Disruption of SM22 promotes inflammation after artery injury via nuclear factor kappaB activation

Abstract

Rationale: SM22 (or transgelin), an actin-binding protein abundant in vascular smooth muscle cells (VSMCs), is downregulated in atherosclerosis, aneurysm and various cancers. Abolishing SM22 in apolipoprotein E knockout mice accelerates atherogenesis. However, it is unclear whether SM22 disruption independently promotes arterial inflammation.

Objective: To investigate whether SM22 disruption directly promotes inflammation on arterial injury and to characterize the underlying mechanisms.

Methods and results: Using carotid denudation as an artery injury model, we showed that Sm22 knockout (Sm22(-/-)) mice developed enhanced inflammatory responses with higher induction of proinflammatory genes, including Vcam1, Icam1, Cx3cl1, Ccl2, and Ptgs2. Higher expression of these genes was confirmed in primary Sm22(-/-) VSMCs and in PAC1 cells after Sm22 knockdown, whereas SM22 recapitulation in primary Sm22(-/-) VSMCs decreased their expression. NFKB2 was prominently activated in both injured carotids of Sm22(-/-) mice and in PAC1 cells after Sm22 knockdown and may mediate upregulation of these proinflammatory genes. As a NF-kappaB activator, reactive oxygen species (ROS) increased in primary Sm22(-/-) VSMCs and in PAC1 cells after Sm22 knockdown. ROS scavengers blocked NF-kappaB activation and induction of proinflammatory genes. Furthermore, Sm22 knockdown increased Sod2 expression and activated p47phox, reflecting contributions of mitochondria and NADPH oxidase to the augmented ROS production; this may result from actin and microtubule cytoskeletal remodeling.

Conclusions: Our findings show that SM22 downregulation can induce proinflammatory VSMCs through activation of ROS-mediated NF-kappaB pathways. This study provides initial evidence linking VSMC cytoskeleton remodeling with arterial inflammation.

Figure 1. Enhanced inflammatory response of Sm22^−/− mice two weeks after carotid denudation

(A) H&E staining showed pronounced carotid swelling, edema, prominent fibrotic adhesion and cell infiltration in denuded carotids of Sm22^−/− mice. Bar, 100 μm. (B, C) IHC using a macrophage marker, F4/80 (B) and a T cell marker, CD3 (C). Top panels, 100X; middle panels, 400X; bottom panels: quantification of positive signals from images at 100X magnification of carotids from five Sm22^−/− and their littermates Sm22^+/+ mice. Values are means ± SE. The asterisk, *, indicates p < 0.05 versus Sm22^+/+ mice. Representative positive signals (brown) are indicated by arrows. Bars: upper panels, 100 μm; middle panels, 20 μm. Abbreviations: N, neointima; M, media; A, adventitia.

Figure 2. Expression of cell adhesion proteins is higher in Sm22^−/− mice two weeks after carotid denudation

(A) Relative mRNA level of Vcam1 and Icam1 in carotids without injury (left panel) and expression level induced by injury (right panel) was evaluated using rtRT-PCR. (B, C) IHC analyses of VCAM1 (B) and ICAM1 (C). Top panels, 100X; middle panels, 400X; bottom panels, quantification of positive signals at 100X magnification in the media of carotids from five Sm22^−/− and their littermates Sm22^+/+ mice. Values (A, B and C) are means ± SE from five pairs of mice. The asterisk, *, indicates p < 0.05 versus Sm22^+/+ mice. Representative positive signals (brown) are indicated by arrows. Bars: top panels, 100 μm; middle panels, 20 μm.

Figure 3. Higher expression of chemokines in Sm22^−/− mice two weeks after carotid denudation

(A) Left panel: relative mRNA expression of Cx3cl1 and Ccl2 in carotids of mice without injury evaluated using rtRT-PCR. Right panel: “Injury induction fold”: the ratio between Cx3cl1 and Ccl2 mRNA expression in the injured carotid and the corresponding expression in the uninjured carotid of each mouse using rtRT-PCR. (B, C) IHC analyses of CX3CL1 (B) and CCL2 (C). Top panels, 100X; middle panels, 400X; bottom panels, quantification of positive signals in images at 100X magnification in the media of carotids from five Sm22^−/− and their littermates Sm22^+/+ mice. Values (A, B and C) are means ± SE from five pairs of mice. The asterisk, *, indicates p < 0.05 versus Sm22^+/+ mice. Representative positive signals (brown) are indicated by the arrows. Bars: top panels, 100 μm; middle panels, 20 μm.

Figure 4. Transcriptional upregulation of pro-inflammatory genes in primary Sm22^−/− VSMCs and in PAC1 cells after Sm22 knockdown

(A) In primary Sm22^−/− and Sm22^+/+ VSMCs, relative mRNA expression of Vcam1, Icam1, Cx3cl1, Ccl2 and Ptgs2 was examined using rtRT-PCR and WB (inserted panel). Values are means ± SE from primary VSMCs of four pairs of mice. The asterisk, *, indicates p < 0.05 versus Sm22^+/+ VSMCs. (B - C) In PAC1 cells, Sm22 knockdown efficiency and the expression of pro-inflammatory genes were determined by rtRT-PCR 1 day, 2 days and 3 days after transfection, and by WB (B, inserted panel) and IF (C) 3 days after transfection. Green: SM22 and CCL2; blue: DAPI. Values in (B) are means ± SE from three independent experiments. The asterisk, *, indicates p < 0.05 versus the scr group. Bar in (C): 20 μm. Abbreviations: scr, scrambled siRNA; si, Sm22 siRNA; d1/2/3, 1/2/3 days after transfection into PAC1 cells.

Figure 5. Activation of NF-κB pathway in Sm22^−/− mice two weeks after carotid denudation

Activation of both canonical NF-κB RELA (A), and non-canonical NF-κB NFKB2 (B) was evaluated by IHC. Top panels, 100X; middle panels, 400X. In the middle panel representative positive signals (brown) are indicated by arrows, and nuclear NFKB2 is indicated by arrow heads. Bottom panels: the ratio of signal positive nuclei versus all nuclei in the same area of the media of carotids from five Sm22^−/− and their littermates Sm22^+/+ mice as quantified using Image-Pro software. Values are means ± SE. The asterisk, *, indicates p < 0.05 versus Sm22^+/+ mice. Bars: top panels, 100 μm; middle panels, 20 μm.

Figure 6. Sm22 knockdown activated NF-κB pathway and NF-κB inhibitors blocked induction of pro-inflammatory genes

Cytoplasmic and nuclear lysates were isolated from PAC1 cells transfected with scrambled RNA (scr) or Sm22 siRNA (si) in the absence or presence of Bay-11-7082 (Bay), an NF-κB pathway inhibitor. Activation of both canonical NF-κB pathway (A) and non-canonical NF-κB pathway (B) was investigated by WB using antibodies against RELA, IKB and NFKB2 as indicated. P100 is the precursor of NFKB2. NF-κB binding activities of nuclear extracts was investigated using EMSA assay using a NF-κB consensus binding site as the probe (C), showing increased NF-κB binding activities after Sm22 knockdown. NS: nonspecific. Effect of Bay-11-7082 on induction of inflammatory genes was investigated using rtRT-PCR (D) and WB (E). Values in (D) are means ± SE. The asterisk, *, indicates p < 0.05 versus si group when comparing the fold changes. Data are from three independent experiments. scr, scrambled siRNA; si, Sm22 siRNA.

Figure 7. Elevated ROS production after SM22 disruption contributed to NF-κB activation and induction of inflammatory genes

(A) ROS levels were quantified using DHE (for superoxide) and DCFDA (for peroxide) based fluorescence microscopy in primary VSMCs (left panel) and PAC1 cells (right panel). Two ROS scavenger, Tiron (5 mM) and Tempol (5 mM), were used. Thirty images for each group were used for quantification. Values are means ± SE. The asterisk, *, indicates p < 0.05 versus primary Sm22^+/+ VSMCs (left panel) and scr group (right panel) respectively. (B-C) In PAC1 cells, activation of both canonical NF-κB pathway and non-canonical NF-κB pathway was investigated respectively by WB using antibodies against RELA, IKB and NFKB2 as indicated. P100 is the precursor of NFKB2. Dashed line, images of non-adjacent lanes on the same gel. (D-E) In PAC1 cells, Tiron or Tempol blocked pro-inflammatory gene induction by Sm22 siRNA using rtRT-PCR and WB (inserted panel). Values were means ± SE. The asterisk, *, indicates p < 0.05 versus si group. Abbreviations: scr, scrambled RNA; si, Sm22 siRNA; Ti, Tiron; Te, Tempol. Data were from three independent experiments.

Figure 8. Mitochondria reorganization and NADPH oxidase activation as a result of cytoskeleton remodeling contributed to ROS production after SM22 disruption

(A) Expression of Sod2 in PAC1 cells, primary VSMCs (P) and injured mouse carotids (M) was investigated using rtRT-PCR, WB using lysates from PAC1 (inserted panel) and IF (lower panel). Values are means ± SE. The two asterisks, *, indicate p < 0.05 for si versus scr group in PAC1 cells, and for Sm22^−/− carotids versus Sm22^+/+ carotids in mice. (B) Mitotracker Red (Mito) was used to visualize mitochondria (upper panel). Mitochondria aggregation and megamitochondria are indicated by arrows and arrow heads respectively. Cell peripheral assembly of p47phox after Sm22 knockdown was investigated with a p47phox antibody (lower panel) and is indicated by arrows. (C) Cytoskeleton structure was investigated using IF with a smooth muscle alpha actin antibody (SMA) and a tubulin antibody (TUB), and the aggregation of tubulin is indicated by arrows. (D) A schematic representation of proposed cellular and molecular events associated with enhanced arterial inflammation after disruption of SM22 in VSMCs. Abbreviations: scr, scrambled RNA; si, Sm22 siRNA; P, primary VSMCs; M, mouse carotids.