Altered proteome turnover and remodeling by short-term caloric restriction or rapamycin rejuvenate the aging heart

Abstract

Chronic caloric restriction (CR) and rapamycin inhibit the mechanistic target of rapamycin (mTOR) signaling, thereby regulating metabolism and suppressing protein synthesis. Caloric restriction or rapamycin extends murine lifespan and ameliorates many aging-associated disorders; however, the beneficial effects of shorter treatment on cardiac aging are not as well understood. Using a recently developed deuterated-leucine labeling method, we investigated the effect of short-term (10 weeks) CR or rapamycin on the proteomics turnover and remodeling of the aging mouse heart. Functionally, we observed that short-term CR and rapamycin both reversed the pre-existing age-dependent cardiac hypertrophy and diastolic dysfunction. There was no significant change in the cardiac global proteome (823 proteins) turnover with age, with a median half-life 9.1 days in the 5-month-old hearts and 8.8 days in the 27-month-old hearts. However, proteome half-lives of old hearts significantly increased after short-term CR (30%) or rapamycin (12%). This was accompanied by attenuation of age-dependent protein oxidative damage and ubiquitination. Quantitative proteomics and pathway analysis revealed an age-dependent decreased abundance of proteins involved in mitochondrial function, electron transport chain, citric acid cycle, and fatty acid metabolism as well as increased abundance of proteins involved in glycolysis and oxidative stress response. This age-dependent cardiac proteome remodeling was significantly reversed by short-term CR or rapamycin, demonstrating a concordance with the beneficial effect on cardiac physiology. The metabolic shift induced by rapamycin was confirmed by metabolomic analysis.

Keywords: caloric restriction; cardiac aging; dynamics; proteomics; rapamycin.

Figure 1

Summary of experimental design. After 3 weeks on synthetic chow diet for acclimatization, young (4 month old) and old (26 month old) female mice had baseline echocardiography and were placed on a synthetic diet ad libitum (control group), 40% caloric restriction (progressively over 3 weeks), or ad libitum plus microencapsulated rapamycin 2.24 mg kg⁻¹day⁻¹. After 10 weeks, echocardiography was repeated and the mice were switched to the same synthetic diet but with ²H₃-leucine fully substituted for normal leucine. Tissue from 3 mice per treatment were harvested 3, 7, 12, and 17 days thereafter. Protein extraction was followed by LC-MS/MS. Topograph was applied to calculate the fraction of newly synthesized peptides, as well as the relative abundance of all isotopomers. The percentage newly synthesized peptides for each protein were plotted for the triplicate mice at each time point to derive the rate constant and half-lives based on first order kinetics. For abundance analysis, Topograph aligned the chromatograms to normalize the recovery times of the corresponding ions in each sample and obtain the areas under the curve for every peptide identified in any one sample. (see method S1 for further detail).

Figure 2

Echocardiography. (A) Left ventricular mass index (LVMI) is significantly higher in old hearts at baseline (25 months old), when compared to young hearts (3 months old), indicating age-dependent left ventricular hypertrophy. After 10 weeks of CR or RP in old mice, LVMI is significantly lower than old control mice (P < 0.001 and P = 0.004, respectively). It is also significantly reversed when compared with the baseline echocardiography of each mouse (P < 0.01 for both CR and RP), indicating reversal of age-dependent ventricular hypertrophy. (B) Fractional shortening does not significantly change with age or treatment group. (C). Myocardial performance index (MPI) significantly worsened (increased) in old mice at baseline. CR or RP significantly improved the MPI in old hearts when compared with control mice. (P = 0.02 and 0.05, respectively). (D) Diastolic function measured by tissue Doppler imaging Ea/Aa significantly declined in old hearts. While old CL mice have progressive decline of Ea/Aa after 10 weeks, treatment with CR or RP significantly increased Ea/Aa. CL: ad libitum control diet, and CR: caloric restriction. *P < 0.05 vs. YCL; ^#P < 0.05 versus pretreatment. n = 5–8.

Figure 3

Global proteomic half-lives. (A) Histograms of half-lives (days). YCL versus OCL: n.s., OCL versus OCR P < 0.001; OCL versus ORP, P = 0.038; OCR versus OR, P = 0.08. (B) Histograms of half-life ratios for two-group comparisons. P = 0.09 for YCL/OCL >1, P < 0.001 for both OCR/OCL and ORP/OCL >1.

Figure 4

Protein half-lives (days) stratified by (A) cellular compartment, (B) most significant canonical pathways in ingenuity pathway analysis. *P < 0.01 for OCR versus OCL or ORP versus OCL, ^#P < 0.01 for OCR versus ORP.

Figure 5

Heat maps of (A) half-life differences, (B) abundance differences for YCL/OCL, OCR/OCL, and ORP/OCL comparisons, ordered by the rank of significance of differ-ences in ingenuity pathway analysis. In part A, red indicates longer and blue indicates shorter half-lives. In part B, red indicates higher and blue indicates lower abundance. The component protein IDs are listed in Tabls S4 for abundance difference and Table S5 for turnover differences. FAO: fatty acid oxidation, BCAA: branched-chain amino acid metabolism; 2-HB: 2-oxobutanoate; MG: methylglyoxal; and Ub: Ubiquitination.

Figure 6

Metabolic profiling and biochemical assay. (A) Relative abundance of the substrates in the glycolytic pathway and TCA cycle in ORP compared to OCL by targeted metabolic profiling. When compared with OCL heart, ORP hearts have significantly lower glucose-6-phosphate and fructose-6-phosphate (both are glycolytic metabolites), and significantly higher a-ketoglutarate, fumarate, malate, and citrate (all are TCA cycle metabolites). *P < 0.05 compared with OCL. See Table S6 for numerical data. (B) A schematic diagram summarizing the changes in metabolism by rapamycin in old heart. (C) Western blots of autophagic markers show no significant change of LC3 II/I, p62, or beclin-1 in cardiac aging. However, OCR has significantly lower p62 than that in OCL. ^#P < 0.05 compared with OCL. (D) Both CR and RP significantly reduce the age-dependent increase in protein carbonyls (nmol mL⁻¹). ^#P < 0.05 compared with OCL. (E). Both CR and RP significantly reduce the age-dependent increase in protein ubiquitination.*P < 0.05 compared with YCL and ^#P < 0.05 compared with OCL. n = 3–8. G6P: glucose 6-phosphate; G1P: glucose 1-phosphate; F6P: fructose 6-phosphate; F1P: fructose 1-phosphate; F16BP: fructose 1,6-bisphosphate; F26BP: fructose 2,6-biphosphate; G3P: glyceraldehyde 3-phosphate; DHAP:dihydroxyacetone phosphate; 2(3)-PGA: 2- or 3-phosphoglycerate; and PEP: phosphoenolpyruvate. Isomers of same molecular weight, that is, G6P versus G1P, F6P versus F1P, and F16BP versus F26BP, were not distinguishable by the LC-MS/MS-based metabolic profiling method.