Moyamoya disease susceptibility gene RNF213 links inflammatory and angiogenic signals in endothelial cells

Abstract

Moyamoya disease (MMD) is a cerebrovascular disorder characterized by occlusive lesions of the circle of Willis. To date, both environmental and genetic factors have been implicated for pathogenesis of MMD. Allelic variations in RNF213 are known to confer the risk of MMD; however, functional roles of RNF213 remain to be largely elusive. We herein report that pro-inflammatory cytokines, IFNG and TNFA, synergistically activated transcription of RNF213 both in vitro and in vivo. Using various chemical inhibitors, we found that AKT and PKR pathways contributed to the transcriptional activation of RNF213. Transcriptome-wide analysis and subsequent validation with quantitative PCR supported that endogenous expression of cell cycle-promoting genes were significantly decreased with knockdown of RNF213 in cultured endothelial cells. Consistently, these cells showed less proliferative and less angiogenic profiles. Chemical inhibitors for AKT (LY294002) and PKR (C16) disrupted their angiogenic potentials, suggesting that RNF213 and its upstream pathways cooperatively organize the process of angiogenesis. Furthermore, RNF213 down-regulated expressions of matrix metalloproteases in endothelial cells, but not in fibroblasts or other cell types. Altogether, our data illustrate that RNF213 plays unique roles in endothelial cells for proper gene expressions in response to inflammatory signals from environments.

Figure 1

RNF213 is transcriptionally activated by IFNG and TNFA in vitro and in vivo. (a) Relative expressions of RNF213 in HUVECs when stimulated with various ligands for innate immunity and cytokines in comparison to that of control (“No Stim”). (b) Synergistic effects of IFNG and TNFA treatments on RNF213 expression in HUVECs. (c) Western blot analysis for the RNF213 protein induction with IFNG treatments in HUVECs. Quantified results are plotted under the blotting image. (d) Coinstantaneous inductions of RNF213 and other co-expressed genes upon IFNG treatments of HUVECs. (e) Suppression of RNF213 induction after IFNG and TNFA treatments by actinomycin D (ActD) for HUVECs. (f) The steady-state Rnf213 expressions in various tissues of female mice at 4-weeks of age (compared with Brain). (g) Acute induction of Rnf213 transcripts after intraperitoneal injections of IFNG and TNFA in vivo. (a–g) Data are shown as mean ± SD values from 3 or more independent assays and analyzed using Dunnett’s test (a,g) Tukey’s HSD test (b,e) and Student’s t-test (c). **p < 0.01, ***p < 0.001.

Figure 2. Phosphatidylinositol-4,5-bisphospate 3-kinase and double-stranded RNA-dependent protein kinase are the two upstream regulators of RNF213 expression.

(a) Plots show the relative expressions of RNF213 mRNA in the presence or absence of protein kinase inhibitors for MEK (U0126, 20 μM), JAK (AG490, 25 μM), JNK (SP600125, 10 μM) and PI3K (LY294002, 25 μM) and PKR (C16, 0.2 μM). Mean values from two independent assays are shown. (b,c) Dose-dependent inhibition of RNF213 induction by LY294002 (b) and C16 (c). Relative expressions of RNF213 are plotted against various concentrations of PI3K and PKR inhibitors. Data are shown as values of mean ± SD from three independent assays and analyzed using Dunnett’s test (b,c). *p < 0.05, ***p < 0.001.

Figure 3. RNF213 up-regulates cell-cycle and proliferation of endothelial cells.

(a) The heat map shows up (yellow) or down-regulated (blue) genes in siRNF213#1-treated or untreated HCAECs (n = 2 for each condition). Clustering of siRNA-treated cells and expression profiles for each experiment were conducted blindly. (b) Validating qPCR assays for CCNA2, CCNB1 and CCNE1 expressions in HUVECs with or without RNF213 knockdown (mean ± SD, n = 3, using Student’s t-test). **p < 0.01, ***p < 0.001. (c) Flow-cytometry analysis for cell cycles of HUVECs. Used siRNAs (siRNF213#1 or control) are denoted at the top. The left two panels show 2D-plots for fluorescence intensity of FITC-labeled BrdU and that of 7-AAD. Fractions (%) of cells in G0/G1, S, G2+M and sub G0/G1-phases are indicated with squares. Bar plots on the right shows significant decrease in S-phase with siRNA-mediated knockdown of RNF213 (n = 3, using Student’s t-test, ***p < 0.001). (d) MTS assay for HCAECs, HeLa, HCASMCs and fibroblasts in the presence of RNF213-specific siRNA (siRNF213#1) or control siRNA (n = 3 in each group, using Student’s t-test, ***P < 0.001). (e) Western blots for phosphorylated form of AKT (p-AKT), total AKT (t-AKT) and ACTB in HUVECs. Quantitative data from three independent Western blot analyses are shown as plots on the right (mean ± SD) and analyzed using Student’s t-test. **p < 0.01. Full length blots are presented in Supplementary Fig. S7.

Figure 4. RNF213 controls the expression of MMPs in endothelial cells.

(a) Relative expressions of RNF213, MMPs, TIMP1 and 2 genes are plotted. Note the significant up-regulations of all but MMP2, 17, TIMP1 genes with knockdown of RNF213 (white bars) compared to control experiments (black bars). Data are show as values of mean ± SD (n = 3) and analyzed using Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001. (b) Secreted MMP1 proteins in the culture medium. Data from three independent assays with ELISA are shown (mean ±SD). Note that MMP1 proteins are significantly elevated with knockdown of RNF213 regardless of IFNG pretreatments. The elevation of MMP1 with RNF213 knockdown are partly attenuated by IFNG pretreatments (IFNG+) compared to that in untreated cells (IFNG−), using Tukey’s HSD test. ***p < 0.001. (c) Preventative effects of IFNG on aberrantly up-regulation of MMP1 RNA due to RNF213 knockdown. The data from three independent experiments followed by qPCR are shown and analyzed using Tukey’s HSD test. ***p < 0.001.

Figure 5. Comparative analysis of MMP1 expressions in fibroblasts from healthy controls and MMD patients carrying the R4810K variant of RNF213.

(a) Sanger sequences for the target region of RNF213. Note that healthy control (Normal) are have the wild-type allele, while two MMD patients (MMD1 and 2) carry a heterozygous c.14756G>A (R4810K) mutations. (b) Relative RNF213 expressions and their responses to IFNG treatment in the fibroblasts from 4 healthy controls (Normal) and 2 MMD patients. Data are shown as mean ± SD values of qPCR assays in three independent assays. (c) Relative expressions of MMP1 transcripts in fibroblasts from 4 individuals of healthy control and 2 MMD patients. Values are shown as mean ± SD (n = 3), and analyzed using Student’s t-test. N.S., not significant. (d) Relative RNF213 and MMP1 expressions in control fibroblasts with or without treatment by IFNG and siRNF213#1 (mean ± SD, n = 3, using Tukey’s HSD test, ***p < 0.001).

Figure 6. RNF213 links the external signals to angiogenesis through regulating MMP1 expressions in endothelial cells.

(a) The angiogenic responses of HUVECs on matrigels in different conditions. Representative images for tubular formation by trypsinized HUVECs in the absence (upper panels) or presence of siRNA for RNF213 (lower). Effects of IFNG pretreatments (right) on angiogenic response of HUVECs are shown in comparison with those of untreated cells (left). Scale bar = 100 μm. (b) MMP is a key downstream molecule for deficits in angiogenic response of endothelial cells with depleted expression of RNF213. Upper, middle and lower panels show tubular formation of HUVECs on matrigel without siRNA treatment (“Control”), with siRNA-mediated knockdown of RNF213 and with co-administration of GM6001, an MMP1 inhibitor, respectively. Scale bar = 100 μm. (c) The bar plots show quantitative results of % tube area (upper) and % tube length (lower) on matrigels using HUVECs (n = 3) for Fig. 6b. Tukey’s HSD test. *p < 0.05, **p < 0.01. (d) Effects of PI3K and PKR inhibitors for tubular formations of HUVECs on the matrigel and its recovery by GM6001. Scale bar = 100 μm. (e) Bar plots presenting quantitative results of % tube area (left) and % tube length (right) on matrigels using HUVECs (n = 3) for Fig. 6d. Tukey’s HSD test. ***p < 0.001. N.S., not significant. (f) A proposed model for the regulatory roles of RNF213, PI3K and PKR pathways in endothelial response to cytokines and in angiogenesis.