*2.2. Lack of Correlation between microRNA and mRNA in Th17 Cells in Control Groups: OA Patients and Healthy Subjects*

The aberrant expression of microRNAs through affecting the expression of target genes involved in Th17/Treg differentiation/function may contribute to the inflammatory responses in RA patients. To better understand the role of the examined miRNAs in Treg/Th17 cells balance, we conducted a correlation analysis between studied miRNA and expression of transcriptional factors playing a central

role in the development of T cells. Data indicated that in OA Treg cells, a strong positive correlation was observed between miR-24 and SMAD3 (r = 0.66, p = 0.03), miR-31 and SMAD4 (r = 0.64, p = 0.037), miR-146a and SMAD3 (r = 0.62, p = 0.048), as well as between miR-155 and SOCS1 (r = 0.71, p = 0.02), STAT3 (r = 0.70, p = 0.02) and SMAD3 (r = 0.78, p = 0.006; Figure 4). In OA Th17 cells, we did not observe the correlation between examined miRNA and mRNA.

**Figure 4.** Correlation analysis of microRNAs to SMAD3, SMAD4, SOCS1, and STAT3 in OA patients in Treg cells. (**A**) correlation between miR-24 and SMAD3; (**B**) correlation between miR-31 and SMAD4; (**C**) correlation between miR-146a and SMAD3; (**D**) correlation between miR-155 and SOCS1; (**E**) correlation between miR-155 and STAT3; (**F**) correlation between miR-155 and SMAD3.

We found that the expression of miR-24 was positively correlated with SOCS1 (r = 0.73, p = 0.02), expression of miR-126 with SOCS1 (r = 0.77, p = 0.013), and expression of miR-155 with SOCS1 (r = 0.65, p = 0.049) in Treg cells from healthy subjects. In Treg cells from healthy subjects, we observed a high, negative correlation between miR-155 and SMAD3 (r = 0.70, p = 0.031) and STAT3 (r = 0.66, p = 0.044; Figure 5). In HCs' Th17 cells, we also did not observe the correlation between examined miRNA and mRNA.

**Figure 5.** Correlation analysis of microRNAs to SMAD3, SOCS1, and STAT3 in HCs in Treg cells. (**A**) correlation between SOCS1 and miR-24; (**B**) correlation between SOCS1 and miR-26; (**C**) correlation between SOCS1 and miR-155; (**D**) correlation between miR-155 and SMAD3; (**E**) correlation between miR-155 and STAT3.

### *2.3. Correlation Analysis Of Related Target Genes Based on the mRNA–miRNA Interaction Network*

Subsequently, we established miRNA–mRNA coexpression network analysis to identify the possible modulating mechanisms of the mRNAs. The miRNA–mRNA coexpression network analysis was constructed in Cytoscape software. This coexpression network analysis revealed that in RA Th17 cells, expression of miR-126 is correlated with expression of miR-26 and both may regulate expression of STAT3, SMAD3, and SOCS1 (Figure 6). In RA Treg cells, we did not observe this relationship.

**Figure 6.** The network of high, significant correlation between selected microRNAs and transcriptional factors in RA patients (n = 14) in Th17 and Treg cells constructed in Cytoscape (coefficient of correlation higher than 0.6, significance < 0.05).

As it is shown in correlation networks (Figures 7 and 8), we did not find any correlation between miRNA and mRNA in Th17 cells from OA patients and healthy subjects. In OA Treg, we were able to observe correlation mostly between miR-155 and STAT3, SOCS1, and SMAD3. Moreover, SMAD3 was also correlated with miR-146a and miR-24. Additionally, we noticed a correlation between miR-31 and SMAD4.

**Figure 7.** The network of high, significant correlation between selected microRNAs and transcriptional factors in OA patients (n = 11) in Th17 and Treg cells constructed in Cytoscape (coefficient of correlation higher than 0.6, significance < 0.05).

**Figure 8.** The network of high, significant correlation between selected microRNAs and transcriptional factors in HC (n = 15) in Th17 and Treg cells constructed in Cytoscape (coefficient of correlation higher than 0.6, significance < 0.05).

Similar to RA and OA patients, we prepared a correlation network to show how tested microRNAs may affect mRNAs in our healthy controls. We did not find any regulation of microRNA and mRNA in Th17 cells from healthy subjects. On the other hand, in Treg cells from healthy subjects, we observed that miR-126 and miR-24 were positively correlated with SOCS1 and that miR-155 was negatively correlated with SMAD3 and STAT3, but positively with SOCS1 (Figure 8).
