A Novel Function of NaV Channel β3 Subunit in Endothelial Cell Alignment Through Autophagy Modulation

Abstract

Endothelial cells (EC) play a pivotal role in vascular homeostasis. By sensing shear stress generated by blood flow, EC endorse vasculoprotection through mechanotransduction signaling pathways. Various ion channels are involved in mechanosignaling, and here, we investigated the endothelial voltage-gated Na⁺ channels (Na_V channels), since their mechanosensitivity has been previously demonstrated in cardiomyocytes. First, we showed that EC from aorta (TeloHAEC) behave as EC from umbilical vein (HUVEC) under laminar shear stress (LSS). For both EC models, cell alignment and elongation occurred with the activation of the KLF2/KLF4 atheroprotective signaling pathways. We found that LSS decreased the expression of SCN5A, encoding Na_V1.5, while LSS increased that of SCN3B, encoding Na_Vβ3. We demonstrated that the KLF4 transcription factor is involved in SCN3B expression under both static and LSS conditions. Interestingly, SCN3B silencing impaired EC alignment induced by LSS. The characterization of Na_Vβ3 interactome by coimmunoprecipitation and proteomic analysis revealed that mTOR, implicated in autophagy, binds to Na_Vβ3. This result was evidenced by the colocalization between Na_Vβ3 and mTOR inside cells. Moreover, we showed that SCN3B silencing led to the decrease in LC3B expression and the number of LC3B positive autophagosomes. Furthermore, we showed that Na_Vβ3 is retained within the cell and colocalized with LAMP1 and LC3B. Finally, we found that resveratrol, a stimulating-autophagy and vasculoprotective molecule, induced KLF4 together with Na_Vβ3 expression. Altogether, our findings highlight a novel role of Na_Vβ3 in endothelial function and cell alignment as an actor in shear stress vasculoprotective intracellular pathway through autophagy modulation.

Keywords: SCN3B; NaVβ3; autophagy; endothelial cell; mechanosignaling; shear stress.

FIGURE 1

Phenotypic changes induced by LSS in TeloHAEC and HUVEC. Representative phase contrast microscopy images of TeloHAEC (A) and HUVEC (B) cultivated under static condition (Static, 0 dyn/cm²) or LSS (20 dyn/cm²) at different times (0, 24, and 96 h). TeloHAEC (C) and HUVEC (D) alignment quantification using local gradient orientation method with Fiji software at 24 and 96 h of LSS or under static condition. Cell amount is expressed over cell orientation in degree. A peak at angle value of 0° corresponds to a cell alignment with the flow direction. The histogram illustrates cell amount between angle values of −5° and +5°. Phenotypic modifications of TeloHAEC (E) and HUVEC (F) were assessed with the cell length (left panel) and width (middle panel), giving the elongation factor (Length/Width, right panel) after 96 h of LSS or of static culture. Values are mean ± SEM are shown (n = 7–13). Significance was analyzed using nonparametric Kruskal–Wallis (C, D) or Mann–Whitney (E, F) tests. *p < 0.05, ***p < 0.001.

FIGURE 2

Activation of the atheroprotective signaling pathway in TeloHAEC and HUVEC by LSS. The transcriptional and protein modification of KLF2/KLF2, KLF4/KLF4, and NOS3/eNOS expression induced by LSS (20 dyn/cm², 96 h) was measured in TeloHAEC and HUVEC, by RT‐qPCR and western blotting, followed by densitometry analysis (A–C). The histogram bars illustrate the fold change values (mRNA and protein) as mean ± SEM (n = 3–8), compared to static condition, given the arbitrary value of 1 for RT‐qPCR data or normalized to HSC70 (loading control protein) and compared to static condition, given the arbitrary value of 1 for western blotting. Nonparametric Mann–Whitney tests were performed. ns, nonsignificant, *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 3

LSS modulates Na_V expression profiles in both TeloHAEC and HUVEC after 96 h of flow. mRNA expression profile of Na_V channel subunits (SCN1A to SCN11A and SCN1B to SCN4B) in TeloHAEC and HUVEC (A, B). The mRNA relative expression of Na_V channel subunits was determined by relative RT‐qPCR. The data are mean of mRNA relative expression ± SEM (n = 3) with the 2^−ΔCq method. ND, not detected (Cq > 35). The protein expression of Na_V channel subunits detected at transcript level was evaluated by western blot in comparison to control tissues as mouse heart (for Na_V1.5 and Na_Vβ3), brain (for Na_V1.6 and Na_Vβ1) or Dorsal Root Ganglion neurons (DRG) (for Na_V1.7) (C, D). Fifty μg of cell protein extracts (TeloHAEC and HUVEC) and twenty μg of protein from control tissue (heart, brain and DRG) were loaded on 8% SDS‐PAGE. HSC70 protein was used as a loading control (C, D). In TeloHAEC or HUVEC, Na_V1.5 and Na_Vβ3 share the same HSC70 loading control as well as Na_V1.6 and Na_Vβ1 since the protein samples immunoblotted with these antibodies (Na_V1.5/Na_Vβ3 and Na_V1.6/Na_Vβ1) are the same after cutting the nitrocellulose membrane. SCN5A (E), SCN8A (F) and SCN9A (G), SCN1B (H), SCN3B (I) mRNA expression and Na_Vβ3 protein expression (J) in LSS (20 dyn/cm² for 96 h) compared to static condition in TeloHAEC (left panel) and HUVEC (right panel). Values of fold change (mRNA and protein) are expressed as mean ± SEM for at least 3 until 6 independent experiments and compared to static condition, given the arbitrary value of 1 using the 2−^ΔΔCq method for RT‐qPCR experiments or normalized to HSC70 used as loading control protein and compared to static condition, given the arbitrary value of 1 for the densitometry analysis. Immunoblot representative images of Na_Vβ3 and HSC70 protein expression in TeloHAEC and HUVEC (J) cultivated under static or LSS (20 dyn/cm²) condition for 96 h. Nonparametric Mann–Whitney tests were performed. ns, nonsignificant, *p < 0.05, **p < 0.01.

FIGURE 4

KLF4 is involved in SCN3B mRNA and protein expression in TeloHAEC. (A) Graphical representation of the KLF4 based on the position frequency matrix (PFM). The height of the column at each position is the information content (bit) and the individual base heights are in proportion to their frequencies. (B) Schematic representation of the location of the seven putative KLF4 (green triangles) DNA binding sites in human SCN3B promoter, with a score greater than 10.00, given by JASPAR analysis, named S1 to S7. (C, D) The effects of KLF4 overexpression using pKLF4 in comparison with empty vector (pCTL) were evaluated on KLF4 mRNA and protein (C) and SCN3B mRNA and Na_Vβ3 protein (D) expression in TeloHAEC. (E, F) The effects of LSS (24 h at 20 dyn/cm²) on KLF4 (E) and SCN5A, SCN8A, SCN9A, SCN1B, and SCN3B (F) expression were evaluated in TeloHAEC transfected with SiRNA Control (SiCTL) or SiRNA targeting KLF4 (SiKLF4). The histogram bars illustrate the fold change values (mRNA and protein) as mean ± SEM (n = 3–4), compared with pCTL or SiCTL, given the arbitrary value of 1 for RT‐qPCR data using the 2^−ΔΔCq method or normalized to HSC70 (loading control protein) and compared to pCTL, given the arbitrary value of 1 for western blotting. Values of mRNA fold change are expressed as mean ± SEM (n = 3) and compared to SiCTL condition, given the arbitrary value of 1. Significance was analyzed using nonparametric Mann–Whitney tests. ns, nonsignificant, *p < 0.05.

FIGURE 5

SiRNA SCN3B decreased cell alignment and increased cell width after 24 h of LSS in TeloHAEC. (A) mRNA expression of SCN3B in TeloHAEC transfected with SiRNA Control (siRNA CTL) or SiRNA targeting SCN3B (SiRNA SCN3B) by ddPCR. Values of SCN3B mRNA fold change are expressed as mean of copy number of SCN3B per μL. (B) The histogram illustrates TeloHAEC amount between angle values of −5° and +5°, after transfection with SiRNA Control (siRNA CTL) or SiRNA targeting SCN3B (siRNA SCN3B). (C) Elongation of HUVEC was determined as the cell length along flow direction (left panel) divided by cell width (right panel). TeloHAEC were submitted to static or LSS (20 dyn/cm²) for 24 h. Values are mean ± SEM are shown (n = 5). Significance was analyzed using one‐way ANOVA statistical analysis, followed by Tukey multiple comparisons test or nonparametric Mann–Whitney tests. ns, nonsignificant, *p < 0.05, ***p < 0.001, ****p < 0.0001.

FIGURE 6

Identification of Na_Vβ3‐interacting partners in TeloHAEC and HUVEC by proteomic analysis. (A, B) Schematic representation of the Na_Vβ3 immunoprecipitation (IP) procedures. TeloHAEC (A) and COS‐7 (B) were transfected with pMyc‐SCN3B, and pCTL (negative control). (A) Myc tag‐Na_Vβ3 with its partners were coimmunoprecipitated using anti‐Myc tag magnetic beads from TeloHAEC protein extracts. (B) Myc tag‐Na_Vβ3 from COS‐7 protein extracts were first immunocaptured with anti‐Myc tag magnetic beads and subjected to pull‐down using HUVEC whole cell protein lysate. Below the schemes are shown representative control immunoblots (n = 3) with anti‐Na_Vβ3 (top blot) and anti‐myc tag (bottom blot). Input (Inp) was the whole protein extract after cell transfection and the flow throw (FT) corresponded to the supernatant recuperated after immunocapturing with the anti‐Myc tag magnetic beads. Wash and IP correspond to the washing step and a sample from the solution used to boil the beads (IP) to elute Na_Vβ3 and its partners. IP from TeloHAEC and HUVEC were subjected to MS/MS analysis. (C) Venn Diagram indicating the number of Na_Vβ3 interacting proteins identified in TeloHAEC and HUVEC after proteomic analysis. (D) Bubble plots of Gene ontology (GO)‐term enrichment analysis of the 76 putative Na_Vβ3 interactants. The four first GO‐terms belonging to the biological process and cellular component categories are indicated on the left. Below the bubble chart, is shown the protein ratio (number of proteins in our dataset compared to number of proteins in annotation). The scale of the bubble plot size (black, upper right) reflects the number of proteins. The significancy of the identification score is illustrated by a color intensity scale (Log10 Bonferroni, lower right). (E, F) Representative Na_Vβ3 and mTOR immunoblots (n = 3) of immunoprecipitated Na_Vβ3 with Myc tag antibody in TeloHAEC (E) and HUVEC (F).

FIGURE 7

Na_Vβ3 modulates LC3B expression in TeloHAEC. (A) Protein expression of P‐mTOR at serine 2448 (mTORC1 complex) or at serine 2481 (mTORC2 complex) and mTOR in static or LSS (20 dyn/cm²–24 h) conditions in TeloHAEC transfected with SiRNA Control (siCTL) or SiRNA targeting SCN3B (siSCN3B). The expression of proteins was evaluated by western blot where ten μg of cell protein extracts were loaded on 4%–20% SDS‐PAGE and where GAPDH protein was used as a loading control. (B) mRNA expression of LC3B in the same conditions. (C) LC3B protein expression in static or LSS (20 dyn/cm²–24 h) conditions in TeloHAEC transfected with SiRNA Control (siCTL) or SiRNA targeting SCN3B (siSCN3B). (D) Densitometry analysis following western blot of the LC3B‐I and LC3B‐II bands in static and LSS with or without siSCN3B in TeloHAEC. Values of fold change (mRNA and protein) are expressed as mean ± SEM for at least 4 or 5 independent experiments and compared to static condition, given the arbitrary value of 1 using the 2^−ΔΔCq method for RT‐qPCR experiments or normalized to GAPDH for western blot experiments. One‐way ANOVA was performed. ns, nonsignificant, *p < 0.05, **p < 0.01, ***p < 0.001. (E) Representative immunofluorescence images of LC3B staining (red) in TeloHAEC transfected with siCTL or siSCN3B and cultivated under LSS (20 dyn/cm² for 24 h). Nuclei (blue) were stained with DAPI. (F) Histograms represent the number of puncta (left) and fluorescence intensity (right) of LC3B staining. Data are presented as mean ± SEM (n = 3) and normalized to SiCTL, given the arbitrary value of 1. Significance was analyzed using one sample t‐test and Wilcoxon test, *p < 0.05.

FIGURE 8

Na_Vβ3 partially colocalizes with LC3B, LAMP1, and mTOR in TeloHAEC under static or LSS conditions. (A) Representative images illustrating the immunostaining of LC3B, LAMP1 or mTOR (red) together with Na_Vβ3 (green) in TeloHAEC which were transfected with pmyc‐SCN3B and cultivated under static condition. Nuclei (blue) were stained with DAPI. (B) Quantification of the colocalization of Na_Vβ3 with LC3B, LAMP1 or mTOR was calculated using Pearson's correlation coefficient (PCC). Histogram bars represent the PCC value as mean ± SEM for each protein (LC3B, LAMP1 or mTOR) in TeloHAEC cultivated under static or LSS conditions for three independent experiments. Significance was analyzed using one‐way ANOVA statistical analysis, followed by Tukey multiple comparisons test. **p < 0.01, ***p < 0.001.

FIGURE 9

Resveratrol induces KLF4, LC3B, and SCN3B mRNA expression in both TeloHAEC and HUVEC and in KLF4, LC3B, and Na_Vβ3 protein in TeloHAEC. (A‐C) The effects of RSV were evaluated on the expression of KLF4 (A), LC3B (B), and SCN3B (C) mRNA in TeloHAEC and HUVEC. (D) The effect of resveratrol was assessed on KLF4, LC3B and Na_Vβ3 protein expression in TeloHAEC. (A–D) The cells were treated with 100 μM of RSV and 0.1% DMSO (CTL) as negative control. The histogram bars illustrate the fold change values (mRNA and protein) as mean ± SEM (n = 3–4), compared to CTL, given the arbitrary value of 1 for RT‐qPCR data using the 2−^ΔΔCq method or normalized to GAPDH (loading control protein) and compared to CTL, given the arbitrary value of 1 for western blotting. Nonparametric unpaired Mann–Whitney tests were performed, *p < 0.05.