1. Introduction
Aging significantly increases the risk of heart failure [
1]. Heart failure affects approximately 8.4% of people aged 65 and older, whereas it is 0.7% in those aged 45−54 years [
2]. The aged heart exhibit structure and function alterations, including increased thickness of left ventricular (LV) wall, alterations in the diastolic filling pattern with severe LV diastolic stiffness, prolonged diastole and interstitial fibrosis [
3]. At the cellular level, cardiomyocytes acquire a senescence-like phenotype (increased myocyte volume, dysfunctional organelles, enhanced senescence-associated β-galactosidase (SA-β-gal) activity, promoted cell cycle arrest, a proinflammatory response and increased oxidative stress) that contributes to aging-related myocardial dysfunction [
4]. Elimination or amelioration of senescent cardiomyocytes has been found to be effective means to improve aging related disorders [
5]. However, there are no effective therapeutic drugs to ameliorate aging-related heart dysfunction so far. For this reason, exploring the molecular alterations associated with heart aging and identifying novel targeted agents based on the pathological mechanism would be highly beneficial.
Lysosomes are membrane-enclosed organelles that play active metabolism effects in a cell [
6]. They contain various hydrolases and degrade extracellular and intracellular biological polymers during the process of endocytosis and autophagy [
7]. Recent studies have revealed that, in addition to their roles in degradation and recycling, lysosomes also serve as an organizing center that influences signal transduction. This regulation is crucial for lysosome biogenesis, enzyme function, and autophagy, all of which are essential for maintaining cellular and organismal homeostasis [
8]. Dysregulation of lysosomal signals has been shown to be strongly associated with cell survival in aging and aging-related disorders [
9], [
10]. TRPML1 is an inwardly rectifying Ca
2+ permeable cation channel encoded by the gene mucolipin 1 (
MCOLN1). It has the representative TRP six transmembrane segments, N- and C-termini in the cytosol, and is primarily found in the endolysosomal compartment [
11]. TRPML1 is one pivotal protein that regulates the normal lysosomal Ca
2+ handling process and the activity of transcription factor EB (TFEB), a key regulator essential for controlling autophagy and lysosome formation [
12]. The deletion of TRPML1 leads to the abnormal transport of substrates and the degradation of lysosomes, hindering lysosomes from entering the plasma membrane through exocytosis [
13]. In addition, abnormal Ca
2+ release mediated by TRPML1 inhibits the binding of autophagy to lysosomes [
14]. The capacity of TRPML1 in regulating the lysosome function and multiple stages of autophagy is evident. However, how aging modulates TRPML1 expression and the associated changes in cell fate remain unknown.
Long noncoding RNAs (lncRNA) are RNA molecules exceeding 200 nucleotides in length that regulate gene expression without encoding proteins [
15]. These molecules can regulate many necessary biological processes such as cell differentiation, development, and metabolism [
16]. In particular, our group has recently highlighted lncRNAs as novel regulators of cardiovascular diseases (CVDs), including heart aging, myocardial infarction, arrhythmia, and atherosclerosis [
17], [
18], [
19]. In addition, latest studies have reported lncRNAs play a significant roles in aging-related disorders. One study suggested that silencing the lncRNA Zeb2-NAT is beneficial for the reprogramming of senescent fibroblasts and another demonstrated that lncRNA inhibits aging-induced muscle hypertrophy [
20], [
21]. Nonetheless, the functions of lncRNAs and their regulatory mechanisms during heart aging largely remain unknown. Using RNA sequencing (RNA-seq) analysis, we discovered an upregulated lncRNA, Zeb1os1 (known as ZEB1-AS1 in humans), in the hearts of aged mice. Previous studies have extensively investigated the oncogenic role of ZEB1-AS1 in various cancers, including liver, pancreatic, oral squamous cell, and colorectal carcinomas [
22], [
23], [
24], [
25]. However, its role in cardiac function is not well understood. Only one study has reported its ability to mitigate myocardial fibrosis in diabetic mice [
26].
Our research identifies Zeb1os1 as a crucial contributor and potential biomarker of heart aging. Notably, our investigation revealed that Zeb1os1 functions as a causal suppressor of TRPML1, thus inducing lysosomal dysfunction and subsequent heart aging.
2. Materials and methods
2.1. Clinical sample processing
Initially, 120 individuals underwent physical, echocardiographic, and laboratory examinations at the Second Affiliated Hospital of Harbin Medical University upon enrollment. Participants were excluded if they had hypertension (defined as SBP ≥ 140 mmHg or DBP ≥ 90 mmHg), diabetes mellitus with fasting glucose levels of ≥ 126 mg∙dl−1, cardio-cerebrovascular disease, body mass index (BMI) ≥ 25 kg∙m−2, or had chronic obstructive pulmonary disease. 49 participants were enrolled for further analysis. The study’s procedures were authorized by Harbin Medical University’s ethics committee and conducted in accordance with the ethical principles of the Declaration of Helsinki (approval number: IRB5041721). The comprehensive clinical details of the participants are provided in Table S1 in Appendix A.
2.2. Animal aging model and construction of adeno-associated virus (AAV)
Male C57BL/6 mice were sourced from Beijing Vital River Laboratory Animal Technology Co. Ltd. This study adhered strictly to the EU Directive 2010/63/EU for animal experiments. Ethical approval for all experimental procedures was granted by the Harbin Medical University Ethics Committee (approval number: IRB5041721).
(1) AAV has multiple serotypes, of which AAV9 has the highest infection efficiency in heart tissue. Therefore, we chose AAV9 for engineering the constructs for Zeb1os1 knockdown and overexpression. The Zeb1os1 interference plasmid (shZeb1os1) was engineered using a double-stranded adeno-associated virus (dsAAV) plasmid with ampicillin resistance, into which the shZeb1os1 sequence (60 bp) and EGFP sequence (717 bp) were inserted. The recombinant plasmid promoters are CAG (a synthetic promoter combining the cytomegalovirus early enhancer element, chicken actin promoter, and globin intron) and H1 (H1 RNA polymerase III promoter), and there is a poly A sequence of bovine growth hormone (BGH) downstream the subcloning site. The construction of the Zeb1os1 overexpression plasmid and AAV packaging involved the use of AAV plasmid as a vector with Zeb1os1 full-length sequence and EGFP sequence inserted into the subcloning site downstream the cytomegalovirus (CMV) promoter. These constructs were stored at −80 °C for future experiments and underwent sequencing verification to ensure accuracy.
(2) Two- and 18-month-old C57BL/6 mice were selected. Old mice were injected intravenously with Zeb1os1 knockdown adeno-associated virus (AAV-shZeb1os1), and young mice were injected with Zeb1os1-overexpressing Adeno-associated virus (AAV-Zeb1os1) via tail vein. Each mouse received a dose of 2 × 10
11 GC (genome containing particles) with saline (150 µL per mouse) as the vehicle once every two months for a total of three times, with analysis beginning at six months after the first injection. Control group mice received the same dose of saline, and mice in negative control (NC) group were injected with negative control shRNA sequence (AAV-shNC) or empty plasmid adeno-associated virus (AAV-NC). Each experimental group contained six mice. Anesthesia was induced with an intraperitoneal injection of 0.2 g∙kg
−1 avertin and subsequently euthanized by exsanguination, consistent with our prior protocol [
27], which was following AVMA Guidelines. Avertin was selected as the preferred anesthetic agent due to its comparatively diminished impact on cardiac function and heart rhythm, especially in comparison to other anesthetics such as isoflurane or ketamine blends [
28], [
29]. This has made it a favored option in investigations pertaining to cardiovascular homeostasis at the animal experimental level [
30].
2.3. RNA-seq
(1) RNA extraction: RNA was isolated from myocardial specimens (young group or old group) using the RNeasy Mini Kit (Cat#74106; QIAGEN, Germany) as per the manufacturer’s guidelines (Bohao Biology, China) and RNA was evaluated with the Agilent Bioanalyzer 2100 (Agilent technologies, USA). To further purify the RNA, we employed the RNA Clean XP Kit (Cat #A63987; Beckman Coulter, USA) and RNase-Free DNase Set (Cat#79254; QIAGEN, GmBH, Germany). Quantification was done using a NanoDrop ND-2000 (Thermo Scientific, USA).
(2) Transcriptome deep sequencing: For RNA high-throughput sequencing, RNA libraries were constructed from each sample group. Key workflow steps included ribosomal RNA removal, total RNA fragmentation, reverse transcription, synthesis of second-strand complementary DNA (cDNA), end repair, addition of dA tails, and adaptor ligation. Final cDNA libraries were amplified and purified by PCR, followed by sequencing on an Illumina HiSeq2500 platform (Illumina, USA).
(3) Differential lncRNA screening: Differential lncRNA expression was analyzed using EdgeR software [
31]. The read counts were normalized, and the Benjamini-Hochberg method was applied to adjust for multiple hypothesis testing. lncRNAs with |logFC| > 3 and an adjusted
p-value < 0.05 were identified as differentially expressed. A heatmap depicting these differentially expressed genes (DEGs) was created using the heatmap package.
2.4. Echocardiographic evaluation of heart function
Heart function was evaluated through echocardiography using a Visual Sonics Vevo 2100 machine with a 30 MHz high‐frequency transducer (VisualSonics, Canada). The mice were sedated via an intraperitoneal dose of 0.2 g∙kg−1 avertin (Sigma, USA). The parasternal long axis (PLAX), short axis (SAX), and apical four chambers (A4C) views were captured using B-Mode, M-Mode, and P-Mode imaging. Exam accuracy was maintained by conducting the procedure within a rigorously controlled timeframe. Measurements included the ratio of peak E to peak A (early diastolic transmitral flow velocity (E)/late diastolic transmitral flow velocity (A) ratio) and the mitral flow velocity at E peak (MV E peak). The left ventricular ejection fraction (LVEF), fractional shortening (FS), isovolumic relaxation time (IVRT), left ventricular posterior wall diameter (LVPWD), and left ventricular mass (LV mass) were measured. Attention was paid not to exert excessive pressure on the chest and to maintain the heart rate of the mice at 400−500 beats per minute.
2.5. Tissue processing and histopathological assessment
Heart tissues were preserved in 10% formalin, followed by a dehydration process and embedding in wax. The tissues were then cut into 4 µm for microscopic examination. To examine changes in heart architecture, hematoxylin and eosin (H&E) used to color sections. Collagen distribution was evaluated by Masson trichrome staining.
2.6. Immunohistochemistry
TRPML1 expression was evaluated in the myocardium. In summarize, paraffin sections were deparaffinized, blocked with goat serum (Cat#AR0009; Boster, China) at room temperature for 30 minutes and then incubated overnight at 4 °C with a rabbit derived anti-mouse TRPML1 antibody (Cat#ACC-081; Alomone labs, Israel). Following primary antibody incubation, sections were treated with a horseradish peroxidase (HRP)-conjugated goat derived anti-rabbit secondary antibody (1:200, Cat#A0216; Beyotime, China) for 30 min at room temperature. Detection was carried out using 3,3′-Diaminobenzidine (DAB) and counterstained with hematoxylin. Examination of TRPML1-positive cells by a light microscopy showed brown or yellow-brown particles.
2.7. Culturing primary cardiomyocytes and cardiac fibroblasts
Neonatal mice (aged 1−3 d) were used to isolate these cells. The hearts were collected under aseptic conditions, minced and washed, and subjected to digestion with 0.25% trypsin. The resultant cell was centrifuged and resuspended in Dulbecco’s modified eagle medium (DMEM) containing 10% fetal bovine serum, 100 U∙ml−1 penicillin and 100 μg∙ml−1 streptomycin. The cell mixture was incubated in a culture flask at 37 °C for 1.5 h, allowing fibroblasts adhere to surface. Non-adherent cells, primarily cardiomyocytes, were then cultured separately. The cardiomyocytes growth medium was subsequently supplemented with 10 nmol∙L−1 5-Bromo-2-deoxyUridine (5-BrdU, Cat#B5002; Sigma) in order to eliminate fibroblasts. The cells were incubated at 37 °C in a humidified atmosphere containing 5% CO2 for 48 h before being used in further experiments.
2.8. Continued cell line
The AC16 human cell line (#SCC109, Millipore Sigma, USA) exhibits numerous biochemical and morphological characteristics of cardiomyocytes. AC16 cells were maintained in medium composed of DMEM: F12 (#SH30023.01; HyClone, USA) with 10% fetal bovine serum, 1% penicillin-streptomycin, and 1% fungizone (Cat#C0222; Beyotime). To sustain the cells, they were cultured at a temperature of 37 °C within a humidified atmosphere composed of 5% CO2 and 95% air.
2.9. Cell transfection
Zeb1os1 specific siRNA (siZeb1os1) and a negative control siRNA (siNC) were synthesized commercially by Ribobio (China). The siRNA constructs (100 nmol∙L−1) were transfected into cells to achieve Zeb1os1 knockdown. Zeb1os1 cDNA was cloned into the plasmid cytomegalovirus DNA 3.1 (pCDNA3.1) plasmid (OE-Zeb1os1). Cells were then transfected with plasmid vectors (OE-Zeb1os1 for overexpression and OE-NC as a control) at a concentration of 2.5 mg∙L−1 to overexpress Zeb1os1. TRPML1 cDNA was cloned into pCDNA3.1 plasmid. TRPML1 plasmid vectors and NC vectors at a concentration of 2.5 mg∙L−1 were transfected into cells. Transfection was conducted with X-treme GENE Transfection Reagent (#10810500; Roche, Switzerland) following the manufacturer’s guidelines. Cardiomyocytes were harvested 48 h post-transfection for RNA isolation or protein purification. The siZeb1os1 sequences: sense 5′-CGGAUGAACUGUUAAUAAACC-3′ and antisense 5′-UUUAUUAACAGUUCAUCCGGU-3′.
2.10. SA-β-gal staining
β‐galactosidase staining kit (Cell Signaling Technology, USA) was utilized following the kit protocol. Cells or cryosections of heart tissue were first fixed for 10−15 min at room temperature, then washed twice with phosphate-buffered saline (PBS). Under non-CO2 conditions, 1 mL of β-galactosidase staining solution was added to the cells, which were then incubated at 37 °C for 12 h. Following this incubation, the samples were then analyzed under a light microscope (Carl Zeiss Microscopy, Germany).
2.11. Quantitative real-time PCR (qRT-PCR)
To measure Zeb1os1 expression, qRT-PCR was employed. Total RNA was isolated using TRIzol reagent (Invitrogen, USA). For the amplification of RNA, the SYBR Green Realtime PCR Master Mix Kit (Toyobo, USA) was utilized. The 20 µL PCR reaction mix consisted of 1 µL of forward and reverse primers, 1 µL of cDNA, 10 µL of SYBR Green PCR Master Mix, and the remaining volume filled with nuclease-free water. Reactions were run according to the cycling parameters. Each gene was amplified in triplicate. Relative gene expression was calculated using the 2-ΔΔCT (CT stands for cycle threshold, and ΔΔCT represents the normalized difference in CT values between target and reference genes) method, normalizing against glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as reference genes. The specificity and quality of amplification were evaluated through melting curve analysis. The primers utilized for real-time PCR are as Table S1 in Appendix A.
2.12. Western blot analysis
Each protein sample (100 μg) was loaded onto a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel and subsequently transferred to nitrocellulose membranes (PALL, USA). The membranes were blocked with 5% skim milk for 2 h and then incubated with the primary antibody overnight at 4 °C, including p21 (Cat#ab188224; Abcam, USA), p53 (Cat#9282; Cell Signaling Technology), LC3 (Cat#ab192890; Abcam), p62 (Cat#ab109012; Abcam), Lamp1 (Cat#ab208943; Abcam), and TRPML1, or GAPDH (Cat#TA-08; Zhongshanjinqiao, Inc., China) as internal control. Band intensities on Western blots were quantified using the Odyssey Infrared Imaging System (LI-COR, USA).
2.13. Electron microscopic examination
Mouse heart and cardiomyocytes were preserved in 2.5% glutaraldehyde at 4 °C overnight, followed by treatment with 2% osmium tetroxide for 70 min. The samples were then dehydrated through a graded ethanol series and stained with uranyl acetate and lead citrate, and subsequently photographede (JEOL Ltd, Japan).
2.14. Flow cytometric analysis
Neonatal mouse cardiomyocytes (NMCMs) were digested and collected, washed thoroughly with PBS three times, then centrifuged at 1000g for 5 min. The cells were then fixed in 70% pre-cooled ethanol at 4 °C for at least 12 h. Following this fixation, the cells were treated with ribonuclease (RNase) and propidium iodide in the dark at 4 °C for 30 min. Upon detecting a propidium iodide signal, cell cycle analysis was conducted by flow cytometry (CytoFLEX; Beckman Coulter Commercial Enterprise (China) Co., Ltd, China). CytExpert 2.0 software was used to analyze the outcomes (Beckman Coulter Company, USA).
2.15. Immunofluorescence staining
(1) Cells were cultured on glass coverslips pre-treated with poly-L-lysine. Following treatment, the cells were washed twice with PBS, fixed in 3.7% formaldehyde, and then blocked with PBS/0.1% Triton-X/3% bovine serum albumin (BSA) for 30 min at room temperature. Next, the cells were exposed to primary antibodies specific to Actinin (Cat#A7811, 1:200; Sigma) and Lamp-1 (Cat#ab25245; Abcam) or TFEB (Cat#A303-673A, 1:200, Bethyl Laboratories, USA). Detection was carried out using Alexa Fluor 594- or Alexa Fluor 488-conjugated secondary antibodies. After three additional washes with PBS, the cells were stained with 4′,6-diamidino-2-phenylindole (DAPI) DNA stain (Cat#D1306; Thermo Scientific) and rinsed with PBS. Imaging was captured and analyzed using a confocal laser microscope (LSM780; Zeiss Microsystem).
(2) Cells were incubated with 500 nmol∙L−1 LysoTracker (Beyotime) at 37 °C for 30 min. To quantify the intensity of LysoTracker fluorescence, images were randomly taken from three separate coverslips per independent experiment. DAPI staining was used to visualize nuclei. Finally, samples were visualized using a confocal laser microscope (LSM780; Zeiss Microsystem).
2.16. Intracellular reactive oxygen species (ROS) detection
The fluorescent dye 2′7′-dichlorodihydrofluorescein diacetate (DCFH-DA) was employed to measure intracellular ROS levels. Following treatments, the NMCMs were incubated with DCFH-DA (20 μmol∙L−1, Beyotime) at 37 °C for 30 min, then washed with PBS. Then fluorescence images were captured under a fluorescence microscope (Carl Zeiss Microscopy).
2.17. RNA-Protein interaction analysis via RNA immunoprecipitation (RIP) assays
RIP assays were conducted using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Cat#17-701; Millipore, Germany) following its protocol. The isolated RNA was converted to cDNA using the ReverTraAce qPCR RT Kit (Cat#FSQ-101; Toyobo). Target gene quantification was conducted using the SYBR Green PCR Master Mix Kit (Toyobo) on a 7500 fast Real Time PCR system (AB Applied Biosystems, USA). The specificity and quality of amplification were evaluated through melting curve analysis.
2.18. RNA-protein pulldown
Zeb1os1 sequence was inserted into the pcDNA3.1 vector through subcloning. In vitro RNA transcription was performed using the T7 RNA Polymerase Kit (Cat#D7069; Beyotime) following the provided protocol. The RNA transcripts were subsequently labeled using the RNA 3′End Desthiobiotinylation Kit (Cat#20163; Oshkosh, USA). RNA-protein pulldown assays were conducted using the Pierce Magnetic RNA-Protein Pull-Down Kit (Cat#20164; Pierce). Following the assay, the TRPML1 protein, which interacts with Zeb1os1, was identified through western blot analysis.
2.19. Co-immunoprecipitation (CO-IP) assay and ubiquitination (UB) analysis
Following a 12-h treatment with MG132 (#2194; CST, USA), cells were lysed in ice-cold radio immunoprecipitation assay (RIPA) buffer. The protein extracts were then mixed and incubated with Protein A/G PLUS Agarose (Santa Cruz Biotechnology, China) pre-bound to an anti-TRPML1 antibody (Cat#ACC-081; Alomone labs) for at least 12 h at 4 °C with vertical shaking. The immunoprecipitates were washed, resuspended, and analyzed via western blotting for UB using anti-ubiquitin (#58395; Cell Signaling Technology) and anti-TRPML1 antibodies (Cat#ACC-081; Alomone labs).
2.20. Monitoring autophagic flux using live-cell imaging
A fluorescent probe was used to monitor autophagic flux. Cells were transduced with the red fluorescent protein-green fluorescent protein-light chain 3 (RFP-GFP-LC3) fluorescent probe (Hanbio Biotechnology Co., Ltd., China) via an adenovirus system. Cells were then imaged 24 h post-transduction using a confocal laser microscope (Olympus, Japan) under consistent settings to monitor autophagic flux.
2.21. Determination of lysosomal calcium release in cardiomyocytes
(1) Dye preparation: Fluo-3 of 50 mg (Cat#F1242; Invitrogen) was solubilized in 50 μL of DMSO to prepare dye stock solution A. F127 of 2 mg (BASF, USA) powder was mixed into 1 mL of calcium benchtop solution and used as a dye stock solution B. The dye working solution formula was set as: 5 μL stock solution A + 65 μL stock solution B + 930 μL calcium benchtop solution.
(2) Staining of primary cardiomyocytes: The dye working solution was added to a 24-well plate containing primary cardiomyocytes, followed by incubation at 37 °C for 35 min. After discarding the supernatant, the cells were washed with calcium benchtop solution three times by pipetting and then placed in a fresh calcium benchtop solution for later use.
(3) Determination of lysosomal calcium release: After treatment with ionomycin (Sigma) for 5 min, followed by stimulation with ML-SA1 (Sigma), the cardiomyocytes were imaged by a flash4.0 LT camera for about 4 s. The change of fluorescence intensity represents the change of intracellular calcium. ΔF represents the fluorescence value of cells, F0 the background fluorescence value of cells, and (ΔF)/F0 the maximum value of calcium release from cardiomyocytes.
2.22. Extraction of RNA from cytoplasm and nucleus
The NE-PER Nuclear Cytoplasmic Extraction Reagent kit (Pierce) was used. In short, the cells were washed twice with PBS, centrifuged, and then resuspended in cytoplasmic extraction reagent I. The pellet was treated with cytoplasmic extraction reagents, followed by centrifugation to collect the cytoplasmic RNA. The remaining pellet, containing nuclei, was resuspended in nuclear extraction reagent, incubated on ice, and centrifuged to obtain the nuclear RNA fraction.
2.23. Lysosomal enzyme activity assay
The activities of cathepsin B (CTSB) and cathepsin D (CTSD) were quantified using specific assay kits for cathepsin D and B (Ab65302 and Ab65300; Abcam) following the manufacturer’s instructions.
2.24. Isolation of adult mouse cardiomyocytes
Anesthesia in mice was induced via intraperitoneal injection of 2% avertin at a dosage of 0.2 g∙kg−1. Hearts were rapidly excised for retrograde perfusion using a Langendorff apparatus. Following 3−5 min of perfusion with calcium-free Tyrode’s solution, hearts were subsequently perfused by collagenase Type II (1 mg∙mL−1, Gibco, USA) and BSA (1 mg∙mL−1, BioFroxx, Germany). Once the myocardial tissue softened, the left ventricle was isolated into calcium-free Tyrode’s solution, where cardiomyocytes were gently dissociated using pipetting. The separated cardiomyocytes were equilibrated in Tyrode’s solution with 0.18 mmol∙L−1 CaCl2 and 1% BSA.
2.25. Fluorescence in situ hybridization (FISH)
The Ribo FISH kit (#R11060.10; Ribobio) was utilized to conduct FISH. Zeb1os1 FISH probe staining was carried out following the kit manual’s instructions and recommendations. Incubated samples at 37 °C for 30 min, then hybridized at room temperature with 20 μmol∙L−1 Zeb1os1 FISH probe mixture for 12 h. After washing with saline-sodium citrate (SSC) solutions and PBS/Tween-20, DAPI was used to stain the nuclei. Finally, imaging was captured using a confocal fluorescence microscope.
2.26. Molecular docking
The TRPML1 crystal structure (Protein Data Bank (PDB): 5jw5) was obtained from the research collaboratory for structural bioinformatics (RCSB) PDB. Zeb1os1-functional domain (FD)’s three dimensional (3D) structure was predicted using ViennaRNA Web Services and the 3dRNA/DNA Web Server. HEX8.0 software was applied for calculation of binding affinity. The binding complex was modulated by PyMOL software.
2.27. Isolation of nuclear and cytoplasmic proteins
Proteins from the nucleus and cytoplasm were extracted using the ExKine Nuclear and Cytoplasmic Protein Extraction Kit (Cat#KTP4002; Abbkine, China). Adherent cells were collected using a cell scraper and centrifuged at 500g for 5 min. The cells were then resuspended in pre-cooled PBS and centrifuged again. The pellet was processed with cytoplasmic extraction reagents to obtain the cytoplasmic fraction. The remaining pellet underwent nuclear extraction. Both extracts were either kept on ice or stored at −80 °C.
2.28. Data analysis
Data were analyzed using SPSS 25.0 (IBM, USA) and GraphPad Prism 9.0. For multiple comparisons, one-way or two-way ANOVA followed by Tukey’s post-hoc test was applied. A Student’s two-tailed t-test was used for comparing two groups, with p < 0.05 considered significant.
3. Results
3.1. LncRNA Zeb1os1 is upregulated in aged hearts
To identify changes in lncRNA expression patterns in aged hearts, RNA-seq analysis was performed on RNA samples isolated from young (2-month) and old (24-month) mouse hearts. After the data for the lncRNA annotation expression levels were filtered, the differentially expressed lncRNAs with statistical significance were identified. The most upregulated (red) or downregulated lncRNAs (green) in the old group were identified (
Fig. 1(a)). We then selected the five most upregulated lncRNAs with interspecies conservation in their sequences and tested their expression via qRT-PCR. We found that NONMMUT031348.2 (Zeb1os1) was significantly upregulated in aged hearts compared with young controls (
Fig. 1(b)). We utilized analysis tools from a single-cell database
†(
†singlecell.broadinstitute.org.) to extensively investigate the level of ZEB1-AS1 in various cell types in the human heart (Fig. S1 in Appendix A). The analysis revealed that ZEB1-AS1 is highly expressed in cardiomyocytes and fibroblasts. We subsequently focused on elucidating its role in cardiomyocytes during heart aging. Furthermore, Zeb1os1 was found to be distributed in both the nucleus and cytoplasm of cardiomyocytes (
Fig. 1(c)). D-galactose (D-gal) is commonly used to artificially induce senescence
in vivo and
in vitro [
32]. Compared to the control group, D-gal markedly increased the expression of Zeb1os1 in primary NMCMs (
Fig. 1(d)). Similar results were replicated in cell line of AC16 (
Fig. 1(e)). In addition, we isolated cardiomyocytes and detected the expression of Zeb1os1 from young and old mice through FISH. We found a substantial upregulation of Zeb1os1 in cardiomyocytes from old mice compared to young mice (Fig. S2 in Appendix A). To investigate the association between ZEB1-AS1 and heart aging, we measured its messenger RNA (mRNA) levels in the blood of healthy adult human subjects of different ages. Specifically, a total of 49 subjects with or without deteriorated heart diastolic function were included in this study and their clinical characteristics are detailed in
Table 1. The E/A ratio, a known indicator of diastolic function that compares the early LV filling phase (E-wave) to the atrial contraction phase (A-wave), showed a significant decline with age (
Figs. 1(f) and (g)). Moreover, the ZEB1-AS1 expression level was markedly higher in the blood of old individuals than in that of young individuals, with no differences observed between the sexes (
Fig. 1(h); Fig. S3 in Appendix A). Additionally, as shown in
Table 2, analyses using both univariable and multivariable approaches demonstrated a strong positive link between ZEB1-AS1 levels and age, along with an inverse relationship with the E/A ratio. The above results suggest that ZEB1-AS1 (Zeb1os1) has the potential to serve as a circulating biomarker indicative of heart aging.
3.2. Zeb1os1 impairs cardiac diastolic function in mice
We hypothesized that if upregulation of Zeb1os1 has deleterious effects on cardiac function, then overexpression of Zeb1os1 could reproduce the phenotypes of aged hearts. An adeno-associated virus serotype 9 (AAV9) vector carrying the Zeb1os1 gene (AAV-
Zeb1os1) was transfected into young wild-type mice (2 months old) for six months. While the negative control (AAV-NC) did not affect the Zeb1os1 level, AAV-Zeb1os1 significantly increased the expression of Zeb1os1 in mouse hearts (Figs. 2(a); Fig. S4 in Appendix A). Brain natriuretic peptide (BNP) is a biomarker of heart aging and has been identified as a diagnostic indicator for diastolic dysfunction [
33]. Our results revealed that AAV-Zeb1os1 markedly elevated the BNP level in mouse serum (
Fig. 2(b)). Moreover, cardiac function was assessed using echocardiography, which showed that AAV-Zeb1os1 led to a notable decrease in the E/A ratio and MV E peak velocity and increased the IVRT (Figs. 2(c−f)), suggesting the deterioration of LV diastolic function in these mice. However, systolic function parameters such as the ejection fraction (EF) and FS showed no significant alterations following AAV-Zeb1os1 treatment (
Figs. 2(g) and (h)). Diastolic dysfunction is one of the causes of LV hypertrophy [
34]. Consistently, parameters related to hypertrophy including LVPWD and LV mass were significantly increased in AAV-Zeb1os1 treated mice (
Figs. 2(i) and (j)). Moreover, cardiac hypertrophy was confirmed by H&E staining of enlarged cross-sectional cardiomyocyte areas (Figs. 2(k); Fig. S5 in Appendix A). The Masson staining results provided further evidence that AAV-Zeb1os1 induced cardiac fibrosis, a condition commonly observed in aged hearts (Figs. 2(l); Fig. S5 in Appendix A). The above results revealed that AAV-Zeb1os1 induced the typical phenotypes of an aged heart. To further validate the aging of the heart, the SA-β-gal staining was used for detection. The results showed a marked increase of the heart with SA-β-gal (
Fig. 2(m)). In addition, the increased levels of senescence markers p53 and p21 also indicated that cardiac senescence was induced by AAV-Zeb1os1 (
Figs. 2(n) and (o)). Collectively, these results reveal that Zeb1os1 induces heart aging as indicated by impaired diastolic function, cardiac hypertrophy and cardiac tissue senescence.
3.3. Knockdown of Zeb1os1 attenuates heart dysfunction in naturally aged mice
To investigate whether Zeb1os1 could serve as a potential therapeutic target for heart aging, we employed a loss-of-function strategy with an AAV9 vector carrying a Zeb1os1-shRNA fragment (AAV-shZeb1os1) and tested whether Zeb1os1 knockdown could ameliorate heart dysfunction in naturally aged mice. The knockdown effect of AAV-shZeb1os1 on endogenous Zeb1os1 in mice was validated using qRT-PCR and immunofluorescence, which revealed a significant decrease in the Zeb1os1 transcript level in the myocardium (
Fig. 3(a); Fig. S4). The elevated BNP expression level, which indicates cardiac damage in aged mice, was effectively decreased by AAV-shZeb1os1 (
Fig. 3(b)). Cardiac function was then assessed in different groups of mice. Aged mice presented pronounced LV diastolic dysfunction measured by a decreased E/A ratio and MV E peak velocity, as well as increased IVRT, whereas AAV-shZeb1os1 markedly restored these parameters (Figs. 3(c−f)). Neither the EF or FS showed significant alterations in aged mice treated with AAV-shZeb1os1 (
Figs. 3(g) and (h)). Moreover, compared with their young control counterparts, aged mice presented severe cardiac hypertrophy as shown by increased LVPWD and LV mass, which was ameliorated by AAV-shZeb1os1 (
Figs. 3(i) and (j)). HE and Masson staining further confirmed that AAV-shZeb1os1 ameliorated cardiac remodeling (
Figs. 3(k) and (l); Fig. S5). Moreover, SA-β-gal staining revealed severe cardiac senescence in aged mice and this deterioration was diminished by AAV-shZeb1os1 (
Fig. 3(m)). Furthermore, the aging-induced upregulation of senescence markers p53 and p21 was significantly mitigated by AAV-shZeb1os1 (
Figs. 3(n) and (o)). These results demonstrate that knockdown of Zeb1os1 is effective in ameliorating heart aging.
3.4. Zeb1os1 determines cellular senescence in cardiomyocytes
We next sought to investigate the impact of Zeb1os1 on cardiomyocyte senescence by using exogenous transfection of a Zeb1os1 plasmid (OE-Zeb1os1) for Zeb1os1 overexpression (
Fig. 4(a)) and well as siZeb1os1 for Zeb1os1 knockdown (
Fig. 4(b); Fig. S6 in Appendix A). SA-β-gal staining revealed that OE-Zeb1os1 markedly increased the number of senescent cells (
Figs. 4(c) and (e)). However, knockdown of siZeb1os1 significantly diminished SA-β-gal positive cells (
Figs. 4(d) and (f)). ROS are known to play a role in driving cellular senescence [
35]. Our results demonstrated that Zeb1os1 overexpression induced cardiomyocyte senescence, as evidenced by a significant increase in ROS production. Conversely, treatment with siZeb1os1 significantly reduced ROS accumulation in senescent cardiomyocytes (Figs. 4(g)−(j)). To further investigate the role of Zeb1os1 in cardiomyocyte function, we assessed changes in contraction and relaxation functions. We conducted sarcomere shortening and beats per minute (BPM) measurements in cardiomyocytes with Zeb1os1 overexpression or knockdown. The results revealed that treatment with D-gal or overexpression of Zeb1os1 significantly impaired cardiomyocyte contraction and relaxation, as well as decreased BPM. Conversely, Zeb1os1 knockdown mitigated the detrimental effects induced by D-gal (Fig. S7 in Appendix A). Senescence cause cell cycle arrest; thus, senescent cells were evaluated by flow cytometry analysis [
36]. The results indicated that Zeb1os1 overexpression generated phenotypes typical of senescent cardiomyocytes, with an increase in G0/G1 phase cells and a reduction in G2/M phase cells. In contrast, knocking down Zeb1os1 with siZeb1os1 markedly reversed these effects (Figs. 4(k−n)). Moreover, the expression of the senescence-associated secretory phenotype (SASP) was detected and observed to be markedly increased in Zeb1os1 and D-gal treated cardiomyocytes, while it was inhibited by siZeb1os1 (
Figs. 4(o) and (p)).
3.5. Zeb1os1 modulates lysosomal function during heart aging
To elucidate the possible mechanisms by which Zeb1os1 regulates heart aging, we performed Gene Ontology (GO) enrichment analysis to identify dysregulated molecular processes in aged hearts based on the differentially expressed transcripts revealed by our RNA-seq results. Several enriched biological processes and cellular components were related to lysosomes as marked by red arrows (
Fig. 5(a)). For further confirmation, transmission electron microscopy (TEM) was used to examine microscopic morphological alterations in heart tissue that are typically associated with lysosome function. As expected, the results revealed decreased numbers of lysosomes, autolysosomes, or phagosomes, and damaged mitochondria in the heart tissue of the AAV-Zeb1os1-treated mice (
Fig. 5(b)). Given that lysosome-associated membrane protein 1 (Lamp1) is a major component of the lysosomal membrane and serves as a marker for lysosome quantity and integrity, we sought to determine Lamp1 expression in the cardiac tissue of mice across various groups [
37]. The immunofluorescent results showed a reduced localization of Lamp1 as puncta near the cell periphery within the cytoplasm in the hearts of mice treated with AAV-Zeb1os1. Consistently its protein level was also decreased (
Figs. 5(c) and (d)). Autophagy depends on the lysosomal bulk degradation mechanism. We further explored the changes in autophagy, Western blot results revealed increased levels of LC3 II and p62, an adaptor protein for autophagosomes, indicating autophagosome accumulation and inefficient autophagosome-lysosome fusion in the AAV-Zeb1os1 treated group (
Fig. 5(d)). Next, we further validated cardiac lysosomal function in naturally aged mice with or without Zeb1os1 shRNA administration. Obvious lysosomal dysfunction, such as decreased lysosome numbers, swollen mitochondria and ruptured sarcomeres, was observed in the aged mice hearts, similar to those in AAV-Zeb1os1 overexpressed mice, and this dysfunction was markedly attenuated by AAV-shZeb1os1 (
Fig. 5(e)). Moreover, Lamp1 was decreased in aged hearts, as immunofluorescence staining and western blotting shown (
Figs. 5(f) and (g)). Consistent with previously published studies [
17], [
38], blockade of autophagosome and autolysosome formation resulted in decreased LC3II and increased p62 levels in aged hearts, and the deterioration of lysosomal function was significantly reversed by AAV-shZeb1os1 (
Fig. 5(g)).
3.6. Zeb1os1 directly binds to the TRPML1 protein
To further identify the downstream regulatory targets of Zeb1os1 in heart aging, we analyzed dysregulated genes within the top two enriched BPs (endosomal transport and autophagy) and the top two enriched CCs (late endosomes and lysosomal membrane) related to lysosomes. The overlap among these genes is displayed in the Venn diagram in
Fig. 6(a), and three genes (
Clec16a,
MCOLN1, and
Ubxn6) overlapped. Among them, TRPML1 (Gene name:
MCOLN1) is a Ca
2+-conducting channel on the lysosomal membrane that promotes the release of Ca
2+ into the cytosol, regulates lysosome biogenesis, and promotes autolysosome formation [
39]. We investigated whether TRPML1 expression is altered in aged hearts and the effects of Zeb1os1 on lysosomal function. The expression of MCOLN1 protein (TRPML1) was significantly decreased in the hearts of normal mice overexpressing Zeb1os1 (
Figs. 6(b) and (c)). Similarly, the expression of TRPML1 was prominently lower in old mice than in young animals, and AAV-shZeb1os1 restored the level of TRPML1 (
Figs. 6(d) and (e)). Moreover, we assessed the expression of TRPML1 at the cellular level. Overexpression of Zeb1os1 significantly suppressed the protein level of TRPML1, and the siZeb1os1 markedly increased its level specifically upon D-gal treatment, while its mRNA level remained unaffected by Zeb1os1 (
Figs. 6(f) and (g); Fig. S8 in Appendix A). LncRNAs often exert their functions through interactions with certain functional proteins; thus, we tested whether Zeb1os1 could bind to the TRPML1 protein to exert its regulatory effect [
18], [
40], [
41]. To gain insight into this issue, we used the RNA-Protein Interaction Prediction (RPISeq) database to analyze potential protein binding. Our analysis identified a high likelihood of interaction between Zeb1os1 and TRPML1 (Fig. S9 in Appendix A). Next, we performed RIP to confirm this interaction. As shown in
Fig. 6(h), TRPML1 immunoprecipitation (IP) captured a significant amount of Zeb1os1, indicating a strong binding affinity between Zeb1os1 and TRPML1. This finding was further supported by Zeb1os1 pulling down TRPML1 in NMCMs (
Fig. 6(i)). To elucidate the underlying mechanisms of TRPML1 protein reduction, we examined various protein degradation pathways. Our results revealed that MG132, a proteasome inhibitor, could prevent the downregulation of TRPML1 induced by Zeb1os1, whereas the lysosome inhibitor chloroquine did not have a similar effect (
Fig. 6(j)). The UB of TRPML1 was subsequently detected, and the level of TRPML1 UB was markedly increased in Zeb1os1-overexpressing NMCMs. These findings suggest that Zeb1os1 facilitates the UB of TRPML1, leading to its degradation via the proteasome pathway (
Fig. 6(k)).
Since the potential role of TRPML1 has not yet been addressed, we further investigated whether TRPML1 was effective in ameliorating cellular senescence by improving lysosomal function. TRPML1 was downregulated in NMCMs following treatment with D-gal, and transfection of plasmid for TRPML1 overexpression rescued the TRPML1 protein expression to the normal level (Fig. S10(a) in Appendix A). As expected, TRPML1 overexpression increased Lapm1 levels in D-gal pretreated cells (Fig. S10(b) in Appendix A). In addition, autophagy was improved by TRPML1, showed by increased LC3 II and decreased p62 protein levels (Figs. S10(c) and (d) in Appendix A). The levels of p53 and p21 were suppressed by TRPML1 in D-gal-pretreated cells (Figs. S10(e) and (f)). SA-β-gal staining consistently revealed that the increase in the percentage of senescent cells was abolished by TRPML1 overexpression (Fig. S10(g) in Appendix A). These results confirm that TRPML1 is an effective therapeutic target for improving lysosomal function and eliminating senescent cells.
3.7. Zeb1os1 regulates lysosomal function by targeting TRPML1 in cardiomyocytes
We then sought to confirm the impact of Zeb1os1 on lysosomal function by assessing lysosome biogenesis, autolysosome formation, and lysosomal enzyme activity. Additionally, we assessed how Zeb1os1 affects TRPML1-mediated lysosomal Ca
2+ release. The TRPML1 activator ML-SA1 induced lysosomal Ca
2+ release, but the amplitude of the Ca
2+ current was decreased in the OE-Zeb1os1 and D-gal pretreatment groups. Moreover, siZeb1os1 transfection restored the level of Ca
2+ release (
Figs. 7(a) and (b)). Nuclear translocation of TFEB, a crucial regulator of lysosome biogenesis and autophagy, is driven by lysosomal Ca
2+ release [
42]. Immunofluorescence staining and western blot results revealed that OE-Zeb1os1 decreased TFEB nuclear translocation and that siZeb1os1 markedly promoted TFEB nuclear translocation in senescent cardiomyocytes (
Figs. 7(c)and (d) ; Fig. S11 in Appendix A). Lysosome biogenesis was identified by immunofluorescence staining for Lamp1 and LysoTracker. Consistent with the
in vivo results, OE-Zeb1os1 impaired the lysosomal function, as manifested by the markedly decreased fluorescence intensity of Lamp1 and LysoTracker in cardiomyocytes (
Figs. 7(e) and (f)). Furthermore, lysosome-mediated autophagy was also assessed. Using tandem RFP-GFP-LC3 adenovirus transfected of NMCMs, OE-Zeb1os1 led to an increase in yellow (autophagosome) and a decrease in red (autolysosome) fluorescent signals, which demonstrated the blockage of autophagosome-lysosome fusion (
Fig. 7(g)). In D-gal induced senescent cardiomyocytes, the amount of lysosome was remarkably decreased but was increased by siZeb1os1 (
Figs. 7(h) and (i)). Moreover, D-gal exposure decreased both yellow (autophagosome) and red (autolysosome) signals, suggesting inhibition of autophagic flux (
Fig. 7(j)). These deleterious changes were reversed by the knockdown of Zeb1os1 (
Fig. 7(j)). Additionally, we utilized Bafilomycin A1 (BafA1) to block autophagosome-lysosome fusion and further investigated how Zeb1os1 influences the autophagic process. As depicted in Fig. S12 in Appendix A, BafA1 treatment resulted in a decrease in autolysosome formation and accumulation of autophagosomes. Knockdown of Zeb1os1 remarkably reversed the blockage of autophagic flux caused by BafA1 in both control and D-gal treated NMCMs. In addition, lysosomal activity was quantified by determining the activity of CTSB and CTSD, key lysosomal enzymes. The results suggested that OE-Zeb1os1 remarkably inhibited CTSB and CTSD activities, whereas siZeb1os1 reversed the deterioration of CTSB and CTSD activities encountered by D-gal (Figs. 7(k)−(n)). An increase in lysosomal pH is associated with lysosomal dysfunction, including impaired autophagosome-lysosome fusion, lysosomal biogenesis, and lysosomal enzyme activities [
43]. Consistent with these findings, we detected a significant increase in lysosomal pH upon OE-Zeb1os1 and D-gal treatment, whereas Zeb1os1 knockdown restored the pH in D-gal treated NMCMs (
Figs. 7(o) and (p)).
3.8. Zeb1os1 functional domain as a key player in the modulation of TRPML1
Through sequence alignment, we identified a highly conserved region of the Zeb1os1 sequence across species (Fig. S13 in Appendix A). We hypothesized that this FD of Zeb1os1 interacts with the TRPML1 protein. Computational docking using Hex 8.0 and PyMOL supported our hypothesis by demonstrating the binding potential of the FD to TRPML1. Specifically, the results revealed that a stretch of the TRPML1 amino acid chain, encompassing ARG-172, ASP-183, HIS-286, and ASP-176, may serve as the core binding residues for Zeb1os1-FD (
Figs. 8(a) and (b)). We synthesized a TRPML1 gene plasmid with mutations in the aforementioned four amino acids (Mut-TRPML1) and examined the effects of Zeb1os1 on the UB of wild-type TRPML1 (WT-TRPML1) and Mut-TRPML1. Compared with the WT-TRPML1 group, the Zeb1os1 group lost its ability to promote the UB of Mut-TRPML1, highlighting the crucial roles of the core amino acids in mediating the interaction between TRPML1 and Zeb1os1 (Fig. S14 in Appendix A). The C-62, G-66, G-79, and C-81 nucleotides of FD were responsible for the interaction with the TRPML1 protein through the formation of hydrogen bonds. To test this observation and gain insight into the molecular mechanisms of Zeb1os1, we synthesized an oligonucleotide fragment corresponding to Zeb1os1-FD. We introduced nucleotide substitutions to mutate the core motif of the FD to interfere with its ability to bind TRPML1, creating Mut-FD (
Fig. 8(a)). Zeb1os1-FD was transfected into NMCMs to assess its interaction with TRPML1 and its impact on lysosomal function and cellular senescence. As shown in Figs. 8(c−f), FD replicated the impact of the complete Zeb1os1 on TRPML1 expression, lysosome biogenesis and cellular senescence in NMCMs, as well as downregulation of the TRPML1 and Lamp1 proteins (
Figs. 8(c) and (d)) and an increase in SA-β-gal activity (
Figs. 8(e) and (f)). In contrast, Mut-FD did not alter TRPML1 protein expression and lost its ability to affect lysosomal function and cellular senescence (Figs. 8(c−f)). Taken together, these data identify the key functional site of Zeb1os1 and suggest that Zeb1os1-FD may be a potential tool for slowing and alleviating heart aging.
4. Discussion
Our study elucidated the pathophysiological effects and the underlying mechanisms of Zeb1os1 in heart aging. Zeb1os1 was robustly increased in aged mice hearts and in the blood of elderly individuals. Specifically, upregulation of Zeb1os1 negatively impacts on heart function. The finding is supported by the observation that knockdown of endogenous Zeb1os1 partially abrogated aging-induced diastolic dysfunction and cardiac remodeling. Zeb1os1 was artificially overexpressed in otherwise normal mice resulted in cardiac function impairments similar to those observed in aged mice. Furthermore, Zeb1os1 deleteriously altered the function of lysosomes, leading to reduced autophagy and facilitating cardiomyocyte senescence at the subcellular level. On a molecular level, Zeb1os1 binds directly to TRPML1, downregulating its expression and thereby limiting the activity of this key lysosomal cation channel. Another important finding is the identification of the Zeb1os1 function domain (Zeb1os1-FD), which is partially responsible for the effects of Zeb1os1 on TRPML1 and cardiomyocyte senescence. These findings led us to propose a regulatory pathway by which Zeb1os1 modulates cardiac function in heart aging: aging →Zeb1os1↑→TRPML1↓→lysosomal Ca
2+ release↓ →lysosomal function↓→cardiomyocyte senescence↑→heart aging↑, as illustrated in
Fig. 8. On the basis of these findings, we demonstrated that lncRNA Zeb1os1 that contributes to heart aging through targeting of TRPML1 and its associated lysosomal function.
There is growing evidence that ncRNAs play a critical role in CVDs, which enhances their potential as biomarkers and novel regulators of aging-related cardiac dysfunctions [
44]. For example, Zhu et al. [
45] demonstrated that lncRNA MALAT1, delivered via stem cell-derived exosomes, can protect against age-related cardiac dysfunction by reducing inflammation. Yang et al. [
46] revealed that the lncRNA ENSMUST00000134285 increased MAPK activity and regulated aging-related myocardial apoptosis. Li et al. [
47] demonstrated that the lncRNA CPAL regulates cardiac metabolic and cardiomyocyte pyroptosis, suggesting CPAL as a potential target for protecting cardiomyocytes from ischemic injury. Our previous study found the lncRNA LOC105378097 was increased in the blood of elderly individuals and could induce heart aging by targeting Parkin mediated mitophagy [
54]. Additionally, we discovered that microRNA-203 levels are reduced in the blood of elderly individuals and are significantly correlated with the E/A ratio in these patients [
48]. In addition to ncRNAs, Wu et al. [
49] created the immunoglobulin G (IgG) N-glycosylation cardiovascular age (GlyCage) index using IgG
N-glycan profiles combined with machine learning to track cardiovascular health and predict cardiovascular risk in elderly individuals. However, the understanding of lncRNAs in cardiac aging is still in its early stages, and additional research is necessary before therapeutic options can be proposed. Future studies should focus on uncovering the molecular mechanisms involved and identify more specific and sensitive biomarkers for screening functional lncRNAs in different pathological contexts. Therefore, in this study, we first performed RNA-seq on RNA samples isolated from young and old mouse hearts and Zeb1os1 expression was significantly increased in hearts from aging mice. Furthermore, we discovered a significant increase of human Zeb1os1 in blood samples obtained from older individuals compared with those obtained from younger individuals. This increase was positively correlated with age. Interestingly, we found that the E/A ratio declined with age in participants who had a normal left ventricular ejection fraction (LVEF), suggesting a mild form of heart failure with preserved ejection fraction (HFpEF) [
50]. The inverse correlation between Zeb1os1 expression and the E/A ratio further supports its viability as a blood-based indicator of heart aging.
In healthy individuals, aging leads to an increase in the incidence of LV hypertrophy, a decline in LV diastolic function, and the preservation of LVEF and LVFS. In the present study, similar pathological characteristics were observed in naturally aged mice, which were alleviated after Zeb1os1 knockdown. Conversely, Zeb1os1 overexpression induced cardiac dysfunction in a manner similar to that observed in naturally aged mice. These results confirmed the important regulatory role of Zeb1os1 in heart aging. In recent years, research has increasingly concentrated on identifying functional regulators involved in heart aging, leading to the discovery of several new genes. For instance, Ye et al. [
51] found that Sirtuin 2 levels were diminished in aged monkeys and mice hearts, leading to hyperacetylation of STAT3. This modification transcriptionally activates CDKN2B, subsequently initiating mitochondrial damage and cardiomyocyte degeneration. FOXP1, identified as a critically downregulated factor in senescent cardiomyocytes, contributes to hypertrophic and senescent phenotypes that accelerate heart aging [
52]. Additionally, reduced expression of the insulin-like growth factor 1 receptor (IGF1R) in mouse hearts was linked to impaired heart function early in life but showed better cardiac outcomes and extended lifespan during aging. IGF1R activation is associated with suppressed autophagic flux, a finding that is consistent with the continuous deterioration in autophagy observed [
53]. The study of lncRNAs in heart aging remains in its infancy and merits further exploration. Ding et al. [
17] highlighted circHIPK3 as a crucial regulator of heart aging, primarily through its role in enhancing the degradation of HuR and inhibiting p21 activity. Deletion of circHIPK3 resulted in exacerbated cardiomyocyte senescence and diminished cardiac function. Additionally, our previously published study demonstrated that the overexpression of the lncRNA LOC105378097 led to mitochondrial dysfunction and cardiomyocyte senescence in aged hearts [
54]. In the present study, knockdown of Zeb1os1 using siRNA effectively ameliorated cardiomyocyte senescence induced by D-gal treatment at the cellular level. In vivo experiments revealed that shZeb1os1 administration diminished senescence in cardiac tissue from aged mice and ameliorated the disorganization of muscle fibers and cardiac remodeling. These results suggest that shZeb1os1 or siZeb1os1 may be novel candidates for an anti-senescence approach for heart aging.
At the subcellular level, further investigations based on RNA-seq data revealed that lysosome dysfunction, including a reduced lysosome number and impaired autolysosome formation, is a pathological hallmark of senescent cardiomyocytes. Consistent with our results, several earlier studies have confirmed that lysosomal dysfunction can drive aging-related cardiac dysfunction by increasing oxidative stress and inflammation [
55], [
56]. In addition, therapies targeting the preservation of lysosomal function have been found to be effective in treating aging-related neurodegenerative diseases [
57]. Furthermore, functional regulators related to the lysosome have been identified as important in heart aging. For example, mTORC1, a nutrient sensor activated on the lysosomal membrane, plays a key role in heart aging by regulating autophagy, cellular senescence, protein synthesis, mitochondrial function, and inflammation. Our study consistently confirmed that the amelioration of lysosomal function effectively reversed aging induced cardiac dysfunction and cardiomyocyte senescence. In addition, AAV-Zeb1os1 and OE-Zeb1os1 could be used as aging or senescence drivers by impairing lysosomal function.
The mechanism by which Zeb1os1 regulates lysosomal function was previously unclear. Then, we investigated the downstream targets of Zeb1os1. Through RNA-seq analysis, TRPML1 was found to be dysregulated in aged hearts, and it was shown to be regulated by Zeb1os1. Lysosomal calcium efflux mediated by TRPML1 is essential for lysosomal biogenesis and the fusion of late-endosomes and autophagosomes with lysosomes [
58]. Calcium signaling regulates TFEB activity through the phosphatase calcineurin, and TFEB nuclear translocation promotes the expression of genes related to autophagy and lysosome biogenesis [
59]. The overexpression of TFEB could protect against proteotoxicity in cardiomyocytes by improving the autophagy−lysosomal pathway and TRPMIL1 via TFEB has been shown to be required for proteasome malfunction to activate autophagy in cardiomyocytes [
60]. Loss of TRPML1 results in a reduced lysosome number and blocked autophagic flux and lysosome degradation, contributing to cell damage and death, which can induce multiple diseases as kidney injury and neurodegenerative disorders [
61], [
62]. The present study is the first to confirm a functional regulatory role of TRPML1 in heart aging and cardiomyocytes senescence. D-gal significantly induced TRPML1 downregulation, and overexpression of TRPML1 effectively improved cardiomyocyte senescence by restoring lysosomal function. Moreover, restoring TRPML1 expression levels, Ca
2+ flux and TFEB nuclear translocation toward normal levels via shZeb1os1 markedly ameliorated the lysosomal dysfunction and cellular senescence. Notably, abnormal TRPML1 activation could robustly increase TRPML1-mediated Ca
2+ release and subsequently impair lysosomal trafficking and biological functions [
63]. However, the specific effects of shZeb1os1 on TRPML1 expression did not result in the side effects mentioned above. Moreover, recent studies have revealed that in ischemia injury and cancer, overexpression of TRPML1 results in the release of a large amount of Zn
2+ from lysosomes, leading to autophagy blockage, ROS accumulation, and cell apoptosis. Restoring TRPML1 expression to normal levels can eliminate these pathological effects [
64], [
65]. These earlier studies, together with the present study, demonstrate that both too little or too much expression of TRPML1 can cause cell damage. Restoring TRPML1 expression to normal levels to maintain the balance of ion flux is an effective strategy for treating diseases related to TRPML1 dysregulation.
The molecular mechanisms of lncRNAs are diverse, including the regulation of epigenetic and (post) transcriptional gene expression, miRNA sponging, and compartmentalization. These roles are accomplished when lncRNAs bind to RNA-binding proteins [
66]. Our earlier research demonstrated that the lncRNAs ZFAS1 can inhibit protein function through direct binding with their respective target proteins via the functional domains of lncRNAs [
18]. Consistent with the above studies, here we found that Zeb1os1 interacted with the TRPML1 protein as predicated by computational docking analysis, and was verified by RIP and pulldown assays. The downregulation of the TRPML1 protein by Zeb1os1 could be attributed to alterations in protein conformation. These changes subsequently lead to excessive degradation of the TRPML1 protein through the UB−proteasome pathway, similar to what has been reported in other studies [
67].
Given the complexity of lncRNA-mediated gene regulation, ZEB1os1 may have multiple targets and mechanisms of action. Lv et al. [
68] found that downregulation of ZEB1-AS1 weakens Wnt/β-catenin signaling and induces apoptosis in colorectal cancer. ZEB1-AS1 also functions as a molecular sponge, specifically for miR-141-3p, leading to suppressed cancer cell proliferation upon ZEB1-AS1 knockdown [
69]. Similar oncogenic roles of ZEB1-AS1 have been observed in hepatocellular carcinoma (HCC), where it sequesters miR-299-3p, increasing E2F1 levels and contributing to HCC cell malignancy [
70]. Additionally, Su et al. [
71] demonstrated that ZEB1-AS1 bound and recruited the histone methyltransferase MLL1, activated ZEB1 transcription, subsequently promoting prostate cancer cells proliferation and migration. The data obtained in our study represent the first reliable evidence demonstrating the role of Zeb1os1 in heart aging. By employing bioinformatic analysis, molecular biological techniques, and gain- or loss-of-function strategies, we elucidated that Zeb1os1 functions as a direct suppressor of TRPML1. These findings pave the way for exciting future research into additional therapeutic targets associated with Zeb1os1 in the context of heart aging.
In summary, our study revealed that Zeb1os1 plays an essential role in regulating aging cardiac function. Zeb1os1 impairs lysosome biogenesis and autophagy by directly targeting and decreasing TRPML1 protein levels. These effects lead to the senescence of cardiomyocytes, resulting in heart aging. Moreover, our data confirmed that knockdown of Zeb1os1 reverses cardiomyocyte senescence and improves cardiac function in older mice, indicating the potential therapeutic effects of targeting Zeb1os1 and TRPML1 in the management of heart aging.
Acknowledgments
This study was funded by the National Natural Science Foundation of China (82273919, 82270396, and U21A20339) and the China Postdoctoral Science Foundation (2023T160176).
Compliance with ethics guidelines
Heng Liu, Haiying Zhang, Han Lou, Jennifer Wang, Shengxin Hao, Hui Chen, Chen Chen, Lei Wang, Huimin Li, Ziyu Meng, Wenjie Zhao, Tong Zhao, Yuan Lin, Zhimin Du, Xin Liu, Baofeng Yang, and Yong Zhang declare that they have no conflict of interest or financial conflicts to disclose.
PDF
(7088KB)
