In this study we systematically explored pathways involved in T cell activation by molecular profiling. We showed that TCR (both CD3/28 and PMA/CD3) driven stimulation profiles are truly distinct from co-stimulatory profiles mediated via PMA/CD28. Secondly, using selective inhibitors and siRNA we found that the proximal kinase Lck is involved in CD3 and not PMA/CD28 activation, whereas PKCθ appears to be a crucial central signaling kinase in both TCR and PMA/CD28 (co)-stimulatory activation of T cells. Finally, stimulations involving TCR/CD3 appear to preferentially induce a Th1-like fingerprint, whereas lack of TCR/CD3 signaling in the presence of PMA/CD28 stimulation diverts T cells towards a Th2-like state.
It has been suggested that the strength of TCR-signaling can regulate the fate determination of naive T cells; high-potency signals skew towards Th1 differentiation, whereas low potency signals promote Th2 differentiation [6, 22]. Although TCR and co-stimulatory pathways have been the focus of many studies in the previous decades, the direct contribution of TCR stimulation vs. co-stimulatory signals towards Th differentiation is not fully understood. By stimulating T cells with PMA/CD3 and PMA/CD28 we dissected signaling pathways and explored the activation profiles. CD3-mediated signaling rapidly increased intracellular Ca2+, a second messenger to activate many enzymes including Calcineurin, which resulted in an increased nuclear translocation of NFATc1. Interestingly, PMA/CD28 stimulation did not result in a Ca2+-mediated response (and was therefore marked a calcium-independent stimulation) but enhanced many of the co-stimulatory mediators including MAPK/AP1 and NFκB signal transduction. These results are in line with earlier studies that showed differential effects of cyclosporine A (CsA) and dexamethasone on CD3 vs CD28-mediated signaling, which revealed that PMA/CD28 stimulation was insensitive towards CsA-mediated Calcineurin inhibition in contrast to PMA/CD3 stimulation [20, 23].
Gene expression induced by combinations of stimulatory signals revealed pathway-specific biomarkers or fingerprints. PMA/CD3-induced gene profiles included IL-2, IFNγ, XCL1, granzyme B, and FASL, which have been associated with a Th1 type of response [24, 25]. Also, sustained NFAT signaling, which is also induced via PMA/CD3 stimulation, has been shown to promote Th1-like gene transcripts, including IFNγ, FasL and P-selectin glycoprotein ligand 1 . Our results are further substantiated by the finding that T-bet (TBX21), the Th1 master transcription factor , and RUNX3, which together with Tbet are crucial for inducing IFNγ and repressing IL-4 , were highly expressed under PMA/CD3-stimulatory conditions.
PMA/CD28 stimulation does not induce a Ca2+ flux nor does it increase nuclear translocation of NFAT. However it provides the cell with a high level of co-stimulatory signaling, and induces a completely distinct genomic fingerprint compared to PMA/CD3 stimulation. Following PMA/CD28 stimulation, Jurkat T cells highly expressed CCL1/I309, a chemokine which is highly expressed during a Th2-eosinophil response in allergic airway diseases [29, 30]. Lymphotoxin (LT), a cytokine which is associated with a Th2-type of response controlling IgE production , was also highly expressed under PMA/CD28 stimulation. In conjunction with this finding, the master transcription factors for Th2, GATA3  and the Retinoid X Receptor (RXR) , were induced under the PMA/CD28 stimulatory condition. Notably, Th2-associated cytokines like IL-4, IL-5 and IL-13 were not induced in Jurkat T cells after PMA/CD28-stimulation, this in contrast with PMA/CD28 stimulation of human whole blood and purified CD4+ T cells, which could be due to the developmental blockage of Jurkat T cells. Additional file 6: Figure 2 shows a schematic overview summarizing the involvement of the signaling pathways and genes induced under differential stimulation as observed in this study and highlights their relation towards T helper 1 and -2 development.
Our results are in line with the notion that high calcium levels drive Th1 and CTL responses and low calcium levels drive Th2 responses [7, 34, 35], which was further substantiated by our results using inhibitors for Lck and Cn, which modulate Calcium signaling in T cells. These inhibitors repressed Th1-associated genes under PMA/CD3-stimulation, but induced Th2 transcription factors GATA3 and RXRA, revealing a skewing of Th1 towards Th2 profiles. In contrast, PMA/CD28 stimulation in the presence of Lck and Cn inhibition, Th2-associated genes, e.g. CCL1 or IL-13 in CD4+ T cells, were not affected or even induced. The crucial role of Calcium and Lck in driving Th1 response is in line with the observation that knock down of Lck affects the virus-specific Th1/CTL response in mice and Lck deficiency increases Th2 associated cytokine production [36, 37]. Interestingly, lack of Calcium signaling can give rise to an anergic T cell phenotype (reviewed in ). Therefore it would be of interest to further explore the role of Lck in calcium-dependent activation via PMA/CD3 on Th1/CTL responses and calcium-independent activation of T cells via PMA/CD28 on the induction of anergy in more detail.
CD28 signaling has been functionally linked with PKCθ induced activation of NFκB , which was also validated using PMA/CD28 as stimulus . Previously it has been reported that CD28-costimulation induces GATA3 expression and Th2 differentiation via the activation of NFκB [41, 42]. Additional studies in mice revealed that PKCθ is involved in mounting both Th2- and Th1-mediated lung inflammation, although Th2-mediated inflammation is more PKCθ-dependent . Our studies show that inhibition of PKCθ can indeed inhibit a PMA/CD28 stimulation, which was reflected by the effect of PKCθ inhibition on the PMA/CD28-induced Th2-like gene expression profile. These observations are in line with the results from CD28 knock-out mice and inhibition of CD28 signaling using CTLA4Ig, showing that the CD28 co-stimulatory signaling is crucial for mounting a proper Th2 response. In contrast, Th1 and CTL responses were found to be less dependent on CD28 signaling [44, 45]. Of interest, PKCθ inhibition in our hands, also affected PMA/CD3-induced Th1-like expression profiles. These results underline the duality of PKCθ in the integration of TCR and CD28-mediated signaling events which is evident from PKCθ KO mice experiments.
Finally, our results also show that this differential stimulation does not only occur in Jurkat T cells, but also plays a role in primary human T cells. These cells were found to secrete a Th1-like response (Tbet-IFNγ) via PMA/CD3 stimulation, whereas PMA/CD28 stimulation led to a Th2 activation profile (GATA3-IL-5/IL-13). In these cells inhibition of the Lck/Cn/NFAT pathway was only effective after PMA/CD3 stimulation whereas inhibition of PKCθ inhibited both PMA/CD3-induced IFNγ production and PMA/CD28-induced IL-13 production. These results illustrate that the findings in the Jurkat T cell line were successfully translated and relevant to a human primary cellular setting. Interestingly, PMA/CD3 stimulation also enhanced IL-17 production in the primary human whole blood assay and increased the expression of the IL-21 receptor, which is crucial for Th17 induction [46, 47], in Jurkat T cells. These results suggest that additional signals, like IL-21 in conjunction with TGFβ and IL-6, might be necessary to differentiate from a Th1-like phenotype towards a Th17 phenotype, whereas the absence of TGFβ in the presence of high levels of IL-2 will favor Treg development or stabilization. Therefore further exploration of these differential stimulations in the presence of defined/different cytokine stimuli could further elucidate T helper cell differentiation and establish sub-set specific genome profiles. The findings described in this paper offer a robust platform for in vitro activation of T cells, in which observed responses can be easily translated form Jurkat T cells, towards purified CD4+ T cells and even human whole blood. This can be of interest for efficiency and selectivity profiling of kinase inhibitors or for pathway-specific biomarker identification for future drug development and clinical studies.