TRPC3 shapes the ER-mitochondria Ca2+ transfer characterizing tumour-promoting senescence

Abstract

Cellular senescence is implicated in a great number of diseases including cancer. Although alterations in mitochondrial metabolism were reported as senescence drivers, the underlying mechanisms remain elusive. We report the mechanism altering mitochondrial function and OXPHOS in stress-induced senescent fibroblasts. We demonstrate that TRPC3 protein, acting as a controller of mitochondrial Ca²⁺ load via negative regulation of IP₃ receptor-mediated Ca²⁺ release, is down regulated in senescence regardless of the type of senescence inducer. This remodelling promotes cytosolic/mitochondrial Ca²⁺ oscillations and elevates mitochondrial Ca²⁺ load, mitochondrial oxygen consumption rate and oxidative phosphorylation. Re-expression of TRPC3 in senescent cells diminishes mitochondrial Ca²⁺ load and promotes escape from OIS-induced senescence. Cellular senescence evoked by TRPC3 downregulation in stromal cells displays a proinflammatory and tumour-promoting secretome that encourages cancer epithelial cell proliferation and tumour growth in vivo. Altogether, our results unravel the mechanism contributing to pro-tumour behaviour of senescent cells.

Fig. 1. Senescent prostate stromal cells have an altered calcium homoeostasis.

a Heatmap of mRNA levels for the indicated ion channels and accessory proteins after chemotherapy-induced senescence in HPrFs or CAFs, oxidative stress-induced senescence in HPrFs and oncogene-induced senescence in MRC5 fibroblasts. Mean values of the fold changes relative to controls (n = 3) are presented in red-green colour scale (see scalebar). IL-8 mRNA expression was used as a positive control for senescence induction. b–e TRPC3 protein levels are shown in the four models of senescence. Data are expressed as mean ± SEM (n = 3). Western blot images are representative of three independent experiments. f–i Cytosolic Ca²⁺ imaging (Fura-2) was performed on HPrFs (CIS, Docetaxel, 5 nM) or MRC5 (OIS, 4-OHT, 250 nM). f and h Averaged time courses (mean ± SEM) of fluorescence ratio with excitation wavelengths 340/380 nm. g and i Means values of (1) basal cytosolic Ca²⁺ levels, (2) peak amplitudes of the Histamine (100 µM)-induced Ca²⁺ release from the ER and (3) peak amplitudes of SOCE, respectively (n_Veh = 212, n_Doce = 234; n_-4OHT = 66, n_+4OHT = 61. Data are presented as mean ± SEM (n = 3). ***P < 0.001 (Student’s t test, two-sided). See also Supplementary Fig. 1.

Fig. 2. TRPC3 controls senescence.

a Cumulative population doublings (PD) of HPrFs either stably expressing doxycycline-inducible shRNA against TRPC3 or treated with siTRPC3 and their siControls (siCTL, scrambled sequences). Results are presented as mean ± SEM; n = 3). b EdU staining on MRC5 cells treated 10 days with siCTL or siTRPC3. % of EdU positive cells is shown (mean ± SEM; n = 3). Scale bar = 150 µm. ***P < 0.001 (Student’s t test, two-sided). c Effect of siRNA-mediated TRPC3 knockdown in HPrFs on: (Top) the cell structural parameters revealed with flow cytometry analysis—cell size (forward scatter; FSC) and internal complexity (side scattered, SSC), (Middle) cell growth reported by crystal violet staining and (Bottom) β-gal activity revealed with SA-β-gal staining (% of ß-galactosidase positive cells is shown, mean ± SEM). Readings were performed 10 days after TRPC3 knockdown. n = 3 biologically independent experiments. d p53 and p16 protein levels are shown for cells treated as in (c). Data are expressed as mean ± SEM (n = 3). Western blot images are representative of three independent experiments. e yH2AX staining and fluorescence microscopy of MRC5 cells treated as in (c) showing signs of DNA damage after 10 days of TRPC3 knock-down. % of yH2AX positive cells is shown (mean ± SEM, n = 3 biologically independent experiments). ***P < 0.001 (Student’s t test, two-sided). f Effect of TRPC3 overexpression in OIS. (Top) colony formation assay reported by crystal violet staining and (bottom) SA-β-gal activity revealed with SA-β-gal staining (% of ß-galactosidase positive cells is shown, mean ± SD). Readings were performed from ER:hRAS-MRC5 cells (bearing doxycycline-inducible TRPC3) in which senescence was induced by 4-OHT treatment over 10 days. To overexpress TRPC3, the cells were kept in doxycycline-containing medium for 3 weeks after senescence induction. g Cyclin D1, PCNA1 and p16 protein levels are shown for cells treated as in (f) and normalised against Calnexin expression used as loading control. Western blot images are representative of three independent experiments. Scale bar = 100 µm.

Fig. 3. TRPC3 inhibits ER Ca²⁺ release.

a Cytosolic Ca²⁺ imaging (Fura-2) performed on HPrFs 10 days after shRNA induction. Left: averaged time courses (mean ± SEM) of fluorescence ratio with excitation wavelengths 340/380 nm. Right: panels comparing the mean values of (1) basal cytosolic Ca²⁺ levels, (2) peak amplitudes of the Histamine (100 µM)-induced ER Ca²⁺ release and (3) peak amplitudes of SOCE, respectively (n_shCTL = 194, n_shTRPC3 = 173). Data are mean ± SEM (n = 3). ***P < 0.001 (Student’s t test, two-sided). b Representative traces (left) of IP₃-AM (250 nM)-induced [Ca²⁺] oscillations reported by confocal time-series imaging (at 5 Hz) of Fluo-4 fluorescence in fibroblasts 5 days after shRNA induction: shCTL and shTRPC3, respectively. The bar diagram plot (right) compares corresponding mean signal temporal densities (n_shCTL = 25; n_shTRPC3 = 26). ***P < 0.001 (Student’s t test, two-sided). See also Supplementary Figs. 2 and 3. c The plots show (mean ± SEM) traces of the self-normalised fluorescence ratio (R/R₀, where R = F_470–500/F_>520 at any given time while R₀ is R averaged before digitonin application) measured in HPrFs 5 days after shRNA induction (n_shCTL = 60) (n_shTRPC3 = 76) or overexpressing TRPC3 (n_over = 52), as indicated. d The bar diagram plot compares mean rates of Ca²⁺ release from the ER, calculated as signal mass per second in fibroblasts 5 days after shRNA induction (shCTL, blue), (shTRPC3, red) or overexpressing TRPC3 (green). ***P < 0.001 relative to shCTL (Student’s t test, two-sided). e The same as in (c) but measured in wild-type fibroblasts pre-treated with either XeC (grey; n_XeC = 132) or vehicle only (black; n_Ctl = 107). The bar diagram plot compares (XeC vs. Ctl) mean rates of Ca²⁺ release from the ER, calculated as signal mass per second. ***P < 0.001 (Student’s t test, two-sided). f The pie chart (right) illustrates IP₃R-mediated Ca²⁺ release as a percentage of the total ER release in non-stimulated cells.

Fig. 4. TRPC3 reduces the rate of IP₃-induced Ca²⁺ release from the ER.

a, b Mean traces of the self-normalised GEM-CEPIA1_er fluorescence ratio (R/R₀) measured at different concentration of IP₃ (as depicted above plots in corresponding colours) 5 days after shRNA induction (b; n = 76, 80, 101, 117 and 80, respectively, for [IP₃] = 0, 10, 100, 1000 and 10000 nM) or 3 days after TRPC3 overexpression (c; n = 52, 113, 51, 104, and 59, respectively for [IP₃] = 0, 10, 100, 1000 and 10,000 nM). c Bar diagram plot of mean rates of Ca²⁺ release from the ER (derived from data in panels a, b), calculated as signal mass per second at different concentrations of IP₃, after induction of shRNA (red) or TRPC3 overexpression (green) as described above. ***P < 0.001, *P < 0.05 (Student’s t test, two-sided). d and e Dependence of the mean rate of Ca²⁺ release on [IP₃] plotted on logarithmic scale for total (d) and IP₃-dependent Ca²⁺ release (e). Note that the expression of TRPC3 suppresses IP₃-mediated Ca²⁺ release at concentrations of IP₃ < 100 nM. f Left: sample traces of IP₃R activity obtained using organelle membrane-derived patch clamp from the ER fractions extracted from cells either transduced with an empty vector (top, n = 9) or TRPC3 (bottom, n = 7). Right: average NP_open summarised in the barplot. *P < 0.05 (Student’s t test, two-sided). g Plots (left) relate traces (mean ± SEM) of the IP₃-induced responses (self-normalised GEM-CEPIA1_er fluorescence ratio; R/R₀) obtained in fibroblasts transduced with empty vector (n = 108) to those transduced with wtTRPC3 (n = 168) or the TRPC3 CIRB mutant MKR/AAA (n = 179) and corresponding bar diagram plots (right) comparing mean rates of Ca²⁺ release. h The same as in (g) but illustrating the effect of the transduction with TRPC3 pore mutants E630K (n = 153) or E630Q (n = 124). ***P < 0.001 for wtTRPC3 (n = 156) vs. empty vector (n = 131), ^#P < 0.001 for mutants vs. wtTRPC3 (one-way ANOVA, Tukey’s multiple comparisons test). See also Supplementary Fig. 4.

Fig. 5. The level of TRPC3 in fibroblasts affects basal mitochondrial Ca²⁺ load.

a The gallery (left) shows (from top to bottom) the transmitted light image of a single fibroblast, confocal image of GEM-GECO1_mito fluorescence captured at wavelengths 470–500 nm, at wavelengths >520 nm and their overlay. The gallery (middle) compares confocal images (overlay of 2 channels: BP 470–500 nm plus LP > 520 nm) of GEM-GECO1_mito fluorescence from the fibroblasts transduced with shCTL and shTRPC3, as indicated. The bar diagram plot (right) compares mean resting [Ca²⁺]_mito (estimated as F_470–500/F_>520) in the fibroblasts transduced with shCTL (n = 96) and shTRPC3 (n = 96). ^***P < 0.001 relative to shCTL (Student’s t test, two-sided). b Elevation of mitochondrial Ca²⁺ load upon hRAS overexpression is markedly prevented by simultaneous overexpression of TRPC3. The gallery (left) compares confocal images (overlay of 2 channels: BP 470–500 nm plus LP > 520 nm) of GEM-GECO1mito fluorescence from control MRC5, MRC5 with hRAS-induced senescence (hRAS) and MRC5 overexpressing hRAS and TRPC3 simultaneously (hRAS + TRPC3) or TRPC3 only (TRPC3). Images of three representative fields of view per condition are shown. The bar diagram plot (right) compares mean resting [Ca²⁺]_mito (estimated as F_470–500/F_>520) in control MRC5 (Control: n = 390) with that in MRC5 with hRAS-induced senescence (+hRAS: n = 232) and MRC5 overexpressing hRAS and TRPC3 simultaneously (+hRAS + TRPC3: n = 266) or TRPC3 only (+TRPC3: n = 251). ^***P < 0.001 relative to control, unless stated otherwise. One-way ANOVA, Tukey’s multiple comparisons test.

Fig. 6. Silencing of TRPC3 promotes spontaneous Ca²⁺ oscillations and transient membrane depolarisations in mitochondria.

a–c Ratiometric confocal [Ca²⁺]_mito imaging (at 0.2 Hz) was performed in fibroblasts expressing GEM-GECO1mito and transduced with either shCTL (a) or shTRPC3 (b). The plots (top) show temporal profiles of the GEM-GECO1mito self-normalised fluorescence ratio (ΔR/R₀; where R = F_470–500/F_>520 at any given time, while R₀ is a mean of at least 50 sample minimums of R) for four cells. The galleries (bottom) show every 50th image (overlay of 2 channels: BP 470–500 nm plus LP > 520 nm) captured during entire observation period (traces: top). c The bar diagram plot compares corresponding mean signal temporal densities, calculated as signal mass (∫(ΔR/R₀)) per second, in shCTL (n = 9) and shTRPC3 fibroblasts (n = 11). ***P < 0.001, Student’s t test, two-sided. d–f Changes in the mitochondrial membrane potential (at 0.1 Hz) were monitored in TMRE-loaded fibroblasts transduced with either shCTL (d) or shTRPC3 (e). The plots (middle) show temporal profiles of the TMRE self-normalised fluorescence ratio (ΔF/F₀) at 5 boxed (top) regions of interest (ROIs), as indicated. The galleries (bottom) show every 2nd image (left to right, top to bottom) captured at corresponding ROIs. f The bar diagram plot compares corresponding mean signal temporal densities in shCTL (n = 6; 75 ROIs) and shTRPC3 (n = 5; 92 ROIs) fibroblasts. ^***P < 0.001 (Student’s t test, two-sided).

Fig. 7. Relief of the TRPC3-mediated inhibition of IP₃Rs promotes senescence via facilitation of ER-mitochondria Ca²⁺ transfer, alteration of mitochondrial metabolism and ROS production.

a The galleries (from left to right) show the confocal images of the GO-ATeam probe fluorescence excited at 488 nm and captured at wavelengths of 500–520 nm, >550 nm and their overlay. The bar diagram plot (right) compares mean ATP levels (estimated as F_500––520/F_>550) in the MRC5 transfected with siCTL (n = 162) or siTRPC3 (n = 178). ^***P < 0.001 relative to siCTL (Student’s t test, two-sided). b Oxygen consumption rate (OCR in pmol O₂ min⁻¹ 50,000 cells⁻¹) measured using the Seahorse XFe24 technology (mean ± SD; n = 4) in fibroblasts transfected with either siCTL, siTRPC3 or siTRPC3/siIP3R3. Inhibitors have been injected as indicated by arrows: Oligomycin A (Oli.), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP), and rotenone and antimycin A (Rot/AA). c and d States of mitochondrial respiration are “basal” (c) for basal respiration and “ATP Turnover” (d) for ATP-linked OCR (basal measurement minus oligomycin response). Mean ± SD; n = 4 (one-way ANOVA followed by Tukeys’s post-test *P < 0.05). See also Supplementary Fig. 6e–g. e Confocal images of Mitosox Green and Mitotracker Deep Red fluorescence (top) and bar diagram plot (bottom) of Mitosox/Mitotracker fluorescence ratio comparing mitochondrial ROS production in fibroblasts transduced either with shCTL (n = 74) or shTRPC3 (n = 143) and in siIP₃R3-transfected fibroblasts transduced with shTRPC3 (n = 74). Mean ± SD ^***P < 0.001, ^**P < 0.01 (one-way ANOVA, Tukey’s multiple comparisons test).

Fig. 8. TRPC3 knockdown-induced senescence is responsible for the pro-tumour behaviour of prostate stromal cells.

a Heatmap showing the log2 fold change in secreted SASP, as revealed by antibody arrays, by CAFs treated with siTRPC3 and compared to siCTL cells (n = 3, multiple T-test, two-sided, Holm–Sidak correction for multiple comparisons). b Bar diagram plot comparing cell proliferation measured with MTS proliferation assay in PC3 (left) and DU145 (right) cell lines cultured in: (1) non-conditioned medium (NCM), (2) conditioned medium from the siCTL-transfected CAFs (CAFs siCTL CM) and (3) conditioned medium from the siTRPC3-transfected CAFs (CAFs siTRPC3 CM). The readings (light absorption) from PC3 or DU145 cells obtained in CAFs siCTL CM and CAFs siTRPC3 CM were normalised to corresponding readings obtained in NCM and expressed as % (mean ± SEM, n = 3: one-way ANOVA, Tukey’s multiple comparisons test). c Same as in (b) but illustrating cell lines cultured in conditioned medium from: (1) control cells, (2) hRAS-overexpressing cells (hRAS), (3) hRAS + TRPC3-overexpressing cells or (4) cells overexpressing TRPC3 alone (mean ± SEM, n = 3: one-way ANOVA, Tukey’s multiple comparisons test). d Bar diagram plot comparing survival of PC3 and DU145 cells (as indicated) after docetaxel treatment between the cell groups described in (b). The same measurements were done on the cells exposed to DMSO (vehicle) instead of docetaxel. Chemotherapy-resistance was expressed as a ratio of the cell survival observed after treatment with docetaxel to that observed with vehicle only and expressed in % (mean ± SEM, n = 3: one-way ANOVA, Tukey’s multiple comparisons test). e Time course (6 weeks) of the bioluminescence mean (10 mice; ±SD) total flux (top) and underlying individual traces (bottom) obtained 15 min after luciferin administration. The mice, continuously admitting doxycycline, were injected with PC3-Luc cells mixed with either siCTL-transduced (black line) or siTRPC3#1-transduced (red line) HPrFs. ^***P < 0.001, ^**P < 0.01 (one-way ANOVA, Tukey’s multiple comparisons test). f Bioluminescence images of oxidised luciferin from 3 mice of the two groups (panel e) on the 6th week after tumour challenge. g Comparison of sizes from tumours collected 6 weeks after tumour challenge. Scale bar: 0.5 cm.