FNDC4 Prevents Aging-Related Cardiac Dysfunction

Visual Abstract

Highlights

• •Aging mice exhibit a sizable decline in cardiac and plasma FNDC4 levels.
• •Cardiac-specific FNDC4 overexpression alleviates aging-related cardiac remodeling and dysfunction.
• •Cardiac-specific FNDC4 knockdown facilitates aging-related cardiac remodeling and dysfunction.
• •FNDC4 activates PPARα signaling pathway to improve mitochondrial dysfunction and lipotoxicity in aging hearts.
• •AMPKα is required for FNDC4-mediated activation of PPARα.

Summary

Mitochondria play critical roles in maintaining oxidative metabolism and cardiac homeostasis; however, their function is compromised in aging hearts. Fibronectin type III domain-containing 4 (FNDC4) is involved in regulating mitochondrial biogenesis, energy expenditure, and metabolic balance. The present study found that aging mice exhibited a sizable decline in cardiac and plasma FNDC4 levels, and that lower FNDC4 expression also correlated with a poor cardiac function. Cardiac-specific FNDC4 overexpression alleviated, while cardiac-specific FNDC4 knockdown facilitated aging-related cardiac remodeling and dysfunction. The unbiased transcriptome analysis and untargeted metabolomics revealed that FNDC4 activated AMP-activated protein kinase α/peroxisome proliferator-activated receptor α signaling pathway to improve mitochondrial dysfunction and lipotoxicity in aging hearts.

The heart is an organ with high energy demand for continuous contraction, and its function depends heavily on mitochondrial oxidative metabolism.^1,2 However, the biogenesis, ultrastructure, and function of mitochondria are dramatically compromised in aging hearts, eventually leading to metabolic disorder, energy deficit, and aging-related cardiac dysfunction.3, 4, 5 Accordingly, Kates et al⁶ found that human aging hearts exhibited a sizable decline in the utilization and oxidation of fatty acids, with comparable glucose utilization. The decreased consumption of fatty acids results in excessive deposition of lipids within the cardiomyocytes, which is linked to chronic low-grade inflammation of the heart and subsequently results in inflammaging of cardiac cells.^7,8 Meanwhile, these cardiac lipids also enter alternative nonoxidative pathways to produce toxic reactive lipid species that facilitates oxidative stress, telomere shortening, DNA damage, and premature senescence.⁹ In addition, the release of mitochondrial DNA (mtDNA) from injured mitochondria and the defective electron transport chain (ETC) subunits also contribute to cardiac inflammation, oxidative damage and impairment.^10,11 Therefore, restoring mitochondrial function and lipid metabolism is vital for preventing aging-related cardiac dysfunction.

Peroxisome proliferator-activated receptors (PPARs) belong to the nuclear hormone receptors superfamily and consist of 3 members (PPARα, PPARδ/β, and PPARγ), which play critical roles in mitochondrial bioenergetics, energy metabolism, and cardiac homeostasis.12, 13, 14 PPARα is mainly expressed in oxidative tissues like the heart and skeletal muscle, and coordinates with PPARγ coactivator 1 (PGC1) to regulate the transcription of genes involved in mitochondrial biogenesis, fatty acid transport, and oxidation. Emerging evidence has revealed that inhibiting PPARα/PGC1α remarkably disrupted mitochondrial structure and substrate oxidation, and led to lipid accumulation as well as cardiac dysfunction.15, 16, 17 In contrast, activating PPARα/PGC1α improved mitochondrial respiration, fatty acid oxidation, and cardiac function in aging hearts.^18,19 AMP-activated protein kinase α (AMPKα) acts as a nodal energy sensor and is implicated in controlling metabolic homeostasis. Zhao et al²⁰ previously demonstrated that AMPKα activation elevated PPARα expression and activity, thereby restoring fatty acid metabolism and cardiac function. Moreover, AMPKα is also essential for the preservation of cardiac structure and function under different stresses. Our previous findings have determined that AMPKα activation dramatically ameliorated pressure overload-, diabetes-, sepsis- and aging-related cardiac dysfunction through inhibiting inflammation, oxidative damage, and cardiac remodeling.21, 22, 23, 24 In contrast, AMPKα deficiency notably disrupted mitochondrial ultrastructure, increased reactive oxygen species (ROS) accumulation, and eventually exacerbated cardiac dysfunction in aging mice.²⁵ However, specific strategies for manipulating these molecular targets in aging hearts remain undefined.

Fibronectin type III domain-containing 4 (FNDC4) is a type I transmembrane glycoprotein and can be cleaved to a soluble bioactive fragment. Bosma et al²⁶ identified soluble FNDC4 as an anti-inflammatory factor, and found that administration of recombinant FNDC4 (rFNDC4) protein effectively reduced the disease severity of colitis. Meanwhile, FNDC4 is also involved in regulating energy expenditure and metabolic balance. Georgiadi et al²⁷ revealed that liver-derived FNDC4 promoted insulin signaling and insulin-mediated glucose uptake in white adipocytes, thereby improving glucose tolerance and insulin resistance in prediabetic mice. FNDC4 also decreased lipogenesis and increased fat browning in human visceral adipocytes by facilitating mitochondrial biogenesis.²⁸ Moreover, FNDC4 displays a high homology with FNDC5 that plays critical roles in mitochondrial homeostasis and cardiac function under different stresses.29, 30, 31 Interestingly, we previously demonstrated that FNDC5 overexpression or supplementation dramatically alleviated doxorubicin- and aging-related cardiac dysfunction.^24,32 Here, we aimed to identify the cardiomyocyte-autonomous role of FNDC4 in mitochondrial function and metabolic homeostasis during cardiac aging.

Methods

Reagents

GW6471 (#S2798, a specific PPARα inhibitor), GlaxoSmithKline3787 (#S8025, a specific PPARδ inhibitor), and GW9662 (#S2915, a specific PPARγ inhibitor) were obtained from Selleck Chemicals. Palmitic acid (#P0500), oleic acid (#O1008), 2′,5′-dideoxyadenosine (2′5′-ddAdo) (#D7408), and H-89 dihydrochloride hydrate (H89) (#B1427) were obtained from Sigma-Aldrich. Senescence-associated β-galactosidase (SA-β gal) staining kit (#9860) was purchased from Cell Signaling Technology. Oil red O (#G1015) was obtained from Servicebio. Nile red (#HY-D0718) was obtained from MedChemExpress. Dihydroethidium (DHE) (#KGAF019) was obtained from Jiangsu KeyGEN BioTECH Corp., Ltd. Mitochondrial membrane potential assay kit with JC-1 (#C2006), 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) (#S0033), and ATP Assay Kit (#S0026) were obtained from Beyotime. Complex I Enzyme Activity Microplate Assay Kit (#ab109721), Complex II Enzyme Activity Microplate Assay Kit (#ab109908), Mitochondrial Complex III Activity Assay Kit (#ab287844), Complex IV Rodent Enzyme Activity Microplate Assay Kit (#ab109911), 3-Nitrotyrosine (3-NT) ELISA Assay Kit (#ab116691) and cAMP Assay Kit (#ab65355) were purchased from Abcam. Triglyceride (TG) Assay Kit (#A110), Malondialdehyde (MDA) Assay Kit (#A003) and 4-Hydroxynonenal (4-HNE) ELISA Kit (#H268) were purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). NE-PER Nuclear and Cytoplasmic Extraction Reagent (#78833), Interleukin-6 (IL-6) Mouse ELISA Kit (#KMC0061), Tumor Necrosis Factor-α (TNF-α) Mouse ELISA Kit (#BMS607-3), Alexa Fluor 488 conjugated Wheat Germ Agglutinin (WGA) (#W11261), and SlowFade Gold Antifade Reagent with DAPI (#S36939) were purchased from Invitrogen. Seahorse XF Cell Mito Stress Test Kit (#103015-100) was purchased from Agilent. Mouse FNDC4 EIA Kit (#EK-067-90) was purchased from Phoenix Pharmaceuticals, Inc. Human FNDC4 ELISA Kit (#MBS9332722) and Mouse n-terminal pro-brain natriuretic peptide (NT-proBNP) ELISA Kit (#MBS2124067) were purchased from MyBioSource, Inc. Plasmids encoding full-length human FNDC4 cDNA (#W2481) were obtained from GeneCopoeia, Inc (Rockville, MD, USA), and then packaged to the cardiotropic adeno-associated virus serotype 9 (AAV9) vectors under a cTnT promoter to produce AAV9-hFNDC4, while AAV9-CTRL was used as the negative control. Two independent mouse FNDC4 shRNA plasmids (#MSH102118 and #MSH102119) were purchased from GeneCopoeia, Inc, and sequences were cloned under a cTnT promoter to produce AAV9-shFndc4 or AAV9-shFndc4#, while AAV9-shCtrl was used as the negative control.

Animals and experiments

Male C57BL/6 mice were fed in a SPF condition with controlled temperature (20-25 °C) and humidity (45-55%) under 12/12-hour light/dark cycles, and allowed free access to food and water. After 1 week acclimation, 6-month-old young and 18-month-old aging mice received a single intravenous injection of AAV9-hFNDC4 or AAV9-shFndc4 at a dose of 1 × 10¹¹ viral genome to specifically overexpress or knock down FNDC4 in the myocardium, and were sacrificed after 2 months.³² Before AAV9 injection, the mice were firstly intraperitoneally injected with 3% pentobarbital sodium (Sigma-Aldrich, P3761) and kept on a heating pad to maintain the body temperature as close as 37 °C. To verify the involvement of PPARα and AMPKα, global knockout mice were used. The source and validity of AMPKα2 global knockout (AMPKα2^−/−) mice have been described in our previous studies.^13,24 PPARα global knockout (PPARα^−/−) mice (#008154) in the C57BL/6 background were provided by Jackson Laboratories, and genotyping was performed using the following primers: primer 17026: CACACCAAGCAGCAGACACT; primer oIMR1100: GCTATCAGGACATAGCGTTGG; primer oIMR8076: CCCATTTCGGTAGCAGGTAGTCTT. To explore the necessity of PGC1α, cardiac-specific PGC1α knockout (cKO) mice were established by crossing Ppargc1a floxed (Flox) mice with α-Mhc-MerCreMer transgenic mice (#005657, Jackson Laboratories). Ppargc1a floxed mice (#009666) in the C57BL/6 background were provided by Jackson Laboratories, and genotyping was performed using the following primers: primer 8041: TCCAGTAGGCAGAGATTTATGAC; primer 8491: TGTCTGGTTTGACAATCTGCTAGGTC. To induce the specific ablation of PGC1α in cardiomyocytes, Ppargc1a cKO mice were intraperitoneally injected with tamoxifen (25 mg/kg/day, dissolved in corn oil) for consecutive 5 days as we previously described.^33,34 For PPARα inhibition, aging mice with or without FNDC4 overexpression were intraperitoneally injected with 3 mg/kg GW6471 every other day for 6 weeks beginning at 2 weeks post-AAV9 injections.³⁵ For PPARδ inhibition, aging mice were fed a diet containing 200 mg/kg GSK3787 for 7 weeks beginning at 1 week post-AAV9 injections.³⁶ For PPARγ inhibition, aging mice were daily treated with 0.35 mg/kg GW9662 in drinking water for 7 weeks beginning at 1 week post-AAV9 injections.^37,38 For PPARα overexpression in the myocardium, aging mice received a single intravenous injection of AAV9-Ctrl or AAV9-Ppara at a dose of 1 × 10¹¹ viral genome. To estimate the therapeutic value of FNDC4, 7- month-old young and 19-old aging mice were intraperitoneally injected with 0.2 mg/kg rFNDC4 protein every other day for 4 weeks. The rFNDC4 protein was prepared by cloning the extracellular fragment of FNDC4 according to previous studies, and the protein was produced in mammalian cells and as such being free of endotoxin.^26,27 At the end of the study, mice were sacrificed by performing cervical dislocation under deep anesthesia with 3% pentobarbital sodium or 2.5% isoflurane. All animal experiments were authorized by the Animal Care and Use Committee of Renmin Hospital of Wuhan University, and adhered to the Guidelines for the Care and Use of Laboratory Animals published by the US National Institutes of Health.

Cells and culture

Neonatal rat cardiomyocytes (NRCMs) were isolated from the neonatal rats as we previously described, and 100 μmol/L bromodeoxyuridine was used to inhibit the proliferation of cardiac fibroblasts.^39,40 The purity of cardiomyocytes was validated by α-actinin staining, and those with the purity over 90% were selected for further study. To stimulate lipotoxicity in vitro, NRCMs were treated with 0.5 mmol/L palmitic acid and 1.0 mmol/L oleic acid mixture (PAOA) for 12 hours, while bovine serum albumin (BSA) was used as the negative control.⁴¹ For FNDC4 overexpression in vitro, NRCMs were preinfected with AdhFNDC4 or AdCTRL for 4 hours at a multiplicity of infection (MOI) of 40, and then incubated in fresh medium containing 10% fetal bovine serum (FBS) for an additional 24 hours before PAOA stimulation. For FNDC4 knockdown in vitro, NRCMs were preinfected with shFndc4 or shCtrl for 4 hours at a MOI of 80, and then incubated in fresh medium containing 10% FBS for an additional 24 hours before PAOA stimulation. To knock down PPARα or PGC1α, NRCMs were pretransfected with 50 nmol/L siPpara or siPpargc1a for 4 hours using a Lipo6000 transfection reagent, and then cultured in fresh medium containing 10% FBS for an additional 12 hours before FNDC4 overexpression. In addition, NRCMs were also treated with 10 μmol/L GW6471 to inhibit the activity of PPARα.³⁵ For AMPKα knockdown, NRCMs were preinfected with shAmpkα2 at a MOI of 150 for 4 hours, and then cultured in fresh medium containing 10% FBS for an additional 24 hours before FNDC4 overexpression.²⁴ To inhibit adenylyl cyclase or protein kinase A (PKA), NRCMs were incubated with 200 μmol/L 2′5′-ddAdo or 10 μmol/L H89 as we previously described.^24,33 For exchange protein directly activated by cAMP (EPAC) knockdown, NRCMs were transfected with 50 nmol/L siEpac using Lipo6000 transfection reagent for 4 hours.²⁴

Echocardiography

Echocardiography was performed in conscious mice using a Vevo 3100 high-resolution Preclinical Imaging System (FUJIFILM VisualSonics) with a 30-MHz MX 400 linear ultrasound transducer as we recently described.^24,39 Briefly, mice were lightly anesthetized by 1.5% isoflurane, and then the standard 2-dimension-guided M-mode echocardiography was performed to identify the left ventricle internal diameter at diastole (LVIDd), the left ventricle internal diameter at systole (LVIDs), the interventricular septal (IVS) thickness at systole, and the IVS thickness at diastole. Fractional shortening (FS) was calculated using the following formula: (LVIDd−LVIDs)/LVIDd × 100%. In addition, tissue Doppler imaging was employed to evaluate the diastolic function, as assessed by the ratio of the early (E) to late (A) ventricular filling velocities (E/A).

Histology

The cross-sectional area of cardiomyocytes was evaluated by hematoxylin-eosin and WGA staining as we previously described.^34,39 Briefly, murine hearts were harvested and fixed in 10% formalin solution for 48 hours, which were then processed to 5 μm slices after dehydration and paraffin-embedding. Next, paraffin-embedded heart slices were exposed to hematoxylin-eosin staining according to standard protocols. The cross-sectional area of cardiomyocytes was counted blindly by 2 independent authors (K.L. and Y.J.Y.) using Image-Pro Plus 6.0, and calculated from 30 fields per group, with at least 5 cardiomyocytes per field analyzed. For WGA staining, heart slices were incubated with WGA working buffer (1:200) at 37 °C for 1 hour, and at least 200 cells per group were counted. To evaluate collagen deposition, heart slices were exposed to standard picrosirius red staining, and at least 60 fields per group were blindly included for the analysis of interstitial fibrosis.^42,43

SA-β gal staining

To evaluate cellular senescence, SA-β gal staining was performed using heart slices and cell coverslips as we recently described.²⁴ Briefly, fresh frozen heart slices and cell coverslips were fixed at room temperature for 15 minutes with the fixation buffer, and then reacted with the staining solution at 37 °C for 24 hours. The percentage of SA-β gal⁺ cells was blindly counted using a light microscope from at least 5 high-magnification fields per mice.

Transmission electron microscopy

For ultrastructural analysis of the heart, fresh hearts were processed into <1 μL pieces, which were then fixed in 2.5% glutaraldehyde at 4 °C overnight and postfixed in 1% osmium tetroxide for 1 hour on ice. Next, the samples were dehydrated in the ethanol gradients, incubated with acetone and embedded in ethoxyline resin. Ultrathin sections (90 nm) were prepared and counterstained with the uranyl acetate and lead citrate. Samples were then analyzed with a HITACHI HT 7800 transmission electron microscope at an acceleration voltage of 120 kV, and the mitochondrial number, area, and cristae number were blindly measured using an Image J software.

Oil red O and Nile red staining

To detect lipid accumulation in the myocardium and cardiomyocytes, oil red O staining was performed as previously described.^20,44,45 Briefly, fresh frozen heart slices or cell coverslips were fixed in 4% formaldehyde at room temperature for 20 minutes, washed with 60 % isopropanol, and then stained with freshly prepared 0.5% oil red O working solution dissolved in isopropanol at room temperature for 15 minutes to detect intracellular lipid droplets, followed by a counterstaining with hematoxylin. In addition, Nile red staining was also used for the evaluation of lipid accumulation in vivo and in vitro as previously described.^46,47 Briefly, heart slices or cell coverslips were incubated with 10 μg/mL Nile red at 4 °C overnight, and cell nuclei were visualized by the SlowFade gold antifade reagent with DAPI. Next, the images were captured using a DP74 fluorescence microscope (OLYMPUS).

Immunofluorescence staining

For immunofluorescence staining in vivo, heart slices were deparaffinized, hydrated, and exposed to high-pressure antigen retrieval in citrate (pH = 6.0). For immunofluorescence staining in vitro, cell coverslips were fixed in 4% formaldehyde and permeabilized in 1% Triton X-100. Next, heart slices or cell coverslips were incubated with 10% goat serum to block the nonspecific reaction, and then stained with the primary antibodies (Supplemental Table 1) at 4 °C overnight, followed by an incubation with the Alexa Fluor secondary antibodies (1:200 dilution) at 37 °C for an additional 1 hour. Cell nuclei were visualized by the SlowFade gold antifade reagent with DAPI, and the images were captured using a DP74 fluorescence microscope (OLYMPUS).^48,49

Measurements of telomere length

Telomere length was measured as we previously described.²⁴ Briefly, genomic DNA was extracted from the heart samples, and then the ratio of telomere repeat copy number to a single-gene, acidic ribosomal phosphoprotein PO forward (36B4) copy number was calculated as the telomere length.

Measurements of lipofuscin content

Lipofuscin in the myocardium was extracted and evaluated as we recently described.²⁴ Briefly, fresh heart samples were homogenized in chloroform-methanol (1:20, w:v), and then lipofuscin content in the chloroform-rich layer was measured at an excitation/emission wavelength of 350/485 nm.

DHE and DCFH-DA staining

To measure superoxide anion (O₂⁻) in the myocardium, DHE staining was performed as we previously described.⁵⁰ Briefly, fresh frozen heart slices were stained with 5 μmol/L DHE solution at 37 °C for 30 minutes, and then the fluorescent images were captured by a DP74 fluorescence microscope. To evaluate ROS in vitro, NRCMs were incubated with 5 μmol/L DCFH-DA solution at 37 °C for 30 minutes, and then the fluorescent images were captured by a DP74 fluorescence microscope.^33,39

Measurements of mitochondrial function

Mitochondrial complex I, II, III, and IV activities were measured using commercial kits according to the manufacturer’s instructions. To determine mitochondrial complex I, II, and IV activities, fresh heart samples were homogenized on ice, and the protein concentration of the tissue homogenates was adjusted to 5.5 mg/mL, which were then incubated with 1/10 volume of Detergent solution on ice for 30 minutes and centrifuged for 20 minutes at 4 °C to collect the supernatants. Next, the supernatants were diluted to a desired concentration according to the manufacturer’s instructions. For the measurement of mitochondrial complex I activity, 200 μL of the supernatants were added to the 96-well microplates pre-coated with a specific complex I capture antibody and incubated for 3 hours at room temperature, and the wells were washed with 300 μL of 1 × Wash Buffer solution for 3 times after removing the diluted samples. Finally, 200 μL of the Assay Solution were added to each well, and the absorbance was repeatedly measured at 450 nm at 30 second-intervals for 30 minutes at room temperature. For the measurement of mitochondrial complex II activity, 50 μL of the supernatants were added to the 96-well microplates precoated with a specific Complex II capture antibody and incubated for 2 hours at room temperature, and the wells were washed with 300 μL of 1 × buffer solution for 2 times after removing the diluted samples. Subsequently, 40 μL of phospholipids were added to each well and incubated for 30 minutes. Finally, 200 μL of activity solution were added to each well, and the absorbance was repeatedly measured at 600 nm at 20 second-intervals for 1 hour at room temperature. For the measurement of mitochondrial complex IV activity, 200 μL of the supernatants were added to the 96-well microplates precoated with a specific complex IV capture antibody and incubated for 3 hours at room temperature, and the wells were washed with 300 μL of solution 1 for 2 times after removing the diluted samples. Finally, 200 μL of the Assay solution were added to each well, and the absorbance was repeatedly measured at 550 nm at 1 minute-intervals for 2 hours at 30 °C. To determine mitochondrial complex III activity, mitochondria were isolated from fresh heart samples, and 1 μL of the mitochondrial samples were added to wells containing reaction mix. Subsequently, 6 μL of oxidized cytochrome c/cytochrome c were added to each well, and the absorbance was repeatedly measured at 550 nm at 30 second-intervals for 10 minutes at room temperature. The mitochondrial complex activities were normalized to total protein concentrations. To further evaluate mitochondrial function, oxygen consumption rate of NRCMs were measured using the Seahorse Bioscience XFe24 Extracellular Flux Analyzers as previously described. Following basal measurement, 1.5 μmol/L oligomycin (Oligo), 0.5 μmol/L cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP), and 1 μmol/L rotenone/antimycin A (Rot/AA) were injected, respectively. Maximal oxygen consumption rate was calculated as: (maximum rate measurement after FCCP injection) − (minimum rate measurement after Rot/AA injection).

Measurements of mtDNA content

To evaluate mtDNA content in the myocardium, fresh heart samples were used for the isolation of mitochondrial and genomic DNA with a commercial DNeasy kit as previously described by us and others.^49,51 Next, relative mtDNA copy number was calculated from Ct values of MT-CO2 normalized to the nuclear DNA (nDNA).

Measurements of mitochondrial membrane potential

Mitochondrial membrane potential was measured by JC-1 staining as we previously described.⁵² Briefly, NRCMs were stained with 5 μg/mL JC-1 in the dark at 37 °C for 20 minutes. The fluorescent images were captured at the excitation/emission wavelength of 525/590 nm (red, polymer) and at the excitation/emission wavelength of 490/530 nm (green, monomer) using a DP74 fluorescence microscope.

Measurements of TG and ATP levels

The level of TG in the myocardium was measured using a commercial kit according to the manufacturer’s instructions. Briefly, fresh heart samples were homogenized and centrifuged to obtain the cell-free supernatants, which were then incubated with 250 μL working solution at 37 °C for 10 minutes. Next, the absorbance was measured at 500 nm, and the level of TG was normalized to total protein contents. To measure cardiac ATP levels, fresh heart samples were homogenized in the lysis buffer and centrifuged at 12,000 g for 5 minutes at 4 °C to obtain the cell-free supernatants. Next, 100 μL of ATP detection working solution were added to each well and incubated at room temperature for 5 minutes to deplete the background ATP, and then 20 μL of the samples or standards were added. Subsequently, the luminescence values were recorded and normalized to total protein contents according to the standard curve method.

Measurements of oxidative stress

The levels of O₂⁻ and hydrogen peroxide (H₂O₂) in the myocardium were measured as we recently described.³⁹ Meanwhile, the levels of MDA, 3-NT, and 4-HNE were also detected to assess the peroxidation of lipids and proteins using commercial kits as we previously described.^39,53

Western blot

Total proteins were extracted, exposed to SDS-PAGE, and transferred onto PVDF membranes as we previously described.^42,43 Next, the membranes were incubated with primary antibodies (Supplemental Table 1) at 4 °C overnight, and then probed with HRP-conjugated secondary antibodies at room temperature for 1 hour, followed by a visualization with the electrochemiluminescence reagent using a Bio-Rad ChemiDoc XRS+ System. To measure PPARα in the nucleus, nuclear extracts were prepared by a commercial kit according to the manufacturer′s instructions, with lamin B1 used as the internal control.³⁹

Quantitative real-time PCR

Total RNA was extracted and exposed to cDNA synthesis using a Transcriptor First Strand cDNA Synthesis Kit (Roche, Basel, Switzerland).^48,49 Next, quantitative real-time PCR was performed on the Roche LightCycler 480 system with the SYBR Green I Master Mix (Roche). The primer sets used are provided in Supplemental Table 2.

Transcriptome analysis

Murine hearts with or without FNDC4 overexpression were harvested and used for total RNA purification and cDNA synthesis. Next, sequencing libraries were constructed using the DNBSEQ-T7 (BGI) following standard protocols. Raw reads were filtered using FASTP to remove reads containing sequencing adapters, N proportion more than 10%, and base proportion over 50%, which were then exposed to the quality control by FastQC to obtain the clean reads. Transcriptional reads were mapped to the GRCm38/mm10 mouse genome using HISAT2. To identify differentially expressed genes (DEGs), the fragments per kilobase million (FPKM) values of all mapped genes were calculated by DESeq2, and those genes with |Fold Change| ≥2 and an adjusted P value (Benjamini-Hochberg method) <0.05 were identified as DEGs. In addition, KEGG and GO enrichment analysis were performed for functional measurements. Heatmaps were created by Z-score scaling.

Untargeted metabolomics

To perform untargeted metabolomics, murine hearts were homogenized and extracted with an extraction solution (acetonitrile: methanol: water = 2:2:1) as we previously described.³⁹ Next, the samples were sonicated on ice and centrifuged at 4 °C for 10 minutes at 12,000 rpm to obtain the supernatants. Ultra-high performance liquid chromatography (UHPLC)-tandem mass spectrometry was then performed using a LC20 UHPLC System (Shimadzu) with a Waters ACQUITY UPLC HSS T3 C18 (2.1 mm × 100 mm, 1.8 μm) coupled to AB Sciex Triple TOF-6600 Mass Spectromete. Ultrapure water containing 0.1% formic acid was used as the mobile phase A, while acetonitrile containing 0.1% formic acid was used as the mobile phase B (400 μL/min). Typical ion source parameters (positive mode) were as follows: ionspray voltage = +5,500 V, ion source gas1 = 50 psi, declustering potential = +60 V, curtain gas = 35 psi, temperature = 550 °C, ion source gas2 = 60 psi, and collision energy = +30 V. Typical ion source parameters (negative mode) were as follows: ionspray voltage = −4,500 V, ion source gas1 = 50 psi, declustering potential = −60 V, curtain gas = 35 psi, temperature = 550 °C, ion source gas2 = 60 psi, collision energy = −30 V. Raw data were converted to the mzXML file by ProteoWizard and exposed to the XCMS software for peak matching, retention time alignment, and peak area extraction. Metabolites were unambiguously annotated using an in-house established, public, AI-predicted, and metDNA databases. Significantly different metabolites were selected according to the following criteria: 1) variable importance in projection (VIP) ≥1; 2) adjusted P value (Benjamini-Hochberg method) <0.05; and 3) |fold change| ≥2. Principal component analysis, heatmaps, and volcano plots were used for the presentation of differentially expressed metabolites.

Biochemical analysis

Circulating FNDC4 levels in mice and human were measured using a mouse FNDC4 EIA Kit and a human FNDC4 ELISA Kit according to the manufacturer’s instructions. Circulating NT-proBNP levels in mice were measured using the commercial kit as we previously described.⁴³ The levels of IL-6 and TNF-α in the myocardium were measured using commercial kits according to the manufacturer’s instructions.³⁹ To measure IL-6 levels, fresh heart samples were homogenized and centrifuged to obtain the cell-free supernatants, which were then added to the 96-well plates pre-coated with a specific IL-6 antibody and incubated for 2 hours at room temperature. Next, 100 μL of Ms IL-6 Biotin Conjugate solution were added and incubated for 30 minutes at room temperature after washing the wells with 1 × wash buffer for 4 times. Subsequently, 100 μL of 1 × Streptavidin-HRP solution were added and incubated for 30 minutes at room temperature after washing the wells with 1 × wash buffer for 4 times, and then 100 μL of stabilized chromogen were added and incubated for 30 minutes at room temperature in the dark after washing the wells with 1 × wash buffer for 4 times. Thereafter, 100 μL of stop solution were added, and the absorbance were recorded at 450 nm. To measure TNF-α levels, fresh heart samples were homogenized and centrifuged to obtain the cell-free supernatants, which were then added to the 96-well plates precoated with a specific TNF-α antibody in the presence of Biotin-Conjugate and incubated for 2 hours at room temperature. Next, 100 μL of diluted Streptavidin-HRP were added and incubated for 1 hour at room temperature after washing the wells with 1 × Wash Buffer for 6 times. Subsequently, 100 μL of TMB Substrate Solution were added and incubated for 30 minutes at room temperature in the dark after washing the wells with 1 × Wash Buffer for 4 times. Thereafter, 100 μL of Stop Solution were added, and the absorbance were recorded at 450 nm. The levels of cAMP were measured as we previously described.³³ Fatty acid oxidation was measured according to previous studies.^54,55 Serum levels of alanine transaminase, aspartate transaminase, blood urea nitrogen, creatinine, and creatine kinase in mice were measured using an ADVIA® 2400 automatic biochemical analyzer (Siemens Healthcare Diagnostics).

Statistical analysis

Results were shown using individual data points and the mean ± SD. The normality of the data distribution was assessed using the Shapiro-Wilks test. Unpaired 2-tailed Student′s t test was performed to compare differences between 2 groups with a normal distribution and homogeneity of variance, and 1-way analysis of variance in combination with Tukey’s post hoc test was performed for multiple comparisons. Associations between continuous variables were evaluated using Pearson′s correlation coefficient (r). A P value <0.05 was considered statistically significant. Analyses were done using GraphPad Prism (version 7.0) as we recently described.^39,56

Results

FNDC4 expression declines during cardiac aging and positively correlates with cardiac function

We recently demonstrated that FNDC5 overexpression or supplementation dramatically alleviated aging-related cardiac dysfunction.²⁴ FNDC4 shows the strongest homology with FNDC5 in the fibronectin type III domain family of proteins, and the present study aims to investigate its cardiomyocyte-autonomous role during cardiac aging. As shown in Figure 1A, FNDC4 protein level was significantly reduced in aging hearts. Immunofluorescence staining suggested that cardiac FNDC4 was abundantly localized to cardiomyocytes, and declined in aging mice (Figure 1B). Meanwhile, PCR data also identified a decrease of Fndc4 mRNA in cardiomyocytes (CMs), rather than cardiac fibroblasts (CFs), macrophages (Mφs), or endothelial cells (ECs), of aging hearts (Supplemental Figure 1A). In addition, Fndc4 mRNA in cardiomyocytes declined in an age-dependent manner (Supplemental Figure 1B). Interestingly, Fndc4 mRNA level in aging hearts negatively correlated with the mRNA level of Bnp, a biomarker of cardiac remodeling and dysfunction (Supplemental Figure 1C). In contrast, Fndc4 mRNA level positively correlated with the systolic function of aging hearts (Supplemental Figure 1D). Lehallier et al⁵⁷ previously established a human plasma proteome profile across the lifespan using the SomaScan aptamer technology, and measured 2,925 plasma proteins during aging. Using this public database, we found that FNDC4 levels in human plasma declined with increasing age (Figure 1C). Compared with young mice, FNDC4 levels were also decreased in the plasma of aging mice (Figure 1D). To enhance the clinical impact and translational value of our findings, we compared the differences in plasma FNDC4 levels between the elderly over 60 years of age and young people (<60 years of age), and the characteristics of these participants are listed in Supplemental Table 3. As shown in Figure 1E, the elderly exhibited a significant decrease of plasma FNDC4 levels. Of note, plasma FNDC4 levels in aging mice negatively correlated with the levels of plasma NT-proBNP, a biomarker of cardiac dysfunction, but positively correlated with FS (Figures 1F to 1G). Consistently, the increased plasma FNDC4 in the elderly also predicted a decrease of plasma NT-proBNP and an increase of ejection fraction (Figures 1H to 1I). Collectively, these findings suggest that FNDC4 expression declines during cardiac aging and positively correlates with cardiac function.

Figure 1.

FNDC4 Expression Declines During Cardiac Aging and Positively Correlates With Cardiac Function

(A) The expression of fibronectin type III domain-containing 4 (FNDC4) protein in heart samples from 8-month-old young and 20-month-old aging mice was detected by Western blot (n = 6). (B) Heart samples from young and aging mice were collected for immunofluorescence staining of sarcomeric α-actinin (green) and FNDC4 (red) (n = 6). (C) Human circulating FNDC4 levels were measured from 4,263 young adults to nonagenarians with the SomaScan aptamer technology, and analyzed using a public database. (D) Plasma FNDC4 levels were measured in young and aging mice using a commercial kit (n = 20). (E) Human circulating FNDC4 levels were measured from young participants (<60 years of age) and old participants (>60 years of age) using a commercial kit (n = 20). (F and G) Pearson’s correlation between plasma FNDC4 levels and plasma N-terminal pro-brain natriuretic peptide (NT-proBNP) levels or fractional shortening (FS) in aging mice (n = 20). (H and I) Pearson’s correlation between plasma FNDC4 levels and plasma NT-proBNP levels or ejection fraction in old participants (n = 20). All data are expressed as the mean ± SD, and analyzed using an unpaired 2-tailed Student′s t-test. For the analysis in F to I, Pearson’s correlation analysis was used. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs the matched group.

Cardiac-specific FNDC4 overexpression alleviates aging-related cardiac remodeling and dysfunction

To investigate the cardiomyocyte-autonomous role of FNDC4 during cardiac aging, mice were intravenously injected with AAV9-hFNDC4 to specifically overexpress FNDC4 in the myocardium. As shown in Supplemental Figures 2A and 2B, AAV9-hFNDC4 injection resulted in robust expression of FNDC4 protein in the myocardium, without affecting FNDC5 level. In addition, we also found no alterations of FNDC4 protein in the liver, skeletal muscle, or adipose tissue from mice with or without AAV9-hFNDC4 injection, indicating the effective and restrict overexpression of FNDC4 in the myocardium (Supplemental Figure 2C). Moreover, FNDC4 expression in cardiomyocytes was dramatically increased, while its expression in cardiac fibroblasts or endothelial cells in the heart was unaffected (Supplemental Figure 2D). Of note, FNDC4 overexpression in hearts did not affect circulating FNDC4 levels in mice (Supplemental Figure 2E). As shown in Supplemental Figures 3A and 3B, FNDC4 overexpression did not affect the levels of mean arterial pressure (MAP) and heart rate in either young or old mice. Yet, cardiac-specific FNDC4 overexpression dramatically restored the functional and structural alterations of aging hearts, as evidenced by the increased FS, and decreased LVIDd, LVIDs, and IVS thickness at systole (Figures 2A and 2B and Supplemental Figures 3C and 3D). Diastolic dysfunction is a hallmark of cardiac aging, and we found that FNDC4-overexpressed aging mice exhibited better diastolic function, as determined by the increased E/A (Figure 2C). SA-β gal staining revealed that cardiac-specific FNDC4 overexpression dramatically inhibited cellular senescence in aging hearts (Figure 2D). Meanwhile, the decreased telomere length and increased lipofuscin accumulation in aging hearts were restored by FNDC4 overexpression (Figures 2E and 2F). Moreover, FNDC4 overexpression also reduced the protein levels of senescent markers in aging hearts, including p16, p19 and p21 (Supplemental Figure 3E). Cardiac remodeling is a crucial determinant of aging-related cardiac dysfunction.^2,24 As shown in Supplemental Figure 3F, aging-related increases of heart mass were decreased in FNDC4-overexpressed mice, as evidenced by the reduced heart weight-to-tibia length (HW/TL). Histologic analysis revealed that cardiac-specific FNDC4 overexpression dramatically attenuated hypertrophic growth and interstitial fibrosis of aging hearts (Figure 2G). In addition, aging hearts with FNDC4 overexpression also displayed lower Anp, β-Mhc, Col1α1, Col3α1, and higher α-Mhc mRNA levels (Supplemental Figures 3G and 3H). Consistent with the PCR data of hypertrophic and fibrotic markers, transcriptome analysis also identified a beneficial effect of FNDC4 against aging-related cardiac remodeling (Figure 2H, Supplemental Figure 3I). Taken together, our results imply that cardiac-specific FNDC4 overexpression alleviates aging-related cardiac remodeling and dysfunction.

Figure 2.

Cardiac-Specific FNDC4 Overexpression Alleviates Aging-Related Cardiac Remodeling and Dysfunction

(A and B) 6-month-old young and 18-month-old aging mice received a single intravenous injection of AAV9-hFNDC4 to specifically overexpress FNDC4 in the myocardium, and AAV9-CTRL was used as the negative control. Two months post-AAV9 injection, FS, left ventricle internal diameters at diastole (LVIDd), and left ventricle internal diameters at systole (LVIDs) were determined by echocardiography to evaluate the systolic function (n = 6). (C) Tissue Doppler imaging was employed to measure the ratio of the early (E) to late (A) ventricular filling velocities (E/A) to evaluate the diastolic function (n = 6). (D) Representative images and quantitative results of senescence-associated β-galactosidase (SA-β) gal staining in hearts from young and aging mice with or without FNDC4 overexpression (n = 6). (E) Relative telomere length in hearts (n = 6). (F) Relative lipofuscin accumulation in hearts (n = 6). (G) Heart samples were collected for wheat germ agglutinin (WGA), hematoxylin-eosin (HE), and picrosirius red (PSR) staining to quantify the cross-sectional area of cardiomyocyte and interstitial fibrosis 2 months post-AAV9 injection (n = 6). (H) Heart samples from aging mice with or without FNDC4 overexpression were exposed to unbiased transcriptome analysis, and the expressions of cardiac remodeling-related genes were presented using a heatmap (n = 4). All data are expressed as the mean ± SD, and analyzed using 1-way analysis of variance followed by Tukey post hoc test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs the matched group. Abbreviations as in Figure 1.

Cardiac-specific FNDC4 knockdown facilitates aging-related cardiac remodeling and dysfunction

To further decipher the role of FNDC4, we specifically knocked down its expression in the myocardium using AAV9 vectors. As shown in Figure S4A-B, FNDC4, but not FNDC5, protein expression was effectively decreased in the myocardium with AAV9-shFndc4 injection. In addition, the expression of FNDC4 in other tissues (liver, skeletal muscle or adipose tissue) and cell types in the heart (cardiac fibroblasts and endothelial cells) was unaffected (Supplemental Figures 4C and 4D). Interestingly, cardiac-specific FNDC4 knockdown dramatically aggravated the systolic and diastolic dysfunction in aging hearts, without affecting MAP and heart rate (Figures 3A to 3D, Supplemental Figures 4E to 4G). Meanwhile, the ratio of SA-β gal⁺ senescent cells in aging hearts was also elevated by FNDC4 knockdown, accompanied with decreased telomere length and increased lipofuscin accumulation (Figures 3E to 3G). Accordingly, FNDC4 knockdown also elevated the expressions of p16, p19, and p21 in aging hearts (Supplemental Figures 4H to 4I). Moreover, we found that aging-related cardiomyocyte hypertrophy and collagen deposition were dramatically exacerbated by FNDC4 knockdown, as evidenced by the increased cross-sectional area of cardiomyocytes, collagen volume and HW/TL (Figure 3H, Supplemental Figure 4J). The mRNA levels of hypertrophic and fibrotic markers in aging hearts were further disturbed by FNDC4 knockdown (Supplemental Figure 4K-L). To further validate the role of FNDC4, we used an independent shFndc4# to confirm specificity and exclude any off-target effects. As shown in Supplemental Figures 5A to 5C, aging-related systolic and diastolic dysfunction were further compromised in mice with shFndc4# injection, as evidenced by increased LVIDd, LVIDs and decreased FS, E/A. Cellular senescence was also aggravated in shFndc4#-injected aging hearts (Supplemental Figure 5D). In addition, aging mice with shFndc4# injection displayed more severe cardiomyocyte hypertrophy and interstitial fibrosis (Supplemental Figures 5E and 5F). Our findings indicate that cardiac-specific FNDC4 knockdown facilitates aging-related cardiac remodeling and dysfunction.

Figure 3.

Cardiac-Specific FNDC4 Knockdown Facilitates Aging-Related Cardiac Remodeling and Dysfunction

(A-B) 6-month-old young and 18-month-old aging mice received a single intravenous injection of AAV9-shFndc4 to specifically knock down FNDC4 in the myocardium, and AAV9-shCtrl was used as the negative control. Two months post-AAV9 injection, FS, LVIDd and LVIDs were determined by echocardiography to evaluate the systolic function (n = 6). (C-D) Tissue Doppler imaging was employed to measure E/A to evaluate the diastolic function (n = 6). (E) Representative images and quantitative results of SA-β gal staining in hearts from young and aging mice with or without FNDC4 knockdown (n = 6). (F) Relative telomere length in hearts (n = 6). (G) Relative lipofuscin accumulation in hearts (n = 6). (H) Heart samples were collected for WGA, HE, and PSR staining to quantify the cross-sectional area of cardiomyocyte and interstitial fibrosis 2 months post-AAV9 injection (n = 6). All data are expressed as the mean ± SD, and analyzed using 1-way analysis of variance followed by Tukey post hoc test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs the matched group. Abbreviations as in Figures 1 and 2.

FNDC4 prevents mitochondrial dysfunction and lipotoxicity in aging hearts

We next analyzed the unbiased transcriptome data from aging hearts with or without FNDC4 overexpression to decipher the molecular basis mediating the cardioprotective effects of FNDC4 against aging. As shown in Figure 4A and Supplemental Figure 6A, the up-regulated KEGG pathways and GO terms by FNDC4 overexpression were closely related with metabolism, especially fatty acid metabolism. Aging hearts exhibit a sizable decline in the utilization and oxidation of fatty acids, which leads to lipid accumulation and subsequent lipotoxicity within the cardiomyocytes.^7,11 In our study, results from oil red O and Nile red staining implied that lipid accumulation in hearts was dramatically increased during aging (Figures 4B and 4C). Interestingly, FNDC4 overexpression effectively inhibited lipid accumulation, accompanied with a decreased TG level in aging hearts (Figures 4B and 4D). The decreased utilization and oxidation of fatty acids significantly diminish ATP production in the myocardium, and subsequently contribute to aging-related cardiac dysfunction. As shown in Figure 4E, aging-related decreases of ATP content in the myocardium were restored by FNDC4 overexpression. In line with the improved fatty acid metabolism, the transcriptome data revealed that cardiac-specific FNDC4 overexpression dramatically elevated the expression of genes related with the utilization and oxidation of fatty acids (Supplemental Figure 6B). A set of target genes observed in transcriptome analysis were also validated by quantitative real-time PCR (Supplemental Figure 6C). To get further insights into the metabolic rearrangements, untargeted metabolomics analysis was performed on aging hearts with or without FNDC4 overexpression. Principal component analysis results, volcano map, and heatmaps of the differentially expressed metabolites were provided in Supplemental Figures 6D to 6F. Indeed, the increased fatty acid metabolism in aging hearts with FNDC4 overexpression was paralleled by decreased levels of acyl-carnitines, especially the long-chain-carnitine conjugates (Figure 4F). Fatty acid oxidation assay also identified an increased extent of fatty acid oxidation in aging hearts with FNDC4 overexpression (Supplemental Figure 6G). Mitochondria are the primary site for fatty acid oxidation and ATP generation, and play critical roles in the preservation of cardiac function during aging. In contrast, the decreased mitochondrial content and defective ETC subunits contribute to the progression of aging-related cardiac dysfunction.^3,9 The ultrastructural analysis demonstrated that cardiac-specific FNDC4 overexpression restored mitochondrial number, cristae density, and reduced mitochondrial swelling of aging hearts (Figure 4G, Supplemental Figure 6H). Increased mitochondrial number in aging hearts with FNDC4 overexpression was also validated by the elevated mtDNA/nDNA levels (Supplemental Figure 6I). The ameliorative mitochondrial structure and function in FNDC4-overexpressed aging hearts were accompanied with increased expressions of mitochondrial ETC genes and mitochondrial complex activities (Figures 4H and 4I). Mitochondrial damage results in mtDNA release and lipid accumulation that synergistically activate inflammation in the myocardium.^58,59 In line with the beneficial effects of FNDC4 on mitochondrial function, the downregulated KEGG pathways and GO terms by FNDC4 overexpression were mainly enriched in inflammation, including inflammatory response, positive regulation of interleukin-6 production, cytokine-cytokine receptor interaction, and NF-kappa B signaling pathway, etc (Supplemental Figures 7A to 7B). Accordingly, aging-related increases of IL-6 and TNF-α in the myocardium were also reduced by FNDC4 overexpression (Supplemental Figures 7C to 7D). Mitochondrial damage, defective ETC function, and the subsequent lipid accumulation also elicit the overproduction of toxic ROS, eventually resulting in oxidative damage to aging hearts.^1,3 As shown in Supplemental Figure 7E, oxidative stress-related genes in aging hearts were regulated by FNDC4 overexpression. Accordingly, cardiac-specific FNDC4 overexpression dramatically inhibited the generation of free radicals in aging hearts, accompanied with decreased levels of MDA, 3-NT and 4-HNE (Supplemental Figures 7F to 7I). In contrast, aging-related increases of lipid accumulation and decreases of ATP generation in the myocardium were further compromised by FNDC4 knockdown (Supplemental Figure 8A to 8C). Fatty acid oxidation assay also identified a decreased extent of fatty acid oxidation in aging hearts with FNDC4 knockdown (Supplemental Figure 8D). The defective structure and function of mitochondria in aging hearts were also exacerbated by FNDC4 knockdown, as evidenced by the decreased mitochondrial number, cristae density, mtDNA/nDNA, and increased mitochondrial area (Supplemental Figures 8E to 8G). Meanwhile, cardiac-specific FNDC4 knockdown further aggravated aging-related inflammation and oxidative damage in the myocardium (Supplemental Figures 8H to 8L). Moreover, the deleterious effects of FNDC4 knockdown on mitochondrial function and lipotoxicity were further validated using an independent shFndc4# (data not shown).

Figure 4.

FNDC4 Prevents Mitochondrial Dysfunction and Lipotoxicity in Aging Hearts

(A) Heart samples from aging mice with or without FNDC4 overexpression were exposed to unbiased transcriptome analysis, and significant KEGG pathways were targeted by upregulated DEGs (n = 4). (B-C) Heart samples were collected for oil red O and Nile red staining to quantify lipid deposition (n = 6). (D) triglyceride (TG) levels in hearts (n = 6). (E) ATP levels in hearts (n = 6). (F) Heart samples from aging mice with or without FNDC4 overexpression were exposed to untargeted metabolomics analysis, and the expressions of significantly different metabolites were presented using a heatmap (n = 6). (G) Representative transmission electron micrographs of aging hearts with or without FNDC4 overexpression (n = 6). (H) Heart samples from aging mice with or without FNDC4 overexpression were exposed to unbiased transcriptome analysis, and the expressions of mitochondrial electron transport chain (ETC)-related genes in aging hearts with or without FNDC4 overexpression were presented using a heatmap (n = 4). (I) The activities of mitochondrial complex I, II, III and IV were measured using commercial kits (n = 6). All data are expressed as the mean ± SD, and analyzed using one-way analysis of variance followed by Tukey post hoc test. For the analysis in I, an unpaired 2-tailed Student′s t-test was used. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs the matched group. Abbreviations as in Figure 1.

To further confirm the beneficial effects of FNDC4 on lipotoxicity and associated cellular senescence in vitro, NRCMs with or without FNDC4 overexpression were stimulated with PAOA to imitate lipotoxicity (Supplemental Figure 9A). As shown in Supplemental Figure 9B, FNDC4 overexpression dramatically inhibited PAOA-induced cellular senescence, as determined by the decreased SA-β gal⁺ senescent cells. Meanwhile, the elevated protein levels of p16, p19, and p21 in PAOA-stimulated NRCMs were reduced by FNDC4 overexpression (Supplemental Figure 9C). In line with the in vivo findings, FNDC4 overexpression preserved the expression of genes related with the utilization and oxidation of fatty acids in PAOA-treated NRCMs (Supplemental Figure 9D). Oil red O and Nile red staining suggested that PAOA-induced lipid accumulation in NRCMs was also inhibited in those with FNDC4 overexpression (Supplemental Figures 9E and 9F). In addition, the decreased mitochondrial membrane potential and increased inflammation as well as oxidative stress in PAOA-stimulated NRCMs were restored by FNDC4 overexpression (Supplemental Figures 9G to 9J). Seahorse analysis showed that FNDC4 overexpression rescued mitochondrial respiratory defects in PAOA-stimulated NRCMs (Supplemental Figure 9K). In contrast, FNDC4 knockdown further aggravated PAOA-induced cellular senescence and lipid accumulation in vitro (Supplemental Figures 10A to 10E). Accordingly, mitochondrial dysfunction, inflammatory response and oxidative damage in PAOA-stimulated NRCMs were further compromised in those with FNDC4 knockdown (Supplemental Figures 10F to 10J). Based on these in vivo and in vitro findings, we conclude that FNDC4 prevents mitochondrial dysfunction and lipotoxicity in aging hearts.

FNDC4 restores mitochondrial function and lipid metabolism by up-regulating PPARα/PGC1α

PPAR signaling pathway, the nuclear receptor superfamily responsible for regulating mitochondrial function and lipid metabolism, was dramatically activated in aging hearts by FNDC4 overexpression; therefore, we determined whether FNDC4 attenuated aging-related cardiac dysfunction by restoring PPAR signaling pathway (Figure 4A).¹² As shown in Figure 5A and Supplemental Figure 11A, both transcriptome analysis and PCR data revealed that the mRNA level of Ppara, rather than Ppard or Pparg, was dramatically elevated in FNDC4-overexpressed aging hearts. Some coactivators, such as estrogen-related receptors (ERRs) (encoded by Esrra and Esrrb), PGC1 (encoded by Ppargc1a and Ppargc1b), and mitochondrial transcription factor A (TFAM), are also required for the role of PPAR proteins in controlling mitochondrial function and lipid metabolism.¹⁷ Interestingly, cardiac-specific FNDC4 overexpression also increased the mRNA level of Ppargc1a in aging hearts, without affecting Esrra, Esrrb, Ppargc1b, and Tfam (Figure 5A, Supplemental Figure 11A). In addition, Fndc4 mRNA level positively correlated with the mRNA levels of Ppara and Ppargc1a in aging hearts (Supplemental Figure 11B). Moreover, we found that FNDC4 overexpression facilitated the expression and nuclear accumulation of PPARα protein in aging hearts, accompanied with an elevated PGC1α protein level (Figures 5B and 5C). In contrast, the protein expressions of PPARα/PGC1α and nuclear accumulation of PPARα in aging hearts were further inhibited by FNDC4 knockdown (Supplemental Figure 11C). In line with the in vivo findings, FNDC4 overexpression restored, while FNDC4 knockdown further reduced the protein levels of PPARα/PGC1α and nuclear accumulation of PPARα in PAOA-stimulated NRCMs (Supplemental Figures 11D and 11E). To validate the necessity of PPARα, PPARα^−/− and PPARα^+/+ mice were used. As shown in Figure 5D and Supplemental Figure 12A, cardiac-specific FNDC4 overexpression failed to restore the ultrastructure of mitochondria in aging hearts, as determined by the unaffected mitochondrial number, cristae density, mitochondrial swelling, and mtDNA/nDNA. Accordingly, the decreased lipid accumulation and increased ATP generation in FNDC4-overexpressed aging hearts were also blocked in PPARα^−/− mice (Figures 5E and 5F, Supplemental Figure 12B). Meanwhile, PPARα deficiency also blunted FNDC4 overexpression-mediated inhibitions on aging-related inflammation and oxidative stress, as evidenced by the increased IL-6, TNF-α, H₂O₂, O₂⁻, MDA, 3-NT, and 4-HNE levels (Supplemental Figures 12C to 12F). In addition, the inhibitory effects of FNDC4 on aging-related cellular senescence and cardiac remodeling were offset by PPARα ablation (Figures 5F and 5G, Supplemental Figure 12G). Moreover, the improved systolic and diastolic functions in FNDC4-overexpressed aging hearts were impaired in PPARα^−/− mice (Figures 5H to 5J). The efficiency of PPARα deletion was verified by genotyping and western blot (Figure 5K, Supplemental Figure 12H). To further determine the role of PPARα in vitro, NRCMs were transfected with siPpara to knock down endogenous PPARα expression (Supplemental Figure 13A). As shown in Supplemental Figure 13B, the decreased SA-β gal⁺ senescent cells in FNDC4-overexpressed NRCMs upon PAOA stimulation were increased by PPARα knockdown. In addition, PPARα knockdown also blunted the inhibitory effects of FNDC4 on PAOA-induced lipid accumulation and mitochondrial dysfunction (Supplemental Figure 13C). Moreover, FNDC4 failed to attenuate PAOA-induced oxidative stress and inflammatory response in NRCMs with PPARα knockdown, as determined by the increased levels of MDA, 3-NT, 4-HNE, Il-6 mRNA, and Tnf-α mRNA (Supplemental Figures 13D and 13E). To test whether PPARα activity contributes to the cardioprotective effects of FNDC4, aging mice or PAOA-stimulated NRCMs with or without FNDC4 overexpression were treated with GW6471. Interestingly, we found that PPARα inhibition by GW6471 also blunted FNDC4 overexpression-mediated protections against aging-related lipotoxicity, cellular senescence, and cardiac dysfunction in vivo and in vitro (data not shown). However, administration with the inhibitors of PPARδ or PPARγ did not affect the cardioprotective effects of FNDC4 in aging mice (Supplemental Figures 14A to 14D). In addition, we also re-expressed PPARα in FNDC4-silenced aging hearts. The results indicated that reintroducing PPARα in aging hearts dramatically prevented FNDC4 knockdown-mediated aggravation of cardiac aging and dysfunction (Supplemental Figures 15A to 15E). To clarify the involvement of PGC1α, Ppargc1a floxed and cKO mice were used. As shown in Supplemental Figures 16A to 16F, the inhibitory effects of FNDC4 on aging-induced lipid accumulation, energy deficit, cellular senescence and cardiac dysfunction were dramatically abolished by cardiac-specific PGC1α deficiency. Similar results were found in PAOA-stimulated NRCMs with PGC1α knockdown (data not shown). Collectively, we demonstrate that FNDC4 restores mitochondrial function and lipid metabolism by up-regulating PPARα/PGC1α.

Figure 5.

FNDC4 Restores Mitochondrial Function and Lipid Metabolism by Up-Regulating PPARα/PGC1α

(A) Heart samples from aging mice with or without FNDC4 overexpression were exposed to unbiased transcriptome analysis, and the expressions of PPAR signaling pathway-related genes were presented using a heatmap (n = 4). (B) Aging hearts with or without FNDC4 overexpression were collected for immunofluorescence staining of sarcomeric α-actinin (red) and peroxisome proliferator-activated receptor α (PPARα) (green) (n = 6). (C) The expressions of PPARα and peroxisome proliferator-activated receptor γ coactivator 1 α (PGC1α) proteins in whole or nuclear lysates from hearts were detected by western blot (n = 6). (D) Representative transmission electron micrographs of PPARα-deficient aging hearts with or without FNDC4 overexpression (n = 6). (E) The levels of TG and ATP in hearts (n = 6). (F) PPARα-deficient aging hearts with or without FNDC4 overexpression were collected for oil red O, Nile red, and SA-β gal staining to measure lipid deposition and cellular senescence (n = 6). (G) Quantitative results of SA-β gal staining in hearts (n = 6). (H-I) FS, LVIDd and LVIDs were determined by echocardiography to evaluate the systolic function (n = 6). (J) Tissue Doppler imaging was employed to measure E/A to evaluate the diastolic function (n = 6). (K) Genotyping results of PPARα^−/− and PPARα^+/+ mice (n = 6). All data are expressed as the mean ± SD, and analyzed using 1-way analysis of variance followed by Tukey post hoc test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs the matched group. Abbreviations as in Figures 1 and 2.

FNDC4 facilitates PPARα signaling pathway by activating AMPKα

We then explored the underlying mechanism through which FNDC4 facilitated PPARα signaling pathway. AMPKα emerges as an attractive target to restore metabolic disorder and cardiac dysfunction through regulating PPARα.²⁰ In addition, our previous studies have shown that AMPKα activation dramatically attenuated pressure overload-, diabetes-, sepsis- and aging-induced cardiac dysfunction.21, 22, 23, 24 Moreover, FNDC4 was defined as an endogenous activator of AMPKα to prevent hyperlipidemia-induced insulin resistance.⁶⁰ Interestingly, our transcriptome data revealed that various downstream targets of AMPK signaling pathway in aging hearts were elevated by FNDC4 overexpression (Supplemental Figure 17A). In addition, cardiac-specific FNDC4 overexpression restored, while cardiac-specific FNDC4 knockdown further compromised the phosphorylation of AMPKα in aging hearts (Figures 6A and 6B). Meanwhile, aging-related inhibition of AMPKα activity was preserved by FNDC4 overexpression, but exacerbated by FNDC4 knockdown, as determined by the phosphorylated levels of ACC, a downstream target of AMPKα (Figures 6A and 6B). Subsequently, AMPKα2^−/− mice were used to validate the involvement of AMPKα in FNDC4-mediated preservations on PPARα expression and cardiac homeostasis during cardiac aging. As shown in Figure 6C, the increased protein levels of PPARα and PGC1α in FNDC4-overexpressed aging hearts were dramatically decreased by AMPKα deficiency. In addition, AMPKα ablation also disturbed the beneficial effects of FNDC4 on lipid metabolism and ATP generation in aging hearts (Figures 6D and 6E). Meanwhile, the inhibitory effects of FNDC4 on aging-related inflammation and oxidative damage were abrogated in AMPKα2^−/− mice (Supplemental Figures 17B to 17D). Accordingly, cardiac-specific FNDC4 overexpression failed to ameliorate aging-related cellular senescence and cardiac remodeling in AMPKα2^−/− mice, as determined by the increased levels of SA-β gal⁺ senescent cells, cross-sectional area and collagen volume (Figures 6F and 6G). The improved cardiac function by FNDC4 overexpression in aging mice was also abolished by AMPKα2 deficiency (Figures 6H and 6I). To clarify the necessity of AMPKα in vitro, we knocked down the expression of AMPKα2 in NRCMs using adenoviral vectors as we previously described.²⁴ In line with the in vivo findings, we determined that FNDC4 overexpression failed to elevate the protein levels of PPARα and PGC1α in AMPKα-silenced NRCMs under PAOA stimulation (Supplemental Figure 18A). The inhibitory effects of FNDC4 on PAOA-induced cellular senescence were also abrogated by AMPKα knockdown (Supplemental Figure 18B). Meanwhile, the decreased inflammation and oxidative damage in PAOA-stimulated NRCMs with FNDC4 overexpression were exacerbated in those with AMPKα knockdown, as determined by the increased levels of Il-6 mRNA, Tnf-α mRNA, MDA, 3-NT, and 4-HNE (Supplemental Figures 18C and 18D). FNDC4 is released to the extracellular space as a soluble bioactive fragment, and displays a high homology with FNDC5. The second messengers are required for the transmission of extracellular signals to intracellular molecular network. The cAMP acts as an important second messenger upstream of AMPKα signaling pathway, and various studies by us and the others showed that extracellular FNDC5 could increase cAMP/AMPKα, thereby exerting cardioprotective effects.^24,61 As shown in Supplemental Figure 18E, FNDC4 overexpression dramatically elevated, while FNDC4 knockdown reduced the level of intracellular cAMP in PAOA-stimulated NRCMs. Both protein kinase A and EPAC are identified as the downstream effectors of cAMP signaling, and play critical roles in AMPKα activation and cardioprotection.^22,24,33 Interestingly, AMPKα activation in PAOA-stimulated NRCMs with FNDC4 overexpression was blocked by 2′5′-ddAdo or siEpac, but not H89, indicating the involvement of cAMP/EPAC axis in FNDC4-mediated AMPKα activation (Supplemental Figure 18F). Taken together, we prove that FNDC4 facilitates PPARα signaling pathway by activating AMPKα.

Figure 6.

FNDC4 Facilitates PPARα Signaling Pathway by Activating AMPKα

(A-B) The phosphorylated levels of AMP-activated protein kinase α (AMPKα) and acetyl-CoA carboxylase (ACC) proteins in whole heart lysates with FNDC4 overexpression or knockdown were detected by western blot (n = 6). (C) The expressions of PPARα and PGC1α proteins in AMPKα^−/− or AMPKα^+/+ aging hearts with or without FNDC4 overexpression were detected by western blot (n = 6). (D) TG levels in hearts (n = 6). (E) ATP levels in hearts (n = 6). (F) AMPKα-deficient aging hearts with or without FNDC4 overexpression were collected for SA-β gal, HE and PSR staining to measure cellular senescence, cardiomyocyte hypertrophy and interstitial fibrosis (n = 6). (G) Quantitative results of SA-β gal, HE and PSR staining in hearts (n = 6). (H) FS was determined by echocardiography to evaluate the systolic function (n = 6). (I) Tissue Doppler imaging was employed to measure E/A to evaluate the diastolic function (n = 6). All data are expressed as the mean ± SD, and analyzed using 1-way analysis of variance followed by Tukey post hoc test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs the matched group. Abbreviations as in Figure 1, Figure 2, Figure 3, Figure 4, Figure 5.

Therapeutic administration of rFNDC4 protein ameliorates aging-related cardiac dysfunction

To enhance the clinical impact and translational value of our findings, we finally investigated whether administration of rFNDC4 protein could provide therapeutic effects against cardiac aging. As shown in Figure 7A, rFNDC4 treatment dramatically reduced the ratio of SA-β gal⁺ senescent cells in aging hearts. Accordingly, the protein levels of senescent markers were also decreased in hearts from rFNDC4-treated mice, accompanied with the improved telomere shortening and lipofuscin accumulation (Supplemental Figures 19A to 19C). Meanwhile, we found that rFNDC4 treatment suppressed aging-related cardiomyocyte hypertrophy and interstitial fibrosis (Figure 7A). The increased HW/TL in aging mice was also decreased by rFNDC4 treatment (Figure 7B). In addition, aging-induced disorders of lipid metabolism and ATP generation were dramatically restored in rFNDC4-treated mice, as determined by the decreased TG accumulation and increased ATP content (Figure 7C and Figure S19D). Moreover, the systolic and diastolic dysfunction in aging hearts were ameliorated by rFNDC4 administration (Figures 7D to 7F). We also evaluated whether therapeutic administration of rFNDC4 protein would result in hepatic, renal, or muscular injuries as fibrates, the clinically used TG-lowering reagents through activating PPARα. As shown in Supplemental Figures 19E to 19G, therapeutic administration of rFNDC4 protein did not lead to hepatic or muscular injuries, and even provided renal protection during aging. In line with the phenotypic alterations, rFNDC4 administration also restored AMPKα/PPARα/PGC1α axis in aging hearts (Supplemental Figures 19H to 19J). In addition, mitochondrial complex activities in the heart were enhanced by rFNDC4 treatment (Supplemental Figure 19K). Collectively, we conclude that therapeutic administration of rFNDC4 protein ameliorates aging-related cardiac dysfunction.

Figure 7.

Therapeutic Administration of rFNDC4 Protein Ameliorates Aging-Related Cardiac Dysfunction

(A) 7-month-old young and 19-month-old aging mice were intraperitoneally injected with recombinant FNDC4 (rFNDC4) protein every other day for 4 weeks, and heart samples were collected for SA-β gal, HE, and PSR staining to measure cellular senescence, cardiomyocyte hypertrophy, and interstitial fibrosis (n = 6). (B) Quantitative results of heart weight-to-tibia length (HW/TL) (n = 6). (C) TG levels in hearts (n = 6). (D-E) FS, LVIDd, and LVIDs were determined by echocardiography to evaluate the systolic function (n = 6). (F) Tissue Doppler imaging was employed to measure E/A to evaluate the diastolic function (n = 6). All data are expressed as the mean ± SD, and analyzed using 1-way analysis of variance followed by Tukey post hoc test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs the matched group. Abbreviations as in Figure 1, Figure 2, Figure 3.

Discussion

Aging is an independent determinant of cardiovascular diseases, and aging-related cardiac dysfunction contributes to increased disability and mortality in the elderly.⁶² In the present study, we demonstrate that FNDC4 expression is decreased in the heart and plasma of aging mice, and that lower FNDC4 expression correlates with a poor cardiac function. Cardiac-specific FNDC4 overexpression alleviates while cardiac-specific FNDC4 knockdown facilitates aging-related cardiac remodeling and dysfunction. The unbiased transcriptome analysis and untargeted metabolomics reveal that FNDC4 activates PPARα signaling pathway to improve mitochondrial dysfunction and lipotoxicity in aging hearts, and that PPARα deficiency abolishes the protective effects of FNDC4 against cardiac aging in vivo and in vitro. Mechanistically, AMPKα is required for FNDC4-mediated activation of PPARα. Moreover, therapeutic administration of rFNDC4 protein dramatically ameliorates aging-related cardiac dysfunction, without resulting in significant side effects. Collectively, our findings highlight a crucial role of FNDC4 to control mitochondrial function and lipid metabolism during cardiac aging, and for the first time identify FNDC4 as an attractive predictive and therapeutic target of cardiac aging.

Accumulative evidence has revealed that mitochondrial dysfunction and subsequent metabolic disorder, especially lipid accumulation, are essential for the progression of cardiac remodeling and dysfunction upon different pathological stimuli, including aging.^3,4 Zhao et al²⁰ found that restoring fatty acid metabolism and ATP synthesis dramatically prevented pressure overload-induced cardiac hypertrophy and heart failure. In contrast, reducing the oxidation of fatty acids resulted in increased TG accumulation, decreased ATP generation, and significant cardiac dysfunction.⁶³ Meanwhile, inhibiting lipolytic activity by genetic inactivation of adipose triglyceride lipase also facilitated lipid deposition in hearts, eventually triggering cardiac fibrosis and dysfunction.⁶⁴ Generally, cardiac lipotoxicity is especially related to the accumulation of multiple reactive lipid intermediates, including acyl-carnitines. In a global PPARα-deficient mouse model, PPARγ overexpression in cardiomyocytes did not affect the levels of TG or other lipid intermediates, but reduced acyl-carnitine levels and subsequently improved cardiac function.⁶⁵ Despite exhibiting increased TG levels in the myocardium, malonyl CoA decarboxylase-deficient obese mice displayed decreased levels of long-chain acyl-carnitines, accompanied with an improved cardiac function.⁶⁶ In our study, cardiac-specific FNDC4 overexpression dramatically restored fatty acid metabolism in aging hearts, which was paralleled by decreased levels of acyl-carnitines, especially the long-chain-carnitine conjugates. Lipid accumulation can induce deleterious effects to the heart through numerous direct or indirect manners, of which, ROS overproduction and inflammation may play indispensable roles.^1,67 Jaishy et al⁶⁸ found that lipids and their intermediates directly elevated the expression of NADPH oxidase to enhance ROS production. The ultrastructure and function of mitochondria are dramatically compromised during the progression of heart diseases, which also increase the generation of free radicals caused by the defective ETC process. ROS generated from lipid intermediates, in turn, contributes to mitochondrial damage and dysfunction, which creates a vicious cycle to provoke the development of oxidative damage to the heart. It is well-accepted that nutrient overload can activate inflammatory responses in both intracardiac and extracardiac tissues, termed metabolic inflammation or metainflammation.⁶⁹ Mouton et al⁷⁰ demonstrated that pro-inflammatory fatty acids directly promoted NF-kappa B activation and M1 polarization, thereby exacerbating inflammatory injury to the heart. Toll-like receptor 4 was identified as a potential pattern recognition receptor of free fatty acids to activate NF-kappa B signaling pathway.⁷¹ Additional metabolic intermediates also contributed to the regulation of gene expression in inflammatory cells through transcriptional or epigenetic mechanisms.^72,73 In addition, Chiu et al⁷⁴ found that the accumulation of long-chain acyl CoA also stimulated cytochrome c release and lipid-induced programmed cell death, thereby initiating cardiac hypertrophy, ventricular dysfunction and premature death. Accordingly, facilitating fatty acid uptake by CD36 overexpression led to increased TG and long-chain acyl CoA levels in the myocardium, imposed lipotoxicity to cardiomyocytes, and subsequently aggravated aging-related cardiac dysfunction.⁵ In our study, we prove that cardiac-specific FNDC4 overexpression restored lipid metabolism in aging hearts and alleviated inflammation as well as oxidative stress, eventually improving aging-related cardiac dysfunction.

Mitochondria are the primary site for fatty acid oxidation and provide 90% of the ATP in hearts. There are 2 populations of mitochondria in the myocardium, including subsarcolemmal mitochondria (SSM) and interfibrillar mitochondria (IFM).^3,67 Emerging studies have shown that IFM population is the main defective mitochondria in aging hearts and primary source of ROS. The amount of IFM was decreased in aging hearts, whereas SSM content was unaffected with age.¹¹ IFM from aging hearts displayed an aberrant inner membrane folding pattern characterized by the presence of concentrically swirling cristae and mitochondrial swelling.⁷⁵ Meanwhile, Fannin et al⁷⁶ previously found that IFM from aging hearts exhibited lower activities of OXPHOS complexes, while respiration in SSM was unaltered. Accordingly, Lemieux et al⁷⁷ demonstrated that IFM population was essential for the decline of global respiration in aging cardiomyocytes. In our study, we also found that the number of IFM was decreased, whereas mitochondrial area was increased in aging hearts. Mitochondrial dysfunction results in defective fatty acid oxidation and increased fatty acid deposition. PPARα is abundantly expressed in the myocardium, and plays critical roles in the regulation of mitochondrial homeostasis through coordinating with other co-activators, including ERRα/β, PGC1α/β, and TFAM.¹² In addition, most of the proteins involving in the process of fatty acid oxidation (fatty acid uptake, TG formation, storage, lysis, fatty acid transport to mitochondria and oxidation) are transcriptionally regulated by PPARα. Findings from Drosatos et al⁶³ revealed that PPARα down-regulation inhibited the expressions of fatty acid metabolism-related genes, thereby compromising fatty acid oxidation, ATP production and cardiac function. Accordingly, our recent study showed that enhancing the expression and activity of PPARα/PGC1α dramatically preserved mitochondrial function as well as metabolic balance in diabetic hearts.¹⁵ In addition, up-regulating PPARα/PGC1β also facilitated fatty acid oxidation and ATP production, and eventually prevented sepsis-induced inflammation and cardiac dysfunction.⁷⁸ Pioneering studies have established an interplay between FNDC4 and metabolic homeostasis. Frühbeck et al²⁸ previously revealed that plasma FNDC4 level was reduced in obese populations and related with systemic inflammation. FNDC4 treatment facilitated mitochondrial biogenesis and subsequently reduced intra-cytosolic lipid accumulation. Consistently, the levels of FNDC4 were also decreased in the plasma and liver of obese mice, and treatment with rFNDC4 protein dramatically attenuated metabolic disorder and systemic inflammation.²⁷ In our study, we found that cardiac-specific FNDC4 overexpression restored the content and ultrastructure of mitochondria, and increased the activities of OXPHOS complexes, thereby facilitating fatty acid oxidation in aging hearts. Mechanistically, FNDC4 elevated the expression and nuclear accumulation of PPARα in an AMPKα-dependent manner.

Study limitations

In addition to inflammaging, excessive inflammation can lead to inflammatory cell death (eg, pyroptosis), which is associated with the progression of aging-related cardiac dysfunction. Whether the cardioprotective effects of FNDC4 against cardiac aging involves pyroptotic inhibition remains further determination. In addition, the specific manner mediating AMPKα/PPARα activation by FNDC4 is also not determined in the present study. The potential receptors of FNDC4 in the myocardium deserves further investigations in the future.

Conclusions

Our findings reveal that FNDC4 prevents aging-related cardiac dysfunction by restoring AMPKα/PPARα-dependent mitochondrial function, and for the first time identify FNDC4 as an attractive predictive and therapeutic target of cardiac aging.

Perspectives.

COMPETENCY IN MEDICAL KNOWLEDGE: Mitochondria play critical roles in maintaining oxidative metabolism and cardiac homeostasis; however, their function is compromised in aging hearts. In this study, we identified the cardiomyocyte-autonomous role of FNDC4 in mitochondrial function and metabolic homeostasis during cardiac aging. Compared with young mice, aging mice exhibited a sizable decline in cardiac and plasma FNDC4 levels, and lower FNDC4 expression also correlated with a poor cardiac function. Cardiac-specific FNDC4 overexpression alleviated, while cardiac-specific FNDC4 knockdown facilitated aging-related cardiac remodeling and dysfunction. Mechanistic studies revealed that FNDC4 activated AMPKα/PPARα to improve mitochondrial dysfunction and lipotoxicity in aging hearts.

TRANSLATIONAL OUTLOOK: FNDC4 expression positively correlates with cardiac function during aging, and FNDC4 overexpression or supplementation effectively prevents aging-related cardiac dysfunction by restoring AMPKα/PPARα-dependent mitochondrial function. Collectively, our study for the first time identified FNDC4 as an attractive predictive and therapeutic target of cardiac aging.

Funding Support and Author Disclosures

Sequencing service was provided by Bioyi Biotechnology Co, Ltd (Wuhan, China). This work was supported by grants from the Natural Science Foundation of Hubei Province (No. 2023AFB099, 2024AFB092), the Fundamental Research Funds for the Central Universities (No. 2042023kf0046, Clinical Medicine+ Youth Talent Support Program of Wuhan University), Free Innovation Pre-Research Fund of Wuhan Union Hospital (2023XHYN034), Undergraduate Training Programs for Innovation and Entrepreneurship of Wuhan University (202410486110, S202410486326), the Open Project of Hubei Key Laboratory (No. 2023KFZZ028) and Excellent and new plan of Wuhan Union Hospital. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.