T Cell Receptor-mediated Activation of CD4⁺CD44^hi T Cells Bypasses Bcl10

Abstract

Previous studies have demonstrated that Bcl10 (B-cell leukemia/lymphoma 10) is essential for T cell receptor-mediated NF-κB activation and subsequent proliferation and interleukin 2 (IL2) production. However, here we demonstrate that, contrary to expectations, Bcl10 is differentially required for T cell activation, including for both proliferation and cytokine production. When CD4⁺ and CD8⁺ T cells were divided based on expression levels of CD44, which distinguishes naïve cells (CD44^lo) versus those that are antigen-experienced (CD44^hi), IL2 production by and proliferation of CD4⁺CD44^lo naïve cells and both subpopulations of CD8⁺ T cells were clearly Bcl10-dependent, whereas these same functional properties of CD4⁺CD44^hi T cells occurred largely independent of Bcl10. As with the other subpopulations of T cells, CD4⁺CD44^hi T cells did not activate the NF-κB pathway in the absence of Bcl10; nevertheless, these CD4⁺CD44^hi antigen-experienced T cells efficiently secreted IL2 after T cell receptor stimulation. Strikingly, therefore, T cell receptor-mediated IL2 production in these cells is NF-κB-independent. Our studies suggest that antigen-experienced CD4⁺ T cells differ from their naïve counterparts and from CD8⁺ T cells in their ability to achieve activation independent of the Bcl10/NF-κB pathway.

B-cell leukemia/lymphoma 10 (Bcl10) is a protein that is essential for NF-κB activation in both the T cell receptor (TCR)2 and B-cell receptor signal transduction cascades (1, 2). TCR engagement initiates sequential activation of several families of tyrosine kinases, including the Src, Syk, and Tec families of kinases, which results in tyrosine phosphorylation of a series of adapter and effector molecules and eventual activation of several critical transcription factors (3). Among these transcription factors, NF-κB stands out as a key mediator of T cell activation, proliferation, survival, and effector functions (4). TCR stimulation triggers the recruitment of protein kinase Cθ to the immunological synapse and results in the recruitment and activation also of a tripartite protein complex composed of CARMA1 (CARD domain- and MAGUK domain-containing protein-1), Bcl10, and Malt1 (mucosa-associated lymphoid tissue lymphoma translocation protein 1), which is required for NF-κB activation (5).

The essential role of Bcl10 in TCR-mediated NF-κB activation has been demonstrated in Bcl10-deficient mice. Deficiency of Bcl10 does not affect TCR-induced overall tyrosine phosphorylation, calcium flux, or activation of ERK and AP-1 but specifically affects TCR-induced NF-κB activation. As a result, Bcl10 deficiency impairs TCR-mediated cellular proliferation and IL2 production (1). Similar phenotypes have also been observed in T cells deficient for either protein kinase Cθ, CARMA1, or MALT1, supporting the notion that the CARMA1·Bcl10·MALT1 complex functions in a linear pathway to bridge protein kinase Cθ and NF-κB activation (6–10).

Numerous studies using T cells deficient for molecules in the signaling pathways leading to NF-κB activation have demonstrated the indispensable role that NF-κB plays in IL2 production (1, 6–12). It is, therefore, not surprising that IL2 production is impaired in Bcl10-deficient T cells (1). Unexpectedly, however, we observed that the dependence on Bcl10 for IL2 production is T cell subpopulation-specific. When CD4⁺ and CD8⁺ T cells were further divided based on their expression levels of CD44, which distinguishes naïve cells (CD44^lo) versus those that are antigen-experienced (CD44^hi), IL2 production by CD4⁺CD44^lo and both subpopulations of CD8⁺ T cells was clearly Bcl10-dependent, whereas IL2 production by CD4⁺CD44^hi T cells was surprisingly found to occur independent of Bcl10. This variability in IL2 production was accompanied also by differences in the proliferation responses of the corresponding T cell subpopulations. As with the other subpopulations of T cells, CD4⁺CD44^hi T cells did not activate the NF-κB pathway in the absence of Bcl10; therefore, strikingly, TCR-mediated proliferation and IL2 production by CD4⁺CD44^hi T cells can bypass the requirement for NF-κB activation.

Our studies demonstrate that CD4⁺CD44^hi T cells differ from their naïve counterparts and CD8⁺ T cells in their ability to achieve activation independent of the Bcl10/NF-κB pathway. Currently, high expression of CD44 appears to be the most reliable marker for both CD4⁺ and CD8⁺ memory T cells in mice (13); our studies suggest that the Bcl10/NF-κB pathway is differentially required for the activation of CD4⁺ naïve and CD8⁺ memory versus CD4⁺ memory T cells, as distinguished based on their expression of CD44. The implications of this property of CD4⁺ memory T cells are discussed.

EXPERIMENTAL PROCEDURES

Mice—Bcl10-deficient mice (C57BL/6 background) were generated as described previously (2). Two-to-four-month old mice were used in the experiments. The mice were maintained in the Biological Resource Center at the Medical College of Wisconsin. All animal protocols were approved by the Medical College of Wisconsin Institutional Animal Care and Use Committee.

Cell Purification—Single-cell suspensions were prepared from spleens and/or lymph nodes (pooled from inguinal, cervical, axillary, and mesenteric lymph nodes) from wild-type and Bcl10-deficient mice. Total T cells were purified by negative selection using a combination of antibody-conjugated magnetic beads including anti-B220, anti-Mac-1, and anti-DX-5 (Miltenyi Biotec). The purity of the isolated total T cells was >90%. These bead-sorted T cells were then used for further purification of CD44^lo and CD44^hi T lymphocytes. Briefly, lymph node or splenic T cells were stained with CD44-fluorescein isothiocyanate, CD4-phosphatidylethanolamine, and CD8-Cychrome (all from eBioscience), and CD44^lo naïve and CD44^hi T cells in the CD4⁺ and CD8⁺ populations were separated with a FACSAria Cell-sorting system (BD Biosciences).

Proliferation and Cytokine Assays—For [³H]dT incorporation assays, round-bottom 96-well plates were coated with anti-CD3 (1 μg/ml) and anti-CD28 (2 μg/ml) in phosphate-buffered saline at 4 °C overnight. The coated plates were washed 3 times with phosphate-buffered saline before adding 2 × 10⁴ purified T cells. Total, CD4⁺CD44^lo, CD4⁺CD44^hi, CD8⁺CD44^lo, or CD8⁺CD44^hi T cells from the spleen or lymph nodes (pooled from inguinal, cervical, axillary, and mesenteric lymph nodes) of wild-type and Bcl10-deficient mice were cultured in T cell culture medium (RPMI1640 complete medium containing 100 μg/ml penicillin, 100 μg/ml streptomycin, 2 mm l-glutamine, 10 mm HEPES, pH 7.0, 10 mm nonessential amino acids, 1 mm sodium pyruvate, 50 μm β-mercaptoethanol, 10% fetal bovine serum). In some cases cyclosporine A (CsA) (1 μg/ml) was added. Cell cultures were maintained in a humidified incubator at 37 °C with 5% CO₂. After 48 h of culture the cells were pulsed with 1 μCi/well of [³H]dT for 12–16 h, harvested using a Micro96™ Harvester (Skatron Instruments), and read with a TriLux MicroBeta plate reader (PerkinElmer Life Sciences).

For determinations of cytokine production by ELISA, sorted CD4⁺44^lo, CD8⁺44^lo, CD4⁺44^hi, and CD8⁺44^hi T cells (10⁵/well) from the spleen and lymph nodes of wild-type and Bcl10-deficient mice were cultured in T cell culture medium in ELISA plates coated with anti-CD3 (5 μg/ml) in the presence or absence of anti-CD28 (2 μg/ml). Twenty-four hours thereafter, the supernatants were collected, and ELISA assays were performed according to the manufacturers' recommendations. Rat anti-mouse IL4 (capture antibody), biotinylated-anti-mouse IL4 (detection antibody), and mouse IFNγ ELISA kits were purchased from eBioscience, whereas mouse OptEIA™ IL2 ELISA kits were obtained from BD Pharmingen. For determinations of cytokine production by intracellular staining, cells were sorted as described and stimulated in T cell culture medium in a 96-well plate (2 × 10⁵/well) coated with anti-CD3 (5 μg/ml) plus anti-CD28 (2 μg/ml) in the presence or absence of CsA (1 μg/ml) for 36 h, then treated with monensin (2 μm) for 6 h followed by intracellular staining.

Western Blotting—CD4⁺CD44^lo, CD8⁺CD44^lo, CD4⁺CD44^hi, and CD8⁺CD44^hi T cell subsets from pooled splenocytes and lymph node cells of wild-type and Bcl10-deficient mice were examined in these studies. The purified T cell subpopulations (∼5 × 10⁵ per sample) were stimulated with phorbol 12-myristate 13-acetate (50 ng/ml) and ionomycin (50 ng/ml) for 10 min, and the cells were then collected and lysed in cell lysis buffer (1% Triton X-100, 10 mm Tris-HCl, pH 7.6, 5 mm EDTA, 50 mm NaCl, 0.1 mm Na₃VO₄, 50 mm NaF, 30 mm Na₄P₂O₇, and protease inhibitors). The cell lysates were resolved on a 10% SDS-PAGE gel, transferred to a polyvinylidene difluoride membrane (Amersham Biosciences) and blotted sequentially with anti-phospho-IκBα, anti-IκBα (both from Cell Signaling Technologies), anti-phospho-ERK, anti-ERK (both from Santa Cruz), and anti-actin (Chemicon) antibodies.

NF-κB Electrophoretic Gel Mobility Shift Assay—Analysis of NF-κB activity by electrophoretic gel mobility shift assay was performed as described previously (2). Briefly, purified T cell subsets (about 5 × 10⁵) from wild-type or Bcl10-deficient mice were stimulated in 0.5 ml of T cell culture medium in 48-well plates with plate-bound anti-CD3 (5 μg/ml) plus anti-CD28 (2 μg/ml) for 6 h followed by nuclear extraction. Nuclear extracts (1 μg) were incubated with a ³²P-labeled NF-κB DNA binding consensus sequence oligonucleotide probe. The DNA-protein complexes were resolved on a native 4.5% polyacrylamide gel followed by autoradiography.

RESULTS

Bcl10-deficient Mice Display Specific Reduction in the CD4⁺CD44^hi T cell Population—Previous studies using knock-out mice have demonstrated an essential role for Bcl10 in T cell proliferation and IL2 production (1). Impaired thymic T cell development has also been observed in Bcl10-deficient mice (1). We have independently generated Bcl10-deficient mice and reported that these animals exhibit impaired development and function of their follicular, marginal zone, and B1 B cells as well (2). Likewise, these mice demonstrated defects in thymic T cell development similar to those previously reported (data not shown) (1). However, in contrast to previous studies (1), we observed that although Bcl10-deficient T cells display markedly reduced TCR-mediated proliferation, they nevertheless do retain the capacity to proliferate to an extent (10–30% compared with the response of wild-type T cells) (Fig. 1).

FIGURE 1.

Bcl10-deficient T cells can mount low level proliferation in response to TCR stimulation in vitro. Total splenic T cells from 2–4-month-old wild-type and Bcl10-deficient mice were purified by negative selection using a combination of antibody-conjugated magnetic beads including anti-B220, anti-Mac-1, and anti-DX-5 (Miltenyi Biotec). These purified T cells (2 × 10⁴/well) were stimulated with plate-bound anti-CD3 (1 μg/ml), anti-CD3 (1 μg/ml) plus anti-CD28 (2 μg/ml), or anti-CD3 (1 μg/ml) plus IL2 (20 units/ml), and proliferation of the cells was determined by [³H]dT incorporation assay as described under “Experimental Procedures.” The data are representative of five independent experiments. Error bars represent the S.D. of triplicate measurements of each data point.

Peripheral T cells are a heterogeneous population. Decreased proliferation of Bcl10-deficient T cells might reflect a change in the composition of the T cell subpopulations in the Bcl10-deficient mice in which the highly proliferative subpopulation is decreased relative to that in wild-type mice. Alternatively, but not exclusively, this reduction in T cell proliferation might reflect Bcl10-dependent proliferation of one or more T cell subpopulations. Expression of CD44 is a marker to distinguish naïve (CD44^lo) and antigen-experienced (CD44^hi) T cells in both the CD4⁺ and CD8⁺ T cell populations; in fact, the CD44 molecule is currently the most reliable marker to phenotypically identify naïve versus memory T cells, with CD44^lo cells being considered naïve and CD44^hi cells regarded as memory cells (13). Importantly, it appears that CD44^hi cells can mount a more rapid proliferative response upon stimulation than their naïve counterparts (13–15). It is, therefore, possible that a reduction of the CD44^hi T cell population could account for the reduced proliferation of Bcl10-deficient T cells.

A previous study reported that the absolute number of splenic CD4⁺CD44^hi T cells is reduced in Bcl10-deficient mice relative to wild-type animals (16), suggesting that alterations do occur in the composition of T cell subpopulations in the absence of Bcl10. To further address this issue, we examined in detail the effect of Bcl10 deficiency on the composition of the T cell subpopulations using the Bcl10 knock-out mice that we had previously generated (2). Our analyses confirmed a significant reduction in the percentage of CD4⁺CD44^hi T cells in these mice (Fig. 2A). Examination of CD62L expression in the CD44^hi population, with CD44^hiCD62L^hi representing resting memory T cells and CD44^hiCD62L^lo expression representing recently activated memory T cells, demonstrated a reduction of both populations in the CD4⁺ T cells in the spleen and lymph nodes from Bcl10-deficient animals compared with those in wild-type mice (Fig. 2A). In addition, we analyzed CD43 expression, which can also be employed to define resting (CD43^–) versus recently activated (CD43⁺) memory T cells in the CD44^hi T cell populations (17); both CD44^hiCD43⁺ and CD44^hiCD43^– T cells in the CD4⁺ population were also reduced in the Bcl10-null mice compared with wild-type animals (Fig. 2B). In striking contrast, however, the percentage of CD8⁺CD44^hi T cells, according to both CD43 and CD62L expression patterns, was normal in Bcl10-deficient mice (Fig. 2, A and B). Consistent with these findings, the absolute number of CD4⁺CD44^hi but not of the CD8⁺CD44^hi T cells in the spleen and lymph nodes was significantly reduced in Bcl10-deficient mice compared with wild-type mice (Fig. 2C). Collectively, these data demonstrate that only CD4⁺CD44^hi memory cells and not either CD4⁺CD44^lo or CD8⁺CD44^lo naïveorCD8⁺CD44^hi memory T cell subpopulations are reduced due to Bcl10 deficiency. Given that CD44^hi T cells can mount a more robust proliferative response than their naïve counterparts (13–15), this specific reduction of the CD4⁺CD44^hi T cell subpopulation without significant effects on the remainder of the T cell subpopulations in Bcl10-deficient mice may at least partially account for the decreased proliferation of Bcl10-deficient T cells relative to wild-type T cells.

FIGURE 2.

The CD4⁺CD44^hi T cell population is reduced in Bcl10-deficient mice. Spleen or lymph node cells from wild-type or Bcl10-deficient mice were stained with fluorescent-conjugated anti-CD4, anti-CD8, and anti-CD44 in combination with anti-CD62L (A) or anti-CD43 antibodies (B). The numbers presented on the dot plots are the percentage of each corresponding cell subset gated on the CD4⁺ or CD8⁺ populations. The data are representative of at least six mice per genotype. C, absolute cell numbers of CD4⁺CD44^hi and CD8⁺CD44^hi T cell subsets from the spleens or lymph nodes (pooled from inguinal, cervical, mesenteric, and axillary lymph nodes) of wild-type and Bcl10-deficient mice. The horizontal bars represent the average number in each group. The p values shown represent the comparisons between the corresponding cell populations from wild-type and Bcl10 knock-out mice.

Only CD4⁺CD44^hi T Cells Can Proliferate after TCR Stimulation in the Absence of Bcl10 Expression—To determine whether Bcl10 deficiency affects the TCR-mediated proliferation of selected T cell subpopulations, we next compared the TCR-mediated proliferative capability of CD4⁺CD44^lo, CD8⁺CD44^lo, CD4⁺CD44^hi, and CD8⁺CD44^hi T cells purified from the spleen or lymph nodes of wild-type and Bcl10-deficient mice by FACS. Each sorted cell population was stimulated with anti-CD3, anti-CD3 plus anti-CD28, or anti-CD3 plus exogenously added IL2. As expected, all of the splenic T cell subpopulations from wild-type mice were able to proliferate in response to anti-CD3 stimulation (Fig. 3). The addition of IL2 substantially enhanced the proliferation of both CD4⁺ and CD8⁺ naïve T cells compared with anti-CD3 stimulation alone (p = 0.003 and 0.010, respectively). CD28 co-stimulation also substantially enhanced the proliferation of CD4⁺ but not CD8⁺ naïve T cells (p = 0.011), consistent with the observation that expansion of CD4⁺ T cells is more dependent on costimulation (18). Both wild-type CD4⁺CD44^hi and CD8⁺CD44^hi T cells exhibited stronger proliferation than the corresponding naïve T cells at the concentration of anti-CD3 antibody used in these assays (1 μg/ml) (p = 0.003 and <0.001, respectively).

FIGURE 3.

Bcl10 is differentially required for the proliferation of T cell subpopulations in vitro. CD4⁺CD44^lo (A), CD8⁺CD44^lo (B), CD8⁺CD44^hi (C), and CD4⁺CD44^hi (D) splenic T cell subsets from wild-type and Bcl10-deficient mice were sorted by FACSAria. The cells (2 × 10⁴/well) were then stimulated as described in Fig. 1, and the proliferation of the cells was assessed by [³H]dT incorporation assay. The data are representative of five independent experiments. E, CD4⁺CD44^hi T cells were stimulated as described in Fig. 1 in the presence or absence of CsA (1 μg/ml). The data are representative of two independent experiments. Error bars represent the S.D. of triplicate measurements of each data point within one experiment. p values presented in the text describing Fig. 3 were calculated based on these triplicate measurements.

In contrast to the results obtained for wild-type T cell subpopulations, splenic CD4⁺ and CD8⁺ naïve as well as CD8⁺CD44^hi T cells from Bcl10-deficient mice were not able to proliferate upon anti-CD3 stimulation (Fig. 3, A–C). The addition of IL2 partially rescued this proliferation defect, whereas CD28 co-stimulation did not. Unexpectedly, and quite different from these other T cell subpopulations, Bcl10-deficient splenic CD4⁺CD44^hi T cells displayed strong proliferative capability after anti-CD3 stimulation (more than 50% of that observed for wild-type T cells) (Fig. 3D). The addition of IL2 or CD28 co-stimulation did not substantially enhance the proliferation of Bcl10-deficient CD4⁺CD44^hi T cells (Fig. 3D). Similarly, CD4⁺ and CD8⁺ naïve as well as CD8⁺CD44^hi T cells isolated from lymph nodes also exhibited dependence on Bcl10 for their proliferation, whereas lymph node-derived CD4⁺CD44^hi T cells displayed Bcl10-independent proliferative responses (data not shown). These data demonstrate that Bcl10 is essential for TCR-mediated proliferation of naïve T cells and CD8⁺CD44^hi T cells but expendable for TCR-mediated proliferation of CD4⁺CD44^hi T cells.

Given the important role of Bcl10 in NF-κB activation, these data imply that NF-κB activation is not essential for CD4⁺CD44^hi T cell proliferation. It is possible that CD4⁺CD44^hi T cells are a unique population that can proliferate without the activation of not only Bcl10/NF-κB but also other important transcription factors downstream of TCR signaling such as NFAT. To examine this possibility, we determined CD4⁺CD44^hi T cell proliferation in the presence of CsA, which specifically inhibits the calcium-dependent serine/threonine phosphatase calcineurin and its substrate, the transcription factor NFAT (19). Cyclosporine A strongly inhibited anti-CD3, anti-CD3 plus anti-CD28, and anti-CD3 plus IL2-induced proliferation of wild-type and Bcl10-deficient CD4⁺CD44^hi T cells as well as the other three T cell subpopulations (Fig. 3E and data not shown). Therefore, contrary to the dispensability of Bcl10 for CD4⁺CD44^hi T cell proliferation, calcineurin and NFAT appear to be essential. Taken together, our data indicate that the substantial reduction in TCR-mediated proliferation of Bcl10-deficient T cells is due to at least two causes; 1) Bcl10 deficiency results in a reduction of the more proliferative CD4⁺CD44^hi T cell subpopulation, and 2) Bcl10 deficiency severely impairs the proliferation of naïve and CD8⁺CD44^hi T cells.

Bcl10 Is Dispensable for Cytokine Production by CD4⁺CD44^hi T Cells but Required by Naïve and CD8⁺CD44^hi T Cells—The fact that exogenous IL2 partially rescued the proliferation of Bcl10-null naive and CD8⁺CD44^hi T cells but did not have a significant effect on the proliferation of Bcl10-null CD4⁺CD44^hi T cells (Fig. 3) suggested that the differences in the proliferation of these T cell subpopulations could be at least in part due to a differential dependence on Bcl10 for IL2 production. To examine this possibility, we measured IL2 production by each of the individual T cell subsets after TCR stimulation. CD4⁺ and CD8⁺ CD44^lo or CD44^hi T cell subsets were sorted from wild-type and Bcl10-deficient mice; due to the small population of CD44^hi T cells present in the Bcl10-deficient mice in particular, we pooled spleen and lymph node cells together for the sorting. Each sorted cell population was stimulated with anti-CD3 or anti-CD3 plus anti-CD28, and IL2 production was then evaluated by ELISA.

As shown in Fig. 4A, anti-CD3 alone induced IL2 production only by wild-type memory T cells, whereas anti-CD3 plus anti-CD28 stimulated both CD44^lo and CD44^hi T cells to produce much larger quantities of IL2. This observation is consistent with the notion that naïve T cells require engagement of both the TCR and costimulatory molecules to efficiently produce IL2, whereas engagement of the TCR alone is sufficient for memory T cells to produce IL2 (20).

FIGURE 4.

Bcl10 is differentially required for cytokine production by T cell subpopulations. CD4⁺CD44^lo, CD8⁺CD44^lo, CD4⁺CD44^hi, and CD8⁺CD44^hi T cell subsets were purified from pooled splenocytes and lymph node cells of wild-type and Bcl10-deficient mice. Each T cell subset (10⁵/well) was stimulated with plate-bound anti-CD3 (5 μg/ml) in the presence or absence of anti-CD28 (2 μg/ml). The supernatants were collected 24 h thereafter, and the amounts of IL2 (A), IFNγ (B), or IL4 (C) produced were determined by ELISA. The data shown are representative of three independent experiments; error bars indicate the S.D. of the triplicate measurements of each data point.

In contrast, Bcl10-deficient CD8⁺CD44^hi T cells were not able to secrete IL2 after anti-CD3 stimulation, indicating that Bcl10 is essential for IL2 synthesis by the CD8⁺CD44^hi T cell population (Fig. 4A). Interestingly, Bcl10-deficient CD4⁺CD44^hi T cells produced levels of IL2 upon anti-CD3 stimulation that were equivalent to those generated by the corresponding wild-type T cells (Fig. 4A). These results indicate that CD4⁺CD44^hi T cells comprise a unique T cell population that is independent of Bcl10 for TCR-mediated IL2 production. Co-engagement of the CD28 molecule dramatically enhanced IL2 production by wild-type CD4⁺CD44^hi T cells but not by Bcl10-deficient CD4⁺CD44^hi T cells (Fig. 4A), indicating that Bcl10 is essential for CD28-mediated augmentation of IL2 production. Thus, the impaired proliferation of Bcl10-deficient naive and CD8⁺CD44^hi T cells after TCR stimulation is in part due to the absence of IL2 production by these cells, whereas the unique ability of CD4⁺CD44^hi T cells to proliferate independent of Bcl10 is at least partially a consequence of their ability to secrete IL2 despite the absence of Bcl10.

To further understand the differences in the properties of naïve and antigen-experienced T cells in terms of the Bcl10 dependence of their functional attributes, we next examined the production of IFNγ and IL4, which are cytokines secreted primarily by Th1 or Th2 cells, respectively, that have undergone prior stimulation and differentiation (21). As expected, wild-type CD4⁺CD44^hi and CD8⁺CD44^hi T cells were able to secrete large quantities of IFNγ in response to either anti-CD3 or anti-CD3 plus anti-CD28 stimulation. Unexpectedly, whereas IFNγ production by CD8⁺CD44^hi cells was dependent on Bcl10, the production of IFNγ by CD4⁺CD44^hi cells occurred fully independent of Bcl10 (Fig. 4B). Thus, IFNγ production also displays T cell subpopulation-specific differences with respect to the dependence upon Bcl10 of the various subsets for their functional responses. As expected, only CD4⁺CD44^hi T cells, which include the Th2 T cells, produced IL4. Surprisingly, however, in the absence of Bcl10, CD4⁺CD44^hi T cells actually produced more IL4 than the corresponding wild-type T cells (p < 0.001) (Fig. 4C).

Intracellular cytokine staining of each purified T cell subset after stimulation also revealed that the percentages of IL2-expressing T cells were equivalent in wild-type and Bcl10-deficient CD4⁺CD44^hi T cell subsets, and the percentages of IFNγ- and IL4-expressing T cells were slightly higher in Bcl10-deficient compared with wild-type CD4⁺CD44^hi T cell subsets (Fig. 5).

FIGURE 5.

Cyclosporine A inhibits cytokine production by all four subsets of T cells. CD4⁺CD44^lo, CD8⁺CD44^lo, CD4⁺CD44^hi, and CD8⁺CD44^hi T cell subsets were purified from pooled splenocytes and lymph node cells of wild-type (WT) and Bcl10-deficient mice. Each T cell subset (2 × 10⁵/well) was stimulated with plate-bound anti-CD3 (5 μg/ml) plus anti-CD28 (2 μg/ml) in the presence or absence of CsA (1 μg/ml). Thirty-six hours later, monensin (2 μm) was added, and cells were collected 6 h thereafter. The amounts of IL2 and IFNγ (A) or IL4 (B) produced were determined by intracellular staining. The data shown are representative of two independent experiments.

Therefore, both ELISA assays and intracellular cytokine staining demonstrated the unique ability of CD4⁺CD44^hi T cells to produce cytokines independent of Bcl10. Furthermore, we also showed that CsA inhibits cytokine production by CD4⁺CD44^hi T cells (Fig. 5), demonstrating that calcineurin (and likely NFAT) is crucial for cytokine production by this population of T cells, although Bcl10 is expendable. This observation is consistent with our data indicating that CD4⁺CD44^hi T cell proliferation is dependent on calcineurin/NFAT but not Bcl10 (Fig. 3E). Collectively, these data identify a unique T cell population, the CD4⁺CD44^hi T cell, that is independent of Bcl10 for TCR-mediated cytokine production and can proliferate after TCR stimulation due in part to its ability to produce IL2 despite the absence of Bcl10.

NF-κB Activation Is Defective in All Subpopulations of Bcl10-deficient T Cells—CARMA1, Bcl10, and MALT1 associate as a tripartite protein complex to bridge protein kinase Cθ and IκB kinase in TCR-mediated NF-κB activation (5). NF-κB in turn plays an essential role in IL2 gene transcription (22); for example, T cells deficient for c-Rel, one of the NF-κB family members, have impaired IL2 secretion (23). Furthermore, T cells deficient in signaling proteins that specifically lead to NF-κB activation downstream of TCR stimulation, such as protein kinase Cθ, CARMA1, Bcl10, or MALT1, all exhibit defective TCR-mediated IL2 production (1, 6–10), thus supporting an important role for the NF-κB signaling pathway in IL2 production after TCR activation. The surprising finding in our current study that TCR-induced IL2 production by CD4⁺CD44^hi T cells can occur in the absence of Bcl10 expression suggests that the NF-κB dependence of IL2 production is T cell subpopulation-specific. However, an alternative possibility is that the dependence upon Bcl10 for NF-κB activation is cell subpopulation-specific such that Bcl10 is not required for TCR-mediated NF-κB activation in CD4⁺CD44^hi T cells but is indispensable in naïve and CD8⁺CD44^hi T cells. To distinguish between these two possibilities, we examined TCR-mediated NF-κB activation in the different populations of T cells using IκBα phosphorylation and degradation as indicators of pathway activation. CD4⁺ and CD8⁺ T cells with CD44^lo or CD44^hi expression were sorted from pooled spleens and lymph nodes from wild-type and Bcl10-deficient mice; to efficiently stimulate limited numbers of T cells within a short time period, we incubated the cells with phorbol 12-myristate 13-acetate and ionomycin to mimic TCR stimulation, then examined IκBα phosphorylation and degradation, complementary readouts for NF-κB activation after T cell stimulation. Whereas all the subpopulations of wild-type T cells were able to phosphorylate IκBα and induce IκBα degradation after stimulation, indicating normal NF-κB pathway stimulation, none of the subpopulations of Bcl10-deficient T cells was capable of NF-κB activation (Fig. 6A). By contrast, ERK-1 and -2 were efficiently phosphorylated in all of the T cell subpopulations from both wild-type and Bcl10-deficient mice, indicating specific inactivation of the NF-κB pathway rather than other signaling cascades due to Bcl10 deficiency (as has been previously reported in whole, rather than subfractionated mature T cell populations) (1) (Fig. 6A). The impaired NF-κB activation in all four subpopulations of Bcl10-deficient T cells was also confirmed by electrophoretic gel mobility shift assay using anti-CD3 and anti-CD28 as stimulants (Fig. 6B). These data demonstrate that CD4⁺CD44^hi T cells are dependent upon Bcl10 for NF-κB activation, similar to the other T cell subpopulations, and confirm that CD4⁺CD44^hi T cells represent a unique T cell population that is able to secret IL2 independent of Bcl10 and NF-κB activation.

FIGURE 6.

NF-κB activation is impaired in all T cell subsets derived from Bcl10-deficient mice. A, CD4⁺CD44^lo, CD8⁺CD44^lo, CD4⁺CD44^hi, and CD8⁺CD44^hi T cell subsets were purified from pooled splenocytes and lymph node cells of wild-type and Bcl10-deficient mice. Each T cell subset (∼5 × 10⁵ cells per sample) was resuspended in phosphate-buffered saline supplemented with 2% fetal bovine serum and stimulated at 10⁶ cells/ml with phorbol 12-myristate 13-acetate (PMA; 50 ng/ml) and ionomycin (iono, 50 ng/ml) for 10 min. The stimulated cells were then collected, and total cell lysates subjected to SDS-PAGE and Western blotting. The same blot was sequentially probed with anti-phospho-IκB-α, anti-IκB-α, anti-phospho (p)-ERK, and anti-actin antibodies. The data shown are representative of two independent experiments. B, different T cell subsets were stimulated with plate-bound anti-CD3 (5 μg/ml) plus anti-CD28 (2 μg/ml) for 6 h, then subjected to electrophoretic gel mobility shift assay. The data shown are representative of two independent experiments.

DISCUSSION

Our current study demonstrates that, in contrast to CD4⁺CD44^lo and CD8⁺ CD44^lo naïve T cells and CD8⁺CD44^hi antigen-experienced T cells, CD4⁺CD44^hi antigen-experienced T cells can produce IL2 in the absence of NF-κB activation, which in turn allows their proliferation even without Bcl10/NF-κB pathway signaling. These findings are contrary to the current notion that NF-κB activation is absolutely essential for IL2 production by all T cell subsets; rather, they demonstrate for the first time that the NF-κB-dependence of TCR-mediated IL2 production is T cell subpopulation-specific. It should be noted that, in addition to NF-κB activation, Bcl10 deficiency also impairs JNK2 activation in T cells (24). Nevertheless, JNK2 deficiency does not affect IL2 production (25); therefore, the impaired IL2 production by naïve and CD8⁺CD44^hi Bcl10-deficient T cells can be attributed to impaired NF-κB activation and not to absent JNK2 function.

Whereas IFNγ production by CD8⁺ T cells is dependent on Bcl10, our studies show that IFNγ production by CD4⁺CD44^hi T cells, like IL2 production by this T cell subset, also occurs independent of Bcl10. It is not clear whether the deficient IFNγ secretion by Bcl10-null CD8⁺ T cells is due to defective NF-κB or JNK2 activation, as inactivation of either pathway has been demonstrated to impair IFNγ production (25, 26). Nonetheless, it is evident based on our experiments that CD4⁺CD44^hi and CD8⁺CD44^hi T cells are differentially dependent upon Bcl10 for IFNγ production. IL4 production by CD4⁺CD44^hi T cells is also independent of Bcl10. Interestingly, in contrast to IL2 and IFNγ, the expression of which by CD4⁺CD44^hi T cells can occur independent of Bcl10, IL4 expression by CD4⁺CD44^hi T cells actually increased dramatically in the absence of Bcl10. As noted above, Bcl10 deficiency impairs JNK2 as well as NF-κB activation in T cells (1, 24). JNK2 deficiency is known to impair the differentiation of CD4⁺ T cells into effector Th1 cells in vitro (25). Thus, it could be that the significantly elevated levels of IL4 secreted by Bcl10-deficient CD4⁺CD44^hi T cells reflects a skewed development of CD4⁺ T cells toward Th2 at the expense of Th1 cells as a consequence of impaired JNK2 activation associated with Bcl10 deficiency. This possibility is supported by the slightly increased percentage of IL4-expressing T cells in the Bcl10-deficient CD4⁺CD44^hi T cell subset compared with those in wild-type mice. However, the fact that CD4⁺CD44^hi T cells generated normal, if not higher levels of IFNγ in the absence of Bcl10 and the slightly increased percentage of IFNγ-expressing T cells in Bcl10-deficient compared with wild-type CD4⁺CD44^hi T cells argues that Th1 cell differentiation was not impaired. Thus, it is also possible that Bcl10 may negatively regulate IL4 generation by T lymphocytes via some yet-to-be defined mechanism that alters cellular signaling, not Th1 versus Th2 differentiation status. Unfortunately, our efforts to provide a more complete experimental elucidation of the reason(s) for the observed IL4 production responses have been limited by practical considerations; for example, successful in vitro studies of Th1 versus Th2 differentiation were not possible given that Bcl10 deficiency impaired the proper activation of naïve T cells, thus making it difficult to differentiate the cells in vitro. Collectively, our data demonstrate that CD4⁺CD44^hi T cells comprise a unique T cell population that is independent of Bcl10 for IL2 and IFNγ production; furthermore, the cell subpopulation-specific dependence upon Bcl10 for IL2 production is due to differential dependence of these cells on NF-κB activation.

In contrast to naïve T cells, CD4⁺CD44^hi T cells can proliferate and secrete cytokines in the absence of Bcl10 and NF-κB activation. Nevertheless, CD4⁺CD44^hi T cells are still dependent on NFAT for proliferation and cytokine production, for CsA can block both of these activities. This finding is consistent with a previous publication demonstrating that CD4⁺ memory (CD44^hi) T cells require NFAT for IL2 expression (27).

High expression of CD44 is an indication of previous T cell activation and is currently the most reliable marker for both CD4⁺ and CD8⁺ memory T cells in mice (13). Our study suggests that antigen-experienced CD4⁺ T cells are different from their naïve counterparts in terms of their requirements for further activation. It has been shown that CD4⁺ memory T cells can respond to relatively low doses of antigen stimulation (28). However, the molecular mechanism by which this occurs is not fully clear. It is possible that memory T cells express more adhesion molecules than naïve cells (28), which assists them to better interact with antigen-presenting cells and, therefore, enables them to mount similar levels of TCR signaling to weaker stimulation as naïve T cells do to strong stimuli. However, this explanation does not reconcile with previous findings that demonstrated CD4⁺ memory T cells (isolated based on CD45 isoform expression) to display less overall TCR-dependent protein tyrosine phosphorylation, inositol triphosphate generation, and calcium mobilization than naïve T cells when stimulated similarly (29, 30). Rather, our study provides an alternative explanation; CD4⁺ memory T cells can mount levels of response similar to strongly stimulated naïve cells without fully activating the entire complement of TCR-responsive signaling pathways. Thus, we posit that strong stimulation that can activate all of the signaling pathways downstream of the TCR is required for naïve T cell activation, whereas weak TCR stimulation that may not induce NF-κB function is nonetheless sufficient for CD4⁺ memory T cell activation.

The reason why antigen-experienced CD4⁺ and CD8⁺ cells differentially require Bcl10 and NF-κB activation for IL2 production remains unknown. Obviously, CD4⁺ and CD8⁺ memory T cells display many different properties (13). In addition, naïve versus antigen-experienced T cells exhibit disparate characteristics; for example, a recent study by Tewari et al. (31) demonstrated differential requirements for the Lck tyrosine kinase during primary and memory CD8⁺ T cell responses and proposed the concept that the TCR signaling apparatus may be rewired from an Lck-dependent state in naïve CD8⁺ T cells to an Lck-independent state in CD8⁺ memory T cells. Our results suggest that TCR signaling is also rewired when CD4⁺ naïve T cells differentiate to memory cells; however, the route of this rewiring appears to be at least in part different (specifically, with respect to the necessity for Bcl10 function and NF-κB activation) from that of CD8⁺ memory T cells.