Fig 1.
Down-regulation of Cisd2 in the naturally aged hearts and the cardiac pathology present in Cisd2KO and naturally aged WT mice.
(A) Western blot analysis of myocardial Cisd2 protein levels using heart tissue samples obtained from young (3M) and old (26M) WT mice (n = 4). Gapdh was used as an internal control. (B) Representative photographs of hearts obtained from WT and Cisd2KO mice at 3M and 6M, as well as those obtained from old WT mice at 26M. Scale bar, 2 mm. (C) Myocardial weights of the LV (n ≥ 5 per group). (D) Long-axis low-power section of Sirius Red/Fast Green staining of the hearts with the aim of detecting collagen. Scale bar, 1 mm. (E) Representative MRI (7-T MRI) of a separate cohort of mice (n ≥ 5 per group) is presented. (F and G) Analysis of the MRI of diastolic segment 7+8 (anterior LV free wall, F) and diastolic segment 10+11 (interventricular LV wall, G). A significant decrease in left ventricular wall thickness can be noted in the old WT mice at 26M and in the Cisd2KO mice at 3M and 6M. The data are presented as mean ± SD and are analyzed by Student t test. *p < 0.05; **p < 0.005. All the mice used in this study are males. Values for each data point can be found in S1 Data. 3M, 3 months old; 6M, 6 months old; 26M, 26 months old; Cisd, CDGSH iron-sulfur domain-containing protein; Gapdh, glyceraldehyde 3-phosphate dehydrogenase; KO, knockout; LV, left ventricle; MRI, magnetic resonance imaging; WT, wild type.
Fig 2.
Cisd2KO mice exhibit cardiac electromechanical dysfunctions that are similar to those observed in the naturally aged WT mice.
(A-D) Representative ECG tracings and continuous 5-minute waterfall plots recorded following anesthesia of the mice. A normal sinus rhythm ECG recorded from a young (3M) WT mouse (A). The irregular PR interval and missing beats recorded from an aged (26M) WT mouse (B). Representative dysrhythmic ECGs, namely a missing beat (C), AV block with irregular and progressive prolongation of PR interval, APCs, and VPCs (D), which were found in the Cisd2KO mice at 3M. (E and F) Measurements of the Tpeak–Tend intervals (E) and the QTc intervals (F) were obtained from 5 minutes of sequential beats obtained from whole ECG tracings. (G) Peak developed pressures and rates of pressure increase and rates of pressure decrease during left ventricular contraction; these were measured by Millar’s catheter in the mice. The dP/dt max is the maximal rate of pressure development, and the dP/dt min is the maximal rate of decay of pressure. The WT mice exhibited an age-dependent progression with respect to cardiac dysfunction with a reduced dP/dt Max and dP/dt min. (H and I) Left ventricular ejection fraction (H) and cardiac output (I) were measured and quantified by cardiac MRI. Data are presented as mean ± SD and are analyzed by Student t test. *p < 0.05; **p < 0.005. Mouse number n ≥ 5 for each group. Values for each data point can be found in S1 Data. 3M, 3 months old; 6M, 6 months old; 26M, 26 months old; APC, atrial premature complex; AV, atrioventricular; Cisd2KO, CDGSH iron-sulfur domain-containing protein 2 KO; ECG, electrocardiography; KO, knockout; MRI, magnetic resonance imaging; QTc, corrected QT; VPC, ventricular premature complex; WT, wild type.
Fig 3.
Cisd2 deficiency causes defects in the intercalated disc and leads to ultrastructural abnormalities in the cardiac muscle.
(A) Lateralization of gap junctions in the cardiac muscle of Cisd2KO mice. Representative IF images of heart sections stained with antibodies against Cx43 (green) in order to localize gap junctions, against pan-cadherin (red) to localize the intercalated discs, and against α-actinin (purple) to stain muscle fibers. The sections were also stained with Hoechst (blue) to identify nuclei. White arrows indicate lateralization of gap junctions. (B) Colocalization coefficient of gap junction protein (Cx43) and intercalated disc protein (pan-cadherin) was analyzed by Pearson’s correlation. The computed value is presented as the Cx43/pan-cadherin colocalization coefficient. Data were collected from 10 randomly selected fields for each heart sample. There are at least three individual mice in each group. (C and D) Relative protein levels of Cx43 (C) and phosphorylated Cx43 (Ser368) (D) in the hearts of WT and Cisd2KO mice were analyzed by western blotting and then quantified (n = 4). (E) Maldistribution of desmosomes in the cardiac muscle of Cisd2KO mice. Representative IF images of heart sections stained with antibodies against desmoplakin (green), which is localized within desmosomes, with WGA (red) to stain cell membranes by binding to membrane glycoproteins, with antibodies against pan-cadherin (red) to localize the intercalated discs, and with antibodies against α-actinin (purple) to stain muscle fibers. The sections were also stained with Hoechst (blue) to identify nuclei. (F and G) Relative protein levels of desmoplakin (F) and phosphorylated desmoplakin (G) in the hearts of WT and Cisd2KO mice were analyzed by western blotting and then quantified (n = 4). (H) TEM analysis reveals ultrastructural defects in the intercalated discs of Cisd2KO cardiac muscle. Overt ultrastructural abnormalities (namely, shortening and breaking down of the FA; degeneration of the mitochondria, including mitochondria with ruptured outer and inner membranes and swollen mitochondria with fewer crista; and extension and fragmentation of gap junctions) were easily found in the cardiac muscle of Cisd2KO mice at 3M. In H5 and H6, red stars (★) indicate that parts of the space between the two membranes of the intercalated discs are expanded. In H3, H6, and H7, purple dots indicate the two ends of gap junctions. In H6, yellow dots and arrows indicate several degenerating regions with loss of compactness interlaced into an extended gap junction (between two purple dots). Data are presented as mean ± SD and are analyzed by Student t test. *p < 0.05; **p < 0.005. Values for each data point can be found in S1 Data. 3M, 3 months old; Cisd2, CDGSH iron-sulfur domain-containing protein 2; Cx43, Connexin 43; FA, fascia adherens; IF, immunofluorescence; KO, knockout; SR, sarcoplasmic reticulum; TEM, transmission electron microscopy; TG, transgenic; WGA, wheat germ agglutinin; WT, wild type.
Fig 4.
Cisd2 deficiency disrupts Ca2+ homeostasis via a dysregulation of Serca2a activity and this then results in mitochondrial Ca2+ overload and dysfunction in cardiomyocytes.
(A and B) Spontaneous Ca2+ waves of beating cardiomyocytes isolated from WT, Cisd2KO, and Cisd2TG mice at 3M were measured by confocal microscopy (100 measures per second) using Fluo-4 AM staining. (A) The Ca2+ waves (F1/F0: fluorescence-intensity change normalized against the background fluorescence). Representative videos are provided in the supplemental information. (B) Quantification of maximal Ca2+ release (maximum-basal level). The data are presented as mean ± SD for 13–18 cells from 3–5 mice of each genotype. (C-E) The levels of cytosolic Ca2+ in single adult cardiomyocytes isolated from WT, Cisd2KO, and Cisd2TG mice were measured by fluorescence microscopy using Fura-2/AM staining. After measuring the basal level of cytosolic Ca2+ (first 50 seconds), Tag was added to release Ca2+ from the SR. Alternatively, CCCP was added to release Ca2+ from the mitochondria. Quantification of the basal cytosolic Ca2+ levels (C), Tag-induced Ca2+ elevation from the SR (D), and CCCP-induced Ca2+ elevation from mitochondria (E) of the adult cardiomyocytes. The data are presented as mean ± SEM for 12–42 cells from 5–6 mice of each genotype. (F) Contractility (cell shortening) was measured during the spontaneous beating of cardiomyocytes (representative videos are provided in the supplemental information). (G) Quantification of cell shortening for 21–31 cells from 5–6 mice of each genotype. The data are presented as mean ± SEM. (H) Quantification of ROS and RNS levels in the heart by measuring DCF levels using cardiac tissue samples (n = 4 for each group). The data are presented as mean ± SD. (I) OCRs of the isolated adult cardiomyocytes from different genotypes. The indicated chemicals (oligomycin A, FCCP, R/A) were added sequentially to determine the ATP-coupled respiration rate, the maximal respiration rate (Max), and the nonmitochondrial respiration rate, respectively, using a Seahorse XFe24 analyzer. (J and K) Representative confocal microscopic pictures (J) and quantification of JC-1 flowcytometric staining results (K), which were used to measure the mitochondrial membrane potential of adult cardiomyocytes isolated from WT, Cisd2KO, and Cisd2TG mice. (L) Western blot analysis of Cisd2, Serca2, Calnexin (SR), Atp5b (mitochondria), and Gapdh (cytosol) using different subcellular fractions prepared from the hearts of WT mice. A total of 10 μg of protein from each fraction was loaded. (M) Bimolecular fluorescence complementation assay using split-YFP constructs of Cisd2 and Serca2a. Cisd2 and Serca2a were fused with the C-terminal (VC) or N-terminal (VN) domain of the YFP protein and transfected into HEK-293T cells. The localization of the refolded Venus protein (green), SR (red) or mitochondria (red), and the cell nucleus (blue) are detected by confocal microscopy. Scale bar, 5 μm. (N) Serca2a (Ca2+-ATPase) activity was measured as the decrease in the absorbance at 340 nm using heart tissue from the various genotypes. Specifically, Serca2a activity was measured by the activity difference between 1 mM CaCl2 with and without the Serca-specific inhibitor TBQ. (O) Quantification of the Serca2a activity in heart tissue (n = 3). The data are presented as mean ± SD. (P) Oxidative modifications involving 3-nitrotyrosine and cysteine sulfonation of Serca2a were assessed by IP of the Serca2a protein from heart tissue samples followed by western blot analysis using antibodies against 3-nitrotyrosine, cysteine sulfonate, and Serca2a. (Q) Quantification of oxidative modifications to Serca2a by 3-nitrotyrosine and cysteine sulfonation (n = 4). The data are presented as mean ± SD and are analyzed using Student t test. *p < 0.05; **p < 0.005. Values for each data point can be found in S1 Data. 3M, 3 months old; 26M, 26 months old; Cc, crude cytosolic fractions; CCCP, carbonyl cyanide m-chlorophenyl hydrazone; Cisd2, CDGSH iron-sulfur domain-containing protein 2; Cp, pure cytosolic fractions; DCF, 2′, 7′-dichlorofluorescein; FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone; Gapdh, glyceraldehyde 3-phosphate dehydrogenase; H, homogenates; IP, immunoprecipitation; KO, knockout; Mc, crude mitochondrial fraction; OCR, oxygen consumption rate; R/A, rotenone/antimycin A; RNS, reactive nitrogen species; ROS, reactive oxygen species; Serca, sarco/endoplasmic reticulum Ca2+-ATPase; SR, sarcoplasmic reticulum; Tag, thapsigargin; TBQ, 2,5-Di-tert-butylhydroquinone; TG, transgenic; WT, wild type; YFP, yellow fluorescent protein.
Fig 5.
A high level of Cisd2 delays cardiac aging and attenuates age-related structural defects and functional decline during old age.
(A) Western blot analysis revealed that Cisd2 protein is about 2.6-fold higher in the cardiac muscle of 26M Cisd2TG mice compared with 26M WT mice. (B) Representative waterfall plots and ECG tracings recorded following anesthesia of mice. Abnormal ECG, namely ST-T elevation and irregular PR interval, were recorded from 26M WT mice. However, relatively normal sinus rhythm of the ECG was recorded from the 26M Cisd2TG mice. (C and D) Analysis of the MRI of diastolic segment 7+8 (anterior LV free wall, C) and diastolic segment 10+11 (interventricular LV wall, D). Data are presented as mean ± SD. Mouse number n = 5–6 for each group. (E and F) A high level of Cisd2 alleviates age-related damage to the intercalated discs. Lateralization of gap junctions revealed by Cx43 staining (E) and maldistribution of desmosomes revealed by desmoplakin staining (F) were easily found in the 26M WT mice. However, these defects are barely detectable in the 26M Cisd2TG mice. Method of IF imaging is the same as in Fig 3A and 3E. (G) Colocalization coefficient of gap junctions (Cx43) and intercalated discs (pan-cadherin) was analyzed using Pearson’s correlation. Method is the same as in Fig 3B. (H) A high level of Cisd2 preserves the integrity of cardiac ultrastructure in the Cisd2TG mice at old age. TEM analysis revealed ultrastructural defects in the cardiac muscle of aged WT mice at 26M. The age-related defects include disorganization of intercalated disc; expanded intercellular space (★), which might be caused by breakdown of the intercalated disc; mitochondrial degeneration (e.g., mitochondria with discontinuous or ruptured outer and inner membranes, swollen mitochondria with fewer crista); and disorganization of myofibrils with partial disruption of Z bands and fewer myofilaments. In the Cisd2TG mice, the integrity of mitochondria and intercalated disc, as well as the alignment of myofibrils, is well preserved in the cardiac muscle. (I) Serca2a activity was measured using an enzyme-coupled spectrophotometric assay (n = 3–4). Method is the same as in Fig 4N. (J and K) Western blot analysis (J) and quantification (K) of oxidative modifications to Serca2a by 3-nitrotyrosine and cysteine sulfonation using heart tissues (n = 4). (L) Quantification of ROS and RNS levels in heart by measuring DCF levels using cardiac tissues (n = 4 for each group). The data are presented as mean ± SD and are analyzed by Student t test. *p < 0.05; **p < 0.005. Values for each data point can be found in S1 Data. 3M, 3 months old; 26M, 26 months old; AV, atrioventricular; Cisd2, CDGSH iron-sulfur domain-containing protein 2; Cx43, Connexin 43; DCF, 2′, 7′-dichlorofluorescein; ECG, electrocardiography; Gapdh, glyceraldehyde 3-phosphate dehydrogenase; IF, immunofluorescence; LV, left ventricle; MRI, magnetic resonance image; RNS, reactive nitrogen species; ROS, reactive oxygen species; Serca2a, sarco/endoplasmic reticulum Ca2+-ATPase; TBQ, 2,5-Di-tert-butylhydroquinone; TEM, transmission electron microscopy; TG, transgenic; WGA, wheat germ agglutinin; WT, wild type.
Fig 6.
RNA sequencing analyses to examine the DEGs for the naturally and prematurely aged mice as well as the long-lived Cisd2TG mice.
(A) A pie chart according to their biological processes, molecular functions, and subcellular localization based on Gene Ontology annotation. (B) A Venn diagram illustrating the common and unique DEGs in the cardiac muscle of natural aging (26M WT versus 3M WT) and premature aging (3M Cisd2KO versus 3M WT). Numbers represent the amount of significantly changed genes (fold change > 1.5, p < 0.05) in each pairwise comparison and in their respective overlaps. Complete gene lists are provided in S7 Fig. (C) Classification of cardiac functional pathways by IPA illustrates the significant canonical pathways related to cardiac electric and mechanical functions, as well as cardiac structure and damage. Each column represents an animal sampled at depicted age. Hue represents the Z-score (RPKM minus mean over SD) of each gene (row) of 3M WT, 26M WT, and 3M Cisd2KO mice. The genes highlighted in red boxes (up-regulation) and purple boxes (down-regulation) are genes present in the common DEGs of natural aging and premature aging mice as shown in Fig 6B, respectively. (D) Comparison of the heatmaps of the common DEGs (up- and down-regulation), which were identified by comparing the DEGs of naturally and prematurely aged mice, for the young mice (3M WT), naturally aged mice (26M WT), and long-lived Cisd2TG mice (26M Cisd2TG). Values for each data point can be found in S1 Data. 3M, 3 months old; 26M, 26 months old; Cisd2KO, CDGSH iron-sulfur domain-containing protein 2 knockout; Cisd2TG, CDGSH iron-sulfur domain-containing protein 2 transgenic; DEG, differentially expressed gene; RPKM, reads per kilobase million; IPA, Ingenuity Pathway Analysis; Sln, sarcolipin; WT, wild type.
Fig 7.
Cisd2 plays an essential role to maintain the integrity of intercalated disc and ultrastructure of cardiomyocytes during cardiac aging.
(A) In the young heart of 3M WT mice, the three types of cell junction that make up the intercalated disc—namely, gap junctions (marked in red), desmosomes (marked in orange), and fascia adherens (marked in blue)—can be easily identified. In addition, their SR (marked in purple), mitochondria (“M”), and myofibrils appear morphologically intact. (B) In the prematurely aged heart of 3M Cisd2KO mice, their fascia adherens are shortened and broken down; gap junctions are extended and fragmented; desmosomes are degenerated. Additionally, SR, mitochondria, and myofibrils are partially degenerated. (C) In the naturally aged heart of 26M WT mice, their fascia adherens are disorganized and partially broken down; gap junctions are extended and fragmented; desmosomes are partially degenerated. Furthermore, expanded intercellular space (★) caused by degeneration of intercalated disc was found. SR, mitochondria, and myofibrils are partially degenerated and disorganized. (D) In the long-lived heart of 26M Cisd2TG mice, the integrity of intercalated disc and ultrastructure of organelles and myofibrils are much better preserved in the cardiac muscle. 3M, 3 months old; 26M, 26 months old; Cisd2, CDGSH iron-sulfur domain-containing protein 2; KO, knockout; SR, sarcoplasmic reticulum; TG, transgenic; WT, wild type.