Spermine and spermidine reversed age-related cardiac deterioration in rats

Aging is the most important risk factor for cardiovascular disease (CVD). Slowing or reversing the physiological impact of heart aging may reduce morbidity and mortality associated with age-related CVD. The polyamines, spermine (SP) and spermidine (SPD) are essential for cell growth, differentiation and apoptosis, and levels of both decline with age. To explore the effects of these polyamines on heart aging, we administered SP or SPD intraperitoneally to 22- to 24-month-old rats for 6 weeks. Both treatments reversed and inhibited age-related myocardial morphology alterations, myocardial fibrosis, and cell apoptosis. Using combined proteomics and metabolomics analyses, we identified proteins and metabolites up- or downregulated by SP and SPD in aging rat hearts. SP upregulated 51 proteins and 28 metabolites while downregulating 80 proteins and 29 metabolites. SPD upregulated 44 proteins and 24 metabolites and downregulated 84 proteins and 176 metabolites. These molecules were mainly associated with immune responses, blood coagulation, lipid metabolism, and glutathione metabolism pathways. Our study provides novel molecular information on the cardioprotective effects of polyamines in the aging heart, and supports the notion that SP and SPD are potential clinical therapeutics targeting heart disease.


INTRODUCTION
Age-related diseases are major contributors to high morbidity and mortality rates in aging populations. Cardiovascular diseases (CVDs), such as coronary atherosclerosis, heart failure, and dilated cardiomyopathy, are age-associated diseases and leading causes of death worldwide. Identifying efficacious age-related CVD treatments is essential to extending elderly patient lifespans.

Research Paper: Gerotarget (Focus on Aging)
Oncotarget 64794 www.impactjournals.com/oncotarget reticulum stress during acute myocardial infarction in male rat hearts by inhibiting reactive oxygen species (ROS) and downregulating PERK-eIF2α signaling [13][14][15]. Our previous study suggested that exogenous polyamine protects against reperfusion injury by inhibiting mitochondrial permeability transition pore (mPTP) opening in isolated rat hearts [14]. Additionally, exercise training can increase the polyamine pool in aged rat hearts, restoring ischemic preconditioning protection [15]. Recent work showed that SPD reduces lipid accumulation and necrotic core formation in atherosclerotic plaques by inducing autophagy in apolipoprotein E (ApoE) -/mice [16]. SPD feeding enhanced cardiac mitophagy and mitochondrial respiration, and improved diastolic function to delay cardiac aging in C57BL/6 mice. Oral SPD supplementation inversely correlated with CVD in humans [17]. However, to the best of our knowledge, no definitive information is available regarding the effect of exogenous polyamines on rat heart aging.
Proteins are the primary effector molecules of all living systems, and any adaptive responses to exogenous stresses will be reflected in altered protein activities and concentrations [18]. Proteomics allows for direct study of protein production and function in a cellular context [19]. Isobaric tags for relative and absolute quantitation (iTRAQ) in combination with liquid chromatography tandem MS (LC-MS/MS) analysis is a new high throughput, high repeatability, high sensitivity, and high accuracy proteomics tool for studying biological processes [20,21]. Proteomics analyses of aged rat hearts revealed changes in cell signaling, immune response, and structural proteins, and in proteins mediating oxidative stress responses [22]. Such changes can directly indicate aberrant physiological status [23]. Thus, proteomics and metabolomics technologies, which provide global profile information, are powerful tools for investigating tissue responses to drug treatment.
This study employed iTRAQ proteomic and GC/ MS-based metabolomic approaches to investigate alterations in aging rat hearts treated with SP and SPD. In analyzing the pathways involved in SP and SPD activity, we provide novel insight into the cardioprotective mechanisms of these two PAs in aging rats.

Myocardial histology and morphology changes in SP-or SPD-treated aged rats
Age-related phenotype changes were observed in rat hearts, including myocardial morphology changes, myocardial fibrosis, and cell apoptosis. Compared to young control rats (group Y), aged rat hearts showed loosely arranged cardiac muscle fibers and increased interstitial Oncotarget 64795 www.impactjournals.com/oncotarget distances as evaluated by H&E staining. However, hearts in aged rats treated with SP or SPD maintained myocardial histological structures as compared to untreated controls (group O) ( Figure 1A). Increased myocardial collagen deposition and a disordered collagen fiber network around cardiomyocytes were observed in old rat hearts ( Figure  1B), and collagen volume fraction (CVF) was higher than in young rats (P < 0.05). However, CVF was decreased to a greater extent in SP-(P < 0.05) or SPD-treated (P < 0.05) old rat hearts compared to untreated old rat hearts ( Figure 1C). The ratio of TUNEL-positive cardiomyocytes to total cardiomyocytes was higher in old rat hearts than in young rat hearts (P < 0.05). In contrast, the positive cell ratio in SP-(P < 0.05) or SPD-treated (P < 0.05) rats was lower than in untreated old rats ( Figure 1D-1E). These findings suggested that SP and SPD effectively reduced age-related cardiovascular changes in rats.  GOTERM_BP_FAT GO:0007596 blood coagulation  11  4.69E-11   FGG, C9, FGA, FGB, C3,  F13A1, F2, APOH, SERPIND1,  ENTPD2, PLG   GOTERM_BP_FAT GO:0043065  positive  regulation  of  apoptosis  9  The number of proteins derived from proteomics analysis. c Probability of identified differential protein list associates to the GO term.

Polyamine-regulated proteins in aged rat hearts
To understand the effects of SP and SPD on protein expression in aged rat hearts, iTRAQ in combination with liquid chromatography-electrospray ionization tandem MS (LC-ESI-MS/MS) was applied to investigate differentially expressed proteins in treated and untreated rats. A total of 2586 protein groups were identified and quantified from 14401 unique peptides. Among these, 2565 protein groups were shared among the three complete iTRAQ datasets (SP, SPD, and O), which was proportionally consistent with reports on technical variance between iTRAQ datasets [22]. Of the 2565 proteins, 184 were differentially expressed in the SP and SPD groups compared with the old group (Supplementary Table 1) and were plotted against the old group using the circos plot ( Figure 2). Of these, 131 proteins were differentially expressed in SP-treated rat hearts compared to untreated old rat hearts. Eighty proteins were downregulated, and 51 were upregulated ( Figure 2B). Similarly, 128 of the 184 proteins were differentially expressed in SPD-treated rat hearts; 84 were downregulated and 44 were upregulated compared to untreated old rat hearts ( Figure 2C). Seventyfive of 184 proteins were affected by both SP and SPD treatment ( Figure 2D).
GO analysis was performed for proteins differentially expressed following polyamine treatment. Of the 131 differentially expressed proteins in SPtreated rat hearts, 112 clustered into categories with P < 0.05, including blood coagulation, positive regulation of apoptosis, triglyceride metabolic process, glutathione metabolic process, positive regulation of cholesterol esterification, and regulation of phosphorylation ( Figure  3A, Table 1). In SPD-treated rats, 106 differentially expressed proteins clustered into the coagulation, positive regulation of cholesterol esterification, regulation of blood vessel size, lipid oxidation, glucose metabolic process, and glycolysis categories ( Figure 3B, Table 2).
The MetaCore tool was used to study canonical pathways differentially regulated by SP and SPD in aged rat hearts. Thirty-three and 40 pathways were enriched in SP-and SPD-treated rat hearts, respectively (P < 0.05) (Supplementary Table 2). Crucial SP-altered  The number of proteins derived from proteomics analysis. c Probability of identified differential protein list associates to the GO term.
We then examined protein-protein interactions (PPI) for 131 (STRING annotation 126) proteins differentially expressed following SP treatment and 128 (STRING annotation 122) proteins differentially expressed following SPD treatment compared to the old control group. Proteins belonged to major networks affecting multiple top GO pathways, such as F2, Plg, and Fgg protein networks in SP-treated rats ( Figure 5A), and serpina1, F2, and Plg protein networks in SPD-treated rats ( Figure 5B).

Differential metabolic profiles in polyaminetreated aged rat hearts
We identified characterized metabolites in senescent heart tissues from SP-or SPD-treated rats using GC-MS-based metabolomics. Six hundred sixty-one peaks were acquired using GC-MS, and 635 metabolites were identified. Based on the full metabolic dataset, PCA analysis identified four principal components. PLS-DA and OPLS-DA score plots ( Figure 6A and 6B) separated

Cardiac metabolite screening and identification in polyamine-treated aged rats
In total, 57 (KEGG annotation 38) metabolites were differently regulated in SP-treated rat hearts compared to untreated old rats; 28 were upregulated and 29 were downregulated (Supplementary Table 3). Two hundred   Table 4).

Integrated analysis of polyamine-modulated cardiac proteins and metabolites in aged rats
To investigate whether differentially expressed/ produced proteins and metabolites interacted, a KEGGbased network was generated to elucidate SP-and SPDmodulated signaling pathways. Proteins/metabolites that shared interactive relations with differentially expressed proteins and metabolites identified in this study were added to the set. The hypergeometric test was performed for pathway enrichment analysis, and pathways with P < 0.05 were filtered (Supplementary Table 5 -6). Six pathways ( Figure 8) were altered in both SP-and SPDtreated rats, including coagulation and complement cascades (ko04610), arginine and proline metabolism (ko00330), glycolysis/gluconeogenesis (ko00010), galactose metabolism (ko00052), and glutathione metabolism (ko00480). Pathways altered only by SP involved apoptosis (ko04210) and cancer (ko05200 and ko05212), while SPD treatment affected amino acid metabolism (ko00250 and ko00260) and glycerolipid metabolism (ko00561). F2, C3, and C4 were identified as key coagulation and complement cascade regulators. Coagulation, immune response, lipid metabolism, and glutathione metabolism are associated with cardioprotection and anti-aging in the heart (Figure 9).

DISCUSSION
Age-induced cardiomyopathy in the mammalian heart is characterized by myocardial hypertrophy, fibrosis, and a predisposition towards cardiomyocyte apoptosis. Interstitial fibrosis of the atria, sinoatrial node (SAN), and ventricles adversely affects myocardial function in older adults [24]. Increased cardiomyocyte apoptosis leads to myocardial damage and eventually heart failure in aging C57BL/6 male mice [25]. Thus, preventing fibrosis and apoptosis are important strategies to prevent heart disease [26]. This study showed that six weeks of SP or SPD treatment reversed previously existing age-associated myocardial morphology changes and myocardial fibrosis, and inhibited cell apoptosis in aging rat hearts. SP and SPD both modulated rat heart proteins and metabolites associated with immune response, blood coagulation, lipid metabolism, and glucose metabolism. However, SP was more associated with apoptosis, chemokine signaling, and cancer-related pathways, while SPD was associated with amino acid and glycerolipid metabolism.
Blood coagulation proteins regulate early atherosclerosis and its progression [27][28][29][30]. Otto, et al. reported that aortic sclerosis increases risk of myocardial infarction and cardiovascular-associated death in the elderly [31]. Our proteomic analysis showed that SP and SPD regulate coagulation-related protein expression. Protein-protein interaction analyses revealed that F2 and serpina1 interact within smaller networking groups. F2 is a prothrombin precursor involved in blood homeostasis, inflammation, and wound healing. Aging and cardiovascular risk factors increase plasma levels of hemostatic molecular markers like F2 in the elderly [32]. We found that F2 (SP/O: 0.710 and SPD/O: 0.704) was downregulated in SP-and SPD-treated aging rat hearts. Serpina1 encodes alpha-1 antitrypsin (AAT), an acutephase inflammation marker synthesized predominantly Oncotarget 64801 www.impactjournals.com/oncotarget in hepatocytes that inhibits proteases, such as elastase, trypsin, thrombin, and bacterial proteases. Serpina1 upregulation is associated with multiple CVDs, including heart failure [33][34][35], and levels were lower in nonischemic right ventricles (RV) than in ischemic RVs. Our proteomic analysis showed that serpina1 is downregulated (SPD/O: 0.650) in SPD-treated aging rat hearts. We speculate that SP and/or SPD may reduce morbidity and mortality from myocardial infarction and CVD by inhibiting aortic sclerosis.
Increased immune activity can lead to chronic inflammation and degenerative diseases. The protein and mRNA expression levels of complement C3 were found to be associated with the occurrence and development of coronary heart disease [36], higher C3 was significantly associated with presence and greater extent of arterial calcification, and C3 could be a potential non-invasive biomarker of early diagnosis of atherosclerosis [37]. Activation of the immune response by elevating hydroperoxides levels, further activating NF-kB signaling, may be a key aging feature in the mouse heart [38]. In addition, Lian et al reported that astroglial NF-kB/C3 activation led to impaired synaptic density and dendritic morphology in brain tissue of Alzheimer's disease (AD) Oncotarget 64802 www.impactjournals.com/oncotarget patients [39], and the expressions of C3, C4 and C5 mRNA increased with age in C57BL/6 mice brain [37][38][39]. Here, our proteomic analysis revealed decreased C3 (SP/O: 0.811 and SPD/O: 0.555) and C4 (SP/O: 0.781 and SPD/O: 0.653) expressions in both SP-and SPD-treated rats, which may contribute to the anti-aging effects of these polyamines.
Fatty acids supply 50-75% of the energy to the normal heart, while glucose oxidation and glycolysis provide adenosine triphosphate (ATP) [40]. Under normal conditions, glucose metabolism cooperates with fatty acid metabolism to produce ATP. However, physiological (aging) and pathophysiological (hypertension and diabetes) conditions can alter glucose and fatty acid metabolism, increasing risk of hypertension, insulin resistance, abnormal cholesterol and possible heart disease. During cardiac aging, increased generation of reactive oxygen species (ROS) can damage mitochondria and lead to perturb fuel utilization. The fatty acid β-oxidation impaired has been observed in aged mice heart [41,42]; the amount of total fatty acid concentrations decreased and the arachidonic acid concentrations increased in aged rat hearts [43]. Reverse in the age-related reduction in fatty acid oxidation improves mitochondrial metabolism and energetic profile in old C57BL/6 mice hearts [44]; decreasing arachidonic acid metabolism has neuroprotective role in neurocognitive disorder patients and rat of neuroinflammation [45]. On the other hand, aging individuals are more likely to suffer from lower glucose levels than younger individuals, as the age Oncotarget 64803 www.impactjournals.com/oncotarget increases, glucose tolerance becomes worse and glucose transporters decrease in rat [46]. Galactose is readily converted to glucose, galactose at 15% of daily intake improves hepatic insulin sensitivity in rats; galactose can also improve glucose metabolism by enhancing mitochondrial oxidative phosphorylation in cultured human myotubes [47]. While stimulating mitochondrial glucose oxidation has been suggested as a viable therapeutic strategy to compensate for the energetically 'starved' heart [48]. Our study showed that SP and SPD reversed lipid and glucose metabolism changes induced by aging in the heart, including glycerol (SPD/O: 0.60), arachidonic acid (SP/O: 0.360), and galactose (SP/O: 3.920 and SPD/O: 2.810) levels. Therefore, we speculate that SP and SPD can inhibit metabolic remodeling and decrease age-related CVD risk.
Glutathione is a major anti-oxidant in cells [49], and reductions may result from increased ROS concentrations in liver tissues [50]. Our metabolomic analysis showed that SPD treatment suppressed glutathione metabolism-related metabolites, including glycine, glutathione, L-glutamic acid, pyroglutamic acid, L-cysteine, ornithine, putrescine, SPD, and cysteinylglycine. Decreased glutathione may in turn lead to increased ROS concentrations, possibly indicating increased mitochondrial β-oxidation. This may lead to more robust fat combustion, similar to a "fasted" metabolic state [35], potentially depleting lipids and inhibiting dyslipidemia in aged rat hearts.
Proteomics methods analyze the entire expressed protein profile of a biological sample [19], while metabolomics provides a metabolite fingerprint [23]. Together, proteomic and metabolomic analyses can provide comprehensive mechanistic information about diverse pathologies [51,52]. We integrated proteomic and metabolomic data streams to elucidate mechanisms underlying the anti-aging effects of SP and SPD in the heart. However, our study had limitations. Natural deaths in the 22-24-month-old rats reduced group sample sizes from more than six to only three. Additionally, proteomics technologies have expanded; strong cation exchangereverse phase (SCX-RP) 2D LC-MS-MS applied to a soluble protein lysate from mouse embryonic fibroblast cells identified more than 5000 proteins [53]. This technique could identify more proteins more accurately in future studies.
In summary, using an integrated proteomics and metabolomics approach, we identified differentially expressed proteins and metabolites in aging rat hearts Oncotarget 64804 www.impactjournals.com/oncotarget following polyamine treatment. We systematically analyzed the functions of these proteins and metabolites, and found that exogenous SP and SPD exerted anti-aging effects in rat hearts likely by regulating immune response, blood coagulation, lipid metabolism, and glutathione metabolism pathways. Future studies will assess whether polyamines work against heart aging through F2 and glutathione. Our study provides novel molecular information on the anti-aging effects of polyamines in the rat heart, and supports SP and SPD as potential clinical therapeutics targeting heart disease.

Animal model and treatment protocol
Three-and 22-24-month-old male Wistar rats were obtained from the Animal Center of Harbin Medical University. Three-month-old male rats were considered the young control group (Y) (n = 3), and 22-24-month-old male rats were randomly assigned to the old control (O)  Oncotarget 64805 www.impactjournals.com/oncotarget (n = 3), SP (n = 3), and SPD (n = 3) groups. SP and SPD group rats were treated with SP (2.5 mg/kg/day) or SPD (10 mg/kg/day) intraperitoneally for 6 weeks. SP and SPD were obtained from Sigma Chemical Co. (St. Louis, MO, USA). Rats were treated according to the Guide for Care and Use of Laboratory Animals published by the China National Institutes of Health. All rats were housed under environmental conditions of constant temperature and humidity, with free access to standard rodent chow and water.

Histological and morphological analyses
Left ventricular tissues were excised and fixed in 4% paraformaldehyde at room temperature for >1 d, and then prepared for H&E staining, Masson staining, and cell apoptosis assay. Tissues were embedded in paraffin, sectioned at 5-μm thickness, and stained with Mayer's hematoxylin followed by 1% eosin alcohol solution, Masson's trichrome, and PicroSirius red. Stained sections were visualized at 40× magnification, and volume fraction of collagen (VFC) fibril structures were quantified from three sections per heart. Results were averaged from five high power random fields from each section. A terminal deoxynucleotidyl transferase mediated dUTP nick end-labeling (TUNEL) assay was applied to detect apoptotic myocytes using the Cell Death Detection Kit (Roche, Basel, Switzerland) following the manufacturer's instructions. Deparaffinized and rehydrated heart sections were incubated with TUNEL reaction mixture, including terminal deoxynucleotidyl transferase (TdT) and fluorescein-dUTP after permeabilization. Sections were developed using diaminobenzidine (DAB; Sigma Chemical Co.; St. Louis, MO, USA), and nuclei were counterstained using hematoxylin. Percentage of apoptotic cells was calculated from three sections per block. One hundred cells were counted in each of four section fields, which were randomly chosen at 200× magnification.

iTRAQ proteomics analysis
Details of protein sample preparation, tryptic digestion, and proteomic profiling acquisition are provided in Supporting Information. GO analysis was performed with iTRAQ results using the open access DAVID bioinformatics platform (Database for Annotation, Visualization and Integrated Discovery, http://www.david. abcc.ncifcrf.gov website). DAVID was used to identify enriched biological processes related to differentially expressed proteins identified via iTRAQ analysis. Pathway analysis was performed with differentially expressed proteins using MetaCore software. The following criteria were used to determine differential protein expression in SP-or SPD-treated rat hearts compared with controls (O group): P < 0.05, fold change >1.2 (or < 0.833), and at least two unique peptides.

Protein-protein interaction analysis
Co-expressed genes may be (1) controlled by the same transcriptional regulatory program, (2) functionally related, or (3) members of the same pathway or protein complex. Protein-protein network analysis was performed for differentially expressed proteins identified via iTRAQ analysis using STRING 10.0 (http://www.string-db.org/), an online database of known and predicted protein-protein interactions, based on two types of evidence: experimental (protein-protein interaction databases) and text-mining (scientific literature abstracts). Differently expressed proteins were mapped to the STRING database and known and predicted associations were scored and integrated, with a combined-score threshold of >0.4. Differentially expressed proteins were visualized after KMEANS clustering. The interaction network was constructed by integrating these relationships.

Metabolomic data analysis
Sample preparation, metabolic profiling acquisition, data processing, and quality control details are provided in Supporting Information. Chroma TOF 4.3X software (LECO Corporation, St Joseph, MI, USA) and the LECO-Fiehn Rtx5 database were used to process GC/MS data for extraction of raw peaks, data baseline filtering and calibration, peak alignment, deconvolution analysis, peak identification, and peak area integration.
Missing raw data values were filled in using half the minimum value, interquartile range denoising was performed to detect peaks and filter metabolites, and data were normalized to an internal standard. Peak numbers, sample names, and normalized peak areas were analyzed using the SIMCA-P 11.5 software package (Umetrics, Umea, Sweden) for principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), and orthogonal projections to latent structures-discriminant analysis (OPLS-DA). PCA showed the distribution of the original data. To better separate groups and understand variables responsible for classifications, a supervised PLS-DA was applied. Based on an OPLS-DA, a loading plot was constructed showing the contributions of variables to differences between two groups.
To refine this analysis, the first principal component of variable importance projection (VIP) was obtained. VIP values >1.0 were selected as differentially produced metabolites. The remaining variables were assessed by Student's t-test, P < 0.05. Variables were discarded between two comparison groups. Additionally, commercial databases, including KEGG (http://www.genome.jp/kegg/) and NIST (http://www.nist.gov/index.html), were used to search for metabolites.
Pathway analysis of potential metabolites (including pathway enrichment analysis, pathway topology analysis, and visualization) was performed in MetaboAnalyst 3.0