Fig 2: SRF is important for regulation of IL1B and CXCL8. (A–C) Human coronary SMCs were transduced with MRTF-A along with different titers of Ad-h-shSRF for knockdown of SRF. After harvesting the cells and isolation of RNA, SRF, IL1B, and CNN1 were assayed using RT-qPCR. In (A,B), IL1B and CNN1 were plotted vs. the relative SRF level in the respective samples. In (C), SRF, IL1B, and CNN1 were plotted vs. the titer (multiplicity of infection, MOI) of the short hairpin virus. (D) Shows the remainder of the inflammatory mediators in control vs. SRF-depleted cells (1000 MOI). (E) Is like (D), except that SRF silencing was done without simultaneous overexpression of MRTF-A. (F) Shows the gene locus for human CXCL8 and binding of RelA (light blue oval) 4 kb upstream of the transcription start site. The proximal promoter contained two DNA sequences with two deviations each from the classical SRF-binding sequence. When testing this sequence in a reporter assay, no suppression by MRTF-A was however, seen (G). In (H), Srf and Il1b were assayed in the aorta from smooth muscle specific and inducible Srf knockout mice. Because knockout was induced by tamoxifen, two control groups were included in addition to the knockout group. Vehicle controls (VC) are Cre-positive mice injected with sunflower oil, whereas tamoxifen controls (TC) are Cre-negative mice injected with tamoxifen. Knockouts (TKO) are Cre-positive mice injected with tamoxifen. All mice are homozygous for the floxed Srf allele. (I–L) Show the effect of overexpression of SRF in the absence and presence of MRTF-A. SRF was capable of suppressing CXCL8 on its own but was without effect on CCL2 and IL6. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig 3: MRTF-A interacts with RelA and RelB. To explore if MRTF-A binds RelA and RelB, we performed co-immunoprecipitations followed by dot blotting (A). Lysates from cells where MRTF-A was overexpressed were incubated with control resin and with MRTF-A-antibody-conjugated resin. Flow through (FC: flow through control; FM: flow through MRTF-A) from both columns contained RelA, RelB, and MRTF-A. RelA, RelB and MRTF-A were moreover detectable in the eluate from the MRTF-A column (EM: eluate MRTF-A) but in the in the eluate from the control column (EC: eluate control). Western blotting (B) of the eluates showed the RelA band at the expected molecular weight. These findings support a model where MRTF binding to Rel proteins suppresses inflammation (C). (C) Also depicts the hypothesis that RelA suppression, by glucocorticoid receptor stimulation or silencing, should mitigate inflammatory suppression by MRTF-A. (D,E) Test this hypothesis using dexamethasone (glucocorticoid receptor agonist, 3 µM) or short hairpin silencing (200 MOI of shRELA). For the experiments in (D,E), the fold repression is given. Independent testing showed that the fold repression of CCL2 and IL1B by MRTF-A was reduced by dexamethasone. In contrast, fold suppression of CXCL8 and IL6 was unaffected. For RelA silencing, CXCL8 and IL6 suppression by MRTF-A were significantly reduced, but the overall effect, where MRTF-A suppression of inflammation was reduced by ˜25%, was also significant. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig 4: The anti-inflammatory impact of MRTF-A is cell type-dependent. To compare the impact of MRTF-A on inflammation in different cell types, MRTF-A was overexpressed using an adenoviral vector, and using a null vector as control. (A–D) Show the effect of MRTF-A on CCL2, CXCL8, IL6, and IL1B in human coronary artery SMCs (hCASMC, (A)), human bladder SMCs (hBSMC, (B)), human coronary artery endothelial cells (hCAEC, (C)), and monocytes (THP-1 cells, (D)). ACTA2 was used as positive control in all panels except for THP-1 cells where we instead assayed MRTFA itself. (E) Shows the effect of the MRTF inhibitor CCG-1423 (10 µM) in hCASMC transduced with MRTF-A. (F) Shows the effect of Latrunculin B (100 nM), which depolymerizes actin, on the inflammatory transcripts in hCASMC. (G) Shows inflammatory mediators determined by RT-qPCR after silencing MRTF-A using different titers of a short hairpin virus (shMRTF-A). In this experiment U6 represents the control virus. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig 5: The anti-inflammatory impact of MRTF-A is maintained in pro-inflammatory conditions. (A) Human coronary artery smooth muscle cells in culture were transduced with Null and MRTF-A viruses in control conditions (open bars), or with simultaneous treatment with lipopolysaccharide (LPS, 500 ng/ml), to mimic a pro-inflammatory environment (gray bars). After harvesting the cells, RNA was isolated, and inflammatory transcripts were assayed using RT-qPCR. The fold repression by MRTF-A is indicated below the asterisks that indicate significance, showing that MRTF-A was similarly effective in the absence and presence of LPS. CCL2 for example was reduced eightfold in control conditions and sevenfold in LPS treated cells, and an independent analysis indicated that this difference was not significant. To support MRTF-dependent reductions of inflammatory mediators at the protein level, MCP-1 (CCL2, (B,D)) and IL-8 (CXCL8, (C,E)) were examined using western blotting (B,C) and ELISAs (D,E). All assays indicted significant MRTF-A-dependent reductions at the protein level, both in the absence and presence of LPS. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.