Contribution of the TRPM4 Channel to Osteogenic Differentiation of Human Aortic Valve Interstitial Cells

Abstract

Background: Aortic stenosis due to deleterious remodeling of the aortic valve is a health concern since it can be corrected only by valve replacement due to the poor knowledge of cellular mechanisms involved. Fibroblastic valvular interstitial cells (VICs) play a central role in valve leaflet stiffness by trans-differentiation into osteoblast-like cells leading to calcification. The TRPM4 (transient receptor potential melastatin 4) cation channel was shown to participate in cardiac fibroblast remodeling. It is also involved in radiation-induced aortic valve remodeling in vivo in mice. We hypothesized that TRPM4 might participate in human VIC transition to osteoblastic phenotype.

Methods: Human aortic valves were collected from patients undergoing surgical valve replacement. Isolated VICs were maintained 14 days in culture in standard or pro-calcifying media and submitted to the TRPM4 inhibitor 9-phenanthrol, or small hairpin RNA-TRPM4. Osteogenic differentiation was evaluated by measuring hydroxyapatite crystals by Alizarin red staining and protein expression of osteogenic markers.

Results: Western blot on VICs revealed TRPM4 protein expression and channel functionality was confirmed by patch-clamp recordings exhibiting a cationic current sensitive to voltage and internal Ca²⁺. VICs maintained in pro-calcifying media exhibited a higher mineralization than in standard media, with an increase in osteogenic markers. Mineralization and osteogenic markers (bone morphogenetic protein 2, runt-related transcription factor 2) were decreased when pro-calcifying media contained 9-phenanthrol or small hairpin RNA-TRPM4. Similarly, SMAD1/5 and nuclear factor of activated T-cell pathways were stimulated in pro-calcifying media conditions compared with standard media but reduced by 9-phenanthrol or small hairpin RNA-TRPM4.

Conclusions: TRPM4 participates in osteogenic differentiation of human VICs and thus appears as a target to prevent aortic valve remodeling.

Keywords: TRPM4; aortic stenosis; valvular interstitial cells.

Figure 1. TRPM4 mRNA, protein expression and typical TRPM4 current in isolated hVICs.

A, TRPM4 mRNA in hVICs detected by polymerase chain reaction. H₂O was used as negative control. B, Representative western blot of TRPM4 protein in hVICs. C, Immunostaining of TRPM4 (green) on isolated hVICs. Nuclei were labeled with DAPI (blue). Scale bar=50 μm. Graphic in the right panel represents the profile of TRPM4 staining through the cell (according to the red vertical line in lower left image of TRPM4). D, Single‐channel tracings recorded at various membrane potentials (V _m) from an inside‐out patch from hVIC. Pipette and bath contained 145 mmol.L⁻¹ NaCl standard solution (CaCl₂=10⁻³ mol.L⁻¹). Dashed lines indicate current level when all channels are closed. At least 4 similar channels are present in the patch. E, Current voltage relationship (i/V _m) under the same conditions as in D. Data points were fitted by linear regression indicating a conductance g=20.3±0.3 pS and a reversal potential=−0.11±0.40 mV (n=28, N=7). F, Voltage dependence of open probability (P _o). In conditions described in D. P _o was determined at various V _m (n=12, N=6). Data were fitted to a Boltzmann equation. G, Current voltage relationship (i/V _m) with the 145 mmol.L⁻¹ NaCl standard solution in the pipette and the 145 mmol.L⁻¹ KCl solution (blue squares; n=7, N=5) or 42 mmol.L⁻¹ NaCl solution (orange circles; n=4, N=3) in the bath. Data points were fitted by linear regression. H, Values of P _o at various [Ca²⁺]_i (V _m=+40 mV; symmetrical 145 mmol.L⁻¹ NaCl standard solution). Numbers inside of data points indicate the number of experiments. Typical current traces are provided in the inset for the same patch at 10⁻³ or 10⁻⁷ mol.L⁻¹ [Ca²⁺]_i. I, Single‐channel current recorded from an inside‐out patch illustrating the blocking effect of 9 PHE applied at the inside of the membrane at 3.10⁻⁶ or 3.10⁻⁵ mol.L⁻¹ (V _m=+40 mV; symmetrical 145 mmol.L⁻¹ NaCl standard solution). Corresponding amplitude histograms are provided on the right. Note that it indicates that 9 PHE decreased channel activity with no effect on single‐channel current amplitude. J, P _o determined in each condition for 5 experiments similar to that shown in E. Circles linked by dashed lines correspond to single experiments. Plain black squares linked by plain line correspond to the mean±SEM. Analyzed with mixed‐effects analysis (J) with uncorrected Fisher's least significant difference test. ***P<0.001, ****P<0.0001. 9 PHE indicates 9‐phenanthrol; hVIC, human valvular interstitial cell; NEG CTRL, negative control; P _o, open probability; and TRPM4, transient receptor potential melastatin 4.

Figure 2. TRPM4 participates in hVIC mineralization.

A, Calcium deposits measured by Alizarin red staining on hVICs. Cells were cultured in SM or PM media during 3, 7, or 14 days (D3, D7, and D14). Histograms are mean±SEM (N=6 for D3; N=15 for D7; N=21 for D14) in arbitrary units. B, Alizarin red staining of hVICs cultured in SM, PM, and PM+9 PHE 3.10⁻⁶ mol.L⁻¹ (PM+9 PHE) for 14 days (N=21). Left panel shows representative pictures of the same culture in the 3 conditions. Right panel shows the fold change vs PM of Alizarin red quantification for each condition. Circles connected by lines represent data for each patient measured in the 3 conditions (average of duplicate). Histograms represent the mean±SEM (N=21). C, hVIC viability was evaluated by flow cytometry using propidium iodide experiment after 14 days of culture in SM, PM, or PM+9 PHE (N=11). D, hVIC cell cycle was determined by flow cytometry after 14 days of culture in SM, PM, or PM+9 PHE (N=9). Histograms are mean±SEM. Analyzed with 2‐way ANOVA (A) with uncorrected Fisher's least significant difference test and Friedman test (B) with Dunn's multiple comparisons test, with mixed‐effects analysis (C) with uncorrected Fisher's least significant difference test. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, ns=nonsignificant. 9 PHE indicates 9‐phenanthrol; G0–G1, growth 0 and growth 1 phase; G2‐M, growth 2 phase and mitosis; hVIC, human valvular interstitial cell; PM, pro‐calcifying media; S, synthesis phase; SM, standard media; and TRPM4, transient receptor potential melastatin 4.

Figure 3. Reduction of osteogenic markers under pharmacological inhibition of TRPM4.

A and F, Representative western blots of BMP2, Runx2, αSMA, and TRPM4 protein expression in hVICs after 7 (A) and 14 days (F) of culture. As indicated under panels, cells were cultured in SM conditions or in PM conditions without or with 9 PHE 3.10⁻⁶ mol.L⁻¹. B–E and G–J, Quantification of intensity labeling of BMP2 (N=9 and 7 for B and G); Runx2 (N=9 and 14 for C and H); αSMA (N=10 and 8 for D and I) and TRPM4 (N=8 and 12 for E and J) after 7 days (B–E) and 14 days (G–J) of hVIC maintained in SM, PM, and PM+9 PHE culture conditions. Histograms are mean±SEM of fold change vs PM. Analyzed with Friedman (B, C, G, H, J) with Dunn's multiple comparisons and 1‐way ANOVA (D, E, I) with uncorrected Fisher's least significant difference multiple comparisons test. *P<0.05, **P<0.01, ***P<0.001, ns=nonsignificant. 9 PHE indicates 9‐phenanthrol; BMP2, bone morphogenetic protein 2; hVIC, human valvular interstitial cell; PM, pro‐calcifying media; Runx2, runt‐related transcription factor 2; SM, standard media; αSMA, α‐smooth muscle actin; and TRPM4, transient receptor potential melastatin 4.

Figure 4. shRNA‐TRPM4 are functional and not toxic for hVICs.

A and B, Quantification of TRPM4 protein expression in hVICs transduced with 1 ng/cm² SH1, SH2 or SH1+2 after 7 days (N=6) (A) or 14 days (N=8) (B). Representative western blots are provided below the histograms. Histograms are mean±SEM of fold change vs PM. C, hVIC viability evaluated by propidium iodide experiments in flow cytometry after 14 days of hVICs cultured in PM or PM+SH at 1 ng/cm² (N=5). D, Cell cycle was determined by flow cytometry after 14 days of hVICs cultured in PM or PM+SH at 1 ng/cm² (N=5). Analyzed with 1‐way ANOVA (A) with uncorrected Fisher's least significant difference multiple comparisons test, Friedman test (B) with Dunn's test and mixed‐effects analysis (C) with uncorrected Fisher's least significant difference test. **P<0.01, ns=nonsignificant. hVICs indicates human valvular interstitial cell; PM, pro‐calcifying media; shRNA, small hairpin RNA; and TRPM4, transient receptor potential melastatin 4.

Figure 5. Repression of TRPM4 expression using shRNA‐TRPM4 reduces hVIC mineralization and osteogenic differentiation.

A, Representative experiment showing calcium deposits measured by Alizarin red staining on hVIC after 14 days in culture. Cells were cultured in SM or PM with or without SH1, SH2, or SH1+2 at 1 ng/cm². Scale bar=200 μm. B–D, Quantification of calcium deposits on hVICs after 14 days in culture in SM, PM, PM+SH1 (N=9, B), PM+SH2 (N=11, C) or PM+SH1+2 (N=9, D) as shown in A. Histograms are mean±SEM of fold change vs PM. Circles connected by lines represent data for each patient measured in the 3 conditions (average of duplicate). E–G, Quantification of BMP2 (N=6, E); Runx2 (N=7, F) and αSMA (N=7, G) after 14 days of hVIC culture in SM, PM, PM+SH1, PM+SH2 or PM+SH1+2 culture conditions. Analyzed with 1‐way ANOVA (B–E, G) with uncorrected Fisher's least significant difference multiple comparisons test and analyzed with Friedman test (F) with Dunn's multiple comparisons test. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, ns=nonsignificant. BMP2 indicates bone morphogenetic protein 2; hVICs, human valvular interstitial cells; PM, pro‐calcifying media; Runx2, runt‐related transcription factor 2; shRNA, small hairpin RNA; SM, standard media; αSMA, α‐smooth muscle actin; and TRPM4, transient receptor potential melastatin 4.

Figure 6. Inhibition and repression of TRPM4 reduces SMAD1/5 and NFAT pathways activation induced by PM.

A, Western blots of P‐SMAD1/5 and SMAD1 total protein expression in hVICs after 14 days of culture in SM, PM, PM+9 PHE at 3.10⁻⁶ mol.L⁻¹ or PM+SH1+2 at 1 ng/cm². B, C, Quantification of P‐SMAD1/5 on SMAD1 total protein expression after 14 days of hVIC maintained in SM, PM, PM+9 PHE (N=9, B) or PM+SH1+2 (N=12, C) culture conditions. Histograms are mean±SEM of fold change vs PM. Analyzed with 1‐way ANOVA (B) with uncorrected Fisher's least significant difference multiple comparisons test and Friedman test (C) with Dunn's multiple comparisons test. D, NFAT protein expression (green) was observed by immunofluorescence in hVICs in SM, PM, PM+9 PHE or PM+SH1+2 (N=6) after 14 days of culture. Nuclei were labeled with DAPI (blue). Scale bar=50 μm. Upper line corresponds to the DAPI labeling and middle line corresponds to the NFAT labeling. The lower line corresponds to the merge (DAPI+NFAT) in the area indicated by a red square in the pictures of the middle line. E, NFAT protein expression was quantified by ImageJ software. Analyzed with 1‐way ANOVA with uncorrected Fisher's least significant difference multiple comparisons. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, ns, non‐significant. 9 PHE indicates 9‐phenanthrol; hVICs, human valvular interstitial cells; NFAT, nuclear factor of activated T cells; PM, pro‐calcifying media; SM, standard media; and TRPM4, transient receptor potential melastatin 4.