PKCα and PKCβ cooperate functionally in CD3-induced de novo IL-2 mRNA transcription.

The physiological functions of PKCα and PKCθ isotypes downstream of the antigen receptor have been defined in CD3(+) T cells. In contrast, no function of the second conventional PKC member, PKCβ, has been described yet in T cell antigen receptor signalling. To investigate the hypothesis that both conventional PKCα and PKCβ isotypes may have overlapping functions in T cell activation signalling, we generated mice that lacked the genes for both isotypes. We found that PKCα(-/-)/β(-/-) animals are viable, live normal life spans and display normal T cell development. However, these animals possess additive defects in T cell responses in comparison to animals that carry single mutations in these genes. Our studies demonstrate that the activities of PKCα and PKCβ converge to regulate IL-2 cytokine responses in anti-CD3 stimulated primary mouse T cells. Here, we present genetic evidence that PKCα and PKCβ cooperate in IL-2 transcriptional transactivation in primary mouse T cells independently of the actions of PKCθ.

Introduction

Protein kinase C (PKC) comprises a closely related family of serine/threonine protein kinases that were originally discovered by Nishizuka and colleagues in 1977 [1]. The PKC family consists of nine isotypes that are expressed in a wide range of cell types and tissues [2]. According to both the structures of the variable regulatory domains and the cofactor dependencies, PKCs are classified into the following categories: conventional PKCs (cPKCs; α, β, γ), novel PKCs (nPKCs; ɛ, δ, θ, η) and atypical PKCs (aPKCs; ζ, ι) [3]. cPKCs require Ca2+ and diacylglycerol (DAG) for activation; however, nPKCs are Ca2+ independent and aPKCs do not require Ca2+ or DAG for their activation. In T cells, all isotypes except PKCγ are expressed. The expression of multiple PKC isotypes in a given cell suggests functional redundancy. Therefore, the pharmaceutical industry is focused upon the generation of low molecular weight inhibitors (LMWIs) that specifically inhibit selected PKC isotypes or combinations of multiple PKCs in order to suppress T cell-dependent immune responses that are concomitant with organ transplantation or autoimmune diseases [4,5].

In gene knockout experiments, two isotypes of PKC, PKCθ and PKCα, were shown to have physiological and non-redundant roles in antigen receptor-mediated signalling in primary CD3+ T lymphocytes. Upon T cell engagement, PKCθ is recruited to the immunological synapse [6,7]. PKCθ was shown to function downstream of the antigen receptor as a critical signal strength regulator for the transactivation of the NFκB, AP1, and NFAT transcription factors [8,9]. PKCθ appears to be required particularly for the development of robust immune responses that are controlled by both Th2 [10,11] and Th17 [12,13] cells.

By contrast, PKCα was shown to be essential for Th1 cell-dependent immune responses [14]. PKCα is highly expressed in T lymphocytes and translocates to the plasma membrane following T cell activation [15,16]. Earlier studies, in which transgenic mice that overexpressed PKCα were used, defined an important role for this isotype in T cell function [17]. Using the established PKCα knockout mice [18], our group studied the physiological and non-redundant functions of PKCα in T cell signalling. PKCα−/− mice are born in expected Mendelian ratios, are fertile, appear to be healthy, are anatomically normal and have normal life spans compared to wild type littermates. The development of CD3+ T cells in these mice is normal, as are the subset distributions (CD4+/CD8+ ratio). However, PKCα-deficient CD3+ T cells showed impaired proliferation upon T cell receptor (TCR) stimulation, while IL-2 production and IL-2 signalling were only slightly impaired. PKCα was shown to be essential for T cell dependent IFN-γ production and IgG2a/2b antibody responses in vivo [14].

The functions of PKCβ are well characterised in both the immune receptor and insulin receptor signalling pathways [19,20]. PKCβ knockout mice, which are also viable, fertile and born in the expected Mendelian frequencies, demonstrate deficient B cell development, reduced numbers of B1 lymphocytes and impaired humoral immune responses [21]. In T cells, the role of PKCβ in migration, which is mediated by the β2 integrin receptor LFA-1, has been described [22,23]. Additionally, a role for PKCβ in TCR induced signalling, IL-2 expression and IL-2 secretion was observed in T cell lines [24,25]. However, our group was the first to investigate the role of PKCβ in primary mature CD3+ T cells, and in contrast to the results that were obtained with different cell lines, we showed that PKCβ is not essential for either the proliferation of or IL-2 secretion in T cells, even though this PKC isotype is highly expressed in T cells [26]. These data suggest that other PKC isotypes may compensate for PKCβ deficiency, which results in normal biological T cell responses.

These data prompted us to generate a line of PKCα/β double knockout mice to investigate the potentially overlapping roles of these two cPKC isotypes in CD3-dependent activation processes. The phenotypic characterisation of T cell-dependent immune responses of PKCα/β mice, in comparison to mice that carry single deficiencies and to wild type controls, revealed that PKCα and PKCβ cooperate functionally to mediate IL-2 gene expression.

Materials and methods

Generation and genotyping of PKCα−/−β−/− double knockout mice: C57BL/6 J PKCα−/−β−/− mice were generated by crossing PKCα−/− [18] with PKCβ−/− mice [21]. Tail DNA was extracted and the PKCα and PKCβ alleles were genotyped by PCR, using the following primers: for the alpha allele, PKCα3′ (5′-CCT GGT GGC AAT GGG TGA TCT ACA C-3′) and PKCα5′ (5′-GAG CCC TTG GGT TTC AAG TAT AGA-3′), which yielded either a 600 bp wild type fragment or a 1.7 kb mutant fragment; for the beta allele, PKCβ1 (5′-CAG GGT CGA ATT GCC ATC CTC CA-3′), PKCβ2 (5′-CCC CAC CCC CTC CTT CTT CCT-3′) and MO13 (5′-CTT GGG TGG AGA GGC TAT TC-3′), which yielded a 900 bp wild type fragment and a 1.3 kb mutant fragment. For all experiments, mice were aged 6–10 weeks. All mice were housed under SPF conditions at the mouse facility of the Medical University of Innsbruck. All experiments complied with the current laws of Austria.

FACS analysis of cell subsets and cell surface activation markers

Single-cell suspensions from freshly isolated thymi, spleens and lymph nodes were incubated on ice in staining buffer (phosphate-buffered saline containing 2% foetal calf serum), with FITC-, PE- or APC-conjugated antibodies to identify T cell subsets. CD3, CD19, CD4, CD8 antibodies were obtained from Caltag Laboratories. CD25, CD44, CD62L and CD69 antibodies (BD/Pharmingen) were used to stain activated CD3+ T cells. Surface marker staining was analysed using a FACSCalibur™ flow cytometer (Becton Dickinson) and CellQuestPro™ software. The results are shown as the mean ± SEM of at least 3 independent experiments.

Analysis of proliferation responses

Naive CD3+ T cells were negatively selected from pooled spleen and lymph node cell suspensions with mouse T cell enrichment columns (R&D Systems). T cell populations consisted typically of 95% CD3+ cells, as determined by staining and flow cytometry. For anti-CD3 stimulations, T cells (2.5 × 105) were added in 200 μl of proliferation medium (RPMI supplemented with 10% FCS (Life Technologies), 2 mM l-glutamine (Life Technologies) and 50 U/ml penicillin/streptomycin (Biochrom)), in duplicate, to 96-well plates that were precoated with anti-CD3 antibody (clone 145-2C11, 10 μg/ml). Where indicated, IL-2 (final concentration 40 U/ml) or soluble anti-CD28 (clone 37.51, 1 μg/ml; BD Pharmingen) were added. Alternatively, PDBu (10 ng/ml; Sigma) plus the Ca2+ ionophore ionomycin (125 ng/ml; Sigma) were used. After 48 h, cells were pulsed for 18 h with [3H]thymidine (1 μCi/well) and were harvested onto a filter. The incorporation of [3H]thymidine was measured with a Matrix 96 direct beta counter system. For surface expression analysis of activation markers, cells were incubated for 20 or 48 h as indicated and were subsequently stained for FACS analysis with the above-mentioned antibodies.

Analysis of cytokine production

For cytokine secretion analysis, cells were activated for 20 or 48 h as indicated. After 48 h, supernatants were collected and were frozen in aliquots for later measurement of secreted cytokines. To determine total IL-2 production, cells were frozen and thawed three times to allow for the detection of non secreted IL-2. The concentrations of cytokines were determined with BioPlex technology (BioRad). The results are shown as the mean ± SD of at least 3 experiments.

RNA transcript analysis

Naive CD3+ T cells were negatively selected from pooled spleen and lymph node cell suspensions with mouse T cell enrichment columns (R&D Systems) and were rested for 12 h in serum free X-vivo 20 medium (Cambrex) prior to stimulation with anti-CD3 (clone 145-2C11) antibody (10 μg/ml precoated), with or without soluble anti CD28 antibody (clone 37.51, 1 μg/ml, BD Pharmingen), for 0 to 20 h. Total RNA was isolated using the Qiagen RNeasy kit. The first-strand cDNA synthesis was performed using oligo(dT) primers (Promega) with the Qiagen Omniscript RT kit, according to the instructions from the supplier. The expression analysis was performed using real-time PCR with an ABI PRIM 7000 Sequence Detection System (Applied Biosystems) and TaqMan gene expression assays, and all expression patterns were normalised to that of Gapdh. The expression levels of wild type unstimulated cells were arbitrarily set to 1. To determine IL-2 decay, Actinomycin D (Sigma) was used to stop transcription and cells were harvested at various time points as indicated. Cells were lysed in Nucleic Acid Purification Lysis Solution (Applied Biosystems) according to the protocol and were stored at −80 °C.

Western blot analysis

T cells were incubated either with medium alone or were stimulated with solid-phase hamster anti-CD3 (clone 145–2C11), with or without hamster anti-CD28 (clone 37.51; BD Pharmingen) at 37 °C for various time periods. For short time stimulations, both antibodies were added in soluble form and were crosslinked with anti-hamster IgG1 (clone HIG-632; BD Pharmingen). Cells were lysed in ice-cold lysis buffer (5 mM NaP2P, 5 mM NaF, 5 mM EDTA, 50 mM NaCl, 50 mM Tris (pH 7.3), 2% Nonidet P-40, and 50 μg/ml each aprotinin and leupeptin) and were centrifuged at 15,000 × g for 15 min at 4 °C. Protein lysates were subjected to Western blotting as previously described [8], using Abs against PKCα (Upstate Biotechnology), PKCβ, PKCθ (all from Transduction Laboratories), PKCξ, DNA polymerase (all from Santa Cruz Biotechnology), NFATc (Affinity Bioreagents), (p)S-32 IκB, pan-IκB, (p)Y-783 PLCγ1, (p)ERK, Fyn and PKB/AKT (all from Cell Signalling). All experiments were performed at least twice with similar outcomes.

Gel mobility shift assays

Nuclear extracts were prepared from 1 × 107 cells that were stimulated as indicated. Briefly, purified CD3+ cells were washed in PBS and resuspended in 10 mM HEPES (pH7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT and protease inhibitors. Cells were incubated on ice for 15 min. Nonidet P-40 was added to a final concentration of 0.6%, the cells were vortexed vigorously, and the mixtures were centrifuged for 5 min. The nuclear pellets were washed twice and resuspended in 20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT and protease inhibitors, and the tubes were rocked for 30 min at 4 °C. After centrifugation for 10 min, the supernatants were collected. Extract proteins (2 μg) were incubated in binding buffer with 32P-labelled, double-stranded oligonucleotide probes (NFκB, 5′-GCC ATG GGG GGA TCC CCG AAG TCC-3′; AP1, 5′-CGC TTG ATG ACT CAG CCG GAA-3′; and NFAT, 5′-GCC CAA AGA GGA AAA TTT GTT TCA TAC AG-3′; AP1, 5′-CGC TTG ATG ACT CAG CCG GAA-3′ (Nushift; Active Motif)). In each reaction, 3 × 105 cpm of labelled probe was used, and bandshifts were resolved on 5% polyacrylamide gels. Supershifts were performed using the following antibodies AP1 cFos; NFATc; NFκB p50 (Nushift; Active Motive). All experiments were performed at least twice with similar outcomes.

Apoptosis detection

Thymocyte apoptosis was induced by the addition of the following treatments: Con A (10 μg/ml), PDBu (1 μg/ml), ionomycin (1 μg/ml), camptothecin (1 μM), etoposide (1 μM), dexamethasone (1 μM), staurosporin (100 nM), TCR crosslinking antibodies (anti CD3, 10 μg/ml precoated, ±anti CD28 1 μg/ml) or FasL (100 ng/ml) and Enhancer of FasL (1 μg/ml). FasL and Enhancer for FasL were purchased from Alexis Pharmaceuticals, anti CD28 from BD Pharmingen, all other agents were obtained from Sigma–Aldrich. The percentages of viable cells were determined by propidium iodide staining at various time points, up to 68 h.

Total splenocytes were used to generate activated T cell blasts by incubation with Con A (2 μg/ml) for 48 h, followed by IL-2 stimulation (100 U/ml) for an additional 72 h in IMEM medium (Biochrom) that was supplemented with 10% FCS, 2 mM l-glutamine and 50 U/ml penicillin/streptomycin. After 5 days, activated T cell blasts were washed twice in medium, and viable cells were enriched by Lympholyte™ (Cedarlane) gradient centrifugation (viability >90%), followed by an incubation in IMEM medium (supplemented with 10% FCS, 2 mM l-glutamine and 50 U/ml penicillin/streptomycin). Activation-induced cell death was challenged by the addition of different concentrations of anti-CD3 cross-linking antibodies (clone 145-2C11; 0.1–10 μg/ml) and apoptosis was induced by the addition of cross-linked recombinant FasL (100 ng/ml of FasL, plus 1 μg/ml of the enhancer for FasL). Eight hours after apoptosis induction, the cells were harvested and stained with Annexin V - FITC (Molecular Probes), anti-CD4-PE and anti-CD8-APC (Caltag). The percentage of apoptotic cells in each T cell subset was determined by FACS analysis, using a FACSCalibur flow cytometer (BD) and CellQuestPro Software. The results are shown as the mean±SD of at least 3 experiments.

Results

Generation of PKCα/β double knockout mice

We generated mice that lacked both cPKC isotypes by crossing PKCα−/− mice [18] with PKCβ−/− mice [21]; these mice were then backcrossed onto the C57BL/6 background for more than 10 generations prior to use in experiments. Genotypes were determined by PCR. The null mutations of PKCα or PKCβ were confirmed by immunoblotting of whole cell lysates from naïve thymocytes and peripheral CD3+ T cells, respectively (Fig. 1a). Neither the other PKC isotypes that were normally expressed in CD3+ T cells nor PKB/Akt were upregulated in the cells derived from PKCα/β double-deficient mice. Fertility was reduced in PKCα/β double knockout mice, although the mice were viable, appeared to be healthy and lived normal lifespans. On average, double knockout mice were slightly smaller when compared to wild type mice at 3 weeks of age (mean body weight 83%), although this difference was abrogated by 6 weeks of age.

Fig. 1

Synergistic roles of PKCα and PKCβ in IL-2 production. Impaired cytokine responses of PKCα/β double knockout T cells are not caused by a secretion defect. (A) Western blot analysis of wild type and PKCα/β double knockout cells. Western blot shown is of whole cell lysates, with 5 × 106 naïve thymocytes or peripheral CD3+ T cells per lane. Blots were stained with antibodies against PKCα, β, θ, ɛ, ζ, PKBα and Fyn (loading control). (B) Proliferative responses of PKCα/β and PKCα-deficient CD3+ T cells were analysed in comparison to wild type littermate controls. After incubation using different stimulatory conditions cells were analysed using standard procedures for thymidine incorporation. (C) IL-2 responses of peripheral CD3+ T cells that were isolated from wild type, PKCα single knockout, PKCβ single knockout and PKCα/β double knockout mice. Cells were stimulated with anti-CD3 (precoated), with or without anti-CD28, as indicated. Cytokine levels in cell culture supernatants were analysed after 48 h by Bioplex suspension array technology. (D) Analysis of secreted versus total IL-2 from wild type and PKCα/β cells. Total IL-2 was determined after 3 cycles of freezing and thawing of cultured cells. (E) IL-2 gene expression is dramatically reduced in PKCα/β double-deficient CD3+ T cells after CD3 stimulation. Wild type and mutant CD3+ T cells were stimulated with anti-CD3 for a maximum of 20 h and IL-2 mRNA expression, normalised to Gapdh, was measured via real-time PCR. Results are shown as the mean values of at least 2 independent experiments.

FACS analysis of thymocyte populations in PKCα/β double knockout mice revealed that CD4+CD8+ double positive thymocytes were able to differentiate normally with regards to numbers of CD4+ and CD8+ T cells and CD3 expression (Table 1). Additionally, the composition of peripheral T and B cell subsets was unaltered in PKCα/β double knockout mice (Table 1).

Table 1

Cell subsets.

	Wt	PKCα−/−	PKCβ−/−	PKCα−/−β−/−
Thymus
CD3+	15.46 ± 2.05	16.75 ± 1.83	17.4 ± 1.14	15.54 ± 2.85
CD4+	8.10 ± 2.13	8.91 ± 4.42	10.36 ± 3.15	8.82 ± 2.36
CD8+	2.55 ± 0.43	2.35 ± 1.10	3.74 ± 0.60	3.33 ± 0.73
CD4+/CD8+	83.61 ± 3.15	86.68 ± 6.39	83.74 ± 4.70	84.00 ± 3.87
Lymph nodes
CD3+	76.55 ± 5.63	69.45 ± 13.86	77.80 ± 7.15	79.33 ± 3.15
CD19+	10.63 ± 7.24	13.54 ± 14.92	10.74 ± 9.62	10.96 ± 5.85
CD4+	58.54 ± 10.08	57.35 ± 12.79	57.35 ± 15.04	57.22 ± 4.66
CD8+	20.22 ± 1.85	15.39 ± 3.23	15.39 ± 3.49	23.16 ± 1.61
Spleen
CD3+	30.35 ± 0.12	37.23 ± 1.23	42.26 ± 1.05	40.22 ± 2.58
CD19+	53.06 ± 0.36	50.71 ± 2.36	43.06 ± 3.76	41.13 ± 0.71
CD4+	22.88 ± 0.10	24.13 ± 0.98	25.81 ± 3.17	29.20 ± 2.10
CD8+	12.97 ± 2.00	10.36 ± 1.89	16.01 ± 2.06	16.36 ± 1.86

Impaired TCR induced proliferation of PKCα/β double-deficient CD3+ T cells

To determine any potentially overlapping roles of PKCα and PKCβ we analysed the proliferative responses of freshly prepared CD3+ T cells from wild type, PKCα single knockout, PKCβ single knockout and PKCα/β double knockout mice. The proliferation of PKCα/β deficient cells that were stimulated by anti-CD3, with or without anti-CD28, was impaired to the same degree as in PKCα single knockout cells (Fig. 1b), as measured by 3H thymidine uptake.

To exclude the proliferative defects being caused by deregulated apoptosis we analysed both apoptotic sensitivity of thymocytes following exposure to apoptotic stimuli (PDBu, ionomycin, dexamethasone, etoposide, camptothecin, staurosporin, concanavalin A, anti CD3 or FasL) and AICD and/or Fasl induced apoptosis of CD4+ and CD8+ T-cell blasts derived from wild-type and double knockout animals using CD3 engagement in vitro. Ex vivo survival of freshly isolated PKCα/β deficient thymocytes was unaffected (supplementary Fig. 1a), compared to that of single knockout or wild type control thymocytes. Additionally, no significant differences were observed in the susceptibilities of CD4+ and CD8+ T cell blasts from adult PKCα/β double knockout mice to the ex vivo apoptotic stimuli anti-CD3 or FasL, versus the susceptibilities of cells from PKCα and PKCβ single knockout mice (supplementary Fig. 1b).

Additive roles of PKCα and PKCβ in de novo IL-2 synthesis

The roles of PKCα and PKCβ in IL-2 signalling were assessed by the analyses of IL-2 mRNA expression and IL-2 protein secretion. In contrast to CD3+ T cells that were obtained from PKCα and PKCβ single knockout mice, which showed nearly normal IL-2 production upon TCR stimulation [14,26], we observed a strong defect in IL-2 production in T cells from PKCα/β double knockout mice upon stimulation with anti-CD3 alone (Fig. 1c), whereas the addition of soluble anti-CD28 induced normal IL-2 levels from double knockout T cells that were comparable to levels produced by wild type control cells. These data indicate that the costimulatory CD28 signal can somehow bypass the requirement for these cPKC isotypes during TCR signalling (data not shown).

To investigate whether the significantly reduced amounts of IL-2 in cell culture supernatants from PKCα−/−β−/− T cells were caused by secretion defects, we performed “freeze and thaw” experiments, in which IL-2 concentrations in cell culture supernatants (secreted IL-2) were compared to the cytoplasmic IL-2 contents of the cells plus supernatants (total IL-2), which were obtained by repeated cycles of freezing and thawing of cells in order to disrupt them and thus release stored IL-2. As shown in Fig. 1d, total IL-2 levels were reduced to the same degree as secreted IL-2 levels, indicating normal IL-2 secretion capacities in PKCα/β double knockout cells. However, quantitative RT-PCR analysis revealed significant differences in IL-2 mRNA levels between wild type and mutant samples (Fig. 1e). In contrast, there were no significant differences between wild type and knockout cells in the expression levels of mRNA for IL5, IL-10, GM-CFS, IL-2RA and PKCθ (not shown). Reduced IL-2 mRNA levels in PKCα−/−β−/− CD3+ T cells were not caused by faster degradation, as the half-life of IL-2 mRNA was not decreased in the mutant T cells (wt: 0.65 ± 0.07 h; PKCα−/−β−/−: 1.40 ± 1.27 h). These data suggest that the combined PKCαβ deficiency in T cells selectively affects de novo IL-2 mRNA transcription and IL-2 protein expression but not IL-2 mRNA stability or protein secretion.

No effects of the genetic losses of PKCα and PKCβ on NFκB, NFAT and AP1 transactivation

To understand the impaired IL-2 transcription, the DNA binding capacities of NFλB, NFAT and AP1 to radioactively labelled IL-2 promoter oligos were studied via EMSAs. Interestingly, neither the binding of NFκB (Fig. 2a) nor of AP1 and NFAT (Fig. 2b) to DNA were impaired in nuclear extracts from stimulated PKCα−/−β−/−T cells. Immunoblot analysis of whole cell lysates demonstrated the unimpaired phosphorylation of IκBα, as well as the normal phosphorylation and activation of ERK in PKCα−/−β−/− double-deficient CD3+ T cells (Fig. 2c). Data from the Western blot analysis of nuclear extracts supported the observed normal DNA binding because nuclear translocation of the NFκB subunits p50 and p65 (and of NFAT) was not reduced in mutant cells or in wild type cells upon stimulation with anti-CD3, with or without anti-CD28 (Fig. 2d). Taken together, these data indicate that PKCα and PKCβ are not involved in the control of CD3-induced transactivation of NFκB, NFAT and AP1.

Fig. 2

(A and B) Loss of PKCα and PKCβ in double-deficient CD3+ T cells did not affect NFκB, AP1 and NFAT DNA binding capacities. Nuclear extracts were prepared from CD3+ T cells from wild type, PKCα−/−, PKCβ−/− and PKCα−/−β−/− mice following stimulation with anti-CD3 with or without anti-CD28 for 16 h. Gel mobility shift assays were performed with radioactively labelled probes that contained NFκB, AP1 and NFAT binding site sequences. One representative experiment of 3 is shown. (C) Western blot analysis of whole cell lysates from wild type, PKCα−/−, PKCβ−/− and PKCα−/−β−/− CD3+ cells that were stimulated briefly. Normal phosphorylation kinetics of IκBα, ERK and PKD were observed in single and double-deficient CD3+ T cells. (D) Immunoblot of nuclear extracts from wild type, PKCα−/−, PKCβ−/− and PKCα−/−β−/− CD3+ T cells that were stimulated as indicated for 16 h. Nuclear translocation of the p50 and p65 subunits of NFκB was normal in PKCα−/−β−/− T cells. One representative experiment of 2 is shown.

Consistently, the stimulation-dependent upregulation of the surface markers CD25, CD69 and CD44, when monitored on CD3+ cells by FACS analysis, was not significantly different between wild type, single knockout and PKCα/β double knockout littermates (Fig. 3a and b); interestingly in the CD8+ subset the CD25 median expression was reproducibly reduced in PKCβ−/− and PKCα−/−β−/− T cells (not shown). The effector/memory T cell ratio, defined by the surface expression of CD44 and CD62L, was unaltered by the concomitant loss of PKCα and PKCβ (Table 2).

Fig. 3

Flow cytometric analysis of CD25, CD44 and CD69 surface marker expression on T cells that were stimulated or not with CD3 alone, CD3 + CD28 or CD3 + IL-2. No gross differences were observed between wild type and PKCα−/−β−/− cells. Results are shown as the mean ± SEM of 3 independent experiments.

Table 2

Induction of CD4+ and CD8+ effector memory cells, stimulation 20 h, n = 2 [percentage of naive and effector memory cells].

		Wt	PKCα−/−	PKCβ−/−	PKCα−/−β−/−
CD4+ T cells
Medium	Naive cells (CD62L pos.)	83.58 ± 2.49	83.45 ± 2.99	80.84 ± 0.46	86.92 ± 8.74
Medium	Effector memory cells (CD62L neg. CD44 pos.)	15.36 ± 6.02	15.31 ± 6.03	18.37 ± 6.84	12.10 ± 4.15
CD3 + CD28	Naive cells (CD62L pos.)	31.96 ± 4.27	39.39 ± 4.26	33.87 ± 4.85	33.26 ± 2.94
CD3 + CD28	Effector memory cells (CD62L neg. CD44 pos.)	68.33 ± 1.97	60.58 ± 2.83	66.18 ± 1.78	66.52 ± 8.88
CD8+ T cells
Medium	Naive cells (CD62L pos.)	70.87 ± 16.44	70.29 ± 10.13	61.26 ± 13.90	73.10 ± 11.66
Medium	Effector memory cells (CD62L neg. CD44 pos.)	26.48 ± 17.53	27.18 ± 14.57	38.21 ± 17.10	25.32 ± 12.43
CD3 + CD28	Naive cells (CD62L pos.)	36.49 ± 16.97	42.71 ± 15.68	35.77 ± 18.26	38.63 ± 12.03
CD3 + CD28	Effector memory cells (CD62L neg. CD44 pos.)	60.63 ± 17.87	52.53 ± 12.98	59.01 ± 16.30	57.51 ± 12.67

Discussion

PKCθ has been proposed as a novel and key player in T cell activation via the stimulation of NFAT, AP1 and NFκB, all of which are transcription factors that are essential for IL-2 gene induction [27-31]. This has been confirmed by experiments in which PKCθ-deficient murine T cells showed profound defects in IL-2 production upon TCR stimulation [8,9,32]. While the importance of PKCθ in T cell function is well documented in numerous in vitro and in vivo studies, an understanding of the roles of other PKC family members, which are normally expressed in T cells, in TCR-mediated cell activation remains elusive. Among the conventional PKCs, PKCα was shown to be essential for adequate T cell proliferative responses upon CD3/CD28 stimulation. Interestingly, despite profoundly reduced proliferation, IL-2 secretion was not impaired in CD3+ T cells that were obtained from PKCα knockout mice [14].

In this study, we generated mice that lacked both conventional PKC isotypes, PKCα and PKCβ, and analysed the functions of T cells from these mice to detect possible overlapping roles of the two isotypes. As a result, we observed a strong defect in the production of IL-2 in peripheral CD3+ T cells, independent of PKCθ function, in contrast to the normal levels of IL-2 that were observed in PKCα and PKCβ single knockout cells, indicating that both PKC isotypes are required for the functional upregulation of IL-2. The isotypes have redundant roles, and each can compensate for the loss of the other. The induction of IL-2 production is under strict transcriptional regulation, which depends on the orchestrated interplay of the NFκB, NFAT and AP1 transcription factors. Interestingly, the transactivation of NFκB, NFAT and AP1 were not affected by the combined loss of function of both PKC isotypes.

Of note, the defect in IL-2 production was only observable when T cells were stimulated with anti-CD3 alone, whereas full activation with CD28 costimulation rescued the proliferative and cytokine secretion defects of PKCα/β cells to levels that were comparable to those observed in wild type and single knockout cells. This finding suggests that TCR stimulation via anti-CD3 cross-linking requires the classical PKCα/β isotypes for the proper transcription of IL-2 mRNA, whereas anti-CD28 costimulation, which has been shown to be an important prerequisite for correct PKCθ localisation at the centre of the immunological synapse (IS) [33,34], can bypass the defect via the induction of PKCθ-mediated signalling pathways. IL-2 production can also be regulated at post-transcriptional levels via stabilising or destabilising processes at AU-rich repeats in the 3′untranslated region of IL-2 mRNA.

Because an impact of PKCβ on PMA stimulation-induced IL-2 mRNA stabilisation was postulated by the work of Zhu et al. [35], we assessed the stability of IL-2 mRNA in actinomycin D-treated anti-CD3-stimulated PKCα/β-deficient T cells via RT-PCR. Our analysis showed directly that no increases in the instability or degradation of IL-2 mRNA were detected in PKCα/β T cells (data not shown). It is known that high levels of IL-2 mRNA are transcribed prior to cell division, while the transcription of IL-4 and IFNγ mRNA is not induced strongly until after multiple cell cycles, when epigenetic remodelling occurs [36,37]. Therefore, the slightly reduced IL-4 mRNA levels that we observed seemed to be consequential to impaired proliferation, while the defect in IL-2 synthesis was the cause for reduced proliferation. However, where exactly cPKCs interact in the IL-2 activation pathway remains unclear.

Concluding remarks

Conventional PKCα and PKCβ play redundant roles in the activation of IL-2 transcriptional responses. Compared to PKCα−/− and PKCβ−/− single knockout cells, which show only slight reductions in IL-2 secretion, IL-2 production is decreased dramatically in anti-CD3-stimulated double-deficient CD3+ T cells. In this study, we found that low cytokine levels were caused by the decreased de novo synthesis of IL-2 mRNA and not by increased mRNA instability and degradation. For the first time, we demonstrate that besides PKCθ, the cPKCs α and β are involved in CD3-induced T cell activation processes.