Identification of Prominin-2 as a new player of cardiomyocyte senescence in the aging heart

Abstract

The aging heart is characterized by a number of structural changes leading to ventricular stiffness, impaired resistance to stress and increased risk of developing heart failure (HF). Genetic or pharmacological removal of senescent cells has recently demonstrated the possibility to relieve some cardiac aging features such as hypertrophy and fibrosis. However, the contribution of the different cell types in cardiac aging remains fragmentary due to a lack of cell-specific markers. Cardiomyocytes undergo post-mitotic senescence in response to telomere damage, characterized by persistent DNA damage response and expression of the classical senescence markers p21 and p16, which are shared by many other cell types. In the present study, we used transcriptomic approaches to discover new markers specific for cardiomyocyte senescence. We identified Prominin2 (Prom2), encoding a transmembrane glycoprotein, as the most upregulated gene in cardiomyocytes of aged mice compared to young mice. We showed that Prom2 was upregulated by a p53-dependent pathway in stress-induced premature senescence. Prom2 expression correlated with cardiomyocyte hypertrophy in the hearts of aged mice and was increased in atrial samples of patients with HF with preserved ejection fraction. Consistently, Prom2 overexpression was sufficient to drive senescence, hypertrophy and resistance to cytotoxic stress while Prom2 shRNA silencing inhibited these features in doxorubicin-treated cardiac cells. In conclusion, we identified Prom2 as a new player of cardiac aging, linking cardiomyocyte hypertrophy to senescence. These results could provide a better understanding and targeting of cell-type specific senescence in age-associated cardiac diseases.

Keywords: aging; cardiomyocyte; heart; senescence.

FIGURE 1

Cell‐specific and tissue‐specific expression of the cardiac senescence markers. (a) Experimental design for the isolation of tissues, cardiomyocytes (cardios) or non‐cardiomyocyte cells (non‐cardios) in young (3 months) and aged (20 months) C57Bl6J mice. (b) Heart weight (HW)‐to‐tibia length (TL) ratios in young (3 months) and aged (20 months) C57Bl6J mice. N = 6, *p < 0.05. (c) Echocardiographic analysis showing estimated Left Ventricle (LV) mass (mg), diastolic Interventricular Septum Thickness (IVSTd) and Fractional Shortening (FS). N = 8, *p < 0.05; **p < 0.01. (d) The heatmap of the most upregulated genes (log2‐fold changes) in the RNAseq analysis of purified cardiomyocytes from old mice compared to young mice (N = 5) (p < 0.001). (e) The mRNA expression of senescence markers by RT‐QPCR in seven distinct tissues of old mice (20 months) relative to young mouse tissues (3 months). Results are expressed as mean ± SEM of individual mice: heart N = 7, liver N = 6, kidney N = 6, lung N = 10, brain N = 4, gut N = 6, skeletal muscle N = 5. *p < 0.05; **p < 0.01; ***p < 0.001 in old vs young with Student's t‐test. (f) mRNA expression of senescence markers by RT‐QPCR in different cell fractions of the heart (cardiomyocytes or non‐cardiomyocytes cells) in old mice (20 months) relative to young mice (3 months). Results are expressed as mean ± SEM of individual mice: cardiomyocytes (N = 7) or non‐cardiomyocyte cells (N = 8). Statistical analysis was performed by Student's t‐test **p < 0.01; ***p < 0.001.

FIGURE 2

Prom2 is the most robust and conserved marker in rat and human cardiac aging. (a) mRNA expression of senescence markers by RT‐qPCR in whole hearts from young (3 months, Y) and old (24 months, O) rats. Data are expressed as mean ± SEM of relative gene expression versus Y group (N = 6 rats). *p < 0.05; **p < 0.01 by Student's t‐test. (b) mRNA expression of PROM2 in human atrial samples of young (Y), old (O), old + Heart Failure with preserved Ejection Fraction (HFpEF) and old + HFrEF patients. Data are expressed as mean ± SEM of relative expression versus Y group. N = 8 Y, N = 20 O, N = 12 HFpEF, N = 12 HFrEF. Multiple comparison test Kruskal–Wallis test with Dunn's post‐hoc was performed test for statistical analysis *p < 0.05.

FIGURE 3

Prom2 is associated with cardiomyocyte hypertrophy in the aging heart (a) The protein levels of Prom2 by immunoblot analysis in young and old mice heart homogenates (N = 6–8). Upper panel: Representative immunoblots and lower panel: quantifications of Prom2 low molecular weight (LMW) and high molecular weight (HMW) normalized to GAPDH and expressed as fold over young mice. p < 0.05 by Student's t‐test. (b) Upper panel: Representative confocal acquisition of Prom2 (green) and WGA (red) immunofluorescence staining in young (3 months) and old (20 months) mice hearts. Scale bar = 20 μm. Lower left panel: Quantification of Prom2⁺ cardiomyocytes in young (Y) and old (O) mice expressed as % of total cardiomyocytes. Lower right panel: Cell area measurement of positive Prom2⁺ cardiomyocytes (green + red) or negative Prom2⁻ cardiomyocytes (Red only) in old mice compared to young mice. About 100 cardiomyocytes were manually counted and measured in each heart. Data are expressed as mean ± SEM of N = 6 mice per group. (c) Upper panel: Representative immunoblot expression of Prom2 and GAPDH in H9C2 cells transfected with empty vector (Vec) or Prom2‐GFP (Prom2) for 48 h. Lower panel: The quantifications of Prom2 to GAPDH expression ratio. Data are expressed as mean ± SEM of N = 7. (d) Upper panel: Representative fluorescent images of H9C2 cells transfected with empty Vec or Prom2‐GFP (green) for 96 h and counterstained with the cardiomyocyte marker α‐actinin (red). Bottom panel: Quantitative measurements of cell area of Prom2⁺ cells (Prom2‐GFP, green + red) compared to Prom2⁻ cells (untransfected, red only). Scale bar = 50 μm. (e) Left panel: Representative fluorescent image of H9C2 cells transfected with Prom2‐GFP (green) for 48 h and counterstained with the DNA damage marker γH2AX (red) and DAPI (Blue). Right panel: Quantitative measurements of the number of γH2AX nuclear foci in Prom2⁺ versus Prom2⁻ cells. Data are expressed as mean ± SEM of N = 4 experiments. Statistical analysis was performed by Student's t‐test or one‐way ANOVA for more than two groups. *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 4

Prom2 is upregulated in stress‐induced senescence. (a) SA‐βgal activity in H9C2 cells treated or not with Doxorubicine (Doxo) for 96 h (100 nM). The number of SA‐βgal⁺ (blue) cells relative to total number of cells is displayed in the corresponding graph (right) and expressed as mean ± SEM (N = 3). (b) Representative fluorescence of H9C2 cells co‐transduced with the lentivirus p53RE‐ZsGreen and p21‐Tdt‐tomato, and treated or not with Doxo for 96 h (100 nM). Scale bar = 100 μm. Cell areas of positive cells for p53RE‐ZsGreen or p21‐Tdt‐tomato (pos) and negative (neg) after Doxo treatment were compared with untreated control cells (CT). Data are expressed as mean ± SEM of N = 4 experiments. (c) The evaluation of p21 and p‐p53 (phosphorylated p53) expressions by western blotting in Doxo‐treated cells. Data are expressed as mean ± SEM of n = 4 experiments. (d) mRNA expression of expression of senescence markers (p16, p15, p21) and SASP markers (GDF15, TGFβ2) by RT‐qPCR in H9C2 treated with Doxo for 96 h. Data are expressed as mean ± SEM of Doxo versus untreated CT group (N = 6). (e) mRNA expression of the Prom2 by RT‐qPCR in H9C2 treated with Doxo for 96 h. Data are expressed as mean ± SEM of Doxo versus untreated CT group (N = 6). (f) Representative immunoblot of Prom2 and GAPDH in H9C2 cells treated with Doxo (96 h). Lower panel: Quantifications of Prom2 to GAPDH expression ratio. Data are expressed as mean ± SEM of N = 3 experiments. Statistical analysis was performed by Student's t‐test (for two groups) or one‐Way ANOVA with post‐hoc tukey test (for three groups). *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 5

Prom2 upregulation is induced upon p53 activation. (a) The analysis of Human Prom2 promoter with “MOTIFMAP” showing putative transcription factor binding sites relative to TSS (transcription Start Site). The p53 motif present at position −88 is shown below. (b) Representative fluorescence of H9C2 cells co‐transduced with the lentivirus p53RE‐ZsGreen and p21‐Tdt‐tomato, and treated with Nutlin‐3a (Nut) for 96 h (5 μM) or DMSO 0.1% (CT). Scale bar = 100 μm. Cell area of positive cells for p53RE‐ZsGreen or p21‐Tdt‐tomato (pos) and negative (neg) after Nutlin‐3a (Nut) treatment compared with untreated control cells (CT). Data are expressed as mean ± SEM of N = 4 experiments. (c) SA‐βgal activity in H9C2 cells treated with Nutlin‐3a (Nut) for 96 h (5 μM) or DMSO 0.1% (CT). The number of SA‐βgal⁺ (blue) cells relative to total number of cells is displayed in the corresponding graph (right) and expressed as mean ± SEM (N = 3). (d) The mRNA expression of senescence markers (Prom2, p16, p15, p21) by RT‐qPCR in Nutlin‐3a‐treated cells (5 μM) for 96 h. Data are expressed as mean ± SEM of relative expression vs DMSO 0.1% CT group (N = 6). Statistical analysis was performed by Student's t‐test (for two groups) or one‐Way ANOVA with post‐hoc tukey test (for three groups). *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 6

Prom2 recapitulates the main features of cell senescence. (a) Representative images of SA‐βgal activity in H9C2 cells transfected with empty vector (Vec) or pENTER‐Prom2 (Prom2) for 96 h. scale bar = 500 μm. The area of blue, indicative of SA‐βgal⁺ cells, divided by the total number of nuclei is displayed in the graph and expressed as mean ± SEM (N = 5). (b) Representative immunoblot and the quantitative expression of Prom2, p21 and p15/p16 proteins normalized to GAPDH in H9C2 cells transfected with empty Vec or pENTER‐Prom2 for 72 h. Data are expressed as mean ± SEM of N = 4–5 experiments. (c) GDF15 and IL6 mRNA expression in H9C2 cells transfected with empty Vec or pENTER‐Prom2 for 72 h. Data are expressed as mean ± SEM (N = 10). (d) LDH release in the supernatant of H9C2 cells transfected with empty Vec or pENTER‐Prom2 for 72 h and treated with Veh or 1 mM H₂O₂ for 24 h. Data are expressed as mean ± SEM (N = 4). (e) Representative immunoblot and quantitative expressions of Bcl2 and Bcl‐XL/GAPDH in H9C2 cells transfected with empty Vec or pENTER‐Prom2 for 72 h. Data are expressed as mean ± SEM (N = 6). (f) Representative immunoblot and quantitative expressions of Prom2, LC3 and p62 relative to GAPDH in H9C2 cells transfected with empty Vec or pENTER‐Prom2 for 72 h and treated with Veh or BafA1 (100 nM) for 1 h. (g) Representative immunoblot and quantitative expressions of phospho‐p70S6K to GAPDH ratios in H9C2 cells transfected with empty Vec or pENTER‐Prom2 for 72 h. Data are expressed as mean ± SEM of N = 3–5 experiments. Statistical analysis was performed by Student's t‐test or two‐way ANOVA for more than two groups. *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 7

Prom2 silencing mitigates SIPS. (a) The mRNA expression of Prom2 in H9C2 cells transfected with ShCT or ShProm2 for 72 h. Data are expressed as mean ± SEM of N = 5 experiments. (b) Protein expression of Prom2 in H9C2 cells transfected with ShCT or ShProm2 for 6 days. Data are expressed as mean ± SEM of N = 3. (c) Immunofluorescence staining of WGA in H9C2 cells transfected with ShCT or ShProm2 for 72 h and then treated with Doxo for an additional 72 h. Scale bar = 50 μm. Data are expressed as mean ± SEM (N = 6). (d) p21 and GDF15 mRNA expression in H9C2 cells transfected with ShCT or ShProm2 for 24 h and treated with Doxo for an additional 72 h. Data are expressed as mean ± SEM (N = 7). (e) SA‐βgal activity in H9C2 cells transfected with ShCT or ShProm2 for 24 h and treated with Doxo for 72 h. The area of blue, indicative of SA‐βgal⁺ cells, divided by the total number of nuclei is displayed in the graph and expressed as mean ± SEM (N = 5).