miR-146a-5p mediates atherogenic signalling from immune to vascular cells

Figure 1. miR-146a-5p transfer from Mϕs to VSMCs. (A) Scheme of the coculture system for the study of cell-to-cell communication with cell types in direct contact. PET membrane inserts with a porosity of 1 μm were used, allowing the transfer of small molecules; Created with BioRender.com. (B) Schematic representation of the procedure for the identification of putative miRNAs transferred from Mϕs to VSMCs using miRNA-sequencing; Created with BioRender.com. (C) Interaction coefficient [adjusted P-value (FDR) < 10%] measured as logFC of putative transferred miRNAs. (D) Relative expression of mature miR-146a-5p measured by RT-qPCR in unstimulated Mϕs (n = 8), VSMCs (n = 9), and ECs (n = 6). (E) Relative expression of mature miR-146a-5p (n = 4) measured by RT-qPCR in unstimulated (M0), LPS-stimulated (100 ng/mL, M1), or IL-4-stimulated (20 ng/mL, M2) Mϕs for 48 h. (F and G) Relative expression of mature miR-146a-5p (F; n = 8) and pri-miR-146a (G; n = 6) measured in VSMCs after 36 h of coculture with Mϕs (CC) or in monoculture (MC). (H) Relative expression of mature miR-146a-5p (n = 6) measured in VSMCs after 36 h of coculture with either unstimulated (CC NS) or LPS-stimulated (CC LPS) Mϕs, or…

Figure 2. Transfer route of miR-146a-5p from Mϕs to VSMCs. (A) Scheme of the coculture system for studying indirect communication. In this case, cell types were plated separately: Mϕs in the PET insert, and VSMCs on the bottom of the plate; Created with BioRender.com. (B) Mature miR-146a-5p (n = 6) was measured in VSMCs after 36 h of indirect coculture with Mϕs or alone (CTR). (C) Cx43 expression by RT-qPCR (n = 9) in VSMCs infected with lentiviral vectors targeting a scrambled (shSCR) or Connexin 43 (shCx43) sequence. (D) CX43 expression by fluorescent immunohistochemistry in VSMCs infected with lentiviral vectors targeting a scrambled (shSCR) or Connexin 43 (shCx43) sequence. (E) miR-146a-5p (n = 7) measured in VSMCs after 36 h of direct coculture with Mϕs (CC) or in monoculture (MC). In these experiments, we used VSMCs previously transduced with lentiviral vectors harbouring Connexin43 shRNA (shCx43) or a scrambled sequence (shSCR). (F) Representative images of VSMCs transduced with a vector expressing the GFP gene controlled by the SM22 promoter and with a synthetic 3′UTR carrying scrambled (DSCR) or decoy (D146) sequences of miR-146a-5p cocultured with Calcein-labelled Mϕs (n = 3; at…

Figure 3. Phenotypic analysis of miR-146a-5p-overexpressing VSMCs. (A) Growth curve of VSMCs overexpressing miR-146a-5p (146aOE) vs. cells expressing a scrambled miRNA (SCR) (n = 3). (B) Migration assay on 146OE VSMCs vs. SCR cells (n ≥ 8). (C) Resting membrane potential data (n ≥ 13) recorded from control SCR and 146aOE cells, in the basal condition (NS) or following stimulation with Angiotensin II (AII). (D) Expression of contraction markers in SCR and 146aOE cells, measured by RT-qPCR (n ≥ 3). (E) Representative Western blots showing differentiation markers (ACTA2, actin alpha 2; MYH11, myosin heavy chain 11; and CNN1, calponin 1) in SCR and 146aOE cells. (F) Quantification of collagen contraction after 24 h in SCR and 146aOE cells (n = 6). For evaluation of contraction markers via RT-qPCR, Ppia was used as internal control. For ΔΔct analysis, we selected a single reference sample of SCR with a value of 1. Data represents the mean ± SD. To compare means, unpaired Student's t-test was used in B, C, D, and F, while 2-way ANOVA with Sidak's multiple comparisons test was used in A. * P = 0.002, ** P < 0.0001; dV, voltage difference.

Figure 4. Transcriptional profile of miR-146-overexpressing VSMCs. (A) Heatmap of differentially expressed genes with logFC > |1| and adj.P.Val <0.05 in VSMCs overexpressing miR-146a; the colour bar reflects the scaled expression values, and the dendrograms represent the Euclidean distances among samples and genes. (B) Volcano plot of differentially expressed genes in miR-146aOE VSMCs; up-regulated genes are in orange, and down-regulated ones are in blue (adj.P-value < 0.05); the x-axis shows the logFC values computed by the limma package, and the y-axis shows the -log10 (adj.P.Val). (C) Circos plot of the core enrichment genes with logFC > 1 and adj.P.Val < 0.05, along with the enriched pathway terms. (D) Heatmap of the reconstructed metabolic pathways from METAFlux algorithm of VSMCs overexpressing miR-146a (146aOE) or scrambled sequences (SRC); the colour bar reflects the pathway activity score calculated as mean of fluxes of associated reactions. Only significantly different pathways (P < 0.05) for the Kruskal-Wallis test between the two groups are shown.

Figure 5. Cholesterol synthesis modulation and gene target identification in miR-146-overexpressing VSMCs. (A and B) Gene expression analysis for genes correlated with cholesterol biosynthesis (A; n ≥ 8) and with Mϕ phenotype (B; n ≥ 8), comparing control (SCR) vs. 146aOE VSMCs. (C) Quantification of total cholesterol produced by control (SCR) and 146aOE VSMCs measured by mass spectrometry (n = 6). (D) Representative image of Filipin III-labelled SCR and 146aOE VSMCs for cholesterol and relative quantification (n = 8, for each experiment 30 cells were quantified). Scale bar: 10 µm. (E) Representative Western blot showing KLF4 expression in SCR and 146aOE cells. (F) Target luciferase reporter assays using vectors encoding putative target sites in the 3′UTR (Klf4 wt, Klf4 mut, and Traf 3). VSMCs were transiently co-transfected with either miR-146a-5p (146aOE) or scrambled control (SCR), and with wild-type 3′UTR reporter plasmids for Klf4 or Traf3. Renilla luciferase activity was measured 48 h after transfection. The results are normalized to firefly luciferase values (n = 5). (G) SRF luciferase reporter assays: VSMCs were transiently co-transfected with mimics for miR-146a-5p, miR-214-3p,…

Figure 6. Spatial localization of miR-146a-5p in human plaque. (A) Representative image of an explanted atherosclerotic plaque from human carotid artery, divided in three parts, going from the least to the most stenotic. Scale bar: 0.4 cm. (B and C) Representative images of CD68 and ACTA2 staining of the distal and stenotic sections of the plaque. Scale bar: 100 μm. (D and E) Representative images of multiplex in situ staining using a specific probe for miR-146a-5p and a scrambled (SCR) one as control, and an antibody against ACTA2. Scale bar: 100 μm. (F) Quantification of the percentage (%) of miR-146a-5p+ VSMCs. The data were derived from three plaques. Data represent mean ± SD. To compare means, unpaired Student's t-test was used in F.

Figure 7. In vivo effects of specific miR-146a-5p inhibition in VSMCs during atherosclerosis development. (A) Schematic representation of the experiment performed in ApoE−/− mice. Animals were infused with a lentiviral vector carrying D146 or DSCR as control and then fed a Western diet (WD) for 16 weeks; Created with BioRender.com. (B) Representative red-oil staining images from DSCR and D146 animals. Scale bar: 200 µm. (C) Quantification of lesion considering the ratio lesion area/total plaque areas (DSCR, n = 11; D146, n = 11). (D–F) Quantification of circulating CD3+, CD4+, and CD8+ cells in D146 and DSCR treated animals, measured by FACS analysis (n = 11). (G) Analysis of circulating factors in serum of treated animals, using the Olink technology exploratory panel (n = 9). (H) Representative CD3 staining images of DSCR and D146 animals and relative quantification of the percentage of CD3+ cells (DSCR, n = 11; D146, n = 10). Scale bar: 50 µm. Data represent mean ± SD. To compare means, unpaired Student’s t-test was used.