**2. Results**

#### *2.1. Increased Expression Levels of miR-146a and miR-155 in Treg Cells from RA and OA*

Analysis of miRNAs expression levels in freshly isolated Th17 and Treg cells (Table S1) have shown that expression of miR-24 was 3 times higher in Th17 cells than in Treg cells from healthy subjects (p = 0.04). We did not see noticeable differences in OA and RA patients in the miR-24 expression level between Treg and Th17 cells (Figure 1A). We observed higher miR-26 expression levels in Treg cells from HCs than in RA Treg cells (p = 0.014). There were no differences in the miR-26 expression levels between Th17 cells and Treg cells in RA and OA patients (Figure 1B). We also observed that the miR-31 expression was 2 times higher in Th17 cells than in Treg cells from HCs (p = 0.0046, Figure 1C). For miR-146a, we noticed that its expression was 4 times higher in OA Treg cells than in OA Th17 cells (p = 0.008, Figure 1D), and 6 times higher in RA Treg cells than in RA Th17 cells (p = 0.0031, Figure 1D). MiR-155 expression was significantly higher in RA and OA Treg cells than in RA and OA Th17 cells (both p = 0.04, Figure 1E). In contrast, miR-155 expression was significantly higher in HC Th17 cells than in RA Th17 cells (p = 0.04, Figure 1E).

**Figure 1.** MicroRNA comparison in Treg and Th17 in healthy controls (HCs), osteoarthritis (OA), and rheumatoid arthritis (RA). MiR-24 level normalized to SNORD48 (**A**), miR-26 level normalized to SNORD48 (**B**), miR-31 level normalized to SNORD48 (**C**), miR-146a level normalized to SNORD48 (**D**), miR-155 level normalized to SNORD48 (**E**). Data are presented with the median as the scatterplot graph. Subtypes of Treg vs. Th17 cells within groups were analyzed by the Wilcoxon test. Multiple comparisons of HCs vs. RA vs. OA were conducted with the Kruskal–Wallis test with Dunn's post hoc. HC (n = 15), OA (n = 11), RA (n = 14). \* p < 0.05, \*\* p < 0.01.

In the next step, we examined whether there exists a relationship between the mRNA expression and miRNAs expression. For this purpose, we analyzed correlations between examined miRNA and mRNA in the subpopulation of Th17 and Treg cells in RA patients. We found that the expression of miR-26 was positively correlated with SMAD3 (r = 0.78, p = 0.002), STAT3 (r = 0.63, p=0.02), and SOCS1 (r = 0.74, p = 0.005) expression, and expression of miR-155 was positively correlated with STAT3 expression (r = 0.67, p = 0.012) in Th17 cells from patients with RA (Figure 2).

**Figure 2.** Correlation analysis of microRNA to SMAD3, STAT3, and SOCS1 in RA patients in Th17 cells. (**A**)- correlation between SMAD3 and miR-26; (**B**) correlation between STAT3 and miR-26; (**C**) correlation between STAT3 and miR-155; (**D**) correlation between SOCS1 and miR-26.

In contrast, in RA Treg cells, a strong positive correlation was observed between miR-26 and SOCS1 (r = 0.68, p = 0.02), between miR-31 and SMAD3 (r = 0.66, p = 0.04), and between miR-155 and SMAD3 (r = 0.62, p = 0.06) and SMAD4 (r = 0.66, p = 0.04). Furthermore, a significant, but negative correlation was observed between miR-26 and STAT5A (r = −0.63, p = 0.04) and between miR-126 and STAT5A (r = −0.85, p = 0.0015; Figure 3).

**Figure 3.** Correlation analysis of microRNAs to SMAD3, SMAD4, SOCS1, and STAT5A in RA patients in Treg cells. (**A**) correlation between SOCS1 and miR-26; (**B**) correlation between SMAD3 and miR-31; (**C**) correlation between SMAD3 and miR-155; (**D**) correlation between SMAD4 and miR-155; (**E**) correlation between miR-26 and STAT5a; (**F**) correlation between STAT5a and miR-126.
