Abstract Myocardial energy metabolism disorders are essential pathophysiology in sepsis-associated myocardial injury. Yet, the underlying mechanisms involving impaired mitochondrial respiratory function upon myocardial injury remain poorly understood. Here we identify an unannotated and cardiomyocyte-enriched long non-coding RNA, Cpat (cardiac-protector-associated transcript), that plays an important role in regulating the dynamics of cardiomyocyte mitochondrial tricarboxylic acid (TCA) cycle. Cpat is essential to the mitochondrial respiratory function by targeting key metabolic enzymes and modulating TCA cycle flux. Specifically, Cpat enhances the association of TCA cycle core components malate dehydrogenase (MDH2), citrate synthase (CS), and aconitase (ACO2). Acetyltransferase general control non-repressed protein-5 (GCN5) acetylates CS and destabilizes the MDH2-CS-ACO2 complex formation. Cpat inhibits this GCN5 activity and facilitates MDH2-CS-ACO2 complex formation and TCA cycle flux. We reveal that Cpat-mediated mitochondrial metabolic homeostasis is vital in mitigating myocardial injury in sepsis-induced cardiomyopathy, positioning Cpat as a promising therapeutic target for preserving myocardial cellular metabolism and function. Subject terms: Acetylation, Cardiomyopathies, Long non-coding RNAs __________________________________________________________________ Mitochondrial dysfunction contributes to septic cardiomyopathy and poor outcomes. Here, the authors identify a cardiomyocyte-enriched non-coding RNA that preserves mitochondrial function and reduces mortality in septic mice by preventing citrate synthase acetylation. Introduction Septic cardiomyopathy (SCM) presents a critical challenge in sepsis management, marked by impaired cardiac function and high mortality rates^[48]1–[49]4. This condition is characterized by both structural and functional myocardial alterations, alongside profound disturbances in mitochondrial metabolism^[50]5–[51]8. The energy requirements of the myocardium, primarily fulfilled by fatty acid oxidation and, to a lesser extent, glucose and lactate, become compromised during sepsis^[52]9. Mitochondrial dysfunction leads to reduced ATP production and altered metabolic pathways, contributing to progressive cardiac failure^[53]10–[54]14. Understanding the underlying mechanisms of mitochondrial metabolic dysfunction in cardiovascular injury such as SCM is essential for developing effective therapeutic strategies. Lysine acetylation involves the addition of an acetyl group from acetyl coenzyme A (CoA) to the ɛ-amino group of the lysine residue, thereby effectively neutralizing the positive charges on lysine residue^[55]15–[56]18. Initially recognized for its role in histone modification and chromatin remodeling, this post-translational modification is now understood to influence a wide array of extranuclear proteins^[57]19–[58]22. In the context of mitochondrial metabolism, lysine acetylation critically regulates key processes such as ATP production, oxidative stress response, and overall metabolic flux^[59]15,[60]23,[61]24. Recent proteomic studies highlight the impact of mitochondrial enzyme acetylation on metabolic pathways^[62]19,[63]25. Acetyltransferases play crucial roles in regulating mitochondrial metabolism by adding acetyl groups to the lysine residues on target proteins^[64]19,[65]25,[66]26. These enzymes include the well-characterized histone acetyltransferases (HATs), E1A binding protein P300 (EP300)/cyclic AMP response element binding protein (CREB)-binding protein (P300/CBP), and lysine acetyltransferase 2 A/general control non-repressed protein-5 (KAT2A/GCN5) families that influence various mitochondria metabolic processes^[67]20,[68]25,[69]27–[70]29. For instance, P300/CBP acetyltransferases acetylate key mitochondrial enzymes that are involved in the tricarboxylic acid (TCA) cycle and oxidative phosphorylation (OXPHOS)^[71]30. Such post-translational acetylation directly affects enzyme activity, stability, and interactions with other metabolic regulators^[72]31–[73]33. Therefore, dynamic regulation of mitochondrial acetylation by these acetyltransferases highlights the importance of these acetyltransferases and their substrates in maintaining cellular energy metabolism and homeostasis. Long non-coding RNAs (lncRNAs) have emerged as key regulators of mitochondrial metabolism, influencing various aspects of mitochondrial function and cellular energy balance^[74]34–[75]36. Recent studies have shown that lncRNAs modulate mitochondrial OXPHOS, mitochondrial fatty acid β-oxidation (FAO), and the TCA cycle^[76]37–[77]39. For example, lncRNA homeobox A11 (HOXA11os) interacts with a core subunit of complex I of the electron transport chain (ETC) and promotes complex I activity in colonic myeloid cells as a mechanism to increase OXPHOS and decrease the production of mitochondrial reactive oxygen species (mtROS)^[78]40. LncRNA H19 decreases mtROS production and enhances mitochondrial ATP level by decreasing the expression of voltage-dependent anion channel-1(VDAC1)^[79]41. These lncRNAs exert their effects by interacting with mitochondrial proteins or by influencing their expressions. However, it remains unknown whether lncRNAs can impact mitochondrial function through direct regulation of mitochondrial protein acetylation in cardiovascular injury, such as SCM. In this study, we identified a cardiomyocyte-enriched lncRNA cardiac-protector-associated transcript (Cpat) in the septic mouse heart. Cardiomyocyte-specific overexpression of Cpat in septic mice preserved mitochondrial function and lowered sepsis-induced mortality. Mechanistically, human antigen R (HuR) and chromosome maintenance protein (CRM1) facilitated Cpat translocation from nucleus to cytoplasm and then into mitochondria with the assistance of polyribonucleotide nucleotidyltransferase 1 (Pnpt1). Cpat prevented citrate synthase (CS) acetylation from GCN5 and increased the assembly of TCA cycle core complex aconitase 2 (ACO2)-CS-malate dehydrogenase 2 (MDH2), thereby maintaining the mitochondrial OXPHOS capacity and respiratory function and ameliorating myocardial mitochondrial dysfunction from cardiac septic insult. Results Cpat is a cardiomyocyte-enriched lncRNA downregulated in septic heart To identify essential lncRNAs that are involved in the pathologic process of sepsis, we performed transcriptome sequencing in hearts of the left ventricle from male mice with CLP (cecal ligation and puncture)-induced sepsis. Transcriptome analysis identified downregulation of 104 lncRNAs and upregulation of 105 lncRNAs in CLP mice (Fig. [80]1a). We selected 3 genes (3 genes downregulated) with the highest fold changes and validated by RT-qPCR in hearts from CLP mice (Fig. [81]1b) and in another independent septic mouse model induced with lipopolysaccharide (LPS) (Fig. [82]1c). Notably, the expression of XR4938650 was decreased significantly in both models (Fig. [83]1b, c). Based on the potential function of XR4938650 and its expression level, we termed it cardiac-protector-associated transcript (Cpat). The expression of Cpat was reduced remarkably over time in hearts from LPS-treated mice (Fig. [84]1d) and in mouse neonatal cardiomyocytes treated with the conditioned medium (CM) from in LPS-treated Raw264.7 cells (Fig. [85]1e). The results from RT-qPCR showed that Cpat expression was specifically enriched in mouse heart but negligibly in other tested tissues, including spleen, lung, kidney, muscle, rectum, brain, and liver (Fig. [86]1f). To determine the cell types that expressed Cpat in mouse heart, we conducted single-cell RNA sequencing (scRNA-Seq) and found that Cpat was predominantly expressed in cardiomyocytes (Fig. [87]1g and Supplementary Fig. [88]1a). Taken together, we identified a previously unannotated cardiomyocyte-enriched lncRNA Cpat that was downregulated in mouse septic heart. Fig. 1. Identification of a cardiomyocyte lncRNA Cpat downregulated in septic mouse heart. [89]Fig. 1 [90]Open in a new tab a Volcano map of differentially expressed genes from transcriptional analysis in hearts of CLP-induced septic mice vs sham mice. Fold change > 1.5. n = 3 biological replicates. b, c RT-qPCR analysis of 3 most significantly differentially expressed lncRNAs from CLP-induced (b) or LPS-induced (c) septic mouse hearts. Gapdh was used as the reference gene. n = 3 biological replicates. d Heart Cpat expression in mice treated with LPS (20 mg/kg, i.p.) for different time periods. n = 3 (24 H) or 5 (12 H) or 6 (0 H) or 7 (6 H) biological replicates. e Cpat expression in cardiomyocyte treated with conditioned medium (LPS 1 µg/ml stimulated Raw264.7 for 24 h) for different time periods. n = 3 biological replicates. f Cpat expression levels in different tissues by RT-qPCR. Gapdh was used as a reference gene. n = 5 biological replicates. g scRNA-Seq of hearts from 3 wild type (WT) mice showed Cpat expression in various heart cell types. h Schematic illustration of Cpat and XR004938650.1 originated from the intergenic region within Myh7 and transcribed into Cpat and XR004938650, and comparison between the sequence structures of Cpat and XR004938650. i CPC predicted the Cpat coding probability. The protein-coding mRNA Gapdh was used as a negative control. LncRNAs Neat1 and Hotair1 were used as positive controls. j Representative images of fluorescence in situ hybridization (FISH) showed localization of Cpat in HL-1 cells. Cytoplasm marker 18S and nuclear marker U6 acts were used. Scale bar =10 µm. Statistical significance was determined using the Wald test (two-sided) implemented in DESeq2, and P values were adjusted for multiple testing using the Benjamini-Hochberg method (a). Data are presented as mean ± SD, unpaired two-sided Student’s t-test (b, c) and one-way analysis of variance (ANOVA) with Tukey’s multiple comparison test (d, e). p values are indicated. Source data are provided as a Source Data file. CLP cecal ligation and puncture, LPS lipopolysaccharide; 0, 6, 12, 24 H: 0, 6, 12, 24 h followed by LPS administration. To identify the full-length Cpat in cardiomyocytes, we isolated mouse neonatal cardiomyocytes and performed 5ʹ and 3ʹ rapid amplification of cDNA ends (RACE). The results showed that the Cpat gene contains 2879 nucleotides that covers 3 exons and a poly (A) tail (Fig. [91]1h and Supplementary Fig. [92]1b). We then determined the conservativity of Cpat using the UCSC Genome Browser (GRCm39/mm10). As shown in Supplementary Fig. [93]1c, the genome locus of Cpat is highly conserved across multiple species, including human, mouse, rat, and rabbit. Human MYH7 loci encoded RNA that resembled Cpat in primary sequence, highly homologous to that of mouse (Supplementary Fig. [94]1d). To further assess its coding potential, we applied several widely used computational prediction tools, including CPC2, CPAT, CNIT, and RNAmining. Results of these tools suggest a low coding potential consistent with its annotation. Based on this, we currently refer to Cpat as a transcript with predicted non-coding properties. The coding potential of Cpat was assessed by Coding Potential Calculator (CPC) to determine whether Cpat is indeed a non-coding RNA. We revealed that Cpat has a low coding-potential score and is a lncRNA (Fig. [95]1i). Using U6 and 18S as nuclear and cytoplasm probes, respectively, we performed fluorescence in situ hybridization (FISH) and identified the distribution of Cpat mostly in HL-1 cardiomyocyte nucleus (Fig. [96]1j). RT-qPCR quantification of nuclear/cytoplasmic RNA in HL-1 cardiomyocyte subcellular fraction extracts showed that the β-actin (Actb) mRNA transcript was enriched in the cytosolic fraction, whereas U6 (a nuclear RNA) and Cpat were enriched in the nuclear fraction (Supplementary Fig. [97]1e). Cpat overexpression ameliorates sepsis-induced cardiac dysfunction To define Cpat function in SCM, we employed AAV9-mediated delivery system to overexpress Cpat in sepsis mouse heart. Cardiac troponin-T (cTNT) promoter was used to achieve cardiomyocyte-specific expression of Cpat (AAV9-Cpat) while empty vector (AAV9-Ctrl) served as control (Fig. [98]2a). RT-qPCR analysis confirmed increased cardiac expression of Cpat 8 weeks after subcutaneous injection of AAV-Cpat in 0-2 days postnatal male mice (Fig. [99]2b). Administration of LPS induced myocardial dysfunction in 6 h and 12 h (Fig. [100]2c, Supplementary Tables [101]1 and [102]2) and dramatic mortality in 48 h (Fig. [103]2d) in AAV9-Ctrl mice. AAV9-Cpat infection-mediated overexpression of cardiac Cpat attenuated LPS-induced cardiac dysfunction (Fig. [104]2c) with increased ejection fractions (EF%), fractional shortening (FS%), and decreased left ventricular end systolic volume (LVESV) (Fig. [105]2e–g). Cpat overexpression markedly improved the survival of LPS septic mice (Fig. [106]2d). Blood cardiac creatine phosphokinase isozymes (CK-MB) and troponin T (cTnT) are commonly used as markers of myocardial injury. Based on preliminary experiments identifying 6 h post-LPS injection as an optimal time point to assess early myocardial injury and inflammatory responses, we collected peripheral blood and heart tissues accordingly. We performed ELISA to measure peripheral blood CK-MB and cTnT levels and found that these markers were significantly lower in LPS-treated AAV9-Cpat mice than those in AAV9-Ctrl mice (Fig. [107]2h, i), supporting a cardioprotective role of Cpat. Cardiac overexpression of Cpat also reduced blood levels of proinflammatory cytokines IL-1β and TNF-α (Fig. [108]2j, k), suggesting a role for Cpat in reducing systemic inflammation. Fig. 2. Cpat overexpression ameliorates cardiac dysfunction in mice with LPS-induced sepsis. [109]Fig. 2 [110]Open in a new tab a Experimental design to test a role for Cpat in ameliorating cardiac dysfunction from LPS-induced sepsis in mice. WT mice at 0-2 days postnatal were infected with adeno-associated virus serovar 9 (AAV9)-Cpat (2x10^11v.g., i.h.). Mice at 8 weeks old were administered with LPS (i.p. 10 mg/kg for cardiac functions and 20 mg/kg for survival test) to induce sepsis (Created in BioRender. Fan, Y. (2025) [111]https://BioRender.com/0pt6yyb). b RT-qPCR quantification of heart Cpat expression in AAV-Ctrl and AAV-Cpat mice. n = 6 biological replicates. c–g Echocardiogram representative M-mode (c), survival curves (d), EF% (e), FS% (f), LVESV (g) in AAV9-Cpat and AAV9-Ctrl mice after LPS administration (i.p. 10 mg/kg for cardiac functions and 20 mg/kg for survival test). n = 6 (echocardiographs) or 10 (survival curve) biological replicates. h-k ELISA quantification of peripheral blood myocardial injury markers CK-MB (h), cTnT (i) and cytokines IL-1β (j), TNF-α (k). n = 6 biological replicates. l Survival curves of Tg (Cpat-Cre) and Tg (Cpat) mice following CLP administration. n = 10 biological replicates. m–o Echocardiography analyses of cardiac functions EF% (m), FS% (n) and LVESV (o) between Tg (Cpat) and Tg (Cpat-Cre) mice after CLP or Sham administration. n = 6-10 biological replicates. p–s ELISA quantification of peripheral blood myocardial injury markers CK-MB (p), cTnT (q) and cytokines IL-1β (r) and TNF-α (s) from Tg (Cpat) mice and Tg (Cpat-Cre) mice after CLP or Sham administration. n = 6 biological replicates. Data are presented as mean ± SD, unpaired two-sided Student’s t-test (b), Kaplan-Meier survival curves were compared using the log-rank (Mantel-Cox) test (d, l). two-way (e–g) or one-way (h–k, m–s) ANOVA with Tukey’s multiple comparison test. p values are indicated. Source data are provided as a Source Data file. WT wild type, Tg transgenic, Cre Myh6-Cre, i.p. intraperitoneally, EF ejection fractions, FS fractional shortening, LVESV left ventricular end systolic volume, CK-MB creatine phosphokinase isozymes, cTnT troponin T, IL-1β interleukin-1β, TNF-α tumor necrosis factor-α. To further establish the cardioprotective and anti-inflammatory roles of Cpat in septic heart, we generated cardiomyocyte-specific Cpat transgenic mice Tg (Cpat-Cre) by breeding the R26-LSL-Cpat mice with Myh6-Cre mice (Supplementary Fig. [112]2a, b). We performed RT-qPCR and confirmed the elevated expression of Cpat in Tg (Cpat-Cre) male mouse ventricle heart tissues, Tg (Cpat) male mice were used as experimental control (Supplementary Fig. [113]2c). Cardiac overexpression of Cpat in Tg (Cpat-Cre) mice significantly improved outcomes following CLP surgery. Specifically, survival was monitored over a 10-day period post-CLP and was markedly increased in Tg (Cpat-Cre) mice compared to Tg (Cpat) controls (Fig. [114]2l). In addition, at 24 h after CLP, Cpat overexpression improved cardiac function with increased ejection fractions (EF%), fractional shortening (FS%), and decreased left ventricular end systolic volume (LVESV) (Fig. [115]2m–o, Supplementary Fig. [116]2d and Supplementary Table [117]3). Furthermore, peripheral blood analysis revealed reduced levels of myocardial injury markers (Fig. [118]2p, q) and a significant decrease in IL-1β (Fig. [119]2r), while TNF-α levels also trended lower, although the difference did not reach statistical significance (Fig. [120]2s), suggesting an overall anti-inflammatory effect of Cpat. We obtained consistent results in the LPS-induced sepsis model. Echocardiographic analysis showed that Cpat overexpression mitigated LPS-induced cardiac dysfunction (Supplementary Fig. [121]2e–g and Supplementary Table [122]4) and reduced myocardial injury markers (CK-MB, cTnT) (Supplementary Fig. [123]2h, i) and proinflammatory cytokine IL-1β (Supplementary Fig. [124]2j) in peripheral blood. Although TNF-α levels also showed a downward trend, the reduction did not reach statistical significance (Supplementary Fig. [125]2k). To further investigate the impact of Cpat on cardiac stress responses, we collected ventricular heart tissues 6 h after intraperitoneal LPS injection for gene expression analysis. RT-qPCR results showed that Cpat overexpression significantly attenuated the LPS-induced upregulation of cardiac stress markers Nppa (Supplementary Fig. [126]2l) and Nppb (Supplementary Fig. [127]2m), whereas the expression of Myh7 (Supplementary Fig. [128]2n) was not significantly altered by LPS stimulation or by Cpat overexpression. Cpat interacts with mitochondria-associated protein Pnpt1 through loops 3 and 4 To investigate the mechanism underlying the cardioprotective and anti-inflammatory functions of Cpat on septic cardiac injury, we sought to identify its targets. HL-1 cardiomyocytes were subjected to comprehensive identification of RNA-binding proteins by mass spectrometry. These pull-downs were then analyzed by MS to identify Cpat interacting proteins (Fig. [129]3a). Use of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis found that Cpat interacted with proteins that are involved in multiple pathways, among which the “metabolic pathways” was on the top of the list (Fig. [130]3b). “Protein translocation” was identified as the top-ranked process from the Gene Ontology (GO) biological process analysis (Fig. [131]3c). Mitochondria are vital organelles for “metabolic pathways”^[132]42. We hypothesized a role for Cpat in mitochondrial function upon septic injury. To test this hypothesis, we treated AAV9-Cpat and AAV9-ctrl control mice with LPS or PBS. We isolated mouse heart mitochondria and assessed mitochondrial respiration function by measuring the oxygen consumption rate (OCR). Using ADP and cytochrome C, we measured OCR for complex I (addition of pyruvate, malate and glutamate (PMG) followed by rotenone, an inhibitor of complex I), complex II (succinate), and complex IV [addition of N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) and ascorbate, followed by sodium azide, an inhibitor of complex IV]. Cpat overexpression significantly promoted the complex I OXPHOS capacity (Supplementary Fig. [133]3a), enhanced the complex II OCR (Supplementary Fig. [134]3b), and increased the respiration of complex IV (Supplementary Fig. [135]3c) when septic hearts from AAV9-Cpat and AAV9-Ctrl mice were compared (Fig. [136]3d). Fig. 3. Cpat interacts with mitochondria-associated Pnpt1 using loops 3 and 4. [137]Fig. 3 [138]Open in a new tab a Schematic diagram of RNA pulldown combined with mass spectrometry (MS) (Created in BioRender. Fan, Y. (2025) [139]https://BioRender.com/gfnrubw). b KEGG pathway enrichment analysis of detected proteins from MS. c Gene Ontology (GO) analysis of detected proteins from MS. d Representative respiratory experiment of mitochondrial OXPHOS capacity by using the substrate uncoupler inhibitor titration (SUIT) protocol. AAV9-Ctrl vs AAV9-Cpat mice, with PBS or LPS administration. n = 3 biological replicates. e Representative electron microscopy images of myocardial mitochondria morphology after PBS or LPS treatment. AAV-Ctrl vs AAV-Cpat mice. n = 5 biological replicates. f Top 10 proteins involved in the metabolic pathway from Cpat pulldown with MS detection, ranked by abundance. g Representative immunoblot of Pnpt1 pulled down by Cpat. n = 3 biological replicates. h RNA immunoprecipitation (RIP) of Cpat by Pnpt1 antibody in HL-1 cells. IgG was used as a negative control. n = 3 biological replicates. i Schematic illustration of Pnpt1 domains. j, Immunoblot of different Flag-Pnpt1 mutants in HEK293T cells. k RIP assays in HL-1 cells using Flag-antibody for immunoprecipitation followed RT-qPCR to quantify Flag-Pnpt1-bound Cpat. l Schematic diagram of Cpat loops based on its secondary structure. m Anti-Flag immunoblot to detect Pnpt1 pulled down by different Cpat loop fragments. Data are presented as mean ± SD, one-way ANOVA with Tukey’s multiple comparison test (h, k). p values are indicated. Source data are provided as a Source Data file. OCR test of cardiomyocytes provided further evidence to establish a role for Cpat in mitochondria respiration at the cellular level. We used lentivirus (Lv-Cpat) transduction effectively induced Cpat overexpression in mouse neonatal cardiomyocytes and treated these cells with conditioned medium (CM) from LPS-stimulated Raw264.7 cells (Supplementary Fig. [140]3d). RT-qPCR analysis confirmed that lentiviral (Lv-Cpat) transduction effectively induced Cpat overexpression in cardiomyocytes (Supplementary Fig. [141]3e), which significantly attenuated the LPS-induced upregulation of the myocardial injury marker B-type natriuretic peptide (BNP) (Supplementary Fig. [142]3f) and proinflammatory cytokines IL-1β, TNF-α, and IL-6 (Supplementary Fig. [143]3g–i). As expected, Cpat overexpression increased both spare and maximal respiratory capacities and respirations attributed to basal F1/F0 ATP synthase activity but did not affect basal respiration or H^+ leak under the LPS treatment (Supplementary Fig. [144]3j, k). Decrease of ATP production and NADH/NAD^+ production was both alleviated by Cpat overexpression (Supplementary Fig. [145]3l, m). Transmission electron microscopy (TEM) analysis revealed abnormal mitochondria in mouse hearts after LPS-induced sepsis, whereas Cpat overexpression rescued such mitochondria impairment (Fig. [146]3e). Together, these results demonstrate that Cpat ameliorated mitochondrial dysfunction in septic mouse heart. Cpat was localized predominantly in the nucleus but likely interacted with mitochondria-associated proteins (Figs. [147]1j, [148]3b–f and Supplementary Fig. [149]1e). Therefore, we hypothesized that Cpat transports from the nucleus to the mitochondria, where it exerts its cardioprotective role. Polyribonucleotide nucleotidyl transferase 1 (Pnpt1) attracted our attention due to its ability of transporting RNA from the cytoplasm to the mitochondria (Fig. [150]3f)^[151]43. Use sense and antisense Cpat sequences, we were able to pulldown Pnpt1 only when sense Cpat was used, as determined by Pnpt1 immunoblot analysis (Fig. [152]3g). Conversely, the Pnpt1 antibody precipitated a substantial amount of Pnpt1-bound Cpat, as quantitated by RT-qPCR (Fig. [153]3h). Pnpt1 is an RNA transporter protein that contains multiple structural domains (Fig. [154]3i). In order to define the Pnpt1 domains that interact with Cpat, we constructed a series of Flag-tagged Pnpt1 mutants (Flag-Pnpt1) and assessed the quality of these mutant proteins in HEK293T cells by immunoblot analysis with the Flag antibody (Fig. [155]3j). Using the same Pnpt1 mutant constructs in HL-1 cells, we performed RNA immunoprecipitation (RIP) assay followed by RT-qPCR quantification of Pnpt1-bound Cpat. Mutants Δ289-363 and Δ676-750 showed the lowest Cpat enrichment capacity compared to the full-length WT Flag-Pnpt1 (Fig. [156]3k), suggesting that these two Pnpt1 domains interacted with Cpat. The secondary structure of Cpat was predicted by the RNAfold WebServer ([157]http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi), which is based on the minimum free energy (MFE) method and the partition function (Fig. [158]3l), we also constructed different Cpat fragments using in vitro transcription. We treated HEK293T cells with Flag-Pnpt1 and different Cpat fragments and performed RNA pulldown followed by immunoblot analysis using the Flag antibody. The results indicated that only loop 3 and loop 4 but not loop 1 and loop 2 on Cpat bound to Pnpt1 (Fig. [159]3m). Therefore, Cpat used loop 3 and loop 4 to interact with mitochondria transport protein Pnpt1. HuR and CRM1 contribute to nuclear Cpat transportation to the cytoplasm Since Cpat predominantly localizes in the nucleus of cardiomyocytes (Fig. [160]1) but interacts with mitochondria proteins (Fig. [161]3), we hypothesize that Cpat is transported out from nucleus into the cytoplasm and then enters to the mitochondria before it involves in mitochondrial bioprocess. Typically, nuclear lncRNAs execute such translocations by interacting with nuclear RNA-binding proteins^[162]44,[163]45. HuR is an RNA-binding protein^[164]46 that is known to export nuclear RNAs^[165]47. HuR may assist the export of Cpat from the nucleus into the cytoplasm. To test this hypothesis, we used both sense and antisense Cpat to pull down Cpat-bound HuR in HL-1 cells and used anti-HuR antibody to detect HuR. The results showed that sense Cpat precipitated much more HuR than antisense Cpat did (Fig. [166]4a). In an RIP assay in HL-1 cells, anti-HuR antibody but not control IgG precipitated Cpat as determined by RT-qPCR (Fig. [167]4b). Fig. 4. HuR facilitates Cpat transportation from nucleus into cytoplasm. [168]Fig. 4 [169]Open in a new tab a Immunoblot of HuR pulled down by sense and anti-sense Cpat in HL-1 cells. b RIP of Cpat by HuR antibody in HL-1 cells. IgG was used as negative control. n = 3 biological replicates. c HuR immunoblot analysis of isolated cytoplasm and nuclear fractions or whole cell lysate (WCL) from HL-1 cells transfected with HuR siRNA or negative control siRNA (NC). Cell fraction markers included GAPDH (cytoplasm) and lamin B (nucleus). d RT-qPCR of HuR in isolated cellular fractions or WCL in HL-1 cells after siRNA-HuR transfection (vs. siRNA-Ctrl). n = 3 biological replicates. e RT-qPCR of WCL Cpat after siRNA-HuR transfection in HL-1 cells (vs. siRNA-Ctrl). n = 3 biological replicates. f Schematic diagram of nascent RNA 4SU labelling and detection (Created in BioRender. Fan, Y. (2025) [170]https://BioRender.com/sorgudk). g, h RT-qPCR of total and nascent Cpat in HL-1 nucleus (g) and cytoplasm (h). (siRNA-HuR vs. siRNA-Ctrl). n = 3 biological replicates. i Immunoblot of HuR in HL-1 cells transfected with siRNA-CRM1 or control siRNA. j Immunoblot of HuR in HL-1 nuclear and cytoplasm fractions after siRNA-CRM1 or control siRNA transfection. k Immunoblot of HuR and CRM1 in HL-1 cells to verify the efficiency of siRNA-HuR and siRNA-CRM1. l RT-qPCR of Cpat in HL-1 cell WCL and cytoplasm after siRNA transfection. n = 3 biological replicates. m IP of Cpat by HuR antibody in HL-1 cells measured by RT-qPCR after siRNA-CRM1 or control siRNA transfection. n = 3 biological replicates. n Immunoblot of Pnpt1 in different cellular components after siRNA-HuR or control siRNA transfection. o RT-qPCR of mitochondrial Cpat after siRNA-HuR transfection in HL-1 cells (vs. siRNA-Ctrl). n = 3 biological replicates. p Immunoblot of HuR pulled down by different fragments of Cpat in HEK293T cells. q Graphical illustration of HuR and CRM1 contribute to Cpat transportation from nucleus to cytoplasm (Created in BioRender. Fan, Y. (2025) [171]https://BioRender.com/2cy50tf). Data are presented as mean ± SD, unpaired two-sided Student’s t-test (b, e, o), two-way ANOVA with Tukey’s multiple comparison test (d, g, h, l, m) and Unpaired Student’s t test (j). p values are indicated. Source data are provided as a Source Data file. To test the subcellular locations of Cpat binding on HuR, we used siRNA to block HuR expression in HL-1 cells. Use of immunoblot and RT-qPCR confirmed successful suppression of HuR in HL-1 whole cell lysates (WCL), nucleus, and cytoplasm fractions. Nuclear structural protein lamin B and cytoplasm membrane protein glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were used as nuclear and cytoplasm protein markers (Fig. [172]4c, d). Results from RT-qPCR confirmed that HuR knockdown did not alter Cpat expression in HL-1 WCL (Fig. [173]4e). Successful HuR knockdown allowed us to test whether HuR participated in Cpat export from nucleus to the cytoplasm. We assessed newly synthesized Cpat in HL-1 cells to minimize potential contamination from pre-existing Cpat and to allow specific analysis of newly transcribed Cpat. HL-1 cells were incubated with 4-thiouridine (4SU), which was incorporated into nascent RNAs. 4SU-labeled RNAs were isolated, biotinylated, and affinity-purified using streptavidin beads (Fig. [174]4f). Cpat levels were measured in both total RNA and nascent RNA extracted from nuclear and cytoplasmic lysates by RT-qPCR. Our results showed that HuR knockdown significantly increased nuclear Cpat retention and reduced cytoplasm Cpat transportation when nascent RNA preparations were used. Similar changes of Cpat were also detected when total RNA preparations were used, although such changes were not statistically significant (Fig. [175]4g, h). These observations support a critical role of HuR in nuclear Cpat exportation to the cytoplasm. CRM1 is a HuR export factor^[176]48,[177]49. Any changes in CRM1 expression may affect nuclear HuR and Cpat exportation to the cytoplasm. To test this hypothesis, we performed siRNA knockdown of CRM1 or HuR or both in HL-1 cells. CRM1 knockdown efficiency was examined by immunoblot analysis (Fig. [178]4i). As expected, immunoblot analysis showed that CRM1 knockdown in HL-1 cells increased HuR accumulation in the nucleus and reduced HuR exportation to the cytoplasm (Fig. [179]4j). When WCL was used, we detected successful knockdown of CRM1 and HuR or both and such knockdowns did not affect Cpat expression as quantified by RT-qPCR (Fig. [180]4k, l). When cytoplasm Cpat was considered, however, knockdown of HuR or both HuR and CRM1 significantly reduced cytoplasm Cpat levels (Fig. [181]4l). RIP of Cpat with HuR antibody confirmed that CRM1 knockdown did not affect the binding of HuR to Cpat in HL-1 cells (Fig. [182]4m). Together, these observations suggest that CRM1 and HuR acted together to export Cpat from the nucleus to the cytoplasm. Prior studies indicate that HuR-RNA interactions influence mitochondrial metabolic functions^[183]50,[184]51. We hypothesized that HuR directed cytoplasmic Cpat to the mitochondria as well. To test this possibility, we first determined the presence of HuR in HL-1 cell mitochondria by analyzing the lysates of purified mitochondria (mitoplast) and remaining cytoplasm fractions. Immunoblot analysis showed that HuR was found in WCL and cytoplasm but not in the mitoplast fraction and confirmed the successful HuR knockdown in WCL and cytoplasm in HL-1 cells. As expected, Pnpt1 was detected in WCL, cytoplasm, and mitoplast fractions and HuR knockdown did not affect Pnpt1 expression (Fig. [185]4n). Results from RT-qPCR indicated that HuR knockdown did not affect mitochondria Cpat expression (Fig. [186]4o). In Fig. [187]3l and [188]m, we demonstrated that Cpat used its secondary structures loop 3 and loop 4 to interact with the mitochondria transport protein Pnpt1 in HEK293T cells. In the same cells, incubation of all four Cpat loops along with the full-length Cpat with Flag-tagged HuR followed by RIP and consequent immunoblot analysis with the anti-Flag antibody showed that Cpat loops 1, 2, and 4 served as the major HuR binding sites (Fig. [189]4p). Collectively, these results suggest that HuR facilitates the delivery of nuclear Cpat into the cytoplasm, and CRM1 also participated for this process (Fig. [190]4q). Cpat prevents LPS-induced MDH2-CS-ACO2 complex disassembly Pnpt1 is an RNA import factor mainly locates at the mitochondrial intermembrane space (IMS) where it mediates the translocation of cytoplasmic RNAs into the mitochondria^[191]43. Here we showed that Pnpt1 was a protein capable of interacting with Cpat (Fig. [192]3). Therefore, we hypothesized that Pnpt1 may mediate the import of Cpat from cytoplasm into the mitochondria. We first checked the expression of Pnpt1 in HL-1 cells and found that LPS treatment did not affect Pnpt1 expression by immunoblot analysis and RT-qPCR (Supplementary Fig. [193]4a–c). Cpat knockdown (siRNA-Cpat) or overexpression (Lv-Cpat) also did not affect Pnpt1 expression in these cells (Supplementary Fig. [194]4d–f), suggesting that Pnpt1 is not a direct downstream target of Cpat. We evaluated the consequences of Cpat binding to Pnpt1. To achieve this goal, we knocked down Pnpt1 expression with siRNA-Pnpt1 or overexpressed Flag-Pnpt1 with pGV657 (OE-Pnpt1) in HL-1 cells. Immunoblot analysis validated the efficiency of Pnpt1 knockdown and overexpression in HL-1 cell WCL (Fig. [195]5a). Results from RT-qPCR confirmed the efficient knockdown or overexpression of Pnpt1 in WCL and mitochondria fractions from siRNA-Pnpt1-transfected or OE-Pnpt1-transfected HL-1 cells (Fig. [196]5b, c). Pnpt1 knockdown in HL-1 cells significantly reduced the Cpat levels in mitochondria but not in WCL as determined by RT-qPCR (Fig. [197]5d). However, overexpression of Pnpt1 increased Cpat levels in mitochondria and WCL (Fig. [198]5e). These observations suggest that Pnpt1 in HL-1 cells is an essential vehicle protein to assist Cpat transportation from cytoplasm to mitochondria. Fig. 5. Cpat prevents LPS-induced MDH2-CS-ACO2 complex disassembly. [199]Fig. 5 [200]Open in a new tab a Immunoblot analysis of Pnpt1 in HL-1 cells. b, c RT-qPCR of Pnpt1 in WCL and mitochondria from HL-1 cells transfected with siRNA-Pnpt1 vs siRNA-Ctrl (b) or Pnpt1 overexpression (OE Pnpt1) vs. OE Ctrl (c). n = 3 biological replicates. d, e RT-qPCR of Cpat in WCL and mitochondria from HL-1 cells after siRNA-Pnpt1 (d) or Pnpt1 overexpression (e). n = 3 biological replicates. f Immunoblot analysis of nuclear lamin B, cytoplasm GAPDH, and mitochondria citrate synthase (CS) in purified HL-1 cellular components. g, h RT-qPCR of cytoplasm GAPDH, nuclear U6, and mitochondria Pnpt1, CS, and ATP8 in purified mitochondria from HL-1 cells (g) or Cpat in mitochondria from mouse neonatal cardiomyocytes treated with LPS or PBS (h). n = 3 biological replicates. i RNA FISH and immunofluorescence image of Cpat and mitochondria marker CS in HL-1 cells, showing their subcellular distribution. j Immunoblot analysis of CS pulled down by sense and antisense Cpat in HL-1 cells. n = 3 biological replicates. k RIP of Cpat by CS antibody in HL-1 cells IgG was used as negative control. n = 3 biological replicates. l Immunoblot analysis of Myc-CS after RIP with different Cpat loops or full-length Cpat (FL) in Myc-CS-treated HEK293T cells. m Graphical illustration of MDH2-CS-ACO2 complex in baseline condition. n Immunofluorescence co-localization of CS with ACO2 in mouse neonatal cardiomyocytes. Scale bar = 2 μm. o Immunoblot analyses using Myc or Flag antibodies following Myc antibody-mediated immunoprecipitations (IP) of HEK293T cells treated with or without Myc-CS and Flag-ACO2. p Immunofluorescence co-localization of CS with MDH2 in mouse neonatal cardiomyocytes. Scale bar = 2 μm. q Immunoblot analyses using Myc or Flag antibodies following Myc antibody-mediated IP of HEK293T cells treated with or without Myc-CS and Flag-MDH2. Data are presented as mean ± SD, two-way (b–e) or one-way (g) ANOVA with Tukey’s multiple comparison test, unpaired two-sided Student’s t-test (h, k). p values are indicated. Source data are provided as a Source Data file. To examine an involvement of Cpat in maintaining mitochondria functions in cardiomyocytes following LPS treatment, we first confirmed the purity of our subcellular compartment preparations, including cytoplasm, nucleus, and mitoplast from HL-1 cell by immunoblot analysis. Lamin B and GAPDH were used as nuclear and cytoplasm markers. Citrate synthase (CS) is a mitochondrial matrix-resident protein. Mitochondrially encoded ATP synthase membrane subunit 8 (ATP8) also served as mitochondria marker. Use of immunoblot analysis allowed detection of CS expression in mitoplast and in WCL or cytoplasm, but not in the nuclear fraction (Fig. [201]5f). In purified mitochondria, we used RT-qPCR and detected the expression of CS, ATP8, and Pnpt1 but not nuclear U6 or cytoplasm GAPDH (Fig. [202]5g). In mouse neonatal cardiomyocyte mitochondria, LPS reduced Cpat expression (Fig. [203]5h). Cpat is predominantly found in the nucleus (Fig. [204]1j). RNA FISH, together with immunofluorescent staining demonstrated colocalization of Cpat with mitochondria marker CS in HL-1 cells (Fig. [205]5i). In RNA pulldown and RIP assays, sense Cpat precipitated much more CS than did the antisense Cpat in HL-1 cells (Fig. [206]5j), and anti-CS antibody effectively precipitated Cpat (Fig. [207]5k). Conversely, we used different Cpat fragments along with the full-length Cpat for the RIP assay to determine Cpat domains responsible for Myc-CS interaction in HEK293T cells. Among all tested four loops, loop 4 showed the strongest activity to interact with Myc-CS as determined by anti-Myc antibody immunoblot analysis (Fig. [208]5l). Together, these observations suggest that Cpat modulates mitochondrial function by targeting CS. CS, aconitase 2 (ACO2), and malate dehydrogenase (MDH2) are canonical mitochondrial enzymes that catalyze the main steps of TCA cycle (Fig. [209]5m)^[210]52,[211]53. Cpat interaction with mitochondria CS may be essential to TCA cycle. Using mouse neonatal cardiomyocytes, we performed immunofluorescent double staining with antibodies against ACO2 and CS and demonstrated their colocalizations in mitochondria (Fig. [212]5n). When HEK293T cells were treated with Myc-CS and/or Flag-ACO2, immunoprecipitation using anti-Myc antibody followed by immunoblot analyses with anti-Myc or anti-Flag antibodies revealed immunocomplex formation between Myc-CS and Flag-ACO2 (Fig. [213]5o). Similarly, immunofluorescent double staining detected colocalization of CS and MDH2 in mouse cardiomyocyte mitochondria (Fig. [214]5p). Using immunoprecipitation followed by immunoblot analysis of HEK293T cells treated with or without Myc-CS and/or Flag-MDH2, we again detected the immunocomplex of CS and MDH2 (Fig. [215]5q). These observations indicate physical interaction of CS with AOC2 and MDH2 in mitochondria. In mouse neonatal cardiomyocytes transduced with Lv-Cpat and treated with PBS, we detected CS colocalization with ACO2 and MDH2 (Supplementary Fig. [216]5a, b). Building on these findings, we further extracted protein lysates from neonatal cardiomyocytes and performed immunoprecipitation using a CS antibody, which confirmed the interaction of CS with both ACO2 and MDH2 (Supplementary Fig. [217]5c). We used immunoprecipitation combined with immunoblot analysis to examine the impact of Cpat overexpression or LPS treatment on the interaction of CS with ACO2 or MDH2 in both HEK293T cells and mouse neonatal cardiomyocytes. HEK293T cells were incubated with Myc-CS and Flag-ACO2 or Flag-MDH2 and treated with or without LPS or transduced with Lv-Cpat. Cell lysates were immunoprecipitated with Myc antibody followed by immunoblot analysis with the Flag antibody. The results showed that CS formed immunocomplexes with ACO2 or MDH2. Treatment of LPS reduced the formation of such immunocomplexes. Cpat overexpression with Lv-Cpat blocked this activity of LPS (Supplementary Fig. [218]5d, e). In HEK293T cells incubated with Myc-CS and Flag-ACO2 or Flag-MDH2, Cpat knockdown with siRNA-Cpat inhibited the immunocomplex formation of CS with ACO2 and MDH2 (Supplementary Fig. [219]5f, g). In LPS-treated mouse neonatal cardiomyocytes, we also found that Cpat overexpression with Lv-Cpat elevated the immunocomplex formation of CS with ACO2 and MDH2 (Fig. [220]6a). Cpat knockdown with siRNA-Cpat diminished the formation of these immunocomplexes (Fig. [221]6b). Consistently, co-immunoprecipitation using ventricle heart tissue lysates from Tg (Cpat-Cre) mice confirmed that Cpat overexpression in vivo enhances the interaction of CS with ACO2 and MDH2 under septic conditions (Fig. [222]6c). These observations from HEK293T cells, mouse neonatal cardiomyocytes, and heart tissues unequivocally suggest that Cpat maintains the MDH2-CS-ACO2 complex stability and ultimately alleviates LPS-induced mitochondrial dysfunction. Fig. 6. Cpat maintains MDH2-CS-ACO2 complex stability by reducing CS acetylation. [223]Fig. 6 [224]Open in a new tab a-b immunoblot analysis of ACO2 and MDH2 in mouse neonatal cardiomyocytes after CS pull down. Cells were transduced with Lv-Cpat (a) or transfected with siRNA-Cpat (b) and then treated with LPS. This experiment was repeated three times. c immunoblot analysis of ACO2 and MDH2 in heart tissues from Tg(Cpat-Cre) and Tg(Cpat) transgenic mice treated with LPS or PBS, following CS pull-down. This experiment was repeated three times. d-e immunoblot analysis of CS acetylation in mouse neonatal cardiomyocytes treated with or without LPS and Lv-Cpat (d) or siRNA-Cpat (e). This experiment was repeated three times. f Immunoblot analysis of CS acetylation in heart tissues from Tg (Cpat-Cre) and Tg (Cpat) transgenic mice treated with LPS or PBS, following CS pull-down. This experiment was repeated three times. g-h immunoblot analysis of CS acetylation in HEK293T cells treated with or without LPS, Myc-CS, and transduced with Lv-Cpat (g) or transfected with siRNA-Cpat (h) as indicated. This experiment was repeated three times. i Graphical illustration of Cpat activity in maintaining MDH2-CS-ACO2 complex stability by reducing CS acetylation (Created in BioRender. Fan, Y. (2025) [225]https://BioRender.com/yr42tnh). j-k NAM and TSA blocked the deacetylase activity. Immunoblot analysis of CS acetylation in HEK293T cells (j) and mouse neonatal cardiomyocytes (k) after different times of NAM + TSA treatment. This experiment was repeated three times. l, m IP combined with immunoblot to detect ACO2-CS (l) and CS-MDH2 (m) immunocomplexes in HEK293T cells with or without NAM and TSA treatment. This experiment was repeated three times. n, o IP combined with immunoblot to detect MDH2-CS and CS-ACO2 immunocomplexes in mouse neonatal cardiomyocytes with Cpat overexpression (Lv-Cpat, n) and knockdown (siRNA-Cpat, o) with or without NAM and TSA treatment. This experiment was repeated three times. Source data are provided as a Source Data file. MDH2 malate dehydrogenase, CS citrate synthase, ACO2 aconitase, NAM nicotinamide, TSA trichostatin A. Cpat maintains MDH2-CS-ACO2 complex stability by reducing CS acetylation Accumulating evidence indicates that the mitochondrial cellular metabolic process is associated with the reversible acetylation of metabolic enzymes^[226]19,[227]54. We therefore hypothesized that Cpat ameliorates sepsis (LPS) insult-caused mitochondria dysfunction by enhancing MDH2-CS-ACO2 complex stability and such function of Cpat is associated with CS deacetylation. In mouse neonatal cardiomyocytes, LPS treatment enhanced lysine acetylation on CS as determined by immunoprecipitation with CS antibody followed by immunoblot analysis with anti-acetyllysine antibody. Cpat overexpression after Lv-Cpat transduction brought the CS lysine acetylation to the basal level (Fig. [228]6d). In contrast, Cpat knockdown with siRNA-Cpat elevated the CS lysine acetylation (Fig. [229]6e). We were able to replicate these results in HEK293T cells. When HEK293T cells were incubated with Myc-CS, immunoprecipitation with Myc antibody followed by immunoblot analysis with anti-acetyllysine antibody demonstrated a role for LPS in increasing CS lysine acetylation (Supplementary Fig. [230]5h). Consistently, in cardiac tissues collected from LPS-induced septic Tg (Cpat-Cre) and Tg (Cpat) mice, immunoprecipitation using a CS antibody followed by acetyllysine immunoblot analysis showed that Cpat overexpression markedly attenuated the LPS-induced increase in CS acetylation (Fig. [231]6f). In HEK293T, Cpat overexpression with Lv-Cpat reduced CS basal lysine acetylation as well as LPS-induced CS lysine acetylation (Fig. [232]6g). In contrast, Cpat knockdown with siRNA-Cpat increased CS basal lysine acetylation and LPS-induced CS lysine acetylation (Fig. [233]6h). Together, these results from cardiomyocytes and HEK293T cells suggest that Cpat maintains the TCA flux metabolic process by blocking CS lysine acetylation and stabilizing the MDH2-CS-ACO2 complexes (Fig. [234]6i). Acetyltransferases and deacetylases control protein acetylation and deacetylation^[235]15,[236]16,[237]54. Nicotinamide (NAM; inhibitor of sirtuins) and trichostatin A (TSA; inhibitor of histone deacetylases) are both deacetylase inhibitors that may increase CS acetylation, thereby destabilizing the MDH2-CS-ACO2 complexes. To test this hypothesis, we treated mouse neonatal cardiomyocytes and HEK293T cells with NAM and TSA. Cells were lysed, immunoprecipitated with anti-CS antibody followed by immunoblot analysis with anti-acetyllysine antibody. The results showed that use of NAM and TSA increased CS acetylation in these cells in a time-dependent manner (Fig. [238]6j, k). As expected, inhibition of CS deacetylation impaired MDH2-CS-ACO2 complex formation. In HEK293T cells, use of NAM and TSA greatly reduced the immunocomplex formation of ACO2-CS and CS-MDH2 (Fig. [239]6l, m). In mouse neonatal cardiomyocytes, we obtained the same results. Use of NAM and TSA reduced ACO2-CS and CS-MDH2 immunocomplex formation. Such reductions were muted when cells were transduced with Lv-Cpat (Fig. [240]6n) but not affected when cell were transfected with siRNA-Cpat (Fig. [241]6o). The above observations support a role for CS deacetylation in stabilizing the MDH2-CS-ACO2 complexes. Acetyltransferase and deacetylase regulate these post-translational modifications of CS. Cpat may have a direct or indirect role in inhibiting acetyltransferase activity or in enhancing deacetylase activity. To test these hypotheses, we treated HEK293T cells with Myc-CS with or without a few acetyltransferase candidates, including CBP, GCN5, KAT5, and KAT8. Myc-CS acetylation was assessed by anti-Myc immunoprecipitation and anti-acetyllysine immunoblot analysis. Among the list, only GCN5 showed its activity in acetylating Myc-CS (Fig. [242]7a). Treatment of HEK293T cells with Myc-CS and HA-GCN5 allowed to detect whether CS directly interact with GCN5. Cell lysates were immunoprecipitated with anti-Myc antibody followed by immunoblot analysis with anti-HA antibody. The results showed that Myc-CS and HA-GCN5 formed an immunocomplex, suggesting a direct interaction of GCN5 against its substrate CS. Treatment of LPS increased such interaction. In contrast, such activity of LPS was muted when HEK293T cells were transduced with Lv-Cpat (Fig. [243]7b). Fig. 7. Cpat blocks GCN5-mediated CS acetylation. [244]Fig. 7 [245]Open in a new tab a Immunoblot to test the activities of different acetyltransferases in CS acetylation in HEK293T cells. This experiment was repeated three times. b IP combined with immunoblot to detect CS (Myc-CS) and GCN5 (HA-GCN5) immunocomplexes in HEK293T cells with and without LPS (1 µg/mL) treatment or Cpat rescue. This experiment was repeated three times. c IP combined with immunoblot to assess the impact of GCN5 in affecting MDH2-CS and CS-ACO2 immunocomplex formation in HEK293T cells. This experiment was repeated three times. d Graphical illustration of Cpat activity in preventing GCN5-mediated CS acetylation. e Schematic diagram of CS acetylation site detection by LC-MS/MS (Created in BioRender. Fan, Y. (2025) [246]https://BioRender.com/4c4aqbr). f Acetylation sites identified by LC-MS/MS is displayed on the molecular structure of CS. CS Uniprot ID: [247]Q9CZU6. g IP combined with immunoblot detected CS acetylation in HEK293T cells with or without LPS (1 µg/mL) treatment. 6KR, all six lysine residues were replaced by arginine. This experiment was repeated three times. h, i Representative immunoblots of Flag-ACO2 (h) and Flag-MDH2 (i) pulled down by Myc-CS and Myc-CS K52R, K366R, K370R and 6KR in HEK293T cells with or without NAM and TSA treatment. This experiment was repeated three times. j Schematic diagram summarizing the proposed mechanism by which Cpat overexpression modulates mitochondrial metabolism and cardiomyocyte injury in response to sepsis-induced stress (Created in BioRender. Fan, Y. (2025) [248]https://BioRender.com/v28g941). Source data are provided as a Source Data file. MDH2: malate dehydrogenase; CS: citrate synthase; ACO2: aconitase; GCN5: general control non-repressed protein-5. To validate the effect of GCN5-CS interaction on MDH2-CS-ACO2 complex formation, we treated HEK293T cells with Myc-CS together with or without Flag-ACO2, Flag-MDH2, and HA-GCN5. Cell lysates were immunoprecipitated with Myc antibody to pulldown CS-binding proteins followed by immunoblot analysis using Flag antibody to analyze CS-bond ACO2 and MDH2. As expected, the presence of HA-GCN5 blocked CS-ACO2 and CS-MDH2 immunocomplex formation (Fig. [249]7c). Together, these observations suggest that Cpat inhibited CS acetylation by blocking the GCN5 activity (Fig. [250]7d). CS immunoprecipitation followed by MS analysis allowed identification of exact lysine residues on CS that underwent acetylation in mouse neonatal cardiomyocytes (Fig. [251]7e). By comparing the CS acetylations by MS analysis in Lv-Ctrl- or Lv-Cpat-transduced mouse neonatal cardiomyocytes that were treated with LPS, we identified 6 lysine residues on CS that underwent acetylation, including K52, K103, K107, K366, K370, and K375 (Fig. [252]7f and Supplementary Fig. [253]6a-d). To test the contribution of these lysine residues on CS acetylation, we made Myc-CS mutants by replacing these lysine residues with arginine. We treated HEK293T cells with LPS together with these Myc-CS mutants. Cell lysates were immunoprecipitated with anti-Myc antibody followed by immunoblot analysis with anti-acetyllysine antibody. It appeared that K103R, K107R, and K375R did not affect LPS-induced CS acetylation, but K52R, K366R and K370R efficiently blocked LPS-induced CS acetylation. Combined mutations of all 6 lysine residues on Myc-CS (6KR) yielded lowest CS acetylation in LPS-treated HEK293T cells (Fig. [254]7g). Use of deacetylase inhibitors NAM and TSA reduced the CS-ACO2 and CS-MDH2 immunocomplex formation in HEK293T cells treated with Myc-CS together with Flag-ACO2 or Flag-MDH2, as determined by anti-Myc immunoprecipitation followed by anti-Flag immunoblot analysis (Fig. [255]7h, i). To further investigate whether acetylation at K52, K366, and K370 regulates CS interactions with ACO2 and MDH2, we introduced the respective single mutants (K52R, K366R, K370R) and the combined 6KR mutant into HEK293T cells, along with either Flag-tagged ACO2 or MDH2. Immunoprecipitation of Myc-CS followed by anti-Flag immunoblotting showed that, unlike wild-type CS, these acetylation-resistant mutants displayed minimal changes in their interaction with ACO2 and MDH2 following treatment with NAM and TSA (Fig. [256]7h, i). These findings suggest that lysine acetylation at K52, K366, and K370 is required for the acetylation-sensitive modulation of the CS-ACO2 and CS-MDH2 complexes. The absence of NAM and TSA responsiveness in the mutants further supports a regulatory role of site-specific acetylation in maintaining TCA cycle complex integrity. Our results established a role for Cpat in inhibiting GCN5 transacetylase activity, leading to reduced CS acetylation, increased MDH2-CS-ACO2 complex formation (Fig. [257]7j), and enhanced TCA flux metabolic process. Discussion In this study, we identified a cardiomyocyte-specific lncRNA Cpat that played a protective role to ameliorate cardiac dysfunction in septic cardiomyopathy heart by enhancing cardiomyocyte mitochondrial respiration and energy metabolism and maintaining cardiac homeostasis. This lncRNA was downregulated in cardiomyocytes from the sepsis mouse heart. Overexpression of Cpat in cardiomyocytes by either cTNT promoter-directed and AAV9-driven or in Myh6-Cre transgenic mice significantly improved cardiac functions, reduced the mortality rate, and inhibited systemic cardiac injury markers (CK-MB, cTnT) and proinflammatory cytokines (IL-1β, TNF-α) in both CLP- and LPS-induced septic mice. In cultured mouse cardiomyocytes, Cpat overexpression elevated cardiomyocyte mitochondria OCR, reduced the expression of myocardial injury marker BNP and proinflammatory cytokines (IL-1β, TNF-α, IL-6), and rescued sepsis-caused mitochondria impairment. Combined use of ChIRP-like RNA pull-down and mass spectrometry (MS) analysis, KEGG, and GO analyses predicted a role for Cpat in mitochondria “metabolic pathways” and “protein translocation”. In addition to the discovery of Cpat in sepsis mouse heart cardiomyocytes, this study reports detailed characters of Cpat and its mechanisms in cardioprotection. Cpat is a nuclear RNA but also presents in mitochondria and interacts with mitochondria proteins. Nuclear RNA-binding protein HuR is responsible for Cpat exportation from nucleus to the cytoplasm. Cpat substructures loop 1, loop 2, and loop 4 serve as the major HuR binding sites. Knockdown of HuR or its export factor CRM1 or both in cardiomyocytes caused nuclear Cpat retention and reduced HuR cytoplasm exportation, although CRM1 did not affect the binding of HuR to Cpat. HuR is a pivotal RNA-binding protein that facilitates the export of various RNA species from the nucleus to the cytoplasm, including coding and non-coding RNAs^[258]55,[259]56. Its export is intricately regulated by its interactions with nuclear export machinery components such as CRM1^[260]44,[261]48,[262]49 and transportins^[263]49,[264]57, and by its phosphorylation status^[265]49,[266]58,[267]59 that is modulated by kinases such as cyclin dependent kinase 2^[268]60, protein kinase C^[269]61, and p38 mitogen-activated protein kinase^[270]62,[271]63. Under normal conditions, these mechanisms ensure proper RNA distribution within the cells. However, sepsis induces inflammatory and oxidative stress responses in cardiomyocytes that may disrupt these regulatory mechanisms. Since HuR is involved in regulating mRNAs related to mitochondrial function and stress responses, inappropriate localization of HuR might impact the stability and translation of these mRNAs^[272]50,[273]51. This disruption may exacerbate mitochondrial dysfunction and lead to cardiac impairments found in septic mice. Therefore, it is essential to study how sepsis changes the function of HuR in nuclear RNA exportation, which may affect the availability and function of Cpat in mitochondrial TCA metabolic process. The import of Cpat from cytoplasm into mitochondria is mediated by mitochondrial intermembrane space-resident RNA import factor Pnpt1. Residues 289-363 and 676-750 on Pnpt1 and Cpat substructure loop 3 and loop 4 are responsible for their interaction and cytoplasm to mitochondria translocation. Expression of Pnpt1 directly affected cytoplasmic Cpat transportation to the mitochondria. Pnpt1 is known to play a crucial role in the import of RNA into the mitochondrial matrix, where it facilitates mitochondrial gene expression and function^[274]43,[275]64–[276]66. The import of RNA into the mitochondrion via Pnpt1 necessitates a specific stem-loop secondary structure within the target RNA^[277]43,[278]67,[279]68. Notably, Cpat was predicted to adopt these stem-loop configurations, suggesting a critical role for Pnpt1 in facilitating the import of Cpat into the mitochondrion. In the setting of mitochondrial dysfunction induced by sepsis, oxidative stress and changes in cellular ATP levels may impact Pnpt1 function and its interaction with mitochondrial import targets. Such impairment might result in decreased levels of Cpat in mitochondria and disrupt the regulatory functions of Cpat, crucial for mitochondrial and cardiac stress responses. After Cpat enters to the mitochondria, it stabilizes the MDH2-CS-ACO2 complex by inactivating the acetyltransferase GCN5 as a mechanism to block CS acetylation. This complex in the mitochondria contributes to mitochondria oxygen consumption and TCA flux metabolic process. After a sepsis insult, GCN5 acetylated three lysine residues (K52, K366, and K370) on CS and destabilized this complex. Similar results were found when cardiomyocytes were treated with deacetylase inhibitors NAM and TSA. Enhanced CS acetylation and MDH2-CS-ACO2 complex destabilization caused cardiomyocyte mitochondria dysfunction. Such adverse effects from NAM and TSA treatment and from sepsis were effectively muted by Cpat overexpression in cultured cardiomyocytes and in septic mice. Under sepsis-induced stress, the transport and function of Cpat may be compromised. Disruption in Cpat transportation to mitochondria may lead to sustained CS acetylation that impacts the assembly and function of the MDH2-CS-ACO2 complex and results in reduced TCA cycle activity, diminished mitochondrial energy production, and impaired cellular metabolism. These disruptions may exacerbate heart failure and related pathologies by causing cardiomyocyte mitochondrial dysfunction. GCN5 acetyltransferase participates in modulating mitochondrial function^[280]20,[281]31,[282]69. Our study provided the first example that lncRNAs regulate the function of mitochondria acetyltransferase. This discovery not only enhances our understanding of lncRNA functions in metabolic regulation but also highlights the intricate balance of acetylation and deacetylation processes in mitochondrial dynamics. In this study, we showed that Cpat exhibits high conservation between humans and mice. Therefore, our understanding of how Cpat regulates mitochondrial protein interactions and acetylation processes provided valuable insights into potential therapeutic potential to mitigate mitochondrial dysfunction in inflammatory cardiac diseases. Despite the discoveries of Cpat as a new member of the lncRNA family and its functions in repairing sepsis-induced cardiomyocyte mitochondria dysfunction and cardiac failure, our results left several questions unanswered. For example, sepsis is merely one of the cardiac stressors. It remains unknown whether other cardiac stressors also affect Cpat expression and impair its function in targeting GCN5 or other acetyltransferases in cardiomyocytes or even in other cardiac cell types and diseases. In this study, we were able to identify GCN5 as the acetyltransferase that is responsible for CS acetylation and MDH2-CS-ACO2 complex destabilization. However, using the deacetylase inhibitors NAM and TSA, we were also able to show their activity in increasing CS acetylation and destabilizing the MDH2-CS-ACO2 complexes. These observations suggest the existence of deacetylases that are responsible for CS deacetylation, a hypothesis that merits for future investigation. Furthermore, the long-term effects of modulating Cpat levels on overall mitochondrial function and myocardial health remain to be tested and may lead to unforeseen consequences. Of course, there are more questions than answers. Future in-depth studies stemmed from results of this work may help explore the role of Cpat and other lncRNAs in cardiovascular diseases and other disease types. Methods Mice Mice were housed in a pathogen-free facility in groups (4–5 mice per cage) on a 12h light/dark cycle at room temperature (22°C) with ad libitum access to autoclaved water and irradiated standard chow diet. Mouse health status was monitored daily and only mice that displayed good general health were used for the in-house breeding colonies. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of the Second Affiliated Hospital, Zhejiang University School of Medicine. All experiments involving live animals adhered strictly to the Guide for the Care and Use of Laboratory Animals (U.S. National Institutes of Health, NIH publication no. 85-23, revised 1996). Cpat knockin mice at Rosa26 (R26-LSL-Cpat) and Myh6-Cre mice (Myh6-CreER) were generated and maintained on the C57BL/6 background by Shanghai Model Organisms Center, Inc. For the identification of R26-LSL-Cpat mice, the following primers were used: R26-LSL-Cpat Forward: ATCCTTCGAAAGCCAGCACA R26-LSL-Cpat Reverse: GGAAAGGACAGTGGGAGTGG WT Forward: TCAGATTCTTTTATAGGGGACACA WT Reverse: TAAAGGCCACTCAATGCTCACTAA For the identification of Myh6-CreER mice, the following primers were used. Common: TCTATTGCACACAGCAATCCA Myh6-CreER: CCAACTCTTGTGAGAGGAGCA WT: CCAGCATTGTGAGAACAAGG Mice with R26-LSL-Cpat only were grouped into Tg (Cpat) as wild-type (WT) control. Hybridized mice with both R26-LSL-Cpat and Myh6-Cre were named as Tg (Cpat-Cre) and could specifically elevate Cpat expression in cardiomyocytes after tamoxifen activation. To overexpress Cpat specifically in cardiomyocytes, we used an adeno-associated virus serotype 9 (AAV9) vector (GV618, GeneChem, China) under the control of the cardiac troponin T promoter (cTNTp), which drives cardiac-specific expression. The expression cassette consisted of cTNTp-MCS-SV40 polyA. Recombinant AAV9 particles (AAV9-Cpat and control AAV9-Ctrl) were produced and purified. For in vivo delivery, neonatal C57BL/6 mice (postnatal day 0-2) received a single subcutaneous injection of 2-2.5×10¹¹ viral genomes (vg) per pup. This approach enabled stable cardiac expression by adulthood. Animal experiments Lipopolysaccharide (LPS, L2630, Sigma-Aldrich) was used to induce sepsis in 8-week-old male C57BL/6 mice that were acclimated for 3-5 days. LPS was dissolved in PBS at a concentration of 1 mg/mL. Mice were injected with LPS at a dose of 10 mg/kg body weight intraperitoneally (i.p.). The inflammatory response can be observed within 6 and 12 h. Cecal ligation and puncture (CLP) was also used to induce sepsis in 8-week-old male C57BL/6 mice 48 h. After anesthesia with 5% chloral hydrate (C8383, Sigma-Aldrich) at a dose of 60 µL/10 g body weight, mice were fixed on a surgical board at a supine position. The abdominal surgical area was routinely disinfected and depilated. Under sterile conditions, a scalpel was used to make a 2-cm incision in the abdominal wall. The incision was entered into the abdomen and the distal end of the ileocecal valve was separated and ligated with a 3-gauge silk suture at 1/3 of the cecum. We used an 18-gauge needle to perforate the ligated end and squeeze out a small amount of feces to avoid damage to the blood vessels. We then used a 4-gauge suture to close the peritoneum and skin, followed by subcutaneous injection of 50 ml/kg normal saline to prevent shock. After the modeling, mice started to show increased hair standing, diarrhea, and pyuria, and a decrease in drinking, eating, and activity. Mouse activity decreased significantly 12 h after surgery and the activity lost its circadian rhythm. A minimum of 6 animals per group was required for the study based on a power calculation. All mouse experiments were carried out and analyzed in a blinded manner. Data collection and analysis were performed by two observers unaware of mouse group assignment or treatment. We used male mice for all experiments in this study to eliminate gender influence. RNA-seq and data processing Total RNA was extracted from heart tissue, followed by rRNA removal using a Ribo-Zero™ Magnetic Gold Kit (Human/Mouse/Rat) (MRZG126, Illumina). RNA was fragmented ( ~ 200 nt), reverse-transcribed to single-stranded cDNA, converted to double-stranded cDNA, end-repaired, ligated with adapters, PCR-amplified, purified, and quality-checked before sequencing. Raw reads were filtered to remove adapters, low-quality reads, and rRNA sequences, generating clean reads. Clean reads were aligned to the reference genome using TopHat2, and mapped read distributions were classified as exonic, intronic, or intergenic. Expression levels of identified lncRNAs were quantified, and differentially expressed lncRNAs were selected based on |log2FoldChange | >1 and Q-value < 0.001. Functional annotation included ORF prediction, protein domain analysis, coding potential assessment, secondary structure prediction, and family classification. Transthoracic echocardiography Echocardiography using a Visual Sonics Vevo® 3100 Imaging System (Visual Sonics, Toronto, Canada) with a 40 MHz MicroScan transducer (model MS-550D) was performed to measure cardiac functions. Mice were anesthetized with isoflurane (2.5% isoflurane for induction and 0.5% for maintenance). Heart rate and left ventricular (LV) dimensions were measured from 2-D short-axis under M-mode tracings at the level of the papillary muscle. LV mass and functional parameters such as percentage of ejection fractions (EF%), left ventricular end systolic volume (LVESV), and fractional shortening (FS%) were calculated using the above primary measurements and accompanying software. Cell cultures Human embryonic kidney cell HEK293T, mouse cardiac muscle cell HL-1 and mouse mononuclear macrophage Raw264.7 were obtained from the American Type Culture Collection (ATCC). HEK293T and HL-1 cells were cultured in DMEM (11965118, Gibco) supplemented with 10% fetal bovine serum (FBS, 16000044, Gibco) and 1% penicillin/streptomycin (15070063, Gibco). Raw264.7 cells were cultured in RMPI 1640 (C11875500BT, Gibco) supplemented with 10% FBS and 1% penicillin/streptomycin. Cells were maintained in a humidified, 5% CO[2] atmosphere at 37 °C. Primary cardiomyocytes were isolated from neonatal mice using the Neonatal Heart Dissociation Kit (Miltenyi Biotec, Germany). Mice were briefly immersed in 75% ethanol to prevent contamination, and hearts were extracted via a left sternum incision. After rinsing heart in PBS and removing clots and surrounding tissues, the minced heart was digested in a prewarmed enzyme solution (Enzyme Mix 1: 62.5 μl Enzyme P, 2300 μl Buffer X, Enzyme Mix 2: 12.5 μl Enzyme A, 100 μl Enzyme D, 25 μl Buffer Y) (Miltenyi Biotec, Germany, cat# 130-098-373) at 37 °C for 15 min, followed by gentle pipetting. After a second 15-min incubation, the tissue was dissociated into a cell suspension. The enzyme solution was neutralized with high-glucose DMEM containing 10% FBS, and the suspension was filtered through a 70 μm cell strainer. Cells were pelleted by centrifugation at 1000 × g for 5 min and resuspended in fresh DMEM with 10% FBS. Cardiomyocytes were enriched by differential adhesion, with fibroblasts adhering to the dish during a 20-min incubation at 37 °C in a CO[2] incubator. The suspended cardiomyocytes were collected for further use. Raw264.7 cells were stimulated with LPS at 1 μg/ml for 24 h. Cell culture supernatant was obtained and followed by centrifugation at 1000 × g for 5 min. Purified supernatant was collected. Cell transfection and lentiviral-based gene transduction All short interfering RNA (siRNA) duplexes used in this study are shown in Supplementary Data [283]1. Transfections were carried out using LipofectAMINE™ RNAiMAX (Thermo Fisher Scientific, Germany). Lentiviral Cpat and adenoviral Cpat were obtained from Genechem. Plasmids of mouse-Pnpt1 ([284]NM_027869.2), human-PNPT1([285]NM_033109.5), human-ACO2 ([286]NM_001098.3), mouse-Aco2 ([287]NM_080633.2), human-MDH2 ([288]NM_005918.4), mouse-Mdh2 ([289]NM_008617.2), human-CS ([290]NM_004077.3), mouse-Cs ([291]NM_026444.4), CBP ([292]NM_004380.3), GCN5 ([293]NM_021078.3), KAT5 ([294]NM_182710.3), KAT8 ([295]NM_182958.4) were obtained from MIAOLING (Miaoling Plasmid Biotechnology Co., Ltd, China). The mutant constructs of loop1, loop2, loop3, and loop4 for Cpat were obtained from GenePharma (Shanghai GenePharma Co., Ltd, China); K52R, K103R, K107R, K366R, K375R, and 6KR for CS were obtained from MIAOLING; Δ55-183, Δ189-363, Δ366-501, Δ605-667, and Δ676-750 for Flag-Pnpt1 were obtained from Genechem (Shanghai Genechem Co., Ltd). Plasmids transfections were carried out using LipofectAMINE™ 3000 (Thermo Fisher Scientific, Germany). To achieve Cpat overexpression in vitro, a lentiviral vector (GV513, GeneChem, China) carrying the full-length Cpat sequence under the control of a ubiquitin promoter (Ubi-MCS-CBh-gcGFP-IRES-puromycin) was used. The viral titer was 2.5 × 10⁸ transducing units (TU)/ml. Cultured cells (e.g., primary cardiomyocytes or cell lines) were plated in 6-well plates and transduced at approximately 60-70% confluence with 2–2.5 μl of lentivirus per well, together with 5 μl of transduction enhancer reagent (P solution, GeneChem). Cells were incubated with virus-containing medium lasted for 24 h and gene expression reached its peak between 48 and 72 h. Transduction efficiency was monitored by GFP fluorescence, and stable cells were selected using puromycin when applicable. Rapid amplification of cDNA ends (RACE) and cloning of full-length Cpat The 3’ and 5’ RACE were performed using the GeneRacerTM Kit (Thermo Fisher, L1500-01) following the manufacturer’s instruction. RNA was extracted from neonatal mouse cardiomyocytes using TRIzol® Plus RNA Purification Kit (Thermo Fisher, 12183-555). Primers used for 3’ and 5’ RACE were designed based on the known sequence information: GCAGAGAGGACCAGGTTGCCATGGA (first-round Cpat 3’-prime specific); GAAGCTCAGGCTTTTTATTAGACAGGAT (second-round Cpat 3’-prime specific); CAGGAAGCAGGTCTGACCCTCTGTGT (first-round Cpat 5’-prime specific); GGAGAAGCTGCGCTCTGACCTGTC (second-round Cpat 5’-prime specific). Once we reached the 5’ and 3’ cDNA ends, we used primers F (CGCTGGAGGTTGTCGATCT) and R (CCTGTTCACAGACCTTCTGGAA) to amplify the middle part of Cpat transcript and get full-length Cpat. Immunoprecipitation and immunoblot analysis Cells were harvested in PBS and homogenized in RIPA lysis buffer (50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, and 0.1% SDS) (P0013B, Beyotime, China) with protease inhibitor cocktail and phosphatase inhibitor cocktail. Lysates were cleared by centrifugation at 12,000 × g for 10 min at 4 °C. Supernatants were applied for immunoprecipitation (IP) or immunoblot (IB) analysis with the indicated antibodies. For IP, the required primary antibody and the control IgG were added separately to the prepared lysates. A portion of the lysate was obtained from the isolated supernatant, serving as input samples for later assays. After incubation at room temperature (RT) for 1.5 h with gentle rotation, 20 μl of protein A + G agarose beads (Thermo Fisher Scientific, Germany) were added each to the lysates, followed by incubation for overnight at 4°C with rotation. The protein-captured beads were washed with wash buffer (Thermo Fisher Scientific, Germany) 3 times for 5 min each at RT with rotation. Then, the eluted beads with 50μl of 2xSDS loading buffer, and the eluted protein or protein complexes were analyzed by IB. The blotting signals were detected using Clarity western blot detection system (ChemiDoc MP-BioRad, USA). As for tagged-protein IP, the primary antibody and the protein A/G beads were replaced with FLAG-M2 magnetic beads (Sigma-Aldrich), S-protein agarose beads (Millipore) or HA magnetic beads (Pierce). Blot images were obtained using Image Lab 6.0.1 software (BIO-RAD, US). The antibodies and commercial assays used in this study, along with supplier names and catalog numbers of commercial reagents, are provided in Supplementary Data [296]1. RNA immunoprecipitation, RNA extraction, and RT-qPCR detection The enrichment process for the protein of interest closely resembled the immunoprecipitation (IP) protocol as outlined above with the following modifications: all procedures were conducted in an RNase-free environment; the use of Ribolock RNase Inhibitor (Invitrogen) was incorporated; the lysis buffer was substituted with polysome buffer (10 mM HEPES (pH 7.0), 100 mM KCl, 5 mM MgCl[2], 25 mM EDTA, 0.5% Nonidet P-40, 1% Triton X-100,0.1% SDS, 10% glycerol, 1 mM DTT and 80 U/mL of Ribolock RNase inhibitor); and the wash buffer was changed to NT2 buffer (50 mM Tris-HCl (pH 7.4), 150 mM NaCl,1 mM MgCl[2], 0.05% Nonidet P-40). RNA was extracted using TRIzol reagent (Invitrogen) following the manufacturer’s guidelines. Reverse transcription was performed with Reverse Transcription SuperMix (R333, Vazyme, China). For RT-qPCR analysis on a LightCycler 480 II instrument (Roche, Switzerland), a mixture of SYBR Green master mix (Q711, Vazyme, China), cDNA, and primers (Sangon Biotech Co., Ltd., Shanghai, China) were prepared. The reaction conditions were implemented according to the SYBR Green master mix product specifications. All oligonucleotide sequences used in this study are provided in Supplementary Data [297]1. ChIRP-like RNA pull-down and mass spectrometry analysis To identify proteins interacting with endogenous Cpat, we employed a ChIRP-like RNA pull-down approach, using biotin-labelled sense and antisense RNA probes complementary to Cpat for the capture of endogenous lncRNA-protein complexes. This strategy avoids overexpression artifacts and allows detection of physiologically relevant RNA-protein interactions. Biotin-labelled RNA was generated by GenePharma. Cell lysate was prepared using IP Lysis Buffer (Thermo Fisher, 87787) with complete protease inhibitor cocktail (Roche) and Ribolock RNase Inhibitor (Invitrogen). The RNA pull-down was performed using the Pierce™ Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher, 20164). Briefly, streptavidin magnetic beads were washed with wash buffer according to the manufacturer’s instructions and then incubated with 100 pmol of biotin-labelled RNA (sense and antisense separately) in capture buffer with RNase inhibitor for 30 min at RT. RNA-captured beads were washed once with the wash buffer. RNA-captured beads and non-RNA-captured bead control were incubated separately with prepared cell lysates (30 mg) in the master mix of RNA-Protein binding reaction for 2 h at RT with gentle rotation. Then, beads were washed with Wash Buffer 4 times. Finally, 40 μl of 1xSDS loading buffer was added at 95 °C for 5 min. The product was subjected to mass spectrometry (MS) analysis. MS analysis was performed by HAOGENE Biotechnology Co., Ltd, Hangzhou, China. For immunoblot detection, 0.5-1 mg of cell lysate and 1-3 μg of biotin RNA were used. For the purified protein RNA pull-down assay, 1-2 μg of purified protein and 1-3 μg of biotin RNA were used. Subcellular fractionation Nuclear and cytoplasmic RNA fractions of cardiomyocytes were isolated using RNA Subcellular Isolation Kit (Active Motif, Catalog No. 25501) according to the manufacturer’s instructions with addition of RNase inhibitor. The extracted nuclear and cytoplasmic RNA fractions were used to detect Cpat subcellular content by RT-qPCR. The expression of Actb was used as a cytoplasmic control, and U6 was used as a nuclear control. Isolation of mouse heart mitochondria and cardiomyocyte mitochondria Mouse heart mitochondria were isolated from freshly harvested heart tissue. A 100 mg sample of heart tissue was cut into 1 mm^3 dices, then washed and minced in 1 mL of cold PBS. The minced tissue was subsequently digested in trypsin digestion solution (Beyotime, C3606-3) for 20 min at ice-cold. Digested tissue was transferred to a precooled glass Potter homogenizer along with 1 mL of isolation buffer A (Beyotime, C3606-1). The mixture was homogenized using 20 strokes at medium speed, followed by centrifugation for 5 min at 600 g at 4 °C. The supernatant was then collected into a new tube and centrifuged for 10 min at 3500 g at 4 °C. Post-centrifugation, the supernatant was carefully transferred to a separate tube for cytoplasmic protein analysis, while the remaining mitochondrial pellet was washed with 1 mL of mitochondria storage buffer (Beyotime, C3606-4) and resuspended in 100 µL of the same buffer. The isolated heart mitochondria were maintained on ice for subsequent analyses. Crude mitochondria were first isolated from heart tissues as described above and resuspended in an isotonic buffer (20 mM HEPES [pH7.3], 210 mM mannitol, 70 mM sucrose). To obtain mitoplasts, the crude mitochondrial pellet was incubated with an equal volume of hypotonic buffer (20 mM HEPES [pH7.3], 70 mM sucrose) on ice for 30 min to induce outer membrane rupture. The swelling was stopped by adding an equal volume of hypertonic buffer (20 mM HEPES [pH7.3], 210 mM mannitol, 70 mM sucrose, 1 mM EDTA). After centrifugation at 12,000×g for 10 min at 4°C, the mitoplast pellet was collected. To remove contaminating cytosolic or nuclear RNAs, the pellet was treated with RNase A (2 mg/mL; Thermo Fisher) on ice for 15 min, and then RNase activity was inhibited by the addition of proteinase K (50 µg/mL; MCE). RNAs were extracted using TRIzol reagent (Invitrogen) and purified with DNase I treatment (Roche) following the manufacturer’s protocol. Cardiomyocyte mitochondria were obtained from primary cardiomyocytes using Mitochondria Isolation Kit for Cultured Cells (Thermo Scientific, 89874). Mitochondria isolation was performed according to the manual instruction. Briefly, 2x107 cultured primary cardiomyocytes were harvested, centrifuged, and incubated with Mitochondria Isolation Reagent A. Mitochondria Isolation Reagent B and C were gradually added, vortexed, and incubated after each addition. Cell fractions were deposited and discarded by centrifugation of 700 g for 10 min at 4 °C. Supernatant was transferred to a new tube and centrifuged at 12,000 g for 15 min at 4 °C. Supernatant was collected as cytosol fraction and the pellet contained the isolated mitochondria. Oxygen consumption rate (OCR) Oxygen consumption was measured using an OROBOROS Instruments High-Resolution Respirometer Oxygraph-2k (Innsbruck, Austria). DatLab software (OROBOROS INSTRUMENTS) was used for data acquisition and analysis. Chambers were washed by ddH[2]O three times before use and filled with 2 mL MiR05 respiration buffer (20 mM Taurine, 0.5 mM EGTA, 3 mM MgCl[2].6H[2]O, 60 mM K-lactobionate, 10 mM KH2PO4, 20 mM HEPES, 110 mM D-sucrose, and 1 g/L fatty acid–free BSA). Oxygen concentration was calibrated before the start. For mitochondria extracted from fresh mouse heart ventricle, same number of mitochondria was added to each parallel chamber. The reagents are added in the following order and dosage: pyruvate (PMG, 5 mM; malate, 0.5 mM; glutamate, 10 mM) 20 μl, ADP (2.5, mM) 5 μl, cytochrome c (Cyto.c, 10 μM) 5 μl, Rotenone (Rot, 0.5 μM) 1 μl, succinate (Succ, 10 mM) 20 μl, antimycin a (AA), 2.5 μM) 1 μl, ascorbate (AS, 10 mM) 5 μl + N,N,N’,N’-Tetramethyl-p-phenylenediamine dihydrochloride (TMPD, 100 μM) 5 μl, and azide (AZD, 100 mM) 50 μl. Each reagent was added while oxygen flux was stable. PMG was used to determine complex I respiration; complex I inhibitor Rot and Succ were used to determine complex II respiration; the complex III inhibitor AA and TMPD plus AS were added sequentially to determine complex IV respiration as indicated in the Fig. [298]3D. AZD was used to terminate all respiration including non-mitochondrial respiration. For cultured primary cardiomyocytes, approximately 3x10^6 cells suspended in DMEM were used per parallel chamber. Basal respiration (ROUTINE) was recorded until a steady state was reached. To assess non-phosphorylating respiration (LEAK), 2.5 mM oligomycin was added to inhibit ATP synthase, and the respiration rate were monitored until a steady state was achieved. Maximal electron transfer system (ETS) capacity was determined by titrating FCCP (carbonyl cyanide 4-[trifluoromethoxy] phenylhydrazone, mitochondrial oxidative phosphorylation uncoupler) in 0.5 mM increments until no further increase in respiration rate was observed. Finally, 2.5 mM antimycin A and 1 mM rotenone were added to inhibit complex III and complex I, respectively. ATP measurement An amount of 10^3 ~ 10^4 neonatal mouse cardiomyocytes were directly cultured in the assay microplate. ADP/ATP ratio was measured by ADP/ATP Ratio Assay Kit (Sigma-Aldrich, MAK135) according to the manufacturer’s protocol. For ATP assay, 90 μl of ATP reagent was added to each well in plate and tap to mix after culture medium was removed. The plate was incubated for 1 min at RT. Read luminescence (relative light units) was used on a luminometer for the ATP assay (RLUA). For the ADP assay, the plate was incubated for 10 more min and the luminescence for ATP (RLUB) was obtained. Immediately following this, 5 μl of ADP Reagent was added to each well and mixed. After 1 min, luminescence (RLUC) read was obtained. ATP/ADP ratio was calculated by the formula: ATP/ADP ratio = (RLUC-RLUB)/RLUA. NAD^+/NADH measurement Freshly isolated mouse neonatal cardiomyocytes (1x10^6) were lysed in 200 μl NAD^+/NADH extract buffer (Beyotime, S0175-5). Cell lysates used to measure NAD^+ and NADH by the NAD^+/NADH Assay Kit with WST-8 according to the manufacturer’s protocol (Beyotime, S0175). Briefly, cell lysates were heated at 60 °C for 30 min. Lysates were added to 96-well plate according to manufacturer’s protocol. The plate was incubated at 37 °C for 10min in the dark and then incubated for another 10min at 37 °C in the dark after adding 10μl chromogenic solution (Beyotime, S0175-2) to each well. NAD^+ and NADH abundances were determined by absorbance at 450 nm wavelength. The relative NADH/NAD^+ was determined as the NADH abundance divided by the NAD^+ abundance. Transmission electron microscopy Fresh mouse hearts were collected and immediately fixed in 2.5% glutaraldehyde/ 1.6% paraformaldehyde in 0.1 M pH 7.2 sodium cacodylate buffer at 4 °C overnight. The samples were then rinsed three times in the same fixation buffer and post-fixed with 1% osmium tetroxide for 1 h at RT before rinsing three times with dH[2]O for 10 min. Samples were stained with 1% uranyl acetate for 30 min at RT and dehydrated through a graded ethanol series (10, 30, 50, 70, 85, 95%), and three times with 100% ethanol. Samples were then infiltrated with two changes of 100% propylene oxide before leaving overnight in a 50%/50% propylene oxide/SPI-Pon 812 resin mixture. Over the following 2 days, seven changes of fresh 100% SPI-Pon 812 resin were performed before the samples were polymerized at 68 °C in flat molds. The samples were then reoriented for horizontal sections of the heart tissue layers. The thin sections ( ~ 70 nm) were placed on gold support grids and contrasted with lead citrate and uranyl acetate. Sections were examined using the HC-1 80.0 kV accelerating voltage, and images were captured using a TEM HT7800 (Hitachi TEM system). Biotinylation and purification of 4-SU-labeled RNA For metabolic labeling of nascent RNAs, 4-thiouridine (4-SU, Sigma-Aldrich, T4509) was added to a 100 μM final concentration in the culture medium for 12 h. After RNAs were extracted, 4-SU-labeled RNA was biotinylated and pulled down using streptavidin beads, isolated, and used for RT-qPCR analysis. Immunofluorescent staining and RNA fluorescence in situ hybridization (FISH) Sterile cell coverslips were placed in 24-well plates, and an appropriate number of cells were cultured for treatments. For cellular immunofluorescent staining, cells were washed with PBS and fixed with 4% paraformaldehyde in PBS for 10 min at RT and permeabilized with 0.1% Triton X-100 in PBS for 10 min. Cells were then blocked with 5% BSA in PBS for 30 min. Primary antibodies were diluted in PBS at a ratio of 1:50-1:100, and 25-50 µL of the antibody solution was evenly distributed in a 6-cm dish. Coverslips were inverted onto the antibody solution and incubated overnight at 4 °C. After incubation, coverslips were returned to the 24-well plates with the cell side facing up, washed 2 ~ 3 times with PBS, and then incubated with fluorescent secondary antibodies (Alexa Fluor 594 or 488 conjugate, 1:200) for 2 h using the same method. Following three washes with PBS, cells were incubated with DAPI for 15 min and washed with PBS before mounting. Immunofluorescent images were acquired on a Leica DM3000 (Leica Microsystems, German). RNA fluorescence in situ hybridization (FISH) was performed according to the FISH kit (Ribobio) manufacturer’s instructions with minor modifications. After secondary antibody incubation, cells on coverslips were fixed again with 4% paraformaldehyde/PBS for 10 min and then washed with PBS. Coverslips were then incubated with 20 nM Cpat probes in hybridization buffer (2xSSC containing 100 mg/mL dextran sulfate and 10% formamide). Hybridization was carried out at 37 °C for 8 ~ 12 h. After hybridization, coverslips were washed twice with 2xSSC at RT, incubated with DAPI for 15 min, and then washed with 2xSSC before mounting. Fluorescent signals were visualized using a Leica STELLARIS 5 (Leica Microsystems), and all images for each channel were acquired with identical settings. Liquid chromatography and mass spectrometry CS protein was immunoprecipitated using CS antibody-conjugated magnetic beads at 4 °C overnight. Beads were washed and eluted with loading buffer for immunoblot analysis. Gels were stained with Coomassie blue, and the target CS protein bands were excised for further analysis. Liquid chromatography and tandem mass spectrometry (LC-MS/MS) analysis was then performed by Aimsmass (Shanghai, China). Gel bands were excised and cut into 1 mm³ pieces, followed by de-staining, washing, and dehydration with acetonitrile. Proteins were reduced with dithiothreitol (DTT) at 50 °C for 30 min, alkylated with iodoacetamide (IAA) in dark at 25 °C for 40 min, and washed. After dehydration, gel pieces were digested with trypsin at 37 °C overnight. Peptides were extracted, and the process was repeated for complete recovery. Samples ( ≥ 200 µg) were processed in ultrafiltration units using urea (UA) buffer and centrifuged at 12,000× g. After washing, proteins were reduced with IAA in dark and centrifuged. The washing steps were repeated with UA and ammonium bicarbonate buffers. Proteins were digested with trypsin at 37 °C overnight, and peptides were collected via centrifugation. Peptides were desalted using acetonitrile and 0.1% formic acid washes via centrifugation at 2500 g. The desalted peptides were eluted with 60% acetonitrile and 0.1% formic acid, followed by lyophilization. The samples were dissolved in 0.1% formic acid and prepared for MS analysis. Peptide separation was performed using a C18 column (1.8 µm, 0.15x 100 mm) on a Thermo Fisher Easy-nLC 1000 system, with a 98-min gradient at a flow rate of 600 nL/min. The mobile phases consisted of 0.1% formic acid (Solution A) and 100% acetonitrile (Solution B). MS/MS was conducted using a Thermo Fisher LTQ Orbitrap ETD. Raw MS data were processed using Proteome Discoverer 1.4 (Thermo) for database searching and analysis. Acetylated peptides were identified, and differential acetylation sites between the groups were screened. Statistical analysis The experiments were conducted independently with a minimum of three replications of each. Data normality was assessed using the Shapiro-Wilk test. For data exhibiting normal distribution between two groups, an unpaired two-tailed Student’s t-test was used. In cases where normality was not met, the Mann-Whitney U test was employed. For comparisons involving multiple groups with normally distributed data, one-way or two-way ANOVA followed by Tukey’s multiple comparisons test. Conversely, the Kruskal-Wallis nonparametric test was applied for non-normally distributed data. Kaplan-Meier survival curves were compared using the log-rank (Mantel-Cox) test (d, l). Statistical analyses and data visualizations were performed using the GraphPad Prism 8.0 software (San Diego, CA). Reporting summary Further information on research design is available in the [299]Nature Portfolio Reporting Summary linked to this article. Supplementary information [300]Supplementary Information^ (1.1MB, pdf) [301]41467_2025_64072_MOESM2_ESM.pdf^ (175.7KB, pdf) Description of Additional Supplementary File [302]Supplementary Data 1^ (17.7KB, xlsx) [303]Reporting Summary^ (103.3KB, pdf) [304]Transparent Peer Review file^ (1.4MB, pdf) Source data [305]Source Data^ (3.9MB, xlsx) Acknowledgements