Identification of the miRNomes of in vivo-generated Th1, Th17 and Treg cells

In order to identify miRNA repertoires of CD4⁺ T cell subsets under inflammatory conditions (EAE) in vivo, we established a triple reporter mouse for IFN-γ (YFP⁺), IL-17A (GFP⁺) and Foxp3 (hCD2⁺) (triple Foxp3^hCD2.Ifng^YFP.Il17a^GFP reporter mice Fig 1a), by crossing previously described single reporter lines [25–27]. While Treg cells were isolated based on hCD2 expression, driven by the activity of the Foxp3 promoter, which defines this lineage independently of activation status, Th1 and Th17 cells were purified based on YFP and GFP expression, thus requiring cell activation for cytokine gene expression.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Identification of the miRNomes of Th1, Th17 and Treg cells from EAE-challenged Foxp3^hCD2.Ifng^YFP.IL17a^GFP reporter mice.

(a) Schematic representation of the genomic alterations introduced in the triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mouse strain with the respective CD4⁺ T cell subset. In the loci of each reporter gene, a sequence for an independently translated reporter protein was inserted downstream of the gene endogenous translational stop codon. (b) Experimental setup: triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice were subjected to EAE and at the peak plateau stage Treg, Th1 and Th17 cells were isolated from the spleen and lymph nodes by FACS, according with each reporter gene. After RNA isolation samples were subjected to both mRNA-seq and small RNA-seq. (c, d) Heatmaps depicting differentially expressed miRNAs between Treg and effector Th (up in Treg vs. Th1 or Treg vs. Th17 (c); vs. up in Th1 vs. Treg or Th17 vs. Treg (d)). Colors indicate the direction and magnitude of calculated z-scores, with red representing higher expression of a miRNA in a given cell subset and blue, the opposite. (e) Volcano plots depicting differentially expressed miRNAs between Th1, Th17 and Treg subsets. miRNAs names are indicated when differentially expressed (log2FC > 1; FDR < 0.05) in two comparisons. (f) RT-qPCR analysis of the 9 candidate miRNAs’ expression in Treg, Th1 and Th17 cells isolated and pooled from the lymph nodes and spleen of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice at EAE peak plateau. Box-and-whiskers indicate median, 25% and 75% percentiles, minimum and maximum expression levels of candidate miRNAs. Results are presented relative to miR-423-3p expression from 7 independent experiments. *p < 0.05 and **p < 0.01 (Kruskal–Wallis test). The data underlying this Figure can be found in https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1067547 (c, d) and in S1 Data (e, f).

https://doi.org/10.1371/journal.pbio.3003271.g001

As the response is triggered in peripheral lymphoid organs, we isolated these populations by FACS from the spleen and lymph nodes of EAE-induced triple Foxp3^hCD2.Ifng^YFP.Il17a^GFP reporter mice at the peak-plateau phase of the disease, and extracted RNA to perform both small RNA- and messenger-RNA sequencing (small RNA-seq and mRNA-seq, respectively) (Figs 1b and S1a). Both datasets showed that the isolated YFP⁺, GFP⁺ and hCD2⁺ CD4⁺ T cell populations segregated unambiguously on a principal component analysis (S1b and S1c Fig, respectively), and importantly, differentially expressed the signature genes of each corresponding CD4⁺ T cell subset (S1d Fig), thus attesting the reliability and reproducibility of our experimental system. The small RNA-seq data revealed clearly distinct miRNomes, with 110 miRNAs differentially expressed between Teff and Treg cell populations in vivo (Fig 1c and 1d; S1 Table). These data are a valuable resource for the community, from which we decided to zoom in on miRNAs specifically enriched or depleted in one of the CD4⁺ T cell subsets. Interestingly, within this candidate list of differentially expressed miRNAs, only nine were specifically up- or down-regulated in one subset (i.e., Th1, Th17 or Treg) compared with the other two subsets (Fig 1e): miR-1247-5p was up-regulated in Th1 cells; miR-126a-5p, miR-5108 and miR-122-5p were up-regulated in Th17 cells, miR-211-5p, miR-15b-5p, miR-151-3p and miR-467a-5p were up-regulated in Treg cells; and miR-125a-5p was down-regulated in Th17 cells. We performed validation studies by RT-qPCR analysis of independent Th1, Th17 and Treg samples isolated by FACS from the spleen/lymph nodes of triple reporter mice with EAE. These confirmed the expression profiles of five of the candidate miRNAs (Fig 1f), namely miR-1247-5p, miR-122-5p, miR-125a-5p, mir-211-5p and miR-15b 5p. Notably, two of the validated miRNAs, miR-15b-5p and miR-125a-5p, have already been thoroughly studied and, consistent with our expression data, shown to be involved in Treg or Th17 differentiation and/or function [22,28–30]. Thus, for subsequent “proof-of-concept” functional assays, and since the effector cell populations are the major drivers of EAE, we focused on the candidates miR-1247-5p and miR-122-5p, highly expressed inTh1 and Th17 cells, respectively (Fig 1e and 1f). Of note, analysis of these candidates at pre-symptomatic stages showed that Th17-associated miRNAs did not exhibit significant differences at this earlier phase, while the remaining candidates displayed a consistent pattern with that observed at the peak/plateau phase (S1e Fig).

Overexpression of miR-122-5p or miR-1247 directly impacts Th17 and Th1 cells in vitro

We inquired whether these miRNAs might have a cell-intrinsic role in Th17 (miR-122-5p) or Th1 (miR-1247) cells by performing overexpression experiments in in vitro-differentiated cells (from isolated naïve CD4⁺ T cells—S2 Fig). Upon transduction with retroviral GFP reporter vectors (RVs) encoding the native stem-loop of miR-122 (RV-122) (Fig 2a), we observed an upregulation of miR-122-5p levels in in vitro-differentiated Th17 cells (S3a Fig). While no differences were observed in the frequency of IL-17A⁺ (Fig 2b and 2c), Th17 cells overexpressing miR-122 had a significantly lower proliferation rate when compared with control cells (RV-empty), as observed by the frequency of proliferating GFP⁺ T cells upon CTV labeling as well as their replication index (Fig 2d–2f). Notably, miR-122 overexpression did not impact Th1 or Treg cell differentiation in vitro (S3b and S3c Fig). These results suggest that miR-122-5p negatively regulates Th17 cell function by limiting their proliferative capacity.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. miR-122-5p and miR-1247 intrinsically impact Th17 cell proliferation and Th1 cell function.

(a) Experimental setup: naive CD4⁺ T cells were sorted from WT mice and differentiated in vitro towards Th17 cells for 4 days. At day 1, cells were transduced with retroviral vectors encoding the precursor forms of miR-122 and differentiation was analyzed by flow cytometry. Empty vector was used as control. (b) Representative flow cytometry plots of IL-17A⁺ cells in retrovirally transduced Th17 cells expressing a control vector (RV-empty) or miR-122 (RV-122) and frequency of IL-17A⁺ cells in retrovirally transduced Th17 cells expressing a control vector (RV-empty) or miR-122 (RV-122) from 5 independent experiments (c). Dotted lines link cells from the same mouse. (d) Representative histogram of proliferating CTV⁺ cells in retrovirally transduced Th17 cells expressing a control vector (RV-empty) or miR-122 (RV-122). (e) Frequency of proliferating CTV⁺ cells in retrovirally transduced Th17 cells expressing a control vector (RV-empty) or miR-122 (RV-122). Results are mean ± SD from 2 independent experiments (n = 4). *p < 0.05 and **p < 0.01 (two-way ANOVA followed by Sidak’s multiple comparisons test). (f) Replication index of retrovirally transduced Th17 cells expressing a control vector (RV-empty) or miR-122 (RV-122). Dotted lines link cells from the same mouse. *p < 0.05 (Paired t test). (g) Experimental setup: naive CD4⁺ T cells were sorted from WT mice and differentiated in vitro towards Th1 cells for 4 days. At day 1, cells were transduced with retroviral vectors encoding the precursor forms of miR-1247 and differentiation was analyzed by flow cytometry. Empty vector was used as control. (h) Representative flow cytometry plots of IFN-γ⁺ cells in retrovirally transduced Th1 cells expressing a control vector (RV-empty) or miR-1247 (RV-1247) and frequency of IFN-γ⁺ cells in retrovirally transduced Th1 cells expressing a control vector (RV-empty) or miR-122 (RV-122) from 5 independent experiments (i). Dotted lines link cells from the same mouse. *p < 0.05 (Paired t test). (j) Representative histogram of proliferating CTV⁺ cells in retrovirally transduced Th1 cells expressing a control vector (RV-empty) or miR-1247 (RV-1247). (k) Frequency of proliferating CTV⁺ cells in retrovirally transduced Th1 cells expressing a control vector (RV-empty) or miR-1247 (RV-1247). Results are mean ± SD from 3 independent experiments (n = 5). (l) Replication index of retrovirally transduced Th1 cells expressing a control vector (RV-empty) or miR-1247 (RV-1247). Dotted lines link cells from the same mouse. The data underlying this Figure can be found in S1 and https://doi.org/10.5281/zenodo.21341947.

https://doi.org/10.1371/journal.pbio.3003271.g002

In what refers to miR-1247-5p, its overexpression with retroviral vectors in differentiating Th1 cells (Figs 2g and S3d) led to a decreased frequency of IFN-γ⁺ CD4⁺ T cells (Fig 2h and 2i), but did not affect cell proliferation (Fig 2j–2l), nor Treg or Th17 in vitro polarization (S3c and S3e Fig). These results indicate that miR-1247-5p selectively downregulates IFN-γ production in Th1 cells.

miR-122-5p is a peripheral brake to Th17 pathogenicity that is overruled in the CNS

We next aimed at dissecting the molecular cues that regulate miR-122-5p and miR-1247-5p expression. For this, we sorted naïve CD4⁺ T cells from the spleen and lymph nodes of WT mice and cultured them in the presence of different cytokines involved in Th17 or Th1 cell differentiation, respectively. Starting with the Th17-biased miR-122-5p, its expression reached maximal levels in the presence of IL-6 and TGF-β, but decreased when stimulated with IL-1β and especially with IL-23 (Fig 3a). IL-6 and TGF-β are the key cytokines involved in the polarization of naïve CD4⁺ T cells into the Th17 subset, while IL-1β and IL-23 are critical for lineage maintenance and the acquisition of a pathogenic phenotype [26]. Thus, our data suggests that miR-122 might limit the pathogenicity of Th17 cells but is downregulated in a strong (IL-23/IL-1β-rich) inflammatory environment. To build on this premise in vivo, we characterized miR-122-5p expression levels in CD4⁺ T cell subsets isolated from the CNS of triple Foxp3^hCD2.Ifng^YFP.Il17a^GFP reporter mice challenged with EAE. Interestingly, our data showed that miR-122-5p was markedly downregulated in cells isolated from the CNS compared with peripheral Th17 cells (Fig 3b). To further provide biological context to these observations, we performed mRNA-seq in Th17 cells isolated from both sites, i.e., spleen/lymph nodes versus CNS. This showed that genes associated with a pathogenic phenotype—such as those encoding key cytokines as IFN-γ and GM-CSF (Ifng and Csf2), chemokines (Ccl4 and Cxcl3) and TFs (Tbx21 and Runx1) - were all increased in CNS-derived Th17 cells when compared to peripheral Th17 cells (Fig 3c) [26]. These results link miR-122 expression dynamics with the pathophysiology of EAE, with its downregulation in the CNS associating with enhanced pathogenicity of Th17 cells.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. miR-122 and miR-1247 act as brakes on pro-inflammatory Th17/Th1 phenotypes that are subverted in the CNS.

(a) RT-qPCR analysis of miR-122 expression in CD4⁺ T cells cultured in vitro for 4 days in the presence of indicated cytokines and antibodies. Results are presented relative to miR-423-3p expression and represent mean ± SD (n = 6–7). *p < 0.05 and **p < 0.01 (non parametric Mann–Whitney test). (b) RT-qPCR analysis of miR-122 expression in Treg, Th1 and Th17 cells isolated and pooled from the periphery (lymph nodes and spleen) and the CNS of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice at EAE peak plateau stage. Results are presented as fold vs. Treg in each specific tissue (ddct) and represent mean ± SD (n = 7). miR-423-3p was used as reference gene. *p < 0.05 (Wilcoxon test). (c) Gene expression determined by RNA-seq of pathogenic vs. non-pathogenic signature genes in Th17 cells isolated and pooled from the periphery (lymph nodes and spleen) and the CNS of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice at EAE peak plateau stage. Results are represented as mean of log fold change (logFC) between periphery and CNS Th17 cells. (d) RT-qPCR analysis of miR-1247 expression in CD4⁺ T cells cultured in vitro for 4 days in the presence of indicated cytokines and antibodies. Results are presented relative to miR-423-3p expression and represent mean ± SD (n = 2–12). *p < 0.05 and p < 0.01 (non parametric Mann–Whitney test). (e) RT-qPCR analysis of miR-1247 expression in Treg, Th1 and Th17 cells isolated and pooled from the periphery (lymph nodes and spleen) and the CNS of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice at EAE peak plateau stage. Results are presented as fold vs. Treg in each specific tissue (ddct) and represent mean ± SD (n = 7). miR-423-3p was used as reference gene. *p < 0.05 (Wilcoxon test). (f) Gene expression determined by RNA-seq of self-regulation vs. proinflammatory genes in Th1 cells isolated and pooled from the periphery (lymph nodes and spleen) and the CNS of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice at EAE peak plateau stage. Results are represented as mean of log fold change (logFC) between periphery and CNS Th1 cells. The data underlying this Figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003271.g003

miR-1247 limits Th1 pathogenicity in the periphery but is subverted in the CNS

We performed similar studies for miR-1247, the Th1-biased miRNA that downregulated IFN-γ production in vitro (Fig 2g). Although miR-1247-5p was upregulated on in vivo-generated Th1 cells (Fig 1e and 1f), its expression levels were low in Th1 cells differentiated in vitro in the presence of IL-12 and anti-IL-4 antibody (Fig 3d). We also tested IL-18, IL-15 and IFN-γ, since IL-18 and IL-15 can promote IFN-γ production in the presence of IL-12 [31]; and IFN-γ regulates its own production in an autocrine loop [32]. However, none of these cytokines was able to induce miR-1247-5p expression in Th1 cells (Fig 3d). We then reasoned that, as miR-1247-5p limits IFN-γ production by Th1 cells, its expression might be controlled by anti-inflammatory cytokines. Indeed, we found that TGF-β and IL-10 were the key inducers of miR-1247-5p expression in vitro, even in the presence of IL-12 (Fig 3d). Moreover, when we challenged our triple reporter mice with EAE, we observed that, similarly to miR-122-5p (Fig 3b), the expression levels of miR-1247-5p collapsed in Th1 cells isolated from the CNS when compared to peripheral Th1 cells (Fig 3e). Furthermore, upon mRNA-seq analysis of Th1 cells isolated from both sites (spleen/lymph nodes versus CNS), we established a CNS-specific pro-inflammatory signature, including the signature Th1 cytokine (Ifng) and TF (Tbx21), as well as chemokines (Ccl3 and Ccl4) and receptors (Il12rb2, Il18r1, Il18rap), which contrasted with a set of “self-regulatory genes”, including Il10, that were upregulated in Th1 cells from the secondary lymphoid organs (Fig 3f). These data suggest that miR-1247 is part of an anti-inflammatory (TGF-β/ IL-10-mediated) regulatory loop that is subverted once Th1 cells invade the CNS and shut down its expression.

Identification of putative mRNA targets of miR-122-5p in Th17 cells

Given that miRNAs exert their functions by suppressing mRNAs, and each miRNA can impact the expression of many mRNAs [33], we next aimed at identifying the repertoires of mRNAs directly regulated by miR-122-5p and miR-1247-5p. We devised a strategy based on differential Ago2-immunoprecipitation followed by RNA-seq (Ago2 RIP-seq) in lineage-specific cells overexpressing each candidate miRNA (upon retroviral transduction) versus control cells (mock-transduced) (S4a Fig). Briefly, naïve CD4⁺ T cells were sorted and activated under polarizing conditions (Th1 or Th17 as appropriate), transduced at day 1 with pre-miR RV particles (as used above for the overexpression experiments) and further incubated for 36 h, when GFP⁺ T cells were sorted and subjected to UV-cross-linking and Ago2 immunoprecipitation with high-affinity antibody (S4a Fig).

Starting with miR-122-5p, RT-qPCR analysis after the Ago-IP confirmed its enrichment in the Argonaute complex of retrovirally miR-122-transduced Th17 cells (RV-122) relative to the empty vector control (RV-empty), whereas other miRNAs, such as miR-126a-5p or miR15b-5p, showed similar levels between RV-122 and RV-empty-treated cells (S4b Fig). The results of the differential Ago2 RIP-seq in Th17 cells identified 587 mRNAs overexpressed in miR-122-5p-transduced Th17 cells, of which 339 (57.8%) had predicted miR-122-5p binding sites (S4c Fig). From these, as proof-of-concept, we inspected our mRNA-seq data for those upregulated in the CNS compared to peripheral Th17 cells, based on the previous observation that miR-122-5p had the opposite expression pattern (Fig 3b). This retrieved 27 mRNAs (FC > 1.5, p < 0.05) (Fig 4a), including 16 whose function was of relevance for Th17 cell biology and/or proliferation (S2 Table). Based on literature curation, we further analyzed the expression levels of 5 genes (Pttglip, Nudt4, Il18rap, Spred2 and Sap130) in retrovirally transduced Th17 cells with the native stem loop of miR-122 (RV-122), and observed that Nudt4, Il18rap and Sap130 were specifically downregulated in miR-122 overexpressing (versus control) cells, further reinforcing their potential to be functional targets of miR-122 in the context of Th17 cells in EAE (Fig 4b). Finally, to biochemically validate these 3 mRNAs as direct miR-122-5p targets, we performed luciferase assays. The three putative targets have predicted miR-122-5p binding sites in the 3’UTR of Nudt4 and Il18rap and the CDS of Sap130 (Fig 4c). To determine whether miR-122-5p was directly targeting these mRNAs, we designed reporter constructs in a pmirGLO Dual-luciferase miRNA target expression vector for the 3′UTRs of Nudt4 and Il18rap and the CDS of Sap130. We additionally included a negative control vector containing a 3′UTR without miR-122-5p binding sites and a positive control containing the 3′UTR of Cpeb1, a previously validated miR-122-5p target [34]. We transiently transfected these constructs into human embryonic kidney (HEK) 293 T cells together with an expression plasmid for either miR-122-5p or an unrelated miRNA. Co-transfection of miR-122-5p (but not an unrelated miRNA) showed a significant repression of luciferase activity compared with control for the three mRNA targets (Fig 4c). Mutations in the miR-122-5p binding sites of Nudt4, Il18rap and Sap130 led to significant recoveries of luciferase levels, attesting specificity (Fig 4c). To further confirm that these targets mediate the repressive effect of miR-122 in Th17 cell proliferation, we knocked-out Nudt4, Il18rap and Sap130 in vitro in naïve CD4⁺ T cells undergoing Th17 cell differentiation using CRIPSR-Cas9 gene editing, namely by electroporating them with a ribonucleoprotein (RNP) consisting of Cas9 combined with a single-guide RNA (sgRNA) against each target (S3 Table) or Cas9 alone as a control [35]. Additionally, we included an sgRNA against the surface marker Cd5 as a positive control for the efficiency of the CRISPR approach, and observed a decrease in the frequency of CD5⁺ cells from 99% to around 4% in Cd5 knocked-out cells comparing to controls (S5 Fig). Importantly, we observed a decreased proliferative capacity (as indicated by CTV dilution) of Th17 cells upon Nudt4 and Sap130 knock-out compared to the control condition (Fig 4d and 4e), thus recapitulating the in vitro phenotype of Th17 cells overexpressing miR-122.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Identification of putative mRNA targets of miR-122-5p and miR-1247-5p in Th17 and Th1 cells.

(a) Gene expression determined by RNA-seq of candidate miR-122-5p mRNA targets that are upregulated in Th17 cells isolated from the CNS of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice at EAE peak plateau stage when compared to Th17 isolated from the periphery (lymph nodes and spleen) of the same mice (FC > 1.5; p < 0.05). Results are represented as mean of log fold change (logFC) between periphery- and CNS-derived Th17 cells. (b) RT-qPCR analysis of candidate miR-122-5p mRNA targets in retrovirally transduced Th17 cells expressing a control vector (RV-empty) or miR-122 (RV-122). Results are presented as fold vs. RV-empty (n = 4). Gray dashed lines link cells from the same mouse. Black dashed line represents the baseline expression level of controls. *p < 0.05 (paired Student t test). (c) Luciferase activity in HEK293T cells co-transfected with a pmirGLO Dual-Luciferase miRNA target expression vector containing either the WT or mutated miR-122-5p-binding sites of Sap130, Il18rap and Nudt4 mRNAs plus a vector containing miR-122-5p or an unrelated miRNA. An unrelated construct (without miR-122-5p binding sites) and a positive construct (Cpeb1) were included. Schematic representation of miR-122-5p putative binding sites of in the miR response elements of Sap130, Il18rap and Nudt4 mRNAs are included. Results are presented as fold vs. unrelated miRNA and represent mean + SD (n = 5). *p < 0.05 and **p < 0.01 (paired students t test). Representative histogram of proliferating CTV⁺ cells (d) and replication index (e) of Th17 cells knocked-out (CRISPR-mediated) for target genes: Sap130, Il18rap and Nudt4. CD5 was used as a positive control for CRISPR-Cas9 gene edition. Data (n = 3) are normalized to control cells (Cas9). *p < 0.05 and **p < 0.01 (Paired t test). (f) Gene expression determined by RNA-seq of candidate miR-1247-5p mRNA targets that are upregulated in Th1 cells isolated from the CNS of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice at EAE peak plateau stage when compared to Th1 isolated from the periphery (lymph nodes and spleen) of the same mice (FC > 1.5; p < 0.05). Results are represented as mean of log fold change (logFC) between periphery- and CNS-derived Th1 cells. (g) RT-qPCR analysis of candidate miR-1247-5p mRNA targets in retrovirally transduced Th1 cells expressing a control vector (RV-empty) or miR-1247 (RV-1247). Results are presented as fold vs. RV-empty (n = 3). Gray dashed lines link cells from the same mouse. Black dashed line represents the baseline expression level of controls. *p < 0.05 (paired Student t test). (h) Luciferase activity in HEK293T cells co-transfected with a pmirGLO Dual-Luciferase miRNA target expression vector containing either the WT or mutated miR-1247-5p-binding sites of Rasal3 and Gng2 mRNAs plus a vector containing miR-1247-5p or an unrelated miRNA. An unrelated construct (without miR-1247-5p binding sites) and a positive construct (Trappc9) were included. Schematic representation of miR-1247-5p putative binding sites of in the miR response elements of Rasal3 and Gng2 mRNAs are included. Results are presented as fold vs. unrelated miRNA and represent mean + SD (n = 5). *p < 0.05 and **p < 0.01 (paired Student t test). The data underlying this Figure can be found in S1 and https://doi.org/10.5281/zenodo.21341947.

https://doi.org/10.1371/journal.pbio.3003271.g004

Together, these data suggest that miR-122-5p is able to repress Th17 cell proliferation in vitro by targeting Nudt4 and Sap130. Furthermore, it provides a much larger Ago2 RIP-seq data set that constitutes a resource for those interested in assessing miR-122-5p in future studies on T cell differentiation.

Identification of putative mRNA targets of miR-1247-5p in Th1 cells

We employed the same experimental approach to identify miR-1247-5p targets in the context of Th1 cell differentiation. The differential Ago2 RIP-seq (S4a Fig) showed that the RNA-induced silencing complex (RISC) of miR-1247-overexpressing Th1 cells (as confirmed by RT-qPCR, S4b Fig) was enriched (compared to mock-transduced Th1 cells) in 644 mRNAs, of which 423 (65.7%) had predicted binding sites for miR-1247-5p (S4c Fig). Upon examination of our mRNA-seq data for Th1 cells, we found 41 mRNAs upregulated (FC > 1.5 and p > 0.05) in the CNS (Fig 4d), of which 18 had relevant functions for Th1 cell biology (S2 Table). Based on literature curation, we selected 5 genes—Gng2, Arfip2, C2h2c3, Rasal3 and Wipf1—to further test in vitro, i.e., upon retroviral transduction of the native stem-loop of miR-1247 in Th1 cells (Fig 4e). Of these, Gng2 and Rasal3 were validated as miR-1247-5p targets both in vitro, as shown by their downregulation upon miR-1247-5p overexpression (Fig 4e), and in luciferase assays demonstrating direct binding of Gng2 and Rasal3 using the same strategy as described above for miR-122-5p target genes (Fig 4f). More specifically, the 3′UTR regions of both wild-type and mutant Gng2 and Rasal3 (Fig 4f) were introduced into a pmirGLO Dual-luciferase miRNA target expression vector and co-expressed with either miR-1247-5p or a control unrelated miRNA. These assays showed a decrease in luminescence in the presence of the miRNA that was abolished when the 3′UTR target regions were mutated (Fig 4f). Thus, our results indicate that miR-1247-5p targets Gng2 and Rasal3. Moreover, they provide the community with a comprehensive Ago2 RIP-seq data set for additional studies on miR-1247-5p in the context of effector T cell differentiation.

miR-122-5p inhibition increases the frequency of IL-17A⁺ cells and anticipates EAE onset

Next, to assess if the candidate miRNAs play a role in EAE pathophysiology, we treated mice challenged with EAE with synthetic antagomiRs and evaluated their impact on disease course and on the frequency of Teff and Treg subsets in the lymph nodes of antagomiR-treated mice at the end of the experiment. As depicted in Fig 5a, we injected the antagomiRs of the candidate miRNAs at days 7, 9 and 11 after immunization, which corresponds to a preclinical stage (onset of symptoms usually occurs around day 10). Focusing first on miR-122-5p antagomiR treatment—which efficiently reduced miR-122-5p expression levels (S6a Fig) —we observed an anticipation of EAE disease onset as mice started to develop symptoms and lose weight earlier than those treated with the control antagomiR (Fig 5b). Importantly, all mice that were subjected to miR-122-5p antagomiR injections were symptomatic already at day 12 post immunization (p.i.), whereas the control group needed an additional 5 days (day 17 p.i.) to present symptoms, thus entering the peak plateau stage later on (S6b Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. miR-122-5p inhibition antecipates EAE onset.

(a) Experimental setup: triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice were immunized s.c. in both flanks with 100 µg of myelin oligodendrocyte glicoprotein (MOG)^35-55 peptide emulsified in CFA solution (day 0). On the day of immunization and 2 days later, mice received a retro-orbital injection of 200 ng of pertussis toxin. At days 7, 9 and 11, mice received 5 µmol/kg of specific antagomiRs (retro-orbital injection). AntagomiR for cel-miR-67-3p was used as control. (b) Mice treated with control and miR-122-5p antagomiRs were scored for EAE clinical signs and weighed daily for 28 days. Results are mean ± SD from 3 independent experiments (n = 9–11). *p < 0.05 and **p < 0.01 (two-way ANOVA followed by Sidak’s multiple comparisons test). (c) Representative flow cytometry plots of IFN-γ⁺ and IL-17A⁺ CD4⁺ T cells in the cervical and lumbar lymph nodes of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice with EAE and treated with control and miR-122-5p antagomiRs. (d) Frequency and absolute numbers of IL-17A⁺ CD4⁺ T cells in the cervical and lumbar lymph nodes of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice with EAE and treated with control and miR-122-5p antagomiRs. Results are mean + SD from 3 independent experiments (n = 9–11). Each dot represents an individual mouse. *p < 0.05 (non-parametric Mann–Whitney test). The data underlying this Figure can be found in S1 and https://doi.org/10.5281/zenodo.21341947.

https://doi.org/10.1371/journal.pbio.3003271.g005

Analysis of cervical and lumbar LNs (cLNs) showed no differences in the total CD4⁺ T cell frequency S6c Fig), but an increased percentage of IL-17A⁺ CD4⁺ T cells in mice subjected to miR-122-5p antagomir treatment when compared with controls (Fig 6c and 6d), consistent with the early onset of the disease. We found no differences in the frequency and total cell numbers of Th1 (IFN-γ⁺) and Treg (Foxp3⁺) cell populations in this case (Figs 6d, S6d, and S6e). Moreover, no differences were observed in the frequency of other immune cells between mice treated with miR-122-5p or control antagomiRs (S6f Fig) based on the gating strategy highlighted in S7a Fig (T cell populations) and S7b Fig (other immune cell populations).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. miR-1247-5p upregulation limits EAE severity.

(a) Mice treated with control and miR-1247-5p antagomiRs were scored for EAE clinical signs and weighed daily for 28 days. Results are mean ± SD from 3 independent experiments (n = 10–11). *p < 0.05 and **p < 0.01 (two-way ANOVA followed by Sidak’s multiple comparisons test). (b) Representative flow cytometry plots of IFN-γ⁺ and IL-17A⁺ CD4⁺ T cells in the draining lymph nodes (inguinal, axillary and brachial) of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice with EAE and treated with control and miR-1247-5p antagomiRs. (c) Frequency and absolute numbers of IFN-γ⁺ CD4⁺ T cells in the draining lymph nodes (inguinal, axillary and brachial) of triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice with EAE and treated with control and miR-1247-5p antagomiRs. Results are mean + SD from 3 independent experiments (n = 10–11). Each dot represents an individual mouse. *p < 0.05 (non-parametric Mann–Whitney test). The data underlying this Figure can be found in S1 and https://doi.org/10.5281/zenodo.21341947.

https://doi.org/10.1371/journal.pbio.3003271.g006

To assess the selectivity of these phenotypes, we studied the impact of an anti-miR-126a antagomiR. While this resulted in the downregulation of miR-126a (S8a Fig), it had no impact neither on clinical scores nor on weight loss, since the specific antagomir-treated mice followed a similar disease course as the control group (S8b Fig), and no differences were found in the frequency of IFN-γ⁺ and IL-17A⁺ nor Treg (Foxp3⁺) cell populations (S8c Fig). These data highlight our finding that miR-122-5p inhibition selectively increases the frequency of Th17 cells and anticipates EAE onset.

Upregulation of miR-1247 limits IFN-γ production by Th1 cells and reduces EAE severity

We next analyzed the in vivo effects of miR-1247 manipulation. Unlike the previous antagomiRs (S6a Fig), the miR-1247-5p antagomir paradoxically led to increased levels of miR-1247-5p in various tissues (S9a Fig), thus constituting an overexpression strategy. Consistent with increased miR-1247-5p activity, we observed a trend towards downregulation of Trapc9 and Synj2bp—miR-1247-5p target candidates identified by Ago2 CLIP-seq [36]—in matching organs (S9b Fig). While the progression of symptoms followed a similar course (S10a Fig), we observed decreased EAE severity, as well as reduced weight loss, in mice overexpressing miR-1247-5p compared with the control group (Fig 6a). Furthermore, this phenotype was associated with a selective decrease in the numbers of IFN-γ⁺ CD4⁺ T cells (Fig 6b and 6c), without affecting the total CD4⁺ T cell frequency (S10b Fig), in the draining lymph nodes of miR-1247-5p-treated group. Of note, the frequency of other immune cells also did not change upon miR-1247-5p overexpression (S10c Fig). Thus, the upregulation of miR-1247 expression restricts IFN-γ production by Th1 cells and reduces EAE severity.

Collectively, our in vivo experiments have identified two novel miRNAs, miR-1247-5p and miR-122, that control Th1 and Th17 cell pathogenicity and thus impact the disease course of EAE.

Resource availability

Lead contact: Anita Quintal Gomes, anita.gomes@gimm.pt

Materials availability: All unique/stable reagents generated in this study are available from the lead contact without restriction.

Ethics statement

All experiments involving animals were approved by the Ethics Animal Welfare Body of Gulbenkian Institute for Molecular Medicine (ORBEA-GIMM), set up in accordance with Portuguese law (article 34 of Decreto-Lei 113/2013, transposed from the European Directive 2010/63/EU), and submitted to the local competent authority Direcção-Geral de Alimentação e Veterinária (DGAV) for authorization

Mice

All mice used were females 6–12 weeks of age. C57Bl/6J (B6) WT mice were purchased from Charles River Laboratories. Triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice were generated and bred in-house by crossing the following single reporter strains: B6.Foxp3hCD2 (http://www.informatics.jax.org/allele/MGI:4950682) [56], Great (Ifng-YFP) (http://www.informatics.jax.org/allele/MGI:5317415) [25], and IL17-GFP (http://www.informatics.jax.org/allele/MGI:5426354) [26]. B6.Foxp3hCD2 and Great mice were obtained from the Jackson Laboratory and Il17a-GFP mice were from Biocytogen, LLC. Mice were bred and maintained in the specific pathogen-free animal facilities of Gulbenkian Institute for Molecular Medicine (Lisbon, Portugal). Euthanasia was performed by CO₂ inhalation and anesthesia was performed by isofluorane inhalation.

EAE induction and scoring

Triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice were immunized s.c. in both flanks with 100 µg of MOG 35–55 peptide (MEVGWYRSPFSRVVHLYRNGK) (Eurogentec S.A.) emulsified in CFA solution (4 mg/ml of heat-inactivated M. tuberculosis in IFA) (Difco Laboratories). On the day of immunization and 2 days after, mice received 200 ng pertussis toxin (PTx) (List Biological Laboratories) in 100 µl PBS i.v. Mice were weighed daily and scored for EAE clinical signs. In brief, the score system ranged from 0 to 5, with 0.5 increments, being score 0 attributed to animals with no clinical signs of EAE and 5 representing death. Score 1 consisted in limp tail; score 2 consisted in limp tail together with hind legs weakness; score 3 consisted in complete limb paralysis; and finally, score 4 consisted in complete hind leg paralysis and partial front paralysis. Mice were euthanized if they lose 20% of body weight or if they scored 4 for 2 consecutive days.

AntagomiR treatment

AntagomiRs were custom synthesized according to [56] (Merck). Antagomir sequences are as follows: antagomiR-1247-5p—5′-U*C*CGGGGACGAACGGGACG*G*G*U*-chol; antagomiR-122-5p—5′-C*A*AACACCAUUGUCACACU*C*C*A*-chol; antagomiR-126a-5p—5′-C*G*CGUACCAAAAGUAAUA*A*U*G*-chol; AntagomiR for cel-miR-67-3p was used as negative control: antagomiR-ctl—5′-U*C*UACUCUUUCUAGGAGGUUG*U*G*A*-chol. All ribonucleotides were 2′-O-methyl modified and (*) represents a phosphorothioate modification of the backbone. A cholesterol molecule was added at the 3′-end of the oligonucleotides. Lyophilized antagomiRs were resuspended in RNase-free water at the desired concentration at room temperature for 30 min with slight shaking. Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice with EAE were treated with 5 µg/kg specific antagomiRs at days 7, 9 and 11 after immunization, which corresponds to a preclinical stage (onset of symptoms occurs around day 10) through i.v. injection in the eye. Treatments were randomized per cage and EAE scoring was performed by a researcher blinded to each animal’s treatment group.

Monoclonal antibodies

The following anti-mouse fluorescently labeled monoclonal antibodies (mAbs) were used (antigens and clones): CD3 (145.2C11), CD4 (GK1.5 and RM4-5), CD25 (PC61), CD44 (IM7), CD62L (DREG.55), IFN-γ (XMG1.2), IL-17A (eBio17B7), Foxp3 (FJK-16s), CD5 (53-7.3), CD8 (53–6.7), TCRδ (GL3), F4/80 (BM8), NK1.1 (PK136), CD19 (6D5), Ly6G (1A8), Ly6C (HK1.4), CD11B (M1/70), CD11C (N418) and CD45 (30-F11). The anti-human CD2 (RPA-2.10) antibody was also used. Antibodies were purchased from BD Biosciences, eBiosciences, or BioLegend.

Cell preparation, cell sorting, and flow cytometry and analysis

Cell suspensions were obtained from spleens, lymph nodes (cervical, axillary, brachial, inguinal and lumbar), and the central nervous system (brain and spinal cord). For the preparation of lymph nodes and spleen, tissues were filtered through a 70-µm cell strainer and erythrocytes from the spleen were osmotically lysed in red blood cell lysis buffer (BioLegend). For the preparation of brain and spinal cord, mice were perfused through the left cardiac ventricle with cold PBS. Brain and spinal cord were dissected and cut into pieces, and further digested with collagenase type IV (1.5 mg/ml; Roche) and DNase I (0.10 mg/ml) (Sigma-Aldrich) in RPMI 1640 containing 5% fetal bovine serum (FBS) for 30 min, at 37 °C. Mononuclear cells were isolated by passing the tissue through a 100-µm cell strainer, followed by a 40%–70% Percoll (Sigma-Aldrich) gradient and 30-min centrifugation at 2,400 rpm. Mononuclear cells were recovered from the interphase, resuspended, and used for further analysis. For cell surface staining, single-cell suspensions were incubated in the presence of anti-CD16/CD32 (eBioscience) with saturating concentrations of combinations of the antibodies listed above for 15 min. For viability assessment, cells were stained in the presence of LIVE/DEAD Fixable Near-IR (Thermo Fisher Scientific), Zombie Aqua (BioLegend) or Zombie Violet (BioLegend), according with manufacturer’s instructions.

For intracellular cytokine staining or reporter analysis, cells were stimulated with PMA (phorbol 12-myristate 13-acetate) (50 ng/ml) and ionomycin (1 µg/ml), in the presence of Brefeldin A (10 µg/ml) (all from Sigma) for 3–3.5 h at 37 °C. Cells were stained for the identified above cell surface markers, fixed for 30 min at 4 °C with the Foxp3/Transcription Factor Staining Buffer set (eBioscience), and finally incubated for 45–60 min at 4 °C with identified above intracellular antibodies in permeabilization buffer. Cells were analyzed using FACSFortessa (BD Biosciences) or CytekAurora (Cytek), and FlowJo software (Tree Star).

For cell sorting, single cell suspensions were stained for cell surface markers as mentioned above and then electronically sorted on a FACSAria (BD Biosciences). Cell sorting based on reporter genes, included stimulation with PMA/ionomycin/Brefeldin A before the staining.

For EAE experiments with triple Foxp3-hCD2.Ifng-YFP.Il17a-GFP reporter mice, T cell subsets were isolated based on the expression of reporter proteins as follows: Th1 cells were GFP⁻YFP⁺, Th17 cells GFP⁺YFP⁻, and Treg cells were defined as GFP⁻YFP⁻hCD2⁺.

RNA isolation, complementary DNA synthesis, and RT-qPCR

RNA was isolated from fluorescence-activated cell sorting (FACS)—sorted cell populations using an miRNeasy Mini kit (Qiagen). For total RNA, reverse transcription was performed with random oligonucleotides (Invitrogen) using Moloney murine leukemia virus reverse transcriptase (Promega). For miRNA, reverse transcription was performed with a Universal cDNA Synthesis kit II (Exiqon). For total RNA samples, relative quantification of specific complementary DNA (cDNA) species to endogenous references HPRT, β-2-microglobulin or β-actin was carried out using SYBR on a ViiA7 cycler (Applied Biosystems). Primers were either designed manually or by the Universal ProbeLibrary Assay Design Center (Roche), and their sequences are indicated in the S4 Table. For miRNA samples, relative quantification of specific cDNA species to reference miR-423-3p was carried out using SYBR on a ViiA7 cycler (Applied Biosystems). The respective miRNA LNA primers were used (Exiqon). In both cases, relative quantification was performed using ddCT method as indicated in each section.

RNA- and small-RNA-sequencing

Deep sequencing (small non-coding RNA and RNA-seq) was performed at the GeneCore facility of EMBL (http://www.genecore.embl.de/). For RNA-seq samples were processed with NEBNext Ultra II RNA Library kit for Illumina. Specifically, the protocol used was that for purified mRNA or rRNA-depleted RNA, with an optimization of the following steps to accommodate the low input: Fragmentation time (4.1.3) 10 min, First strand synthesis step 2 (4.2.3) was increased to 50 min for a longer insert, Adaptor was diluted to 1:30 (4.5.4) and final PCR (4.8.3) was increased to 18 cycles. For small RNA-seq samples were processed with the NEBNext Small RNA Library Prep Set for Illumina.

miRNA-Seq data analysis

Sequenced reads were processed using the CAP-miRSeq pipeline [57]. Namely, reads were quality filtered and adaptors removed using cutadapt. Filtered reads were aligned to the mouse genome (GRCm38) using bowtie [58]. miRNA prediction and quantification was performed using miRDeep2 [59], where known miRNAs were obtained from mirbase (version 21) [60]. Differential expression analysis was performed using the edgeR R package [61]. Namely, count data was normalized using the TMM normalization [62], and moderated bayesian statistics [63] was applied to obtain differentially expressed genes (FDR < 0.05 and [Log2FC] > 1) between conditions. Comparisons were performed between Treg and Th (i.e., combining Th17 and Th1), Treg and Th1, and Treg and Th17.

RNA-Seq data analysis

Sequenced reads were quality filtered and adaptors removed using cutadapt. Filtered reads were aligned to the mouse genome (GRCm38) using Hisat2 [64]. Gene count tables were obtained using featureCounts [65] against the mouse gene models (gencode M16).

In vitro CD4⁺ T cell subset polarization

Naive CD4⁺ T cells (CD3⁺ CD4⁺ CD25⁻ CD62L⁺CD44^low) were FACS-sorted from the LNs and spleen of B6 WT mice and differentiated under polarizing conditions for 4 days. For Th1 cell culture conditions, naive CD4⁺ T cells were incubated with plate-bound anti-CD3ε (145.2C11) and soluble anti-CD28 (37.51) antibodies (both at 2 μg/ml) in the presence of IL-12 (10 ng/ml) and neutralizing anti–IL-4 antibody (11B11) (5 μg/ml). For Treg cells, naive CD4⁺ T cells were incubated with plate-bound anti-CD3ε and soluble anti-CD28 (both at 2 μg/ml) in the presence of IL-2 (10 ng/ml) and TGF-β (2 ng/ml). For Th17 cell polarization, naive CD4⁺ T cells were incubated with plate-bound anti-CD3ε (1 μg/ml) and anti-CD28 (10 μg/ml) in the presence of IL-1β (10 ng/ml), IL-6 (20 ng/ml), IL-23 (20 ng/ml), TGF-β (2 ng/ml) and neutralizing anti–IFN-γ antibody (R4-6A2) (10 μg/ml). All cytokines were from PeproTech, except TGF-β and IL-23, which were from R&D systems.

Retroviral overexpression of miR-122 and miR-1247

The retroviral constructs encoding either mmu-miR-122 and mmu-miR-1247 were generated by inserting the respective native pre-miRNA sequences flanked by about 200 bp into a modified pMig.IRES-GFP retroviral vector (Addgene #9,044), as previously reported [66]. For retroviral transduction, CD4⁺ T cells were sorted from LNs and spleen of C57BL/6J mice and polarized towards Th17, Th1 or Treg cells as mentioned above. At polarization day 1, cells were transduced with retrovirus encoding the precursor forms of miR-122 (RV-122), miR-1247 (RV-1247) or the control vector (RV-empty), in the presence with polybrene (8 µg/ml; Sigma-Aldrich). At day 4, differentiation was analyzed by flow cytometry cells and gene expression by qPCR.

Proliferation assay

For analysis of proliferation, naïve CD4⁺ T cells were stained with 1 mM of CellTrace Violet (Thermo Fisher Scientific) in PBS for 20 min at room temperature. Reactions were quenched by washing with ice-cold RPMI medium supplemented with 10% (v/v) FBS. CTV-labeled cells were then polarized towards Th1 or Th17 cells and, on the next day, retrovirally transduced with retrovirus encoding the precursor forms of miR-122 (RV-122), miR-1247 (RV-1247), respectively, or the control vector (RV-empty), as previously described or used for CRISPR studies. After 4 days, proliferation was assessed by flow cytometry.

Differential argonaute 2 RNA immunoprecipitation followed by deep sequencing

Th17 and Th1 were retrovirally transduced with RV-control vector and RV-miR-122 or RV-miR-1247 vector, respectively. Seventy-two hours after transduction, GFP⁺ live cells were FACS-sorted and UV-cross-linked in Stratalinker 2400 (400 mJ cm⁻²). Cell lysates were pulled from independent experiments (~1 million Th17 and ~400,000 Th1 cells per replicate) and lysed on ice for 15 min before immunoprecipitation. Argonaute (Ago) protein immunoprecipitation was performed using Protein G Dynabeads (Invitrogen) and the 2A8 anti-Ago antibody (Millipore). Briefly, Dynabeads were coated with AffiniPure rabbit anti-mouse IgG antibody (Jackson ImmunoResearch) for 50 min at room temperature, and incubated for additional 4 h at 4 °C with the 2A8 anti-Ago antibody (both with spinning). After incubation, cell lysates were added to antibody-coated beads and incubated overnight at 4 °C while spinning. Ago2-bound RNA was then partially digested using RNase T1 (ThermoFisher Scientific) and, after proteinase K treatment, RNA was isolated using phenol chloroform isoamyl (Fisher Scientific) extraction. Deep sequencing was performed as above mentioned.

Ago2 RIP-seq data analysis

Raw Illumina sequence reads obtained by the pair-end sequencing strategy were quality-checked with the FastQC tool [67] (version 0.11.9; Babraham Bioinformatics, UK) and were filtered and trimmed, with Trimmomatic software [68]. Filtered and trimmed sequence reads were aligned with the mouse genome (genome build GRCm38, release 102) using the STAR aligner [69] (version 2.7.10b). Differential gene expression among sample groups was analyzed by the DESeq2 algorithm, implemented in the Nasqar platform for high-throughput sequencing data analysis and visualization. Predicted miRNA targets for mmu-miR-1247-5p and mmu-mir-122 were determined by the miRWalk2 platform.

Dual luciferase reporter assay

Plasmid vectors pMig-miR-122-5p and pMig-miR-1247-5p that allow for the overexpression of miR-122-5p or mir-1247-5p, respectively, were generated as described above. The binding regions of miR-122-5p targets—Nudt4, Il18rap and Sap130 and of miR-1247-5p targets Gng2 and Rasal13 and respective regions with mutations in the predicted target sequences were directly synthesized into pmiRGLO vector (GeneCust Europe). Each luciferase reporter vector carrying one of the sequences described above was co-transfected with either the pMig-miR-122-5p or pMig-miR-1247-5p expression vectors or a control empty pMig vector into HEK293 T cells (ATCC CRL-3216) using X-tremeGENE HP DNA Transfection Reagent (Roche). After 48 h, firefly and Renilla luciferase activity were measured by using the Dual-Glo Luciferase Assay System (Promega). Renilla luciferase activity served as the internal control, and relative luciferase activity was normalized to empty pMirGlo and to empty pMig-IRES-GFP.

CRISPR-mediated knockout of miR-122 candidate targets

sgRNA oligonucleotides (5 µg, Integrated DNA Technologies, designed using CRISPOR [70] and TrueCut Cas9 protein v2 (10 µg, Invitrogen) were assembled into ribonucleoprotein (RNP) complexes by incubating them together for 10 min at room temperature. FACS-isolated naïve CD4⁺ T cells from the LNs of B6 mice stained with CTV were then electroporated with RNPs or TrueCut Cas9 alone as control using the P3 Primary Cell 4D-Nucleofector X kit L (Lonza) in a 4D-Nucleofector (Lonza), set with the pulse code DN-100. Polarization medium was immediately added after electroporation and cells were polarized in vitro into Th17 as described previously until flow cytometry analysis 4 days later. Knockout of the selected targets using each sgRNA was confirmed by Sanger-seq analysis of the CRISPR-edited regions through PCR-amplified fragments with Phusion High-Fidelity DNA Polymerase, New England Biolabs from DNA of CRISPR-edited polarized CD4⁺ T cell subsets extracted using the DNeasy Blood & Tissue Kit (Qiagen) and purified using the QIAquick PCR Purification Kit (Qiagen). Analysis of Sanger-seq results was performed with ICE CRISPR Analysis, 2026 (EditCo Bio).

Statistical analysis

The statistical significance of differences between two groups was assessed using non-parametric two-tailed Mann–Whitney test or, if both groups followed a normal distribution (tested by Shapiro–Wilk normality test), using two-tailed unpaired Student t test. Wilcoxon-matched-pairs or paired t test were used for paired samples, following normality testing as above. When more than two groups were compared, two-way ANOVA followed by Sidak’s post hoc test or Kruskal–Wallis followed by Dunn’s post hoc test were performed. Unless otherwise indicated, individual values and mean are plotted or mean ± SD. p < 0.05 was considered significant and is indicated on the figures.