). Loss of Lck or ZAP70 and Syk results in arrested thymocyte development with a partial or complete block at the DN3 stage. Similarly, genetic inactivation of adaptor proteins regulated downstream of these tyrosine kinases, such as LAT (9) and Shc (8), results in arrested thymocyte development. Our finding that Abl kinase activity is required for maximal activation of ZAP70/Syk, LAT, PLCγ1, and Shc following pre-TCR activation defines a new role for Abl family kinases downstream of the pre-TCR in thymocytes.

As expected from our finding that CD4⁺ and CD8⁺ SP thymocytes were decreased in the Abl/Arg-conditional null mice, mature splenic T cells were also significantly reduced by ∼40–50% in the mutant mice. Although Abl- and Arg-single-null T cells showed a limited defect in proliferation, Abl/Arg-double-null T cells exhibited a more profound proliferative defect upon anti-CD3 and/or anti-CD28 stimulation. Thus, both Abl and Arg kinases are required for maximal immune responses. The basis for the proliferative defect in T cells lacking Abl kinases is most likely due to delayed cell cycle entry into S phase. A role for Abl kinases in cell cycle progression is consistent with previous findings that Abl is required for growth factor-induced DNA synthesis in fibroblasts (29, 30, 31). Furthermore, inhibition of TCR-induced proliferation in cells lacking Abl and Arg is also consistent with a role for Abl kinases in the regulation of mature TCR signaling. Our current findings and previous data (14) suggest that Abl kinases may link TCR engagement and Lck activation to the activation of ZAP70/Syk, LAT, and PLCγ1. Alternatively, Abl kinases may not directly activate ZAP70 and specific signaling molecules, but rather Abl kinases may regulate the assembly of signaling complexes and/or cytoskeletal scaffolds required for proper activation of ZAP70 and downstream signaling targets.

We have also shown that Abl kinases are involved in maximal activation of Shc, ERK, and JNK in mature T cells. Following TCR engagement, the adaptor protein Shc is recruited to the TCR through interactions with ZAP70 and CD3ζ (32, 33). Shc proteins have been reported to be phosphorylated by Lck and Syk/ZAP70 tyrosine kinases (34), and subsequently bind to Grb2 and son of sevenless, leading to activation of the Ras-MAPK pathway (35). We found that phosphorylation of Shc at tyrosines 238/240 following TCR stimulation was greatly reduced in Abl/Arg-double-null T cells. Thus, Abl kinases may regulate Shc activity indirectly through the regulation of ZAP70/Syk phosphorylation, or may directly phosphorylate Shc at Y238/Y240. Tyrosine phosphorylation of Shc is important for both T cell proliferation and IL-2 production (35, 36), and Abl-mediated phosphorylation of Shc may play a role in the regulation of these processes. Furthermore, decreased IL-2 production in Abl/Arg-null T cells is most likely linked to decreased transcriptional activity of factors required for IL-2 gene expression, such as NF-κB. In this study, we showed that TCR-induced NF-κB activity, as measured by the phosphorylation of IκBα, is decreased in the absence of the Abl kinases. Additionally, JNK activity is reduced in Abl/Arg-double-null T cells, which may contribute to the impaired production of cytokines, including IFN-γ, observed in Abl/Arg T cell-conditional knockout mice in response to TCR stimulation and bacterial infection. In this regard, JNK activity was shown to be required for production of IFN-γ in T cells (37).

A most striking finding derived from the analysis of the Abl/Arg T cell-conditional null mice was the discovery that these mice are impaired in the development of Ag-specific CTL and display a marked reduction of IFN-γ-positive CD8⁺ T cells compared with wild-type mice in response to L. monocytogenes infection. Moreover, the Abl/Arg-double-null mice fail to mount a response to T cell-dependent Ags. Thus, Abl kinases are required for both humoral and cellular immunity. The findings presented in this study with the Abl/Arg-conditional knockout mice support the notion that loss of Abl family kinases specifically in T cells underlies the increased susceptibility to infection observed in Abl1 knockout mice (18, 19), as well as the reported immunosuppressive effects of imatinib treatment (38, 39).

L. monocytogenes is a potent inducer of CD8 T cell responses, and once primed to proliferate by the pathogen, CD8 T cells undergo proliferation as well as differentiation into effector and memory T cells (40). Recent work has suggested that asymmetric cell division of CD8 T cells proliferating in response to L. monocytogenes infection results in the segregation of immune receptors and polarity proteins following formation of the immunological synapse, thereby producing effector CTL derived from the daughter cell near the synapse, and long-lived memory T cells derived from the daughter cell distal to the synapse (41). We have previously shown that Abl kinases regulate synapse formation at the neuromuscular junction (42). Moreover, we showed that the Abl targets Abi and Wave are critical for immunological synapse formation and the accumulation of actin of the T cell-B cell interface (43, 44). Thus, loss of Abl kinases may impair adhesive interactions between T cells and APCs, thereby affecting immunological synapse formation and maintenance. The availability of the T cell-conditional Abl/Arg-double-knockout mice described in this work will allow us to examine whether Abl tyrosine kinases have a role in the regulation of the immunological synapse in primary T cells, as well as asymmetric T cell division in response to pathogens.

We thank Weiguo Zhang for advice and reagents, Emily Riggs for technical assistance, and Mike Cook and Lynn Martinek for FACS assistance and advice.

The authors have no financial conflict of interest.