LAUSR

Upon first activation, CD4+ T cells differentiate into specialized T helper subtypes, dependent on the key transcription factors (TFs) T-bet for Th1, GATA3 for Th2, PU.1 for Th9, and RORγt for Th17. Our group established decisive roles for Interferon-Regulatory-Factor 4 (IRF4) in the differentiation of Th2 [1], Th9 [2], and Th17 [3] cells, while others showed an in vivo role for Th1 cells [4]. Thus, IRF4 may serve as an initiation factor for transcribing genes necessary for further differentiation [5, 6]. Our group hypothesizes that IRF4 has functional subunits, which preferably bind to one key TF and determine its activity. If so, IRF4 mutations might selectively affect T helper subtypes depending on which TF binds to that domain of IRF4. Accordingly, IRF4 lacking its autoinhibitory domain (AD) (Fig. 1A) specifically induces Th17 differentiation [7]. PU.1 and IRF4 direct Th9 differentiation. The description of the IRF4/PU.1/DNA crystal structure showed the tight interaction of IRF4 amino acids (AA) V111, L116, and D117, DNA, and the AA R222 and K223 of PU.1 [8]. Importantly, mutations of leucine 116 to arginine have been reported in patients with chronic lymphocytic leukemia (CLL), leading to increased levels of IRF4 and MYC [9] and a higher DNA binding capacity than wt IRF4 [10]. We thus asked if the L116R exchange affects IRF4 function in T cell subtypes, particularly during IL-9 production. Purified naïve CD4+ T cells from IRF4−/− mice were infected with retroviruses overexpressing GFP and variants of IRF4, followed by in vitro differentiation and evaluation of cytokine production. L116R mutations were introduced into either a wildtype (wt) IRF4 background or one lacking AD (dAD), i.e. L116Rwt or L116RdAD. By flow cytometry, retroviral overexpression into IRF4−/− cells created slightly less IRF4 protein amounts than in control IRF4+/+ Th cells (Fig. S1A–C). An empty vector without IRF4 served as a control. Transfected cells express GFP to separate them from IRF4 negative cells within the same sample, as an internal control (Fig. S1C). In a bead-based multiplex assay, we tested the effect of the mutants on cytokines in the supernatant of cells differentiated under Th2, Th9, and Th17 conditions restimulated with PMA/ionomycin. Importantly, we noted a strong IL-9enhancing activity in Th2 cells expressing the L116RWT mutant (Fig. 1B), also on the dAD background. While Th9 cells produced more IRF4-dependent IL-9 than Th2, there was no difference between mutants. Slight IL-9 induction was noted in Th17 cells, but at negligible amounts (Fig. S2A). Importantly, however, in Th17 cells IL-17A and IL-17F were strongly reduced by L116RWT and L116RdAD (Fig. S2A). Enhancement in Th2 cells was limited to IL-9, but not seen in IL-5, IL-13, IL-2, IL6, and IL-10. For IL-6 and IL-22, mutants even caused a drop in production. In Th9 cells, suppressive effects were seen for IL13. IFNγ was already produced by empty vector control cells and IRF4 mutants had little impact, supporting no significant role for IRF4 during in vitro Th1 production [1, 3]. These screening experiments document an inducing effect of L116R mutants specifically on IL-9 production. We then used ELISAs for IL-9 and IL-13 for supernatants from Th2 and Th9 cells (Fig. 2A), harvested after differentiation for 48h (Th9) or 72h (Th2) or after washing and restimulation for 8h. As before, the L116 mutants caused upregulation of IL-9 production, both on the IRF4 wt or dAD background and during primary culture as well as after restimulation. For IL-13, a suppressive effect by the L116R mutants was observed in Th2 and Th9 cells. These results were based on supernatants unable to separate transduced from untransduced cells in the same culture. Furthermore, transfection rates slightly differ between virus preparations encoding different mutations. Finally, these results made no statement on whether this increase was related to few individual or all transfected cells. We therefore restimulated cells differentiated as described above and performed intracellular staining of cytokines followed by flow cytometry. The gating strategy is depicted in Fig. S2B and a representative experiment of Th2 and Th9 cells in Fig. S3. No cytokine production was observed in untransfected GFP− cells (Fig. S2B) or in GFP+ cells transfected with empty control virus (Fig. S3). Transduction with IRF4 or dAD induced IL-9 or IL-13, (Fig. S3) as expected. The data raised in these experiments confirmed the above-described findings, namely significant inducing capacity of L116 mutants for IL-9 but not IL-13 production. This finding was related to a higher frequency of cytokine-producing cells rather than higher cytokine production per cell (Fig. S3A–D). Quantification of flow cytometry for IL-9 and IL-13 producing cell frequencies in Th2 and Th9 differentiation cultures (Fig. 2B) confirmed effects of the mutants for IL-9 production particularly in Th2 cells, but also in Th9 cells, as compared to controls. A suppressive effect was again observed on IL-13 production in Th9 and less so in Th2 cells. The same mutants had again a strongly suppressive effect on IL17 production in Th17 cells, in L116Rwt and L116RdAD transduced cells. The increase in IL-9 production of L116R transduced cells may be related to increased DNA binding capacity [10]. A