Spermidine Enhances Mitochondrial Function and Mitigates Aortic Valve Calcification: Implications for DNA Methyltransferase-1 Activity

Abstract

Aortic stenosis (AS) is a severe heart valve disease marked by calcification, leading to heart failure. This study examined mitochondrial function in human aortic valve interstitial cells isolated from patients with AS and tested spermidine, an autophagy inducer as AS treatment. Spermidine treatment reduced fibrosis and calcification in human aortic valve interstitial cells and improved these features in spermidine-treated mice. The AKT-TP53-DNMT1-PPARG pathway was implicated, and DNA methyltransferase 1 inhibition by 5-azacytidine enhanced mitochondrial biogenesis by reducing mitochondrial DNA hypermethylation. These findings suggest that spermidine or DNA methyltransferase 1 inhibition could prevent aortic valve disease by improving mitochondrial function.

Keywords: DNA methyltransferase 1; aging; aortic stenosis; mitochondrial function; spermidine.

Graphical abstract

Figure 1

Decreased Mitochondrial Function in hAVICs From Patients With AS (A) Mitochondrial respiration was measured using the Seahorse XF24 Analyzer, and the oxygen consumption rate (OCR) was determined in normal hAVICs (n = 10) compared with diseased hAVICs (n = 10). (B) ATP production and coupling efficiency were quantified by normalizing of OCR levels to total protein. (C) Representative confocal microscopy images of hAVICs stained with the mitochondria-specific dye MitoTracker, with integrated density of MitoTracker divided by area quantified (n = 10). (D) Mitochondria in the aortic valve of patients with AS were compared with those in normal valves using transmission electron microscopy images, and measurements of mitochondrial area (μm²) were made. (E) Western blot analysis of the protein expression of mitochondrial-related genes (OPA1, TOMM20, DRP1) and autophagy-related genes (BECN1, ATG5, ATG16L1, ATG7, ATG3) was performed in hAVICs (n = 16 each) and normalized against β-actin. For (A), (B), and (C), data are presented as mean ± SEM as the data followed a normal distribution as confirmed by the Shapiro-Wilk test. Student’s t-test was used for statistical analysis. For (D) and (E), the data showed a non-normal distribution and are presented as scatter plots or box plots with median and IQR. Statistical comparisons were made using the Mann-Whitney U test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. AS = aortic stenosis; DAPI = 4ʹ,6-diamidino-2-phenylindole; hAVIC = human aortic valve interstitial cell.

Figure 2

Spermidine Inhibits Markers of Fibrosis and Calcification in hAVICs Gene expressions were confirmed in hAVICs by real-time polymerase chain reaction after treatment with 10, 50, and 100 μM spermidine for 24 hours. (A) Expression of fibrosis-related genes (FN1 (fibronectin 1), MMP2, MMP9, COL1A1) and calcification-related genes (ALPL (ALP), BGLAP (osteocalcin), RUNX2) were decreased in diseased hAVICs (n = 8). (B) Representative image of Alizarin red staining to visualize calcium deposition. A total of 6 independent experiments were performed. Scale bar = 500 μm. Quantification of Alizarin red staining at 562-nm absorbance was performed by destaining solution with cetylpyridinium chloride. (C) Relative changes in messenger RNA expression of ALPL (ALP), and RUNX2 were quantified after treatment with spermidine. All data showed a non-normal distribution and are presented as box plots with median and IQR. Statistical comparisons were made using the Kruskal-Wallis test with Dunn’s post hoc test for multiple comparisons.∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. Abbreviation as in Figure 1.

Figure 3

Effect of Spermidine Treatment on Aging Ldlr-Deficient Mice (A) During the period from 24 to 62 weeks, Ldlr-deficient mice were fed a high-fat diet (HFD), and spermidine was added to their water. (B) Ex vivo results of the Seahorse assay performed on mitochondria isolated from mouse myocardial tissue. Oxygen consumption rates (OCRs) were altered by oligomycin, FCCP, and antimycin A. Basal respiration, ATP production, proton leak, and maximal respiration values were quantified by monitoring the changing OCR (WT normal diet n = 8, Ldlr^−/−HFD-Water n = 10, Ldlr^−/−HFD-Spermidine n = 11) (C) Western blot results showing the expression of BECN1 and LC3 A/B protein from the myocardium of each group. Images were quantified using ImageJ and normalized to GAPDH. (D) Western blot analysis showing the expression of BECN1, OPA1 (OPA1 mitochondrial dynamin-like GTPase), TOMM20 (translocase of outer mitochondrial membrane 20), and LC3 A/B (autophagy marker light chain 3 isoform A, B) proteins from the aortic valves of each group. All data showed a non-normal distribution and are presented as box plots with median and IQR. Statistical comparisons were made using the Mann-Whitney U test for 2-group comparisons and the Kruskal-Wallis test with Dunn’s post hoc test for multiple comparisons. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. Ldlr = low-density lipoprotein receptor; WT = wild type.

Figure 4

Aortic Valve Thickness and Fibrocalcific Change in the 3 Groups (A, E) Valve thickness was measured by hematoxylin-eosin (H&E) staining. The average thickness of the valve was determined by dividing the area by the length of the valve after H&E staining (WT normal diet n = 8, Ldlr^−/−HFD-Water n = 16, Ldlr^−/−HFD-Spermidine n = 18). (B, F) Representative images of aortic valve Masson's trichrome (MT) staining among the 3 groups. Areas of collagen-positive staining (blue signal) were measured using ImageJ (WT normal diet n = 8, Ldlr^−/−HFD-Water n = 15, Ldlr^−/−HFD-Spermidine n = 20). (C, G) The amount of fibrocalcific deposition was evaluated by immunohistochemistry using a COL1A1 antibody. For quantification, the COL1A1-positive area was measured relative to the valve area using ImageJ (WT normal diet n = 8, Ldlr^−/− HFD-Water n = 15, Ldlr^−/− HFD-Spermidine n = 20) (D, H) Calcium deposits in the valve were quantified by Alizarin red staining in the 3 groups. Representative images are shown together with a graph of the positive area measured relative to the valve area using ImageJ (WT normal diet n = 6, Ldlr^−/−HFD-Water n = 15, Ldlr^−/−HFD-Spermidine n = 16) All data are presented as mean ± SEM as the data followed a normal distribution as confirmed by the Shapiro-Wilk test. One-way analysis of variance with Tukey’s post hoc test for multiple comparisons was used for statistical analysis. Scale bar = 200 or 500 μm. ∗P < 0.05, ∗∗P < 0.01. Abbreviations as in Figure 3.

Figure 5

Echocardiographic Markers in Spermidine- and Water-Treated Mice Transaortic maximum velocity (Vmax), peak pressure gradient (PGmax), and left ventricular mass (LVmass) were measured by echocardiography at 24, 48, and 60 weeks (n = 18, 18, and 17 mice for each time point, respectively) in both spermidine- and water-treated mice. Data are presented as box plots with median and IQR, and mixed-effects generalized linear model analysis was performed. ∗P < 0.05, ∗∗P < 0.01. Abbreviations as in Figure 3.

Figure 6

Results of Proteome Analysis of Aortic Valve Tissue From Animal Experiments and IPA (A) Venn diagram showing the number of common and unique proteins expressed in each group. (B) Volcano plot for total proteins of the proteome analysis. Green and red areas indicate downregulated and upregulated proteins expressed in the spermidine group compared with the water group. (C) Ingenuity Pathway Analysis (IPA) canonical pathways from the proteomics data set. A Z score is an index used to determine the activity of a pathway based on gene expression from the knowledge base, gene expression in our dataset, and their relationship to each other. (D) Interaction network analysis in IPA showing the relationships of different molecules with SIRT1 as the center. (E) Upstream regulator analysis in IPA identifying downstream proteins regulated by PPARG and DNMT1, revealing a relationship with SIRT1.

Figure 7

Protein Expression of Molecules Included in the Hypothesis (A) Western blot analysis was performed on the molecules TP53 and SIRT1, which were included in the hypothesis established by the results of proteome analysis in aortic valve tissue from the animal model. (B) Western blot images were quantified using ImageJ and normalized to β-actin. (C) Immunohistochemical analysis was performed for the molecules DNMT1 and PPARG. Staining was performed on aortic valves from mice fed an HFD with either water or spermidine. The scale bar in images represents 100 μm or 200 μm. (D) Quantitative analysis was performed by measuring the total area and the DAB-positive area of a single valve and distributing them as percentages of the stained area. A total of 36 valves from 12 tissues were measured for each protein. Data are presented as mean ± SEM as the data followed a normal distribution as confirmed by the Shapiro-Wilk test. Student’s t-test was used for statistical analysis. ∗P < 0.05, ∗∗P < 0.01. Abbreviations as in Figure 3.

Figure 8

Spermidine Improves Mitochondrial Function through AKT-TP53-DNMT1-PPARG Pathways in hAVICs hAVICs from patients with aortic stenosis were treated with 10, 50, and 100 μM spermidine for 24 hours. (A) Western blot analysis confirmed an increase in TP53, SIRT1, and PPARG, whereas AKT phosphorylation and DNMT1 were inhibited. (B) Levels of protein expression were normalized to β-actin from total lysate and Lamin B1 from nuclear lysate (n = 10). The data showed a non-normal distribution and are presented as box plots with median and IQR. Statistical comparisons were made using the Kruskal-Wallis test with Dunn’s post hoc test for multiple comparisons. (C) Mitochondrial respiration was measured using the Seahorse XF24 Analyzer after spermidine treatments (n = 3 for each group). (D) ATP production and coupling efficiency were quantified by normalizing the values of oxygen consumption rate to total protein. Data are presented as mean ± SEM as the data followed a normal distribution as confirmed by the Shapiro-Wilk test. One-way analysis of variance with Tukey’s post hoc test for multiple comparisons was used for statistical analysis. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. AVIC = aortic valve interstitial cell; other abbreviations as in Figure 1.

Figure 9

5-Azacytidine as a DNMT1 Inhibitor Induces the Expression of Mitochondrial Biogenesis–Related Genes (A) Immunohistochemistry showed that DNMT1 was expressed significantly higher in human aortic valve leaflets than in normal valves. (B) The stained area of DNMT1 divided by the total area was quantified. The data showed a non-normal distribution and are presented as scatter plots and statistical comparisons were made using the Mann-Whitney U test. (C) mRNA expressions were confirmed in hAVICs by real-time PCR after treatment with 5, 10, and 20 μM 5-azacytidine (5-AzaC) for 24 hours (n = 9 for each group). The mRNA expression of mtDNA (ATP8, ND3, CO1) increased and the mRNA expression of DNMT1 decreased. The expression of mitochondrial biogenesis–related genes (PPARG, CPT2, AMPKA, PGC1A, NRF2, TFAM, SOD2) also increased. The data showed a non-normal distribution and are presented as box plots with median and IQR. Statistical comparisons were made using the Kruskal-Wallis test with Dunn’s post hoc test for multiple comparisons. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. DNMT1 = DNA methyltransferase 1; mRNA = messenger RNA; mtDNA = mitochondrial DNA; other abbreviation as in Figure 1.