PIM kinases regulate early human Th17 cell differentiation

Abstract

The serine/threonine-specific Moloney murine leukemia virus (PIM) kinase family (i.e., PIM1, PIM2, and PIM3) has been extensively studied in tumorigenesis. PIM kinases are downstream of several cytokine signaling pathways that drive immune-mediated diseases. Uncontrolled T helper 17 (Th17) cell activation has been associated with the pathogenesis of autoimmunity. However, the detailed molecular function of PIMs in human Th17 cell regulation has yet to be studied. In the present study, we comprehensively investigated how the three PIMs simultaneously alter transcriptional gene regulation during early human Th17 cell differentiation. By combining PIM triple knockdown with bulk and scRNA-seq approaches, we found that PIM deficiency promotes the early expression of key Th17-related genes while suppressing Th1-lineage genes. Further, PIMs modulate Th cell signaling, potentially via STAT1 and STAT3. Overall, our study highlights the inhibitory role of PIMs in human Th17 cell differentiation, thereby suggesting their association with autoimmune phenotypes.

Keywords: CP: Immunology; PIM kinases; T helper cell differentiation; Th1 cells; Th17 cells; bulk RNA sequencing; single-cell RNA sequencing; transcriptomics.

Graphical abstract

Figure 1

PIMs are upregulated in Th17 cells via the IL6/STAT3 axis (A) Reads per kilobase of transcript, per million mapped reads (Rpkm) of PIM1, PIM2, and PIM3 are depicted at different times of activation (Th0) or Th17 differentiation from three biological replicates, using our published RNA-seq data (GEO: GSE52260). (B) The expression of the three PIMs in Th0 and Th17-polarizing cells over time was analyzed by western blot (bottom). Band intensities of target proteins from four biological replicates were normalized to β-actin (top). (C) Representative western blots of PIM1, PIM2, and PIM3 are shown from naive CD4⁺ T cells cultured for 72 h under activated Th0 condition, Th17 differentiation, or activated Th0 in the presence of Th17 cytokines (IL6, IL1β, and TGFβ) (right). Graphs on the left show band intensities of target proteins from four biological replicates, normalized to β-actin and relative to Th0. Statistical significance was calculated by comparing each condition to Th0. (D) Western blots of STAT3, PIM1, PIM2, and PIM3 protein levels in non-targeting (Scr) vs. STAT3 KD cells, at 72 h of Th17 polarization are shown (left). Protein intensities of STAT3 and PIM kinases from three biological replicates were normalized to β-actin and relative to Scr (right). Graphs in (B)–(D) show mean ± SEM. Statistical significance was calculated using two-tailed Student’s t test (^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001).

Figure 2

PIMs negatively regulate expression of IL17 and RORC (A) Workflow. Naive CD4⁺ T cells were simultaneously transfected with a pool of one LNA and two siRNAs each targeting PIM1, PIM2, and PIM3, respectively, (TKD), or with in vitro transcribed PIM1, PIM2, and PIM3 RNAs (TOE). After 24 h resting, cells were cultured under Th17 conditions for 72 h. (B and C) PIM TKD efficiency was confirmed at RNA level at 6 h, 24 h, and 72 h post-differentiation in four biological replicates using qRT-PCR (B) or at protein level at 72 h of Th17 cell differentiation by western blot (C, left). Band intensities of PIM kinases from five biological replicates were normalized to β-actin and relative to Scr control (C, right). (D) Secreted IL17A cytokine levels in supernatants of PIM TKD Th17 cells are shown at 72 h of polarization. Boxplot represents median and interquartile range, and whiskers extend to maximum and minimum values. Data represent five biological replicates. (E–G) IL17 A/F RNA levels at 72 h (E and F) and RORC RNA levels at 48 and 72 h (G) in PIM TKD Th17 cells were analyzed in four biological replicates using qRT-PCR. (H–L) PIM overexpression was confirmed at 48 h of polarization by western blot (H, left). Band intensities of PIM kinases from three biological replicates were normalized to β-actin and relative to GFP control (H, right). (I–L) IL17 secretion (I), IL17A, and RORC RNA expression (J and K) and CCR6 surface expression (L) in PIM TOE Th17 cells at 72 h of polarization, were assessed by ELISA, qRT-PCR, and flow cytometry analyses, respectively, for three biological replicates. (L) Mean fluorescence intensity (MFI) values were normalized to Scr control. ELISA values in plots (D) and (I) were normalized for cell count (live), and then normalized to Scr or GFP control, respectively. (B), (E), (F), (G), (J), and (K) depict transcript FC normalized to control. Plots in (B, C, E–L) show mean ± SEM. Statistical significance is calculated using two-tailed Student’s t-test (ns not significant, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001). See also Figure S2.

Figure 3

PIMs alters Th17 gene expression (A) Z score heatmaps standardized with a FDR of <0.1 and FC of >1.4 for the differentially expressed genes detected at 6 h and 24 h are shown for four biological replicates. Genes common between the two time points are highlighted in bold. (B) Venn diagram demonstrating the number of overlapping differentially expressed genes in PIM TKD Th17 cells at 6 h and 24 h of polarization with a FDR of <0.1 and a FC of >1.4. (C) IPA was used to identify signaling pathways that are significantly altered upon PIM TKD. For analysis, differentially expressed genes at 6 h and 24 h were merged and the top enriched pathways related to T cell signaling and immune-mediated diseases are shown. (D) Volcano plots highlight the Th17-associated transcripts that are differentially expressed upon co-depletion of PIM1, PIM2 and PIM3, at 6 h (left) and 24 h (right) of Th17 polarization with a FDR of <0.1 and a FC of >1.4. Upregulated genes are in pink, and downregulated genes are in blue. Genes colored in gray are selected by a FDR of <0.25. (E and F) RORA (E) and STAT3 (F) gene expression was analyzed in PIM-depleted Th17 cells at 6 h and 24 h by qRT-PCR. FC normalized to the Scr control was plotted for four biological replicates. Boxplots represent median and interquartile range, and whiskers extend to maximum and minimum values. Statistical significance is calculated using two-tailed Student’s t tests (^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001). See also Figure S3 and Table S1.

Figure 4

PIMs regulate Th1/Th17 signaling axis during early Th cell differentiation (A) Based on bulk RNA-seq data, STAT1 was predicted as a positive upstream regulator of the differentially expressed genes at 6 h upon PIM TKD using the IPA “upstream regulator” prediction tool (Z score < −2, which indicates predicted upstream regulators; gray arrows represent effects not predicted. (B) The downregulation of STAT1 RNA levels was confirmed at 24 h and 72 h in PIM TKD Th17 cells by qRT-PCR. (C and D) The RNA expression of Th1-related factors TBX21 (C) and STAT4 (D) was validated in PIM TKD at 6 h and 24 h of Th17 cell differentiation by qRT-PCR. FC normalized to the Scr control was plotted for four biological replicates. Boxplots represent median and interquartile range, and whiskers extend to maximum and minimum values (B–D). Statistical significance is calculated from four biological replicates using two-tailed Student’s t tests (^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001). (E and F) scRNA-seq was performed at 6 h in PIM TKD and control Th17 cells using the single-cell fixed RNA profiling. Cells double-negative for PIM^–/TBX21 but RORA positive (E) and single cells double-positive for PIM/TBX21 but RORA-negative (F), were colored in blue and projected into a two-dimensional map using UMAP. (G) Differential expression analysis was performed between Scr (control) and PIM^– in PIM TKD samples at 6 h of differentiation (FDR of <0.05, log2FC of >0.24) for one biological replicate. Scaled scRNA-seq dot plot depicting the differentially expressed genes of interest on the x-axis. The color scale represents the average expression of a given gene in the cluster, and the size of the dot represents the percent of cells that express a given gene. (H) Violin plots showing the expression of CD82 and STAT1 in Scr and enriched PIM^– cells of PIM TKD samples at 6 h. See also Figures S4 and S5 and Table S2.

Figure 5

PIM downstream targets STAT1 and CD82 negatively regulate Th17 cell differentiation (A) STAT1 KD efficiency was validated at 72 h in Th17 cells by western blot. Band intensities of STAT1 were normalized to β-actin and relative to Scr control. (B) The surface expression of CCR6 was assessed at 72 h in STAT1 KD Th17 cells. (C) Secreted IL17A cytokine levels in supernatants of STAT1 KD Th17 cells are shown at 72 h of polarization using ELISA. (D–F) IL17 A/F RNA levels at 72 h (D and E) and STAT3 RNA levels at 24 h (F) of differentiation in STAT1-deficient Th17 cells were analyzed by qRT-PCR. Data in (A–F) represent five biological replicates. (G) siRNA-mediated KD of CD82 was validated at 24 h and 72 h in Th17 cells by flow cytometry. Representative histograms are shown (right). (H) Secreted cytokine IL17A levels in supernatants of CD82 KD Th17 cells are shown at 72 h of polarization using ELISA. (I–K) IL17 A/F RNA levels (I and J) and STAT3 RNA levels (K) in CD82 KD Th17 cells, respectively, at 72 h and 24 h, were analyzed using qRT-PCR. (G–K) Data represent four biological replicates. (D–F) and (I–K) depict transcript FC normalized to control. ELISA values in (C) and (H) were normalized for cell count (live), and then normalized to Scr control. Boxplots represent median and interquartile range, and whiskers extend to maximum and minimum values. (B and G, left) Bar plots show the mean fluorescence intensity values normalized to Scr control. Statistical significance is calculated using two-tailed Student’s t tests (^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001). Plots in (A), (B), (D–G), (I), and (K) show mean ± SEM. See also Figures S6, S7, and S8.

Figure 6

Gene expression of PIMs and CD82 in healthy and disease datasets (A) Gene expression of PIMs and CD82 in peripheral naive CD4⁺ and Th17 memory cells are shown from 91 healthy human subjects using the Database of Immune Cell Expression. Transcripts per million (TPM) are represented as dot plots for individual subjects with median. (B and C) The expression levels of PIMs and CD82 are shown as violin plots using the scRNA-seq atlas of cycling T cells of 18 patients with UC (inflamed and non-inflamed adjacent tissue) and 12 healthy individuals (B) and T cells in 21 cancer types from more than 300 patients (C).