CD4⁺ and CD8⁺ T Cell Survival Is Regulated Differentially by Protein Kinase Cθ, c-Rel, and Protein Kinase B¹

Both immature and mature T cells are destined to die upon Ag-specific TCR engagement. In thymocytes, TCR-specific engagement with high-affinity ligands leads to negative selection or clonal deletion (reviewed in Ref. 1). Similarly, in the absence of costimulatory and proinflammatory signals, mature T cells also undergo clonal deletion upon engagement with high-affinity ligands (2). Therefore, one fundamental issue in T cell immunity is to understand the nature of survival signals and the integration of these signals during T cell activation. Although a number of molecules, such as Bcl-x_L and FLIP, have been shown to play a role in T cell survival (3, 4, 5, 6, 7), the signaling pathways that mediate T cell survival remain unclear.

The serine/threonine kinase protein kinase B (PKB)α³ (Akt1) has been shown to promote survival in multiple cell lineages, including T cells (PKB reviewed in Ref. 8). PKB is activated after the generation of lipid second messengers at the plasma membrane. The conversion of phosphatidylinositol-3,4-bisphosphate into phosphatidylinositol-3,4,5-trisphosphate by PI3K prompts the recruitment of PKB to phosphatidylinositol-3,4,5-trisphosphate via its pleckstrin homology domain. Once recruited to the plasma membrane, PKB is subsequently phosphorylated and fully activated (9). Active PKB has been demonstrated to promote survival due to its ability to phosphorylate, and thereby inhibit, the proapoptotic BH3-only protein BAD (10, 11, 12). As well, the forkhead transcription factors are also direct substrates of PKB (13, 14, 15); phosphorylation of the forkhead transcription factors FOXO by PKB prevents the transcription of the proapoptotic BH3-only protein Bim (16), as well as the transcription of Fas ligand (13). Many other molecules such as GSK-3, Cot, and p27 have been shown to be downstream targets of PKBα (17, 18, 19, 20). In addition to its direct substrates, PKB also promotes cell survival through the activation of the NF-κB family of transcription factors (21, 22, 23, 24).

In mammals, the NF-κB transcription factors are composed of homo- or heterodimers of five conserved subunits: p65/RelA, c-Rel, RelB, NF-κB1/p50, and NF-κB2/p52 (NF-κB reviewed in Ref. 25). These dimeric transcription factors are essential for T cell activation and regulate genes associated with proliferation, cytokine production, as well as protection against apoptosis. In T cells, NF-κB signaling has been linked not only to the induction of antiapoptotic members of the Bcl-2 family of proteins, such as Bcl-x_L (26, 27, 28), but also to protection from Fas-mediated apoptosis (29, 30, 31). Initiation of NF-κB-mediated transcription requires the activation of the IκB kinase (IKK) complex. This complex consists of two kinase subunits, IKKα and IKKβ, as well as a regulatory subunit IKKγ/NF-κB essential modulator. Activation of the kinases in this complex results in the phosphorylation and subsequent proteosomal degradation of the IκB protein. The IκB protein functions by sequestering NF-κB transcription factors in the cytosol. Degradation of IκB allows the nuclear translocation of the NF-κB transcription factor and gene transcription.

Protein kinase C (PKC)θ, a member of the calcium-independent subfamily of PKCs (PKCθ reviewed in Ref. 32), has been demonstrated to be an important mediator of NF-κB signaling. T lymphocytes from mice deficient for PKCθ demonstrate not only impaired activation of NF-κB, but also reduced activation of the AP-1 transcription factor (33, 34). The functional consequence of this impaired signaling is that PKCθ^−/− T cells displayed impaired proliferation (35) and reduced production of IL-2 (33) upon activation. Recently, a role for PKCθ in the promotion of survival of activated T cells has been described (36). Moreover, it has been suggested previously that PKB and PKCθ molecules physically associate (37, 38). Together, these reports suggest interplay between the PKB and PKCθ signaling pathways in the promotion of survival of activated T cells. Therefore, we investigated the roles of PKCθ, PKB, and members of the NF-κB family in controlling the survival of activated primary murine T cells. Our results demonstrate that CD4⁺ and CD8⁺ T cell survival are programmed differentially. PKCθ signaling events are important for both CD4⁺ and CD8⁺ T cell survival, which can be modulated by PKB or IL-2. PKCθ- and c-Rel-mediated signals, however, appear to be more important for CD8⁺ T cell survival.

The PKCθ^−/− (33), nf-κb1^−/− (p50^−/−) (39), and the c-rel^−/− (40) mouse lines have all been described previously, as have the human CD2-gag-PKB transgenic (B6/PKB) (24), and the P14 TCR transgenic (41). The PKB/PKCθ^−/− line of mice was generated by breeding B6/PKB mice to PKC-deficient mice. These mice were subsequently crossed to the P14 TCR transgenic mice. C57BL/6J mice were obtained from The Jackson Laboratory. All mice were maintained in a specific pathogen-free environment at the Ontario Cancer Institute, according to institutional guidelines.

Murine rIL-2 was purchased from PeproTech. Purified anti-CD3 (2C11) and anti-CD28 (37.51) were purchased from eBioscience. Anti-Bcl-x_L FITC Ab was purchased from Southern Biotechnology Associates. Purified anti-FcγRIII/II (2.4G2) was purchased from BD Pharmingen. The following Abs purchased from BD Pharmingen were used for flow cytometry (PE conjugated): anti-CD4 and anti-CD8. Th annexin V FITC was also purchased from BD Pharmingen, whereas the 7-aminoactinomycin D (7AAD) was purchased from Sigma-Aldrich. The mouse IgG FITC isotype control Ab was purchased from Jackson ImmunoResearch Laboratories.

To assay Bcl-x_L expression, splenic T cells were purified using the Pan T Cell Isolation Kit on the AutoMACS (Miltenyi Biotec), as per the manufacturer’s instructions. Purified T cells were plated at 1–2 × 10⁶ cells/ml in 24-well flat-bottom plates coated with either purified anti-CD3 or combined anti-CD3 and anti-CD28. After the indicated times, cells were harvested for analysis.

A total of 10⁷ splenocytes was plated at 2 × 10⁶ cells/ml in six-well flat-bottom plates coated with 2 μg/ml anti-CD3 and 2 μg/ml anti-CD28. Cells were stimulated for 72 h in RPMI 1640 medium supplemented with 10% heat-inactivated FCS (Invitrogen Life Technologies), 50 μM 2-ME, 2 mM glutamine, and 0.1% penicillin/streptomycin. After 72 h of stimulation, cells were washed and cell viability was assessed in the CD4 and CD8 compartments from a triplicate of aliquots using flow cytometry (viability at t₀; t stands for time). A total of 2 × 10⁵ cells was then plated in triplicate wells in 96-well flat-bottom plates. After 20 h of incubation, cell viability was assessed again in the CD4 and CD8 compartments using flow cytometry. Cell viability is reported as a percentage of the viability of each culture at t₀, using the formula: percentage viability = ((viability after 20 h)/(viability at t₀)) × 100.

For cell viability assays, 2 × 10⁵ cells were first stained with either anti-CD4 or anti-CD8 at 4°C. Cell viability was then assessed by staining with annexin V FITC and 7AAD. Samples were acquired on a FACScan instrument (BD Biosciences) and were analyzed using CellQuest software (BD Biosciences). For intracellular Bcl-x_L staining, FcγRIII were first blocked with an Ab. A total of 1 × 10⁶ cells was then fixed and permeabilized using BD Cytofix and BD Cytoperm (BD Biosciences) as per the manufacturer’s instructions. Cells were then stained with anti-Bcl-x_L FITC Ab or an isotype control Ab. Samples were then acquired and analyzed.

Single-cell suspensions were lysed by incubation on ice in Gentle Soft Buffer (10 mM NaCl, 20 mM PIPES (pH 7.4), 0.5% Nonidet P-40, 5 mM EDTA, 5 μg/ml leupeptin, 1 mM benzamidine, aprotinin, and 100 μM Na₃VO₄) for 20 min. The lysates were then cleared by centrifugation, and the supernatants were normalized for total protein (Bio-Rad). Protein was resolved by 4–20% SDS-PAGE and then electroblotted onto polyvinylidene difluoride membrane (Costar). The membranes were blocked in 5% nonfat milk in TBST and probed with primary Abs for 12 h at 4°C. Anti-Bcl-x_L was purchased from Cell Signaling Technology, and anti-actin was purchased from Sigma-Aldrich.

Splenocytes from P14 PKCθ^−/−, P14 PKCθ^−/− PKB, and P14 PKCθ^+/− mice (2 × 10⁵ cells/well) were incubated 72 h at 37°C in 96-well flat-bottom plates with increasing concentrations of gp33 peptide. Proliferation was measured by the incorporation of an overnight pulse of 1 μCi of [³H]thymidine after 72 h of incubation.

To examine whether PKCθ conveys survival signals in activated T cells, T lymphocytes from PKCθ^−/−, littermate PKCθ^+/−, and control C57BL/6 mice were stimulated with plate-bound anti-CD3 and anti-CD28 Abs. After stimulation, cells were incubated for 20 h without stimulation, and then viability was assessed by flow cytometry using annexin V and 7AAD. Cell survival was monitored in both the CD4⁺ and the CD8⁺ T cell subsets and was normalized in each culture to the percentage of viable cells after removal from the stimulation (viability at t₀). As compared with the C57BL/6 and littermate heterozygous control cells, which displayed similar viability in both the CD4⁺ and the CD8⁺ T cell compartments, the CD4⁺ T cells lacking PKCθ displayed a decrease in cell viability after 20 h of culture (Fig. 1 A). The average viability of the PKCθ-deficient CD4⁺ T cells was 36.7% as compared with 52.1% for the C57BL/6 cells and 55.2% for the PKCθ^+/− control cells. This difference in viability between the C57BL/6 and knockout CD4⁺ T cells is statically significant (p < 0.0003 by two-tailed Student’s t test). The impaired survival in PKCθ^−/− T cells, however, was much more drastic in the CD8⁺ compartment, because the PKCθ-deficient cells averaged only 12.8% viable cells as compared with 59.2% viability for the PKCθ^+/− CD8⁺ T cells and 64.8% for the C57BL/6 control cells. These results are consistent with a recent report that suggests PKCθ is important for CD8⁺ T cell survival (36). Interestingly, we found no enhanced cell death in PKCθ^−/− CD4⁺ or CD8⁺ T cells that were subjected to activation-induced cell death via TCR restimulation (data not shown). Thus, PKCθ may not be required to protect against all forms of cell death. Our data, however, demonstrate that PKCθ is required to maintain the survival of activated T lymphocytes, particularly those of the CD8⁺ lineage.

In the next experiments, we sought to ascertain the molecules downstream of PKCθ that are responsible for promoting cell survival. PKCθ is known to activate NF-κB (33, 34, 42, 43). We investigated whether T cells lacking only p50 or c-Rel displayed a similar phenotype as T cells lacking PKCθ. When cell viability was assessed in CD4⁺ T cells lacking either p50 or c-Rel, no impairment in survival was noted relative to wild-type controls (Fig. 1 B). This suggests that there might be redundancy between the various subunits of NF-κB that the absence of one is insufficient to impair the survival of CD4⁺ T cells. This does not, however, appear to be the case for CD8⁺ T cells. The CD8⁺ T cells lacking c-Rel displayed a similar phenotype to CD8⁺ T lymphocytes lacking PKCθ. The CD8⁺ c-Rel^−/− T cells averaged only 24.8% viability as compared with 71.7% viability for the C57BL/6 cells. These findings implicate the PKCθ to c-Rel signaling axis as a major survival signaling pathway in CTL.

The specific requirement for the c-Rel subunit of NF-κB downstream of PKCθ for CD8⁺ T cell survival was demonstrated by the lack of impaired survival in the T cells from the p50^−/− mice. The CD8⁺ T cells from these mice displayed a similar profile as the T lymphocytes from the C57BL/6 mice (Fig. 1 B). The CD8⁺ p50^−/− T cells had a marginally impaired average survival relative to the control cells (59.4 vs 71.7%). This difference, however, was not statistically significant (p > 0.1 by two-tailed Student’s t test). These results demonstrate that, unlike c-Rel, the p50 subunit of NF-κB is dispensable for the PKCθ-mediated survival signaling pathway in activated CD8⁺ T lymphocytes.

There exists strong evidence connecting NF-κB with the expression of Bcl-x_L. Mapping of the bcl-x promoter in mouse revealed multiple κB consensus binding sites (44, 45). As well, studies in T cell lines (26) and primary human T cells (27) demonstrated a requirement for NF-κB in the expression of Bcl-x_L. Furthermore, the expression of pattern of Bcl-x_L supports a role in the survival of activated T cells. Naive T cells express relatively high levels of Bcl-2 (46) and almost undetectable levels of Bcl-x_L (47). Upon activation, however, the expression of Bcl-x_L is induced rapidly in T lymphocytes, whereas the expression of Bcl-2 does not change as significantly (4, 47). All of these findings implicate Bcl-x_L as an attractive candidate for a prosurvival molecule regulated by PKCθ and c-Rel. To investigate this possibility, the expression of Bcl-x_L was assessed in T cells lacking PKCθ.

Purified splenic T cells were stimulated by plate-bound Abs to CD3 and CD28 for 24 h and then stained with either anti-CD4 or anti-CD8 as well as for intracellular Bcl-x_L. As illustrated in Fig. 2, CD4⁺ T cells lacking PKCθ displayed reduced expression of Bcl-x_L relative to both the C57BL/6 and the littermate PKCθ^+/− controls. PKCθ^−/− CD8⁺ lymphocytes showed a minor reduction in Bcl-x_L levels. These results demonstrate that CD4⁺ and CD8⁺ T cells lacking PKCθ display impaired expression of Bcl-x_L during the process of T cell activation.

The kinetics of Bcl-x_L expression during T cell activation was also assessed by flow cytometry in both CD4⁺ and CD8⁺ T cells. As can be seen in Fig. 3, stimulation of T lymphocytes from C57BL/6 mice with anti-CD3 Ab alone resulted in ∼2-fold increase in expression of Bcl-x_L after 16 h of stimulation. The level of Bcl-x_L expression increased further after 24 h of stimulation with anti-CD3 in both the CD4⁺ and CD8⁺ T cells. Combined anti-CD3 and anti-CD28 Ab treatment of the wild-type T cells enhanced the expression of Bcl-x_L at all time points assayed in both the CD4⁺ and CD8⁺ T lymphocytes. These data are consistent with previous studies that demonstrated that signals through CD28 augment the expression of Bcl-x_L (4, 48, 49). In contrast to the control cells, both CD4⁺ and CD8⁺ T cells from the PKCθ^−/− mice showed a reduced proportion of cells expressing Bcl-x_L at high levels when stimulated through the Ag receptor alone or with anti-CD3 and anti-CD28 Abs. This difference was most prominent after 24 h. These findings suggest that PKCθ is required downstream of the Ag receptor for optimal expression of Bcl-x_L upon T cell activation.

Our results from Fig. 1 suggest that signals downstream of PKCθ and c-Rel, but not p50, are important for the survival of activated CD8⁺ T cells. If Bcl-x_L is indeed a key survival effector downstream of c-Rel, one prediction would be that the expression of Bcl-x_L should also be impaired in the c-Rel^−/−, but not the p50^−/− T cells. Accordingly, the expression of Bcl-x_L in T cells from c-Rel^−/− and p50^−/− mice was assessed by flow cytometry. Intracellular Bcl-x_L staining confirmed the specific role of c-Rel in the expression of Bcl-x_L (Fig. 4). The mean fluorescence intensity of Bcl-x_L staining after 24 h of stimulation with anti-CD3 Ab was 16.4 for the C57BL/6 T cells, 14.86 for the p50^−/− T cells, and only 8.9 for the c-Rel^−/− T lymphocytes. The MFI of Bcl-x_L staining after stimulation through CD3 and CD28 further confirmed these findings, because the MFI of Bcl-x_L staining was 26.72 for the wild-type T cells, 27.31 for the p50^−/− T cells, and 10.86 for the c-Rel^−/− T cells. Thus, our data obtained by intracellular staining indicate that the c-Rel, but not the p50 subunit of NF-κB plays a nondispensable role in the up-regulation of Bcl-x_L in activated T lymphocytes.

Multiple studies (21, 22, 23, 24, 50) have demonstrated that PKB can enhance the activation of NF-κB. Studies by Kane et al. (23, 50) have suggested that PKB exerts a greater affect upon the transcription of NF-κB-dependent genes, such as IL-2, when it is activated in conjunction with other signaling pathways, particularly PKCθ. Accordingly, we wanted to establish whether transgenic expression of active PKB could rescue the expression of Bcl-x_L in T cells lacking PKCθ, or whether alternatively, PKB requires the presence of PKCθ to affect Bcl-x_L expression. Our laboratory has previously described a transgenic model that expresses active PKB in T cells using the CD2 promoter. We have characterized T cells in these mice and have shown that they have increased survival to multiple stimuli, enhanced NF-κB activation, and increased levels of Bcl-x_L (24). To examine whether PKB influences Bcl-x_L expression in the absence of PKCθ, the human CD2-gag-PKBα transgenic mice were bred onto the PKCθ-null background. The expression of Bcl-x_L in activated T cells was then assessed by Western blot analysis.

Active PKB rescued the expression of Bcl-x_L in the absence of PKCθ. As can be seen in Fig. 5 A, the expression of Bcl-x_L in activated PKCθ^−/− PKB T cells was higher than that of the PKCθ^−/− T cells after both 16 and 24 h of stimulation. The level of Bcl-x_L expression in the PKCθ^−/− PKB T cells, however, was slightly reduced as compared with the amount of Bcl-x_L expressed by the C57BL/6 control cells. Thus, although transgenic expression of PKB could rescue the expression of Bcl-x_L in cells lacking PKCθ, this rescue did not appear to be complete.

Given that we had found a difference in the expression of Bcl-x_L in CD4⁺ and CD8⁺ cells (Fig. 2), we next examined whether PKB had a differential ability to restore Bcl-x_L in CD4⁺ and CD8⁺ T cells that lacked PKCθ. Interestingly, as can be seen in Fig. 5 B, active PKB was unable to rescue the expression of Bcl-x_L in CD4⁺ T cells, because the average MFI of Bcl-x_L staining in the PKCθ^−/− PKB CD4⁺ T lymphocytes was comparable to that of the PKCθ^−/− T cells. Conversely, the average MFI of Bcl-x_L staining in the PKCθ^−/− PKB CD8⁺ T cells was 15.6 as compared with 10.0 for the PKCθ^−/− CD8⁺ T lymphocytes. This difference is statistically significant (p < 0.0006 by two-tailed Student’s t test) and demonstrates that active PKB can rescue Bcl-x_L expression in CD8⁺ T cells lacking PKCθ. This rescue of Bcl-x_L expression, however, is not total, because the MFI of Bcl-x_L staining in the control C57BL/6 CD8⁺ T was 22.4. These data demonstrate that in CD8⁺ T cells, PKB can still promote the expression of Bcl-x_L in the absence of PKCθ. This suggests that there exists a signaling pathway that links PKB to Bcl-x_L expression that does not require PKCθ in CTL.

Next, we examined whether PKB expression was sufficient to rescue cell survival. Like in the previous survival experiments, the viability of in vitro-activated T cells from C57BL/6, PKCθ^−/−, and PKCθ^−/− PKB mice was assessed by flow cytometry using annexin V and 7AAD staining 20 h after their removal from plate-bound Ab stimulation. Fig. 6,A shows the percentage of viable CD4⁺ and CD8⁺ cells in each culture after their removal from stimulation (viability at t₀) as well as after 20 h of reculture. These data show that PKB was able to rescue partially the survival of PKCθ^−/− CD4⁺ T cells. When the survival in each culture was normalized to the viability at t₀ (data not shown), the average viability of the PKCθ^−/− CD4⁺ T cells was 30.0 vs 49.1% for the PKCθ^−/− PKB CD4⁺ T cells. This difference is statistically significant (p < 0.003 by two-tailed Student’s t test) and demonstrates that the PKB transgene can affect cell viability. The effect is through a pathway that does not alter Bcl-x_L expression, because we demonstrated active PKB does not restore Bcl-x_L expression in CD4⁺ T cells (Fig. 5). This rescue of cell survival, however, is not complete. The PKCθ^−/− PKB CD4⁺ T cells were not as viable after 20 h as the C57BL/6 CD4⁺ T cells, which had an average of 69.5% live cells in culture relative to their viability at t₀. Thus, active PKB expression only partially restored T cell survival in PKCθ^−/− CD4⁺ T cells through a Bcl-x_L-independent pathway.

In CD8⁺ cells, the PKB transgene had a reduced effect on cell viability. The PKCθ^−/− CD8⁺ T cells averaged 4.2% living cells in culture vs 10.6% for the PKCθ^−/− PKB CD8⁺ T cells. Although this difference is statistically significant (p < 0.02 by two-tailed Student’s t test), the average viability of the PKCθ^−/− PKB CD8⁺ T cells was still reduced greatly as compared with the C57BL/6 CD8⁺ T cell cultures. The control cell cultures had an average of 61.9% viable cells after 20 h relative to the viability at t₀. Thus, although PKB could rescue Bcl-x_L expression in PKCθ^−/− CD8⁺ T cells (Fig. 5), it was unable to markedly improve the cell survival of activated PKCθ^−/− CD8⁺ T cells in vitro. These findings further emphasize the different mechanisms that modulate CD4⁺ and CD8⁺ T cell survival.

T cells lacking PKCθ display poor proliferative responses in vitro when stimulated with Abs (33, 34) or cognate peptide Ag (35). We also wanted to determine whether active PKB could provide appropriate signals to restore the proliferative defect in vitro. To do these experiments, P14 transgenic mice expressing a TCR specific for the lymphocytic choriomeningitis virus glycoprotein peptide, gp33, presented in the context of H-2D^b were bred with PKCθ^−/− and PKCθ^−/− PKB mice. Spleen cells from P14 PKCθ^+/−, P14 PKCθ^−/−, and P14 PKCθ^−/− PKB mice were then cultured with varying concentrations of the gp33 peptide and proliferation was measured by [³H]thymidine incorporation at different time points. Fig. 6 B shows that the PKB transgene could not rescue the proliferation of the PKCθ^−/− T cells to a level comparable to the PKCθ^+/− after 72 h of stimulation. These findings demonstrate that neither the restoration of Bcl-x_L expression nor the potential alternative signals provided by PKB is sufficient to restore the proliferative response to Ag in PKCθ^−/− T cells in vitro.

Because the expression of IL-2 has been linked to PKCθ signals (33) and to the c-Rel transcription factor (40, 51), we tested whether the addition of exogenous IL-2 was able to promote the survival of PKCθ^−/− and c-Rel^−/− T cells. To do this, in vitro-activated splenic T cells from PKCθ^−/−, c-Rel^−/−, and C57BL/6 mice were cultured for 20 h in the presence or absence of 10 U/ml IL-2. Cell viability was then assessed by flow cytometry using annexin V and 7AAD staining. As can be seen in Fig. 7, the addition of exogenous IL-2 was able to restore the survival of PKCθ-deficient CD4⁺ T cells. The average survival of the PKCθ^−/− CD4⁺ T cells in the presence of IL-2 was 72.0%. Although this was slightly lower than the average survival of the control CD4⁺ T cells, it represents a much more substantial rescue of cell viability than was provided by the PKB transgene (Fig. 6 A). Similarly, in the case of the CD8⁺ lymphocytes, IL-2 had a profound effect upon survival. The addition of IL-2 was able to rescue fully the survival defect in both the c-Rel^−/− and PKCθ^−/− CD8⁺ T cells. The average survival of the c-Rel^−/− T lymphocytes was actually slightly enhanced compared with the average survival of the wild-type CD8⁺ T cells cultured with IL-2. Collectively, these data demonstrate that unlike PKB, IL-2 can restore fully the survival of T cells lacking PKCθ or c-Rel.

Because signaling through the IL-2R has been linked to the expression of Bcl-x_L (52, 53), we wanted to ensure that the addition of exogenous IL-2 was not altering the expression of Bcl-x_L in activated c-Rel^−/− and PKCθ^−/− T cells. To do this, purified splenic T cells from c-Rel^−/− and PKCθ^−/− mice were stimulated for 24 h with plate-bound Ab in the presence or absence of IL-2. After 24 h, the expression of Bcl-x_L was assessed using flow cytometry. The addition of either 10 or 50 U/ml IL-2 did not alter the expression of Bcl-x_L in either the activated c-Rel^−/− or PKCθ^−/− T cells as compared with activation with Abs alone (data not shown). These data suggest that the survival-promoting effects of exogenous IL-2 we observed in the c-Rel^−/− and PKCθ^−/− T lymphocytes were independent of Bcl-x_L expression.

The signaling pathways that govern activated T cell survival have yet to be elucidated fully. This is the first study that directly compares the role of PKCθ, c-Rel, and PKB in the survival of CD4⁺ and CD8⁺ T cells. Our study has shown that both CD4⁺ and CD8⁺ T lymphocytes require PKCθ-derived signals to survive (Fig. 1,A). In the absence of PKCθ, however, CD4⁺ T cell survival can be restored partially by PKB (Fig. 6), whereas PKB does not have a major impact upon the survival of CD8⁺ T cells. These data demonstrate a different requirement for PKB-mediated survival signals in CD4⁺ and CD8⁺ T cells. These findings also indicate that PKCθ and PKB signal independently of each other to promote CD4⁺ T cell survival. The importance of active PKB for the survival of CD4⁺ T lymphocytes is supported by studies by Song et al. (54). These authors demonstrated that signals activating PKB are important in maintaining the long-term survival of CD4⁺ T cells in vivo (55).

Although our results indicate the PKCθ is required for the survival of both CD4⁺ and CD8⁺ T lymphocytes, our data also indicate that PKCθ delivers a more vital survival signal in activated CD8⁺ T lymphocytes than in CD4⁺ T cells. CD8⁺ cells lacking PKCθ display a much more striking impairment in survival than do CD4⁺ PKCθ^−/− T cells (Fig. 1,A). The importance of PKCθ in CD8⁺ T cell survival in vivo has been illustrated previously by work by Barouch-Bentov et al. (36). Our data confirm the findings of Barouch-Bentov et al. (36) and further suggest that c-Rel is required to relay PKCθ-mediated survival signals in activated CD8⁺ T cells, but not in CD4⁺ T cells (Fig. 1 B). This finding is consistent with a previous report that has linked PKCθ to the mobilization of c-Rel (56). It is also consistent with studies that have demonstrated a more profound impact upon activated CD8⁺ T cell survival than activated CD4⁺ T cell survival in mice with impaired ability to activate NF-κB (30, 57). The target genes of c-Rel that are important for the survival of CD8⁺ T lymphocytes remain to be elucidated fully.

A multitude of studies have suggested that Bcl-x_L is the key prosurvival molecule of activated T cells (4, 47, 58). Given that our results indicate that the expression of Bcl-x_L is affected by both PKCθ (Figs. 2 and 3) and c-Rel (Fig. 4), we suspected that the diminished survival of CD8⁺ T cells we observed (Fig. 1) was due to an impaired ability to express Bcl-x_L. Thus, we were surprised to find that rescuing Bcl-x_L expression with the PKB transgene in cytotoxic lymphocytes did not significantly alter their survival (Fig. 6). These data question the importance of Bcl-x_L expression downstream of PKCθ and c-Rel for the survival of activated CD8⁺ T lymphocytes. Consistent with these findings, Zhang and He (59) recently reported that Bcl-x_L is dispensable for the survival of CD8⁺ T cells. Our results would agree with these data and suggest that perhaps IL-2 is the important c-Rel target gene downstream of PKCθ for the survival of CD8⁺ T lymphocytes (Fig. 7). The importance of IL-2 gene expression downstream of PKCθ in supporting the survival of CD8⁺ T cells, however, may be dependent upon the context of T cell activation. Work in our own laboratory (35) and that of others (60) has demonstrated that the absence of the entire PKCθ signaling pathway in CD8⁺ T cells can be overcome via the provision of strong pathogen-derived signals in vivo. PKCθ-mediated survival signals, however, may be more important when pathogen-mediated signals are absent, such as in the context of autoimmunity or during an antitumor response.

Previous work by Zheng et al. (61) demonstrated that CD4⁺ T cells lacking both p50 and c-Rel did not proliferate when stimulated. These cells failed to induce the expression of Bcl-x_L as well as demonstrating impaired c-Myc expression upon stimulation. Thus, these authors concluded that the reasons for failed proliferation were 2-fold. Not only were the cells unable to induce the genes required for survival, but they also did not initiate the appropriate genetic program required to enter cell cycle. Our results support the conclusion that restoring Bcl-x_L expression alone is insufficient to rescue proliferation, because returning the expression of Bcl-x_L in PKCθ^−/− CD8⁺ T cells was unable to enhance proliferation (Fig. 6 B). Similar conclusions have been reached in other systems. In B lymphocytes lacking c-Rel, expression of Bcl-x_L was not sufficient to restore proliferation. Expression of cylin E in conjunction with Bcl-x_L was required to restore proliferation (62). As well, in T cells, retroviral expression of Bcl-x_L in T cells lacking CD28 did not rescue proliferation (63). Collectively, these findings indicate that expression of Bcl-x_L alone is insufficient to promote proliferation. Moreover, because we used the active PKB transgene to increase Bcl-x_L expression, our results also suggest that in the absence of PKCθ, signals downstream of PKB are unable to affect proliferation. This result is interesting because PKB has many downstream effector molecules and has been demonstrated previously to promote proliferation, in part via the phosphorylation of the cell cycle inhibitor p27 (19, 20).

In this study, we have demonstrated that c-Rel is able to provide an important survival signal in CD8⁺ T cells. Recent work in our laboratory has established that p50 is required for protection against Fas-mediated apoptosis, whereas c-Rel is dispensable for this protection (31); now, we have demonstrated a survival signaling pathway engaged during T cell activation in which the converse is true: c-Rel is required, whereas p50 is dispensable. This pattern also holds true for the expression of Bcl-x_L. We demonstrate c-Rel is involved in the expression of Bcl-x_L upon the activation of T cells, whereas p50 is not required (Fig. 4). These data indicate that the signaling cascade leading to the expression of Bcl-x_L in T cells is similar to the one described in B cells.

In B lymphocytes, the PKCβ isoform is required to activate the IKK complex and initiate the expression of Bcl-x_L (64, 65). As well, c-Rel has also been demonstrated to be essential for the expression of Bcl-x_L in B lymphocytes stimulated through the BCR or CD40 (66, 67). Our data, and those from the B cell studies, indicate that a signaling paradigm linking PKC isoforms to Bcl-x_L through the induction of c-Rel may be conserved in cells of lymphoid, and potentially, all of the hemopoietic lineages. This model would fit with the expression of c-Rel, because its expression is confined to the hemopoietic compartment (68, 69).

In summary, this work demonstrates that CD4⁺ and CD8⁺ T cell survival is differentially regulated and is not strictly governed by Bcl-x_L expression. In CD4⁺ T lymphocytes, PKCθ provides a subset of survival signals that can be augmented by PKB and restored by IL-2. In contrast, PKCθ and c-Rel are vital to CD8⁺ T cell survival, whereas PKB plays a limited role in the survival of CTL. Delineation of all of the differences in the CD4⁺ vs CD8⁺ survival programs requires further investigation, but may reveal potential subset-specific therapeutic targets.

We thank Rosa Pileggi and Sandra McGugan for administrative help.

The authors have no financial conflict of interest.