Lipid droplets (LDs) are cellular organelles critical for lipid homeostasis, with intramyocyte LD accumulation implicated in metabolic disorder-associated heart diseases. Here we identify a human long non-coding RNA, Lipid-Droplet Transporter (LIPTER), essential for LD transport in human cardiomyocytes. LIPTER binds phosphatidic acid and phosphatidylinositol 4-phosphate on LD surface membranes and the MYH10 protein, connecting LDs to the MYH10-ACTIN cytoskeleton and facilitating LD transport. LIPTER and MYH10 deficiencies impair LD trafficking, mitochondrial function and survival of human induced pluripotent stem cell-derived cardiomyocytes. Conditional Myh10 deletion in mouse cardiomyocytes leads to LD accumulation, reduced fatty acid oxidation and compromised cardiac function. We identify NKX2.5 as the primary regulator of cardiomyocyte-specific LIPTER transcription. Notably, LIPTER transgenic expression mitigates cardiac lipotoxicity, preserves cardiac function and alleviates cardiomyopathies in high-fat-diet-fed and Leprdb/db mice. Our findings unveil a molecular connector role of LIPTER in intramyocyte LD transport, crucial for lipid metabolism of the human heart, and hold significant clinical implications for treating metabolic syndrome-associated heart diseases.
Lipid droplets (LDs) are highly dynamic cellular organelles ubiquitously existing in prokaryotes and eukaryotic cells. Synthesized in the endoplasmic reticulum (ER), LDs consist of a neutral lipid core, primarily containing triacylglycerols (TAGs) and sterol esters, surrounded by a phospholipid monolayer with associated proteins1. Although initially considered storage reservoir for excess lipids, LDs have been increasingly recognized for their critical roles in intracellular lipid trafficking, lipid homeostasis and membrane synthesis1,2. Under hyperlipidaemic conditions, LDs accumulate in the myocardium (that is, cardiac steatosis) and cardiomyocytes (CMs)2. While increased LD levels afford transient cardiac protection by storing surplus fatty acids (FAs) in CMs3,4, long-term lipid intermediary accumulation results in detrimental outcomes, such as apoptosis, tissue injury and cardiac dysfunction (that is, lipotoxicity)2,5,6. Intramyocyte LD accumulation in humans is associated with cardiomyopathies and heart failure in hyperlipidaemia-related metabolic disorders, including obesity and diabetes mellitus6,7. Animal model studies have also revealed that an expanded pool of LDs can lead to early impairment of cardiac contractility and subsequent heart dysfunction8,9,10. Despite these findings, the CM intrinsic mechanisms determining LD accumulation and mobilization, as well as the aetiology of intramyocyte LD accumulation in obesity and diabetes mellitus, remain poorly understood.
LDs play a vital role in cellular metabolism, particularly under nutrient-deprived conditions. Under starvation conditions, LDs are mobilized and transported to mitochondria for β-oxidation to generate energy for the cell11. TAG within LD core is digested to FAs via lipolysis and lipophagy processes on LDs1,12. The extensive interactions of LDs with various organelles, such as the ER, endosomes and mitochondria, facilitate intracellular lipid trafficking and channelling13. LD-associated proteins mediate LD–organelle membrane contacts. For example, FATP1 and DGAT2 at the LD–ER interface regulate LD expansion14, while LD–mitochondria tethering in skeletal and heart muscle is mediated by PLIN5 for channelling FAs from LDs to mitochondria15. Although LD biogenesis, composition and turnover have been extensively studied, the molecular mechanism underlying LD transport from the ER to other organelles remains elusive. Previous studies have implicated myosin motors, such as Myosin I and V, in the transport of ER-derived vesicles16,17. Additionally, Myh9 depletion has been shown to impair LD dissociation from the ER18. Mammalian CMs predominantly express non-muscle myosin IIB (Myh10) (ref. 19), which can interact with membrane lipids20 and link cytoskeleton with the ER21. While the key roles of protein factors in intracellular LD trafficking have been extensively studied, the direct involvement of RNA molecules in this process remains unknown. Gaining insight into this aspect is crucial for understanding intramyocyte lipid trafficking and channelling, which are essential for maintaining lipid homeostasis in the human heart.
Although RNA is primarily known for carrying and regulating genetic information, recent evidence indicates that RNA interacts with various molecules, such as nucleic acids, ions, proteins and lipids22, to perform diverse functions. RNA–lipid interactions have revealed new regulatory mechanisms in essential cellular processes, implying previously undefined functions of RNA. In this study, we identified a human long non-coding RNA (lncRNA), named Lipid-Droplet Transporter (LIPTER), which facilitates LD transport within human CMs. LIPTER deficiency impairs LD transport and metabolism, mitochondrial function and CM viability. Mechanistically, LIPTER directly binds two phospholipids on LD membrane, phosphatidic acid (PA) and phosphatidylinositol 4-phosphate (PI4P), as well as the MYH10 motor protein, connecting LDs with cytoskeleton for intramyocyte transport. Importantly, LIPTER overexpression mitigates cardiomyopathies and preserves cardiac functions in mouse models of obesity and diabetes, highlighting LIPTER’s potential clinical relevance in treating human metabolic syndrome-associated heart disease and failure. Our findings reveal that RNA can directly participate in intracellular vesicle transport, expanding RNA’s functional dimensions and suggesting a potential ancient role of RNA–lipid interaction in forming the first living system in the primordial ‘RNA world’.
LINC00881 (LIPTER) is specifically expressed in human CMs
Using our previously established method23, human pluripotent stem cells (hPSCs) were differentiated into cardiovascular cells. Transcriptomic profiles of hPSCs, hPSC-derived multipotential cardiovascular progenitors (MCPs), CMs, smooth muscle cells (SMCs), endothelial cells (ECs)24 and left ventricle tissues from individuals with and without type 2 diabetes (T2DM) (3 individuals per group) were compared. We identified the top four CM-enriched lncRNAs, LINC00881, TTN-AS1, SLC8A1-AS1 and NAV2-AS2, which were downregulated in T2DM hearts compared with non-T2DM hearts (Fig. 1a,b and Extended Data Fig. 1a). However, RT–qPCR validation confirmed that only LINC00881 expression was significantly reduced to less than 50% (P < 0.05) in all collected T2DM hearts compared with non-T2DM hearts (Fig. 1c and Extended Data Fig. 1b). Multiple tissue-specific gene expression databases demonstrated that LINC00881 expression is highly enriched in the human hearts (Fig. 1d and Extended Data Fig. 1c–e). Additionally, the analysis of single-cell RNA sequencing (scRNA-seq) data from a 6.5–7 week human embryonic heart25 revealed that LINC00881 expression was specifically enriched in the NKX2.5+ and cardiac Troponin T (cTnT)+ CM clusters (Fig. 1e), with an expression level comparable to the key cardiac transcription factor (TF) NKX2.5 (ref. 26) (Fig. 1f). LINC00881 is conserved in human and non-human primates but not across other species (Fig. 1g). Although three putative open reading frames (ORFs) were predicted in LINC00881 by ORF-FINDER27 (Extended Data Fig. 2a), the precited peptides were not detected using western blotting or immunostaining (Extended Data Fig. 2b,c), albeit increased RNA level of LINC00881 detected in transfected 293T cells (Extended Data Fig. 2d). Altogether, these results demonstrate that LINC00881 is highly and specifically enriched in human CMs and significantly downregulated in human T2DM hearts. In this study, LINC00881 was renamed LIPTER due to its role in LD transport within human CMs.
LIPTER deficiency impairs LD metabolism and transport
During CM differentiation from human induced pluripotent stem cells (hiPSCs) by forming embryoid bodies23 (EBs; Fig. 1h), LIPTER expression rapidly increased by over 10,000-fold (CMs versus hiPSCs), similar to NKX2.5 (Fig. 1i). Next, two LIPTER knockout (LIPTERKO) hiPSC clones were established by CRISPR/Cas-9 (ref. 28) (Fig. 1j,k and Extended Data Fig. 3a). LIPTERKO did not affect the ratios of beating EBs and cTnT+ CMs or the expression levels of CM markers, cTnT and MYH6, at day 20 of CM differentiation (Extended Data Fig. 3b–e), indicating that LIPTER is not required for CM differentiation. However, since LIPTER expression declines in T2DM hearts (Fig. 1c), we cultured hiPSC-EBs with high glucose (22.75 mM) for an additional 20 days. At day 40, significantly reduced ratios of cTnT+ CMs were observed in LIPTERKO compared with wild-type (WT) hiPSC-EBs (Fig. 1l and Extended Data Fig. 3f), suggesting that LIPTER is required for CM survival under high-glucose condition and implicating a potential role of LIPTER in CM metabolism. To investigate global changes in metabolites upon LIPTER deficiency, untargeted metabolomics (Fig. 2a and Supplementary Table 1) were performed on enriched day 40 hiPSC-CMs. The top increased metabolites in LIPTERKO compared with WT hiPSC-CMs were lipid and lipid-like molecules, including phosphatidylinositol (PI), PAs and TAGs. Notably, TAG is the main core lipid of LDs1,29, while PI and PA are phospholipids on LD membranes29. These data suggest increased LD levels in LIPTERKO compared with WT hiPSC-CMs, which was confirmed by Oil Red O and Nile Red lipid staining (LIPTERKO versus WT; Fig. 2b,c and Extended Data Fig. 3g,h). Finally, LIPTER expression was rescued in LIPTERKO (LIPTERKO/OE) hiPSCs using doxycycline-inducible lentivirus. After CM differentiation, LIPTER expression was induced, resulting in a two- to fourfold increase compared with WT hiPSC-CMs (Extended Data Fig. 3i). We observed that rescued LIPTER significantly repressed LD accumulation in LIPTERKO hiPSC-CMs (LIPTERKO/OE versus LIPTERKO; Fig. 2b,c and Extended Data Fig. 3g,h). Collectively, these data demonstrate that LIPTER deficiency disrupts lipid metabolism and enhances LD accumulation in hiPSC-CMs.
Intramyocyte LD accumulation could be due to increased uptake of free FAs via CD36, a membrane receptor to import extracellular FAs30. However, both CD36 expression (Extended Data Fig. 4a) and lipid uptake capability were reduced in LIPTERKO hiPSC-CMs compared with WT hiPSC-CMs (Fig. 2d,e), indicating that LD accumulation was not a result of enhanced FA uptake. We detected no significant changes in genes for TAG synthesis and lipolysis on LDs, except for ~30% reductions of GPAM and ATGL1 (refs. 1,31) (Extended Data Fig. 4a–c), upon LIPTERKO, suggesting that accumulated LDs in LIPTERKO CMs might not be primarily caused by altered LD formation or lipolysis process. Interestingly, LIPTERKO prominently reduced expression and induced accumulation of PLN5 (Extended Data Fig. 4a–d), which was reported to mediate LD–mitochondria tethering in CMs15. These results implied that LIPTERKO could impair LD transport to and/or interaction with mitochondria.
To trace LD transport, hiPSC-CMs were cultured with palmitate (200 μM) for 6 h to induce LD formation (Fig. 2d). In WT hiPSC-CMs, LDs surrounded nucleus and were broadly distributed in the cytoplasm (Fig. 2f, top, arrowheads), while cytosolic transport and distribution of LDs were retarded in LIPTERKO hiPSC-CMs (Fig. 2f,bottom). We then cultured CMs without palmitate for additional 12 h to mobilize the LDs (Fig. 2d). While LDs in WT hiPSC-CMs were fully mobilized, large LDs accumulated in LIPTERKO hiPSC-CMs (Fig. 2g, arrowheads). Statistical results show that, from 6 h to 18 h, LD densities in WT hiPSC-CMs rapidly declined in whole CMs, cytoplasm and the surrounding region of nucleus, whereas high amounts of LDs remained in LIPTERKO hiPSC-CMs (Fig. 2h). Since mobilized LDs could transfer stored lipids to mitochondria for β-oxidation1,2, we conducted live cell imaging to trace LD–mitochondria interaction. Rhodamine-B-labelled palmitic acid was added to CM culture for 6 h and enriched in LDs. In WT hiPSC-CMs, approximately 28% of LDsRh-B fused with mitochondria and then disappeared (Extended Data Fig. 4e, top, yellow arrows; Extended Data Fig. 4f), whereas only about 8% of LDsRh-B fused with mitochondria in LIPTERKO hiPSC-CMs (Extended Data Fig. 4e, middle, red arrows; Extended Data Fig. 4f). Finally, mitochondria were isolated from hiPSC-CMs after 2 h of Rhodamine-B-labelled palmitic acid treatment. Significantly reduced fluorescence levels were detected in LIPTERKO versus WT mitochondria (Fig. 2i), indicating reduced Rhodamine-B-labelled palmitic acid transfer to mitochondria upon LIPTER ablation. Collectively, these results demonstrate that LIPTER deficiency disrupts LD transport and mobilization in human CMs, leading to extensive LD accumulation.
LIPTER deficiency induces mitochondrial dysfunction and apoptosis
Next, we performed whole messenger RNA sequencing (mRNA-seq) (Supplementary Table 2) and discovered that LIPTERKO altered the transcriptome of hiPSC-CMs (Fig. 3a). The genes with significant expression changes were enriched into Gene Ontology (GO) pathways, including elevated apoptotic signalling, aberrant mitochondrial functions and lipid storage (Fig. 3b), as well as toxicity signalling pathways, such as increased heart failure and cardiac dysfunction (Fig. 3c). Consistently, transmission electron microscopy (TEM) revealed compact and rod-shaped mitochondria in WT hiPSC-CMs (Fig. 3d, yellow arrowheads), while ~60% of mitochondria in LIPTERKO hiPSC-CMs displayed giant and/or swollen morphology (Fig. 3d, red arrows, Fig. 3e), indicating mitochondrial dysfunction. Mass spectrometry (MS) of purified mitochondria revealed that LIPTERKO altered the expressions of mitochondrial respiratory proteins, such as reduced levels of NDUFB2, NDUFB11 and COX7B (Supplementary Table 3). Compared with WT hiPSC-CMs, mitochondrial maximal respiration capacity, spare respiratory capacity (Fig. 3f,g) and long-chain fatty acid oxidation (FAO) capacities of LIPTERKO hiPSC-CMs (Fig. 3h and Extended Data Fig. 4g) were significantly reduced. Furthermore, increased apoptosis was detected in LIPTERKO versus WT hiPSC-CMs (Fig. 3i,j and Extended Data Fig. 4h,i). To investigate whether these CM abnormalities were LIPTER dependent, LIPTER expression was rescued in LIPTERKO hiPSC-CMs, which restored the swollen and dysfunctional mitochondria and suppressed apoptosis of LIPTERKO CMs (Fig. 3i,j, LIPTERKO/OE versus LIPTERKO). Consequently, rescued LIPTER enhanced the ratios of cTnT+ CMs at day 50 of EB differentiation compared with LIPTERKO (Fig. 3k,l). Altogether, these results reveal the crucial roles of LIPTER in mitochondrial function and viability of human CMs.
NKX2.5 controls CM-specific LIPTER transcription
Since LIPTER expression is enriched in human CMs and downregulated in T2DM hearts (Figs. 1b,c and 4a) and extensive LDs were observed in both LIPTERKO and T2DM human CMs (Figs. 2b,c and 4b,c), we posited that LIPTER transcription was under the control of a CM-specific regulatory mechanism that could respond to hyperglycaemia. The LIPTER promoter (~2.5 kb upstream of transcriptional start site) was predicted to contain multiple putative binding sites for NKX2.5 RXRA and CEBPB, which are associated with CM-specific gene regulation and lipid metabolism (Fig. 4d). NKX2.5 was the most effective in increasing LIPTER promotor activity, and its co-transfection with RXRA or CEBPB further enhanced the activity (Fig. 4e). Knockdown of NKX2.5 (NKX2.5KD) by NKX2.5-shRNAs led to reduced expressions of NKX2.5 and LIPTER (Fig. 4f), but not CM markers cTnT and MYH10 (Extended Data Fig. 5a). NKX2.5KD also prominently increased LD accumulation and apoptosis of WT hiPSC-CMs, phenocopying LIPTERKO hiPSC-CMs (Fig. 4g,h and Extended Data Fig. 5b–e, shNKX2.5 versus shControl in WT CMs). All these results indicate that NKX2.5 can control LIPTER transcription in human CMs. Next, we found exposure to high glucose levels (11.0 and 22.0 mM) for 14 days significantly reduced LIPTER expression (Fig. 4i), and the mRNA and nuclear protein levels of NKX2.5, RXRA and CEBPB in WT hiPSC-CMs (Fig. 4j,k and Extended Data Fig. 6a,b). Similar reductions in these TFs were observed in T2DM hearts compared with non-T2DM human hearts (Fig. 4l,m and Extended Data Fig. 6c,d). These results are consistent with previous reports that high glucose suppressed NKX2.5 (ref. 32) and RXRA33 expressions in vivo. Finally, LIPTER overexpression was found to repress NKX2.5KD-induced LD accumulation and apoptosis (Extended Data Fig. 5b–e, shNKX2.5 versus shControl in LIPTERKO/OE CMs), suggesting a therapeutic potential of LIPTER for treating diabetes-associated dilated cardiomyopathy (DCM)34. Altogether, these results demonstrate that NKX2.5 is the primarily regulator of CM-specific LIPTER transcription and that hyperglycaemia can repress the NKX2.5–LIPTER axis expression in human CMs.
LIPTER selectively binds phospholipids and MYH10 protein
We sought to elucidate the molecular mechanisms by which LIPTER regulates LD transport. RT–qPCR detected cytosolic LIPTER expression with GAPDH, but not in the nucleus of hiPSC-CMs (Extended Data Fig. 7a). RNA fluorescence in situ hybridization (RNA FISH) revealed that ~74% of LDs co-localized with LIPTER (Fig. 5a,b), suggesting potential binding between LIPTER and LDs. We then isolated total lipids from WT hiPSC-CMs, which contain lipid-binding RNAs, followed with RT–qPCR. Compared with the house-keeping gene ACTB, LIPTER was highly enriched in isolated lipids (>150-fold, Fig. 5c), suggesting a possible direct interaction between LIPTER and lipids. Next, using the RNA–lipid overlay assay35, we identified that LIPTER selectively bound PA and PI4P, whereas antisense-LIPTER (AS-LIPTER) could not bind any lipid (Fig. 5d). Given that PA and PI4P are found on the LD membranes, we asked whether LIPTER could bind PA and PI4P in a membrane context. Giant unilamellar vesicles (GUVs) generated with TopFluor-labelled PI4P/PA were incubated with Alexa-594-labelled LIPTER/AS-LIPTER, respectively. LIPTER bound both PA- and PI4P-GUVs, whereas AS-LIPTER did not (Fig. 5e). Furthermore, neither LIPTER nor AS-LIPTER bound GUVs generated by PI (Extended Data Fig. 7b). The binding affinities of LIPTER-PA and LIPTER-PI4P were assessed using microscale thermophoresis (MST) assay36, with Kd values of 706 and 70 in the liquid phase (Fig. 5f,g), indicating strong and specific LIPTER-PA and LIPTER-PI4P interactions, respectively. Collectively, these data reveal the selective binding of LIPTER with PA and PI4P on LD membranes.
Next, we employed the MS2-BioTRAP system to identify LIPTER-interactive proteins and trace LIPTER in live hiPSC-CMs as previously described37 (Fig. 5h). As shown in Fig. 5i lower panels, control hiPSC-CMs expressed MS2-tag and MS2 binding protein MS2YFP, which was ubiquitously distributed in CMs. In comparison, MS2YFP bound to MS2-tagged-LIPTER, forming green LIPTER-MS2YFP particles observed only in the cytoplasm and co-localized with LDs (Fig. 5i, top right, yellow arrows). Since cytosolic LIPTER–LD co-localization was observed by both MS2-BioTRAP and RNA FISH (Fig. 5a,i), these results indicate that the MS2-tag did not affect cellular location and interaction of LIPTER. LIPTER-interacting proteins were then pulled down in 293T cells and hiPSC-CMs using antibody against MS2FLAG or MS2YFP, respectively, followed by MS analysis and western blotting validation. LIPTER specifically pulled down non-muscle Myosin IIB (NM-IIB, or MYH10) protein, as compared with AS-LIPTER (Fig. 5j and Extended Data Fig. 7c,d). RNA immunoprecipitation further validated LIPTER enrichment by anti-MYH10 antibody in hiPSC-CMs (Fig. 5k). Confocal fluorescence microscopy revealed co-localizations of LIPTER with LDs and MYH10 in hiPSC-CMs (Extended Data Fig. 7e and Supplementary Video 1). Live cell imaging showed LIPTER-MS2YFP co-localizing with migrating LDs in the cytosol of hiPSC-CMs (Supplementary Video 2). We observed the MYH10-ACTIN cytoskeleton in hiPSC-CMs (Extended Data Fig. 7f), as myosin motors move along Actin filaments16,17. Altogether, these results demonstrate that LIPTER can interact with the cytoskeleton via binding MYH10.
To identify LIPTER domains that specifically interact with PA, PI4P and MYH10, we generated truncated LIPTER fragments (Fig. 5l) for RNA–lipid overlay and GUVs assays. Exon 3 of LIPTER was found to contain binding domains with PA and PI4P (Fig. 5m–o). Further analysis revealed the PI4P binding domain on exon 3–1, and PA binding domains on both exon 3–2 and exon 3–3 regions (Fig. 5m). A MYH10-binding domain was identified on exon 3–4 (Fig. 5p). LD–LIPTER–MYH10 interactions were observed in hiPSC-CM using confocal fluorescence microscopy (Fig. 5q). These results demonstrate that LIPTER selectively binds PA, PI4P and MYH10 via distinct RNA domains. Overall, our data demonstrate that LIPTER functions as a molecular linker, connecting LDs with the cytoskeleton for intramyocyte LD transport (Fig. 5r).
MYH10 is indispensable for LIPTER function in CMs
MYH10 was knocked out (MYH10KO) in hiPSCs using CRISPR/Cas-9 (Fig. 6a,b and Extended Data Fig. 8a). Compared with WT hiPSC-CMs, MYH10KO hiPSC-CMs displayed increased LD accumulation (Fig. 6c,d), ~50% swollen mitochondria (Fig. 6e,f), reduced mitochondrial spare respiratory and maximal respiration capacities (Fig. 6g), and decreased FAO capacity (Fig. 6h and Extended Data Fig. 8b). Additionally, MYH10KO hiPSC-CMs exhibited increased apoptosis (Fig. 6i and Extended Data Fig. 8c), retarded LD transport (Extended Data Fig. 8d,d′), reduced LD–mitochondria fusion rate (third row, Extended Data Fig. 4e,f) and decreased Rhodamine-B-labelled palmitic acid transport into mitochondria (Extended Data Fig. 8e). Additionally, inhibition of MYH10 ATPase activity with (S)-(−)-Blebbistatin38 led to enhanced LD accumulation and apoptosis in WT hiPSC-CMs compared with its inactive enantiomer control, (R)-(+)-Blebbistatin (Extended Data Fig. 8f–h). These results demonstrate that loss-of-MYH10 phenocopies LIPTER deficiency in hiPSC-CMs. We then investigated whether MYH10 is required for LIPTER function by inhibiting MYH10 in LIPTER-overexpressing CMs (Fig. 6j). Compared with (R)-(+)-Blebbistatin, (S)-(−)-Blebbistatin treatment enhanced LD accumulation (Fig. 6k,l), reduced FAO capacity (Fig. 6m) and elevated apoptosis (Fig. 6n) in LIPTERKO/OE hiPSC-CMs. These results reveal that inhibition of MYH10 could abolish LIPTER gain-of-function, indicating the critical role of MYH10 in executing LIPTER function.
Unlike LIPTER, MYH10 is conserved in humans and mice. Therefore, Myh10 was conditionally knocked out in mouse CMs (Myh10CKO) by crossing Myh10f/f with Tnnt2Cre mice (Fig. 6o–p and Extended Data Fig. 8i). Fed with high-fat diet (HFD, 45 kcal% fat) for 3 months, Myh10CKO male mice exhibited severe cardiac lipid accumulation, including enhanced LD deposition (Fig. 6q, arrowhead) and elevated concentrations of TAG and FAs, compared with Myh10f/f littermates (Fig. 6r,s). Notably, Myh10 deficiency significantly reduced the FAO rates of whole Myh10CKO hearts compared with Myh10f/f hearts (Fig. 6t). Lastly, left ventricular ejection fraction (EF) and fractional shortening (FS) values significantly declined in Myh10CKO mice compared with Myh10f/f mice (Fig. 6u,v). Collectively, all these in vitro and in vivo results demonstrate a conserved and critical role of Myh10 in LD metabolism of mammalian heart muscle cells.
Gain-of-LIPTER mitigates lipotoxicity of hiPSC-CMs
Palmitate overload has been shown to induce cardiac lipotoxicity in vivo, which compromised mouse heart function and increased CM death39. In vitro, palmitate overload can incite oxidative and ER stress of CMs2. We found that palmitic acid (400 µM) treatment for 4 days significantly induced apoptosis in hiPSC-CMs (Fig. 7a,b and Extended Data Fig. 9a, Vector-Ctrl versus Vector-PA), which was prominently reduced by LIPTEROE (Fig. 7b,c, LIPTEROE-PA versus Vector-PA; Extended Data Fig. 9a,b). These data reveal an in vitro protective role of LIPTER against lipotoxicity.
In vivo LIPTER transgene displays a cardiac protective role
Intramyocyte LD accumulation occurs in human obese and diabetic hearts when cardiac function is still normal5,39, suggesting that LD metabolic disturbance precedes the onset of cardiac dysfunction. We hypothesized that gain-of-LIPTER function could alleviate obesity- and diabetes-associated cardiomyopathies and cardiac dysfunctions. To test this, we generated a LIPTER transgenic (LIPTER(Tg)) mouse line by knocking in LIPTER into the Rosa26 locus, resulting in high LIPTER expression in the heart (Extended Data Fig. 9c,d). Six-week-old male LIPTER(Tg) and WT mice were fed a HFD (45 kcal% fat) to induce obesity, insulin resistance and cardiac abnormalities as previously reported (Fig. 7d)40,41. After 7 months, HFD-fed LIPTER(Tg) mice showed no significant change in heart weight/tibia length ratios compared with WT mice (Extended Data Fig. 9e). However, LIPTER(Tg) significantly reduced LD levels, and FA and TAG concentrations compared with WT hearts (Fig. 7e–g). Notably, LIPTER(Tg) mice exhibited significantly enhanced FAO rates in whole hearts (Fig. 7h) and globally altered transcriptome (Supplementary Table 4). Metabolic process, energy homeostasis, cardiac muscle development and growth, and sarcoplasmic reticulum Ca2+ transport biological processes were significantly over-represented in the LIPTER(Tg) upregulated genes (Fig. 7i). We also observed significantly reduced fibrotic areas (Fig. 7j,k) and apoptosis (Extended Data Fig. 9f,g) in HFD-fed LIPTER(Tg) hearts compared with similarly treated WT hearts. Importantly, LIPTER(Tg) preserved cardiac function in HFD-fed mice. Significantly reduced EF and FS values were observed in WT mice fed with HFD for 10 months compared with WT mice fed with normal chew (NC) (Fig. 7l,m, WT-HDF versus WT-NC). However, EF/FS values of HFD-fed LIPTER(Tg) mice were significantly higher than those of HFD-fed WT mice and were preserved at levels similar to NC-fed WT mice (Fig. 7l,m, LIPTER(Tg)-HFD versus WT-HFD). These results demonstrate that gain-of-LIPTER can mitigate HFD-induced cardiomyopathy and preserve cardiac function. Furthermore, to test whether the transgenic LIPTER could interact with mouse Myh10 protein, we performed RNA immunoprecipitation sequencing (RIP-seq) to pull down and sequence all mouse Myh10-binding RNAs from LIPTER(Tg) and WT mouse hearts. Although no Myh10 binding signals were detected on control Actb and Gapdh mRNAs, mouse Myh10 protein bound LIPTER in LIPTER(Tg) heart (Extended Data Fig. 9h). Interestingly, mouse Myh10 binding signals were enriched on exon 3 of transgenic LIPTER (Fig. 7n, red arrow), which was found to interact with human MYH10 (Fig. 5p). These results suggest that LIPTER–MYH10 interaction might play a conserved role in LD transport of mammalian CMs, although the mouse homologue of LIPTER remains to be identified.
Finally, we examined the impact of CM-specific LIPTER transgene on cardiomyopathy of leptin-receptor-deficient Leprdb/db mouse (Fig. 7o), a model of T2DM and obesity that develops cardiac hypertrophy and dysfunction from 10 weeks of age42,43. We employed an AAV9 virus carrying a chicken cTnT promoter to selectively deliver LIPTER into CMs44. Retro-orbital injection of AAV9-cTnT-GFP virus (2 × 1010 vg/g) into Leprdb/db mice led to robust GFP expression in mouse CMs (Extended Data Fig. 10a,b). Subsequently, AAV9-LIPTER, control AAV9-GFP and AAV9-AS-LIPTER viruses were administered to 6-week-old male B6.BKS(D)-Leprdb/db mice using the same approach. Age-matched WT C57BL/6J male mice without AAV-9 injection also served as controls. Six weeks post AAV-9 injection, transgenic LIPTER, AS-LIPTER and GFP expressions were detected in the mouse hearts (Fig. 7p and Extended Data Fig. 10c). AAV9-LIPTER administration resulted in reduced intramyocyte LD accumulation (Fig. 7q), increased whole heart FAO rates (Fig. 7r), reduced CM size (Fig. 7s and Extended Data Fig. 10d), and preserved EF and FS values in Leprdb/db mice (Fig. 7t–u) compared with control groups. Since AAV9-LIPTER transgene in CMs did not impact blood glucose levels of Leprdb/db mice (Extended Data Fig. 10e), these findings indicate that CM-restricted LIPTER transgene can mitigate cardiomyopathy and preserve cardiac function in Leprdb/db mice.
Our data demonstrate that LIPTER functions as an RNA linker, facilitating LD transport in human CMs by connecting LDs via binding PA/PI4P and the cytoskeleton via binding MYH10. Despite the importance of lncRNAs in regulating cellular processes, the roles of lncRNA–lipid interactions remain largely unexplored. Lin et al. reported that LINK-A bound PI(3,4,5)P3 to hyperactivate AKT in cancer cells35, and SNHG9 was found to bind PA to facilitate LATS1 liquid–liquid phase separation, promoting oncogenic YAP signalling45. Our study reveals that LIPTER binds both PA and PI4P (Fig. 5d,e), which play crucial roles in membrane–membrane interactions that facilitate LD transfer. Specifically, PA on the ER membrane contributes to the formation of newly formed LDs46, while PI4P on the Golgi and plasma membrane can recruit and bind lipid transport carrier proteins on LDs47. Our MS analysis of proteins pulled down by LIPTER (Supplementary Table 5) did not detect typical LD-associated proteins, except TFG (Trafficking From ER To Golgi Regulator). However, we did not detect any interaction between TFG and LIPTER (Extended Data Fig. 10f). These findings, in conjunction with the GUV results (Fig. 5e), indicate that LIPTER can directly bind PA and PI4P on LDs without additional protein co-factors. Currently, factors determining the selective interactions of RNA with different lipids are not well understood. A few studies have suggested that RNA–lipid interaction may depend on RNA length, base pairing and nucleotide content48,49. Tomasz et al. also reported that guanine and G-quadruplex formation are critical for RNA–lipid interactions49.
Our study demonstrates that MYH10 depletions disrupted LD metabolism and transport in hiPSC-CMs, and compromised cardiac function in Myh10CKO mice, highlighting the crucial and conserved role of cytoskeleton in LD trafficking within mammalian CMs. Additionally, we observed enlarged LDs in LIPTERKO hiPSC-CMs compared with WT hiPSC-CMs (Fig. 2g). Myh9 depletion in U2OS cells was found to reduce LD dissociation from the ER and increase LD size18. Since MYH9 is not expressed in human CMs, our data suggest that disconnection of MYH10 from LDs due to LIPTER deficiency may impair the dissociation of newly formed LDs from ER, leading to increased LD size.
Diabetic cardiomyopathy is recognized as a non-ischaemic form of DCM, characterized by progressive muscle loss, global systolic dysfunction and heart failure. It has been reported that approximately 75% of patients with idiopathic DCM are diabetic34. An 85-fold increase in apoptosis and 4-fold increase in necrosis of myocytes have been observed in diabetic hearts compared with non-diseased hearts50. Our data reveal that hyperglycaemia suppresses the NKX2.5–LIPTER axis in hiPSC-CMs, and NKX2.5/LIPTER deficiencies induce LD deposition and CM death (Extended Data Fig. 5). Although the mechanism of NKX2.5 downregulation by hyperglycaemia remains to be investigated, our data suggest that a declined NKX2.5–LIPTER axis could possibility contribute to CM loss in diabetic hearts. We hypothesize that hyperglycaemia could differentially repress NKX2.5–LIPTER expression levels in individual CMs, and CMs with critically low LIPTER expression levels could gradually die, with this slow progress of muscle loss eventually leading to DCM of diabetic hearts. Since intramyocyte LD accumulation precedes onset of cardiac dysfunctions in individuals with obesity or diabetes5,39, our in vivo LIPTER transgene data strongly suggest that targeted LIPTER delivery into heart muscle might be an effective strategy to prevent cardiac dysfunction and heart failure in metabolic syndromes.
Compared with DNA, RNA exhibits a broader range of interacts with various molecules, suggesting more diverse and complex functional dimensions. LDs are ancient organelles that exist in prokaryotes and eukaryotic cells. Our findings reveal a molecular linker role of RNA in mediating the transport of lipid-covered vesicle through RNA–lipid/protein interactions. This may reflect ancient RNA functions in the primordial ‘RNA world’, where RNA–lipid/protein interactions could have play critical roles in forming the first living system, which encompassed lipid membranes and a mixture of RNA, protein and other molecules.
In summary, this study uncovers an lncRNA-mediated LD transport system in human CMs and reveals the crucial role of LIPTER and MYH10 in lipid metabolism of mammalian hearts. The mouse homologue of LIPTER has not been identified, limiting the in vivo loss-of-function analysis of LIPTER in rodents. Moreover, considering that a single lncRNA can interact with multiple proteins28,51,52, it is not yet clear whether LIPTER could regulate LD metabolism through other mechanisms, such as epigenetic regulation of genes in TAG synthesis and lipolysis on LDs (Extended Data Fig. 4a). Lastly, FAO rate measurements were conducted at a 6:1 palmitate-to-albumin ratio, much higher than the physiological ratio (1:1–3:1) (ref. 53), raising a concern of lipid overload during measurement. Additionally, whole heart FAO rate measurement using heart tissue homogenate were not performed under the same oxygen concentration as in vivo conditions, resulting in the majority of radiolabelled palmitate being converted into ASMs rather than CO2. To address these technical challenges, FAO rates will be further evaluated in live animals as previously described54.
Experimental model and subject details
WT hiPSCs were generated from fibroblasts of a healthy donor as previously in this report55 and transferred with an MTA. The hiPSCs were cultured on irradiated mouse embryonic fibroblast cells and maintained in KSR medium containing Dulbecco’s modified Eagle medium (DMEM)/F12 (Gibco, 11330032), 20% KSR (Invitrogen, 10828028), penicillin/streptomycin (P/S; Gibco, 15140163), GlutaMAX Supplement (Gibco, 35050061), non-essential amino acids (Gibco, 11140050), β-mercaptoethanol (Sigma-Aldrich, M3148) and bFGF (R&D Systems, 233-FB-025). For differentiation into CMs as previously described23,56, hiPSC-derived EBs were suspended in StemPro-34 medium (Gibco, 10639011) and treated with the following conditions: days 0–1, BMP4 (2.5 ng ml−1); days 1–4, BMP4 (10 ng ml−1), bFGF (5 ng ml−1) and Activin A (2 ng ml−1); and days 4–8, XAV-939 (Tocris, cat. no. 3748). After 20 days, beating EBs were cultured with DMEM high-glucose medium (Gibco, 11966025) containing 10% foetal bovine serum (FBS) and P/S. For monolayer CM differentiation, hiPSCs were seeded on Matrigel and treated with 12.5 nM CHIR99021 (Sigma, SML1046) in differentiation medium (RPMI-1640 with B27 minus insulin) for 24 h, which were then replaced with differentiation medium without additional factors. Two days later, cells were treated with 5 nM IWP4 (Tocris, 5214) for 48 h, and then the medium was changed every two days until day 14. The beating cells were then selected in DMEM medium containing no glucose with 4 mM lactate for four additional days. After CM selection, CMs were cultured in DMEM high-glucose medium containing 10% FBS and P/S.
Animals and diets
All animal experiments were performed following the approval and guidelines of the Institutional Animal Care and Use Committee of Indiana University. Animals are maintained in accordance with the applicable portions of the Animal Welfare act and the DHHS Guide for the Care and Use of Laboratory Animals. LIPTER(Tg) C57BL/6J mice were generated using CRISPR/Cas9 at the Genome Editing, Transgenic, and Virus Core at UPMC Magee-Women’s Research Institute. Six-week-old male homozygous Leprdb/db mice were obtained from the Jackson Laboratory (strain number 000697). Myh10 floxed mice with C57BL/6J background were obtained from Dr Robert S. Adelstein lab at Laboratory of Molecular Cardiology, National Heart, Lung and Blood Institute57. C57BL/6J Tnnt2Cre mice were obtained from Dr Chenleng Cai lab at Indiana University. Conditional Myh10cKO in mouse CMs was generated by crossing Myh10f/f mice with Tnnt2MerCreMer mice. Six-week-old male Myh10f/f and Myh10f/f/Tnnt2MerCreMer mice were injected with tamoxifen (0.1 mg g−1 body weight) for days 1, 3 and 5 by intraperitoneal injection and then fed with a HFD (45 kcal% fat, Research Diets Inc., D12451i) for 3 months. Six-week-old male LIPTER(Tg) mice and age-matched WT C57BL/6J mice were fed with a HFD (45 kcal% fat, Research Diets Inc., D12451i) or NC (18 protein, Inotiv, 2018sx) for 10 months.
Human heart tissue acquisition
Human left ventricle tissues were obtained from the Duke Human Heart Repository, which is a Duke University Health System institutional review board-approved tissue repository. Samples were procured by the Duke Human Heart Repository in accordance with an approved Duke University Health System institutional review board protocol with written informed consent. All patient information was de-identified. Left ventricle tissues were immediately flash frozen in liquid nitrogen and stored in a −80 °C freezer.
LIPTER and MYH10 knockout hiPSC lines were established using CRISPR/Cas9 as we previously described28. The guide RNAs (gRNAs) targeting the LIPTER or MYH10 gene were designed using the CRISPR design platform (https://zlab.bio/guide-design-resources), and the dual gRNA knock-out method was used. The gRNA was cloned into the pENTR-spCAS9-T2A-EGFP vector, and CRISPR/Cas-9 and gRNA plasmids were co-transfected into human iPSCs using X-tremeGENE 9 DNA Transfection Reagent (Roche, 6365787001). After transfection, GFP+ cells were sorted by FACSAria II cell sorter (BD Biosciences) and seeded on feeder cells with Y-27632 (Selleckchem, S1049). Single hiPSC clones were picked and genotyped by PCR.
LIPTER transcripts were amplified from CMs and cloned into the pHAGE-puro-inducible vector for generating LIPTER overexpression vectors. The LIPTER promoter region (hg38, chr3:157087815-157090367) was amplified and cloned into the pGL3-basic vector. The pBABE-puro human RXR alpha plasmid and pBABE-puro LAP2 (C/EBP beta isoform) plasmid were obtained from Addgene (plasmid numbers 11441 and 15712). NKX2.5 was cloned into the pMXs-puro plasmid. FLAG-tagged ORF1, ORF2 and ORF3 of LIPTER were cloned into a pQCXIP vector.
LIPTER coding potential prediction and verification
Protein coding potential was predicted using Open Reading Frame Finder (https://www.ncbi.nlm.nih.gov/orffinder/). FLAG sequences (GAC TAC AAA GAC GAT GAC GAC AAG) were inserted after each predicted ORF before the stop codon and cloned into a pQCXIP vector. Three LIPTER ORFs with FLAG plasmids were transfected into HEK293T cells (CRL-3216, ATCC) respectively using the X-tremeGENE 9 DNA Transfection Reagent (Roche, 6365787001). A non-relevant protein-coding gene with a FLAG-tag was transfected as a positive control. RNA expression levels were detected using RT–qPCR. FLAG was detected using anti-FLAG antibody (1:1,000, Cell Signaling, 14793S) by immune fluorescent staining and western blotting.
Luciferase assay was performed using the Dual-Glo Luciferase Assay System (Promega, E2920) following the manufacturer’s protocol. Briefly, HEK293T cells were co-transfected with the pGL3-LIPTER-promoter and NKX2.5, CEBPB or RXRa plasmids, or a vector control using the X-tremeGENE 9 transfection reagent (Roche, 6365787001). Cells were collected 48 h post-transfection. Firefly and Renilla luciferase activities in cell lysates were measured using the GLOMAX Explorer microplate reader system (Promega).
Lipid uptake assay
Rhodamine-B-labelled palmitic acid (Avanti, 810104) was dissolved in ethanol and then in 10% bovine serum albumin (BSA). The final concentration of PA was 5 mM. HiPSC-CMs were dissociated into single cells and seeded at a density of 1 × 106 cells per well in a 12-well plate. Cells were treated with 200 μM Rhodamine-B-labelled palmitic acid for 1 h and lysed with 100 μl RIPA lysis buffer. For the mitochondrial lipid uptake assay, hiPSC-CMs were treated with 200 μM Rhodamine-labelled palmitic acid for 2 h. Subsequently, mitochondria were isolated using the Mitochondria Isolation Kit for Cultured Cells (ThermoFisher, 89874) and lysed with 100 μl RIPA lysis buffer. Fluorescence values were detected at an excitation of 520 nm with a 580–640 nm filter on a GLOMAX Explorer microplate reader system (Promega).
WT and LIPTERKO hiPSC-CMs were selected and dissociated into single cells. Mitochondria were isolated from 5 × 106 CMs using Mitochondria Isolation Kit for Cultured Cells (ThermoFisher, 89874). MS and data analyses were performed at the Proteomics Core Facility at the Indiana University School of Medicine.
Flow cytometry was performed to detect cTnT positive cell ratios during EB differentiation, as previously described28. Briefly, day 20 or 40 EBs were collected and dissociated with 1 mg ml−1 Collagenase B for 30 min, followed by 0.25% trypsin–EDTA for 5 min at 37 °C. Dissociated single cells were fixed in 4% paraformaldehyde (PFA) for 15 minutes at room temperature and washed three times with phosphate-buffered saline (PBS). Cells were incubated in a blocking buffer containing 1× PBS, 10% goat serum and 0.1% saponin. Cells were then incubated with primary antibody diluted in 1× PBS with 2% BSA and 0.1% saponin for 1 h at 37 °C, followed by incubation with Alexa Fluor 488-labelled goat anti-mouse secondary antibody (1:200, Thermo Fisher Scientific, A-11001) for 1 h at 37 °C. Flow cytometry analysis was conducted using an Attune NxT flow cytometer (Invitrogen). Data were analysed using FlowJo (Treestar). Supplementary Fig. 1 shows the gate setting.
Nuclear and cytoplasmic RNA detection
A total of 2 × 106 hiPSC-CMs were separated into nuclear and cytoplasmic fractions using NE-PER Nuclear and Cytoplasmic Extraction Reagents (ThermoFisher, 78833). RNAs from both nuclear and cytoplasmic fractions were extracted using the RNeasy Mini Kit (Qiagen, 74106). LIPTER, GAPDH (cytosolic marker) and U6 (nuclear marker) expression levels were detected using RT–qPCR on a QuantStudio 6 Flex system (Applied Biosystems).
Extraction of lipid-associating RNAs
A total of 1 × 106 hiPSC-CMs were treated with 100 μM palmitate acid for 48 h, followed by lipid extraction using the Lipid Extraction Kit (Cell Biolabs, STA-612) according to the manufacturer’s protocol. The isolated total lipids were dried in a vacuum concentrator at 4 °C and resuspended in 700 μl of TRIzol. RNA was then extracted by using the RNeasy mini kit (Qiagen, 74106).
hiPSC-derived EBs at day 40 of differentiation were fixed in 2.5% glutaraldehyde in 1× PBS for 1 h. Then EBs were washed with PBS and post-fixed with 1% osmium tetroxide and 1% potassium ferricyanide for 1 h at 4 °C, dehydrated with a graded series of EtOH (30%, 50%, 70%, 90% and 100%). Subsequently, EBs were embedded and sectioned. TEM images were taken using the JEM-1400 Flash Electron Microscope at the Indiana Center for Biological Microscopy.
Cells were lysed using the Complete Lysis-M EDTA-free kit (Roche, 04719964001). Lysates were run on Mini-PROTEAN TGX Precast Gels (Bio-Rad) and transferred to polyvinylidene difluoride membranes using a Trans-Blot Turbo Transfer System (Bio-Rad). The membrane was blocked with 5% non-fat milk for 1 h at room temperature, incubated in TBXT buffer containing 5% BSA and primary antibodies against MYH10 (1:1,000, Santa Cruz, sc-33729), FLAG (1:1,000, Cell Signaling, 14793S), ATGL1 (1:1,000, Life Technologies, 55190-1-AP), PLIN5 (1:1,000, Life Technologies, 26951-1-AP) or GPAM (1:1,000, Life Technologies, PA520524) at 4 °C overnight. The membrane was then washed with TBXT and incubated with horseradish peroxidase-conjugated anti-rabbit IgG (1:200, Santa Cruz Biotechnology, sc-2357) or anti-mouse IgG (Santa Cruz Biotechnology, sc-2005) secondary antibody, and detected using ECL Western Blotting Substrate (Pierce) on a ChemiDoc Imaging System (Bio-Rad).
Mice were anaesthetized with 1–2% isoflurane, and chest hairs were shaved. The anaesthetized mice were placed on a heated platform in the prone position. Pre-heated ultrasound coupling gel was applied to the chest area, and a linear array transducer (18–23 MHz) was positioned to obtain one-dimensional M-mode images or two-dimensional B-mode parasternal long- and short-axis views on a VisualSonics Vevo 2100 Imaging System. Left ventricle FS percentages and EF percentages were calculated from the M-mode measurements using Vevo LAB ultrasound analysis software at the IU Small Animal Ultrasound Core.
Dr Zhiqiang Lin (Masonic Medical Research Institute, United States) provided the AAV9-CTNT vector. AAV9-CTNT-GFP (green fluorescent protein), AAV9-CTNT-AS-LIPTER and AAV9-CTNT-LIPTER (human full-length LIPTER RNA) contained a chicken troponin T promoter to drive CM-specific GFP, AS-LIPTER and LIPTER expression. AAV9 viruses were packaged by the Genome Editing, Transgenic, and Virus Core at Magee-Women’s Research Institute with a fee-based service. AAV9 was administered by retro-orbital injection into 6-week-old B6.BKS(D)-Leprdb/J mice (stock no. 000697, The Jackson Laboratory) at 2 × 1010 vg/g. WT C57BL/6J mice (stock no. 000664, The Jackson Laboratory) without AAV9 injection were also included as a control.
AAV9-shNKX2.5 were utilized to knock down NKX2.5 in hiPSC-derived CMs. Briefly, AAV9-shRNA-ctrl (Addgene, 85741) was used as a control scramble shRNA. Two shRNAs targeting human NKX2.5 were designed and inserted into the AAV9 plasmid to replace the shRNA-ctrl sequence. The pAdDeltaF6 and AAV2/9 plasmids were used for packaging viruses. Viruses with a concentration of 1 × 107 GC ml−1 were used to infect hiPSC-derived EBs. Media containing AAV9 virus were changed every three days and maintained for 9 days.
Cryosection and CM area measurement
HiPSC-derived EBs were fixed with 4% PFA for 30 min, washed with PBS and dehydrated in sucrose solutions of increasing concentration (10%, 20% and 30%) for 1 h each. Dehydrated EBs were embedded in OCT (Fisher Scientific, 23730571) and stored at −80 °C. Frozen EBs were sectioned at 7 μm on a Leica CM3050 cryostat and stored in a −80 °C freezer. For human heart LV tissues and mouse heart sectioning, the tissues were fixed with 4% PFA overnight, and dehydrated in sucrose solutions of increasing concentration (10%, 20% and 30%) for 3 h each. Fixed tissues were embedded in OCT and stored at −80 °C freezer. To measure myocyte cross-sectional area, the sections were stained with Alexa Fluor 488-conjugated wheat germ agglutinin (Invitrogen, W11261) for membrane and with DAPI for nuclei. A minimum of 50 CMs per section were manually outlined and single-cell cross-sectional area was measured using ImageJ software (National Institutes of Health (NIH)).
Fast Green and Picoro-Sirius Red staining
Mouse heart sections were fixed with Bouins’ solution at 55 °C for 1 h and washed with tap water for 45 min. Until yellow colour disappeared, sections were stained with 0.1% Fast Green for 30 min, followed by 1% acetic acid for 15 s and tap water for 2 min. Sections were then stained with 0.1% Sirius Red for 30 min, dehydrated with EtOH (70–100%), cleared with xylene and mounted with Cytoseal 60 (Fisher Scientific, 23-244257).
Cryosections of hiPSC-EBs and human heart tissues were prepared by following the sample preparation protocol for fixed frozen tissues from ACD Inc. Briefly, the tissue sections were washed with PBS, baked at 60 °C for 30 min, post-fixed with 4% PFA for 15 min at 4 °C, and then dehydrated with 50%, 70% and 100% EtOH for 1 × 5 min at room temperature. Tissues were baked in target retrieval solution at 98–102 °C for target retrieval, and then treated with Protease IV for 30 min at 40 °C. The LIPTER specific probes labelled with C2 and negative control probes were synthesized at ACD Inc. The prepared tissue and EB sections were hybridized with LIPTER or control probes following the RNAscope Fluorescent Multiplex Assay protocol. Briefly, the sections were incubated with Amp 1-FL for 30 min, Amp 2-FL for 15 min, Amp 3-FL for 1 × 30 min and Amp4 AltB-FL for 15 min at 40 °C, and washed with Wash Buffer twice at room temperature between each inculcation. For FISH and lipid double stainings, after FISH, the sections were stained with 1:1,000 LipidTOX Deep Red (ThermoFisher, H34477) in PBS for 30 min at room temperature. After three washes with PBS, samples were mounted with DAPI Fluoromount-G (SouthernBiotech, 0100-20) for imaging.
EBs or tissue sections were washed with PBS to remove OCT and blocked with 5% goat serum in PBST for 1 h at room temperature. CMs were fixed with cold 4% PFA for 15 min at room temperature. The blocked sample sections and fixed CMs were incubated with anti-cTnT antibody (1:200, Thermo Fisher Scientific, MS-295-P), cleaved caspase-3 antibody (1:200, Cell Signaling, 9664), MYH10 antibody (1:200, Santa Cruz, sc-33729), NKX2.5 antibody (1:100, Abcam, ab97355), RXRa antibody (1:200, ABclonal Science, A19105), or CEBPB antibody (1:100, Abcam, ab32358) in blocking buffer containing 2% goat serum and 0.1% saponin overnight at 4 °C. After incubation, samples were washed with PBS three times and incubated with fluorescence-labelled secondary antibody (1:200, Thermo Fisher Scientific, A-21235, A-11001 or A-21422) for 1 h at 37 °C. Finally, cells were washed three times in 1× PBS, then mounted with DAPI Fluoromount-G (SouthernBiotech, 0100-20) for imaging.
Samples were treated with Oil Red O (Sigma, O0625) staining or Nile Red (Sigma, 19123) staining to detect LDs. For Oil Red O staining, samples were washed in 60% isopropanol for 5 min and then incubated with Oil Red O working solution (300 mg Oil Red O in 100 ml isopropanol, and 6:4 diluted with water) for 20 min at room temperature. After washing with 60% isopropanol for 1 min and H2O for 5 min, the slides were mounted with DAPI Fluoromount-G (SouthernBiotech, 0100-20) for imaging. For Nile Red Staining, slides were stained with Nile Red (Sigma, 19123) at 1 μg ml−1 for 10 min at room temperature, washed with PBS and mounted with DAPI Fluoromount-G (SouthernBiotech, 0100-20).
TUNEL staining was performed following the manufacturer’s protocol of In Situ Cell Death Detection Kit (TMR red) (Roche, 12156792910). Briefly, after immunofluorescent staining with anti-cTnT antibody (1:200, Thermo Fisher Scientific, MS-295-P), heart tissue sections or cell samples were permeabilized with permeabilization solution (0.1% Triton X-100 in 0.1% sodium citrate) for 2 min on ice and washed with PBS twice. Samples were then stained with TUNEL reaction mixture for 1 h at 37 °C. After washing with PBS three times for 5 min each, the samples were mounted with DAPI Fluoromount-G medium (SouthernBiotech, 0100-20) for imaging.
RNA isolation, cDNA synthesis and RT–qPCR
Total RNAs were extracted from cells or tissues using TRIzol Reagent (Invitrogen, 15596018) and purified with the RNeasy mini kit (Qiagen, 74106). DNase treatment was performed using RNase-free DNase (Qiagen) to remove any contaminating genomic DNA. Total RNAs were reverse transcribed into complementary DNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, 4374966). RT–qPCR was performed on a QuantStudio 6 Flex system (Applied Biosystems) using SYBR Green Master Mix (Applied Biosystems, 4385614). Results were analysed using the 2−ΔΔCt method and normalized to the expression of ACTB gene. RT–PCR was performed with Thermo Fisher Scientific DreamTaq Green PCR Master Mix. Primer sequences are described in Supplementary Table 6.
RNA–lipid overlay assay
The RNA–lipid overlay assay was performed according to the previously published method58. Briefly, Biotinylated LIPTER RNA, truncated LIPTER RNA fragments and control AS-LIPTER were transcribed using the TranscriptAid T7 High Yield Transcription Kit (ThermoFisher, K0441), and labelled with RNA 3′ End Biotinylation Kit (ThermoFisher, 20160) following the manufacturer’s protocols. Lipid strips membrane (Echelon Biosciences, P-6002) was blocked with RNA–lipid binding buffer and then incubated with biotinylated RNA (1 µg ml−1) in RNA–lipid binding buffer supplemented with 50 U ml−1 RNase inhibitor. The membrane was then washed three times for 10 min with RNA–lipid binding buffer supplemented with 0.05% IGEPAL CA-630 (Sigma, I8896), and signal was detected using the Chemiluminescent Nucleic Acid Detection Module (ThermoFisher, 89880) of a ChemiDoc Imaging System (Bio-Rad).
Giant lipid vesicles assay
To generate giant lipid vesicles59, fluorescent-labelled phospholipids including TopFluor TMR PA (Avanti Polar Lipids, 810240), TopFluor PI4P (Avanti Polar Lipids, 810185) and TopFluor PI (Avanti Polar Lipids, 810187) were dissolved in CHCl3. A total of 250 nmol of each fluorescent-labelled PA, PI4P or PI in 500 µl CHCl3 was transferred to a small glass tube (ThermoFisher, 14-958) and dried using nitrogen air in a chamber. A thin lipid layer was formed on the bottom of the glass bottle. RNA–lipid binding buffer was slowly added to the bottle, and after 48 h of incubation at room temperature in the dark, giant lipid vesicles were collected from the upper layer of the solution. LIPTER, LIPTER-antisense and LIPTER exon 1–2 and exon 3 fragments were labelled with the Alexa Fluor 594 Nucleic Acid Labeling Kit (ThermoFisher, U21654) according to the manufacturer’s protocol. The lipid vesicles and the Alexa594-labelled RNAs were gently mixed at a volume ratio of 5:1, incubated at room temperature for 15 min in the dark, and dropped on glass coverslips. Images were captured using a DMi8 inverted fluorescent microscope (Leica).
MST assays were performed on a NanoTemper Monolith NT.115 with blue/red filters (NanoTemper Technologies) at the Physical Biochemistry Instrumentation Facility at Indiana University, Bloomington. TopFluor TMR PA (Avanti Polar Lipids, 810240) and TopFluor PI4P (Avanti Polar Lipids, 810185) were dissolved in RNA–lipid binding buffer (100 nM), and a final concentration of 25 nM of each was used. LIPTER and LIPTER-antisense RNA were synthesized in vitro and serially diluted (1–9,600 nM) in the RNA–lipid binding buffer into 16 vials with 10 µl per vial. Fluorescent-labelled PA or PI4P (10 µl) was added and mixed well in each vial. Next, 10 µl of the mixed solution from each vial was removed and loaded into a capillary. Fluorescence intensities of the 16 capillaries were measured at 40% MST power and 90% LED power. The Kd value was calculated using MO Affinity Analysis software.
MS2-tagged RNA affinity purification
MS2-tagged RNA affinity purification was performed as previously described60. Plasmids pcDNA3-FLAG-MS2 and pcDNA3-24xMS2-stemloop were obtained from Dr Je-Hyun Yoon Lab (Medical University of South Carolina, United States). LIPTER full-length RNA, LIPTER RNA fragments and control RNA were cloned into the pcDNA3-24xMS2-stemloop plasmid between HindIII and EcoRI sites. HEK293T cells (CRL-3216, ATCC) were co-transfected with 6 µg pcDNA3-LIPTER-24xMS2SL (or different exons-24xMS2SL or pcDNA3-LIPTER-antisense-24xMS2SL) plasmids and 3 µg pcDNA3-FLAG/YFP-MS2 by using X-tremeGENE 9 DNA Transfection Reagent (Roche, 6365787001). After 48 h, transfected cells were collected to pull down RNA binding proteins using the EZ-Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Sigma,17701). Cells were lysed with 100 µl RIP lysis buffer plus RNase inhibitor with protease inhibitor cocktail. Then 100 µl cell lysate was mixed with 860 µl RIP wash buffer, 5 µg FLAG antibody (1:200, Cell Signaling, 14793S)-labelled magnetic beads, 35 μl 0.5 M EDTA and 5 μl RNase inhibitor and incubated. The mixture was rotated overnight at 4 °C. After inculcation, beads were washed with RIP wash buffer six times, resuspended in RIPA lysis buffer plus loading buffer and heated for 5 min at 95 °C. The beads were removed by centrifuge at 200g for 5 min, and the supernatant was collected for SDS–PAGE, followed by silver staining using the Pierce Silver Stain kit (ThermoFisher, 34612) or western blotting.
Live cell imaging
To trace LIPTER RNA in live CMs, LIPTER-24xMS2SL and LIPTER-antisense-24xMS2SL were separately inserted into a pHAGE lentiviral vector with puromycin resistance. The pHR-tdMCP-YFP (MS2-coat binding protein tagged with YFP) lentiviral plasmid was obtained d from Addgene (plasmid number 99151). A total of 1 × 106 hiPSCs were infected with packaged pHAGE-LIPTER-24xMS2SL or pHAGE-LIPTER-antisense-24xMS2SL lentivirus for 72 h. After infection, cells were selected using 1 µg ml−1 puromycin for 4 days. The remaining cells were then infected with pHR-tdMCP-YFP lentivirus. YFP+ cells were sorted by a BD FACSAria II Cell Sorter, expanded and differentiated into CMs. The differentiated CMs were seeded into poly-d-lysine-coated 35 mm glass-bottom dishes (MatTek, P35GC-0-10-C). Live cell video was taken on a Nikon live cell imaging system with an Apo 60× Oil lens at Indiana Center for Biological Microscopy. Consecutive images were taken during 4 h with 3 min intervals, and a total of 81 images were taken per field. The video was exported with 16 frames per second using the NIS Elements software.
Quantifications of FA and TAG in mouse hearts
The concentration of free FA was quantified by using the Free Fatty Acid Assay kit (Abcam, ab65341), and TAG concentrations were quantified using a Triglyceride Quantification Assay Kit (Abcam, ab65336). Approximately 25 mg of mouse heart tissue per heart was collected and homogenized in 500 μl lipid extraction buffer (Abcam, ab211044) using a Beadbug Homogenizer. After centrifugation, the supernatant was removed to a new tube. FA or TAG concentration was measured according to the manufacturer’s protocols. The FA and TAG concentrations in WT and genetically modified mouse hearts were normalized to the collected heart tissue weights.
Whole mouse heart FAO rate assay
Freshly collected mouse heart was minced into small pieces and thoroughly homogenized in ice-cold STE buffer (0.25 M sucrose, 10 mM Tris–HCl and 1 mM EDTA, at pH of 7.4) using a Dounce homogenizer. After homogenization, the specimen was centrifuged at 420g at 4 °C for 10 min. Supernatant containing crude mitochondria was transferred into chilled Eppendorf tubes for protein quantification and FAO assay as previously described61,62,63. In brief, 30 µl of tissue homogenate was added into Eppendorf tubes in triplicate (including STE blank control for background reading), followed by the immediate addition of 370 µl oxidation reaction mixture (100 mM sucrose, 10 mM Tris–HCl, 5 mM KH2PO4, 0.2 mM EDTA, 80 mM KCl, 1 mM MgCl2, 2 mM l-carnitine, 0.1 mM malate, 0.05 mM Coenzyme A, 2 mM ATP, 1 mM dithiothreitol, 0.5 mM palmitate, 0.7% BSA and 0.4μCi 14C-palmitate), and incubated at 37 °C for 1 h. After incubation, the entire reaction mixture was transferred into tubes containing 200 µl of 1 M perchloric acid and the CO2 trap disc, and incubated for 1 h at room temperature. Then the trap disc was transferred to scintillation vials for complete oxidation measurement, whereas the supernatant obtained by centrifugation of the remaining acid solution was used to measure acid-soluble metabolites (ASMs) produced by incomplete oxidation. In addition, the nonincubated reaction mixture was counted in triplicate to get the measurement of the amount of radioactivity input into each reaction.
The specific activity of the reaction mixture was determined by dividing the disintegrations per minute (DPM) of the input by the number of nanomoles of palmitate (cold + hot) per reaction. The specific activity can be used to convert DPM to nanomoles for the other samples.
To determine the rate of conversion of 14C-palmitate to 14CO2, we divide the nanomoles of 14CO2 by the amount of time in the reaction and the amount of protein per action. A similar calculation can be done to determine the rate of conversion of 14C-palmitate to ASMs.
Untargeted metabolomics was performed, and data were analysed by Creative Proteomics. Briefly, 10 million enriched hiPSC-CMs were collected at day 40 and washed with PBS. Metabolites were extracted with 800 μl of methanol, 10 μl of DL-o-chlorophenylalanine (2.8 mg ml−1) and 10 μl of LPC (12:0), followed by ultrasonication for 30 min. The samples were analysed using the Ultimate 3000 HPLC and UHPLC combined with Q Exactive MS systems (ThermoFisher) and screened with electrospray ionization (ESI)–MS. Raw data were acquired and aligned using the Compound Discover (3.0, ThermoFisher) based on the m/z value and the retention time of the ion signals. Ions from both ESI− and ESI+ were merged and imported into the SIMCA-P program (version 14.1) for multivariate analysis.
Seahorse mitochondrial function assays
Seahorse XFp Extracellular Flux Analyzer was used to measure oxygen consumption rate (OCR) and extracellular acidification rate of hiPSC-CMs according to the manufacturer’s protocol. Briefly, 5 × 104 CMs (based on FCCP titration assay) were seeded into Matrigel-coated eight-strip utility plate for 48 h before Mito Stress Test assay. We prepared the pH 7.4 assay medium (with 4 mM l-glutamine and 25 mM glucose),and hydrated the cartridge in calibration medium at 37 °C without CO2 before running the assay. Then 1.5 μM oligomycin was loaded into port A, 0.25 μM FCCP into port B and 0.5 μM Rot/AA into port C on the cartridge. The Mito Stress Test was run according to the standard process of analyzer. After the assay, the protein concentration of each well was measured using BCA protein assay kit (ThermoFisher, 23225). Raw data were processed using Seahorse Wave software.
Long-chain FAO was measured according to the protocol of Palmitate Oxidation Stress Test Kit on a Seahorse XF96 Analyzer. Briefly, 5 × 104 CMs were seeded into one well of Matrigel-coated 96-well utility microplate, cultured in normal culture medium for 48 h and then switched to the substrate-limited growth medium (DMEM plus 0.5 mM glucose, 1.0 mM GlutaMAX, 1% FBS and 0.5 mM l-carnitine) overnight. The FAO assay buffer was prepared according to the protocol and replaced the substrate-limited growth medium 45 min before measurement. FAO buffer or 40 μM etomoxir was added into each well 15 min before measurement. BSA control or palmitate:BSA was added into well before the Mito Stress test was started. FAO level was quantified on the basis of the differences in the values of the maximal oxygen consumption in palmitate:BSA without or with presence of etomoxir.
Illumina RNA sequencing and data analysis
Transcriptional profiles of human ESC-derived CMs, SMCs and ECs were obtained by Illumina mRNA deep sequencing using a service from LC Sciences as we previously reported24. Three non-failure hearts, three non-failure hearts with T2DM, three hearts with DCM, and three hearts with DCM and T2DM were subject to mRNA-seq. Briefly, total RNAs were extracted from human left ventricle tissues by using the RNeasy kit (Qiagen, 74106), and sequenced by using paired-end deep-transcriptome sequencing with the Illumina platform at IU Genomic Core. All lncRNAs were profiled according to their fragments per kilobase of transcript per million mapped reads values. Additionally, mRNAs were extracted from enriched day 40 WT and LIPTERKO hiPSC-CMs, as well as hearts of WT and LIPTER(Tg) mice fed with HFD for 7 months and sequenced by using paired-end deep-transcriptome sequencing with the Illumina platform. The RNA-seq data were collected and analysed as we previously described24,64. In general, the sequencing reads were mapped to the reference genome (either hg38 or mm10) by STAR (version 2.7.2a). Gene expression levels were evaluated by the featureCounts on uniquely mapped reads. Following gene expression normalization based on trimmed mean of M (TMM) values, edgeR (refs. 65,66) was employed to perform differential analysis given the comparison between LIPTERKO and WT for human samples or LIPTER(Tg) and WT for mouse heart samples. The genes with false discovery rate (FDR)-adjusted P values <0.05 after multiple test correction were determined as differentially expressed genes.
Functional enrichment analysis
GO analysis was conducted on differentially expressed genes using DAVID67,68,69. The gene set enrichment analysis (GSEA)70 was performed (http://www.gsea-msigdb.org/gsea/index.jsp, V4.1.0) based on the fold change (log2) of gene expression between either KO versus WT CMs or Tg versus WT hearts by focusing on C5 biological process (bp) from GO database or specific toxicity gene sets retrieved from Ingenuity Pathway Analysis71. The significantly enriched GO terms or toxicity-related gene sets were selected for presentation in figures.
RIP was conducted using the EZ-Magna RIP RNA-Binding Protein Immunoprecipitation kit (Millipore) according to the manufacturer’s instruction (Sigma,17701). Briefly, 1 × 106 hiPSC-CMs were resuspended in 50 μl of RIP lysis buffer with protease inhibitor cocktail and RNase inhibitor. For each sample, 5 μg anti-human MYH10 antibody (1:400, Abcam, ab230823) was added and 1 μg isotype IgG antibody (1:200, Cell Signaling, 5415S) was used as the control. RIP–qPCR results were calculated as fold enrichment from specific antibody versus non-specific IgG (1:200, Cell Signaling, 5415S).
RIP-seq for mouse hearts
RNA immunoprecipitation was performed with mouse hearts as described above. Hearts were collected from WT and LIPTER (Tg) mice. One-hundred milligrams of left ventricle tissue per heart was collected and lysed with RIP lysis buffer according to the manufacturer’s protocol (Sigma, 17701). Anti-Myh10 antibody (1:400, Abcam, ab230823) was used to pull down Myh10-binding RNAs. Normal rabbit IgG was used as negative control. The pulldown RNAs were extracted by using the RNeasy mini kit (Qiagen, 74106). Total RNA sequencing was performed at the IU Genomic Core. The sequence reads were mapped to the mouse genome mm10 using STAR (v2.7.2a) with the the following parameter: ‘–outSAMmapqUnique 60’. Then unmapped reads were mapped to the human genome hg38 by STAR with the same parameter. Peak calling on mapped reads was performed by MACS2 compared with IgG signals with a cut-off of FDR-adjusted P value <0.01.
ScRNA-seq data analysis
The gene count matrix of scRNA-seq data from a foetal human heart was retrieved from a previously published report25. Normalization, dimensionality reduction and clustering of scRNA-seq data were performed using Seurat (v3) (ref. 72) according to the parameters previously utilized25. The t-distributed stochastic neighbour embedding was used to visualize single-cell clusters, and gene expression in a reduced 2D space. Annotation of cell clusters was mapped from the annotation of the clusters presented in this report25.
Statistics and reproducibility
Statistical results were analysed using Prism 8 (GraphPad Software). Unpaired Student’s t-tests (two groups) and two-way analysis of variance were used in data analyses unless specified in the legend. Column bar plots show mean ± standard error of the mean (s.e.m.) from at least three independent experiments. Statistical parameters, including the exact value of n, statistical test method and statistical significance, are reported in the figures, figure legends and related data resources. Data were judged to be significant when P < 0.05. No statistical method was used to pre-determine sample size. No animal or data were excluded from the analyses. The experiments were not randomized. The Investigators were not blinded to allocation during experiments and outcome assessment.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
RNA-seq data of WT and LIPTERKO hiPSC-CMs, and WT and LIPTER (Tg) mouse hearts were deposited in NCBI GEO database (GSE175370). The UCSC Genome Browser view of RNAseq results from T2DM and NF human hearts can be found at https://genome.ucsc.Edu/s/samal/hg38_YangLei_ILMN550. The metabolomics data can be accessed in Supplementary Table 1. The datasets used to determine LINC00881 expression in human hearts can be accessed at NCBI Accession GSE64283 as shown in Extended Data Fig. 1c73, NCBI Accession PRJNA280600 as shown in Extended Data Fig. 1d74 and NCBI Accession GSE30611 (Illumina Human Body Map 2.0 Project) as shown in Extended Data Fig. 1e. Source data are provided with this paper. All other data supporting the findings of this study are available from the corresponding author on reasonable request.
Olzmann, J. A. & Carvalho, P. Dynamics and functions of lipid droplets. Nat. Rev. Mol. Cell Biol. 20, 137–155 (2019).
Goldberg, I. J. et al. Deciphering the role of lipid droplets in cardiovascular disease: a report From the 2017 National Heart, Lung, and Blood Institute Workshop. Circulation 138, 305–315 (2018).
Jacob, S. Lipid droplet accumulation in the heart during fasting. Acta Histochem. 82, 149–152 (1987).
Listenberger, L. L. et al. Triglyceride accumulation protects against fatty acid-induced lipotoxicity. Proc. Natl Acad. Sci. USA 100, 3077–3082 (2003).
Sharma, S. et al. Intramyocardial lipid accumulation in the failing human heart resembles the lipotoxic rat heart. FASEB J. 18, 1692–1700 (2004).
Borradaile, N. M. & Schaffer, J. E. Lipotoxicity in the heart. Curr. Hypertens. Rep. 7, 412–417 (2005).
Lopaschuk, G. D., Ussher, J. R., Folmes, C. D., Jaswal, J. S. & Stanley, W. C. Myocardial fatty acid metabolism in health and disease. Physiol. Rev. 90, 207–258 (2010).
Hankiewicz, J. H., Banke, N. H., Farjah, M. & Lewandowski, E. D. Early impairment of transmural principal strains in the left ventricular wall after short-term, high-fat feeding of mice predisposed to cardiac steatosis. Circ. Cardiovasc Imaging 3, 710–717 (2010).
Chiu, H. C. et al. A novel mouse model of lipotoxic cardiomyopathy. J. Clin. Invest. 107, 813–822 (2001).
Finck, B. N. et al. The cardiac phenotype induced by PPARα overexpression mimics that caused by diabetes mellitus. J. Clin. Invest. 109, 121–130 (2002).
Walther, T. C. & Farese, R. V. Jr. Lipid droplets and cellular lipid metabolism. Annu. Rev. Biochem. 81, 687–714 (2012).
Zechner, R., Madeo, F. & Kratky, D. Cytosolic lipolysis and lipophagy: two sides of the same coin. Nat. Rev. Mol. Cell Biol. 18, 671–684 (2017).
Kilwein, M. D. & Welte, M. A. Lipid droplet motility and organelle contacts. Contact (Thousand Oaks) 2, 2515256419895688 (2019).
Xu, N. et al. The FATP1-DGAT2 complex facilitates lipid droplet expansion at the ER–lipid droplet interface. J. Cell Biol. 198, 895–911 (2012).
Wang, H. et al. Perilipin 5, a lipid droplet-associated protein, provides physical and metabolic linkage to mitochondria. J. Lipid Res. 52, 2159–2168 (2011).
McConnell, R. E. & Tyska, M. J. Leveraging the membrane–cytoskeleton interface with myosin-1. Trends Cell Biol. 20, 418–426 (2010).
Tabb, J. S., Molyneaux, B. J., Cohen, D. L., Kuznetsov, S. A. & Langford, G. M. Transport of ER vesicles on actin filaments in neurons by myosin V. J. Cell Sci. 111, 3221–3234 (1998).
Pfisterer, S. G. et al. Role for formin-like 1-dependent acto-myosin assembly in lipid droplet dynamics and lipid storage. Nat. Commun. 8, 14858 (2017).
Takeda, K., Kishi, H., Ma, X., Yu, Z. X. & Adelstein, R. S. Ablation and mutation of nonmuscle myosin heavy chain II-B results in a defect in cardiac myocyte cytokinesis. Circ. Res. 93, 330–337 (2003).
Schewkunow, V. et al. Thermodynamic evidence of non-muscle myosin II–lipid–membrane interaction. Biochem. Biophys. Res. Commun. 366, 500–505 (2008).
He, Y. et al. Nonmuscle myosin IIB links cytoskeleton to IRE1α signaling during ER stress. Dev. Cell 23, 1141–1152 (2012).
Dai, X., Zhang, S. & Zaleta-Rivera, K. RNA: interactions drive functionalities. Mol. Biol. Rep. 47, 1413–1434 (2020).
Yang, L. et al. Human cardiovascular progenitor cells develop from a KDR+ embryonic-stem-cell-derived population. Nature 453, 524–528 (2008).
Li, Y., Lin, B. & Yang, L. Comparative transcriptomic analysis of multiple cardiovascular fates from embryonic stem cells predicts novel regulators in human cardiogenesis. Sci. Rep. 5, 9758 (2015).
Asp, M. et al. A spatiotemporal organ-wide gene expression and cell atlas of the developing human heart. Cell 179, 1647–1660 e1619 (2019).
Anderson, D. J. et al. NKX2-5 regulates human cardiomyogenesis via a HEY2 dependent transcriptional network. Nat. Commun. 9, 1373 (2018).
Rombel, I. T., Sykes, K. F., Rayner, S. & Johnston, S. A. ORF-FINDER: a vector for high-throughput gene identification. Gene 282, 33–41 (2002).
Liu, J., Li, Y., Lin, B., Sheng, Y. & Yang, L. HBL1 is a human long noncoding RNA that modulates cardiomyocyte development from pluripotent stem cells by counteracting MIR1. Dev. Cell 43, 372 (2017).
Jackson, C. L. Lipid droplet biogenesis. Curr. Opin. Cell Biol. 59, 88–96 (2019).
Shu, H. et al. The role of CD36 in cardiovascular disease. Cardiovasc. Res. 118, 115–129 (2022).
Coleman, R. A. & Mashek, D. G. Mammalian triacylglycerol metabolism: synthesis, lipolysis, and signaling. Chem. Rev. 111, 6359–6386 (2011).
Han, S. S. et al. Investigating the mechanism of hyperglycemia-induced fetal cardiac hypertrophy. PLoS ONE 10, e0139141 (2015).
Singh, A. B., Guleria, R. S., Nizamutdinova, I. T., Baker, K. M. & Pan, J. High glucose-induced repression of RAR/RXR in cardiomyocytes is mediated through oxidative stress/JNK signaling. J. Cell. Physiol. 227, 2632–2644 (2012).
Tarquini, R., Lazzeri, C., Pala, L., Rotella, C. M. & Gensini, G. F. The diabetic cardiomyopathy. Acta Diabetol. 48, 173–181 (2011).
Lin, A. et al. The LINK-A lncRNA interacts with PtdIns(3,4,5)P(3) to hyperactivate AKT and confer resistance to AKT inhibitors. Nat. Cell Biol. 19, 238–251 (2017).
Seidel, S. A. et al. Microscale thermophoresis quantifies biomolecular interactions under previously challenging conditions. Methods 59, 301–315 (2013).
Zhao, R., Nakamura, T., Fu, Y., Lazar, Z. & Spector, D. L. Gene bookmarking accelerates the kinetics of post-mitotic transcriptional re-activation. Nat. Cell Biol. 13, 1295–1304 (2011).
Kovacs, M., Toth, J., Hetenyi, C., Malnasi-Csizmadia, A. & Sellers, J. R. Mechanism of blebbistatin inhibition of myosin II. J. Biol. Chem. 279, 35557–35563 (2004).
Goldberg, I. J., Trent, C. M. & Schulze, P. C. Lipid metabolism and toxicity in the heart. Cell Metab. 15, 805–812 (2012).
Heydemann, A. An overview of murine high fat diet as a model for type 2 diabetes mellitus. J. Diabetes Res. 2016, 2902351 (2016).
Sikder, K., Shukla, S. K., Patel, N., Singh, H. & Rafiq, K. High fat diet upregulates fatty acid oxidation and ketogenesis via intervention of PPAR-γ. Cell. Physiol. Biochem. 48, 1317–1331 (2018).
Leifheit-Nestler, M. et al. Importance of leptin signaling and signal transducer and activator of transcription-3 activation in mediating the cardiac hypertrophy associated with obesity. J. Transl. Med 11, 170 (2013).
Burke, S. J. et al. db/db mice exhibit features of human type 2 diabetes that are not present in weight-matched C57BL/6J mice fed a western diet. J. Diabetes Res. 2017, 8503754 (2017).
Prasad, K. M., Xu, Y., Yang, Z., Acton, S. T. & French, B. A. Robust cardiomyocyte-specific gene expression following systemic injection of AAV: in vivo gene delivery follows a Poisson distribution. Gene Ther. 18, 43–52 (2011).
Li, R. H. et al. A phosphatidic acid-binding lncRNA SNHG9 facilitates LATS1 liquid–liquid phase separation to promote oncogenic YAP signaling. Cell Res. 31, 1088–1105 (2021).
Fei, W. et al. A role for phosphatidic acid in the formation of ‘supersized’ lipid droplets. PLoS Genet. 7, e1002201 (2011).
Bankaitis, V. A., Garcia-Mata, R. & Mousley, C. J. Golgi membrane dynamics and lipid metabolism. Curr. Biol. 22, R414–R424 (2012).
Khvorova, A., Kwak, Y. G., Tamkun, M., Majerfeld, I. & Yarus, M. RNAs that bind and change the permeability of phospholipid membranes. Proc. Natl Acad. Sci. USA 96, 10649–10654 (1999).
Czerniak, T. & Saenz, J. P. Lipid membranes modulate the activity of RNA through sequence-dependent interactions. Proc. Natl Acad. Sci. USA 119, e2119235119 (2022).
Frustaci, A. et al. Myocardial cell death in human diabetes. Circ. Res. 87, 1123–1132 (2000).
Liu, J. et al. LncRNA HBL1 is required for genome-wide PRC2 occupancy and function in cardiogenesis from human pluripotent stem cells. Development 13, dev199628 (2021).
Han, L. & Yang, L. Multidimensional mechanistic spectrum of long non-coding RNAs in heart development and disease. Front. Cardiovasc. Med. 8, 728746 (2021).
Kleinfeld, A. M. et al. Increases in serum unbound free fatty acid levels following coronary angioplasty. Am. J. Cardiol. 78, 1350–1354 (1996).
Li, Y. et al. Determination of fatty acid metabolism with dynamic [11C]palmitate positron emission tomography of mouse heart in vivo. Mol. Imaging 14, 516–525 (2015).
Carvajal-Vergara, X. et al. Patient-specific induced pluripotent stem-cell-derived models of LEOPARD syndrome. Nature 465, 808–812 (2010).
Lin, B. et al. High-purity enrichment of functional cardiovascular cells from human iPS cells. Cardiovasc. Res. 95, 327–335 (2012).
Ma, X. et al. Conditional ablation of nonmuscle myosin II-B delineates heart defects in adult mice. Circ. Res. 105, 1102–1109 (2009).
Lin, A. et al. The LINK-A lncRNA interacts with PtdIns(3,4,5)P3 to hyperactivate AKT and confer resistance to AKT inhibitors. Nat. Cell Biol. 19, 238–251 (2017).
Janas, T. & Yarus, M. Visualization of membrane RNAs. RNA 9, 1353–1361 (2003).
Yoon, J. H. & Gorospe, M. Identification of mRNA-interacting factors by MS2-TRAP (MS2-tagged RNA affinity purification). Methods Mol. Biol. 1421, 15–22 (2016).
Huynh, F. K., Green, M. F., Koves, T. R. & Hirschey, M. D. Measurement of fatty acid oxidation rates in animal tissues and cell lines. Methods Enzymol. 542, 391–405 (2014).
Zhou, H. et al. IL-1 induces mitochondrial translocation of IRAK2 to suppress oxidative metabolism in adipocytes. Nat. Immunol. 21, 1219–1231 (2020).
Naiman, S. et al. SIRT6 promotes hepatic beta-oxidation via activation of PPARα. Cell Rep. 29, 4127–4143 (2019).
Liu, J. et al. Genome-wide studies reveal the essential and opposite roles of ARID1A in controlling human cardiogenesis and neurogenesis from pluripotent stem cells. Genome Biol. 21, 169 (2020).
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140 (2010).
McCarthy, D. J., Chen, Y. & Smyth, G. K. Differential expression analysis of multifactor RNA-seq experiments with respect to biological variation. Nucleic Acids Res. 40, 4288–4297 (2012).
Dennis, G. Jr et al. DAVID: database for annotation, visualization, and integrated discovery. Genome Biol. 4, P3 (2003).
Huang da, W., Sherman, B. T. & Lempicki, R. A. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 37, 1–13 (2009).
Huang da, W., Sherman, B. T. & Lempicki, R. A. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57 (2009).
Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005).
Kramer, A., Green, J., Pollard, J. Jr. & Tugendreich, S. Causal analysis approaches in Ingenuity Pathway Analysis. Bioinformatics 30, 523–530 (2014).
Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902 (2019).
Szabo, L. et al. Statistically based splicing detection reveals neural enrichment and tissue-specific induction of circular RNA during human fetal development. Genome Biol. 16, 126 (2015).
Duff, M. et al. Genome-wide identification of zero nucleotide recursive splicing in Drosophila. Nature 521, 376–379 (2015).
This work was supported by the National Institute of Health grants RO1HL147871 and RO1HL160856 and American Heart Association 20EIA35260114 to L.Y. The mRNA-seq was carried out at the Center for Medical Genomics at Indiana University School of Medicine. We thank Z. Lin for providing AAV9 vector and C. Cai for providing the Tnnt2-MerCreMer mouse line. We thank L. J. Field for critical reading and suggestions.
A provisional patent application from L.Y. is under preparation. The other authors declare no competing interests.
Peer review information
Nature Cell Biology thanks Gary Lopaschuk and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
a, RT-qPCR detection of pluripotency marker OCT4, cardiac progenitor marker PDGFRA, cardiomyocyte marker TNNT2, endothelial cell marker CD31, smooth muscle cell marker SM22, as well as lncRNAs NAV2-AS2, TTN-AS1 and SLC8A1-AS1 expressions in enriched cardiovascular cell types differentiated from human iPS cells. (n = 2 independent experiments). b, RT-qPCR detection of lncRNAs NAV2-AS2, TTN-AS1 and SLC8A1-AS1 expression levels in left ventricle tissues from NF (n = 10 samples), NF+T2DM (n = 4 samples), DCM (n = 12 samples) and DCM+T2DM (n = 5 samples) individuals. All bars are presented as mean ± s.e.m. Unpaired two-tailed t-test is used for comparison. c, Histogram of LINC00881 expressions in the human hearts during fetal development. RPKM, Reads Per Kilobase of transcript (RPKM). Bars are presented as mean ± sd. Data were analyzed from GSE6428373. d, Histogram of LINC00881 expressions across 20 human tissues. Data were analyzed from NCBI Accession PRJNA28060074. e, Histogram of LINC00881 expressions across 16 human tissues. Data were analyzed from GSE30611 (Illumina Human Body Map 2.0 Project). Source numerical data are available in source data.
a, ORFs predicted in LINC00881, showing predicted peptide sequence of each ORF. A FLAG tag was inserted right after each ORF. b, Western blotting detection of peptide expressions in HEK293T cells transfected with positive control, vector, and FLAG tagged ORF1, ORF2 or ORF3. Positive control is an irrelevant FLAG tagged protein coding gene. The experiment was carried out 2 times with similar outcomes. c, Immunofluorescent staining to detect expressions of FLAG tagged peptides in HEK293T cells transfected with positive control, vector, and FLAG tagged ORF1, ORF2 or ORF3. d, RT-qPCR detecting LINC00881 RNA expression levels in HEK293T cells transfected with positive control, vector, and FLAG tagged ORF1, ORF2 or ORF3. Bars are presented as mean ± s.e.m. Unpaired two-tailed t-test is used for comparison. **p < 0.01, ***p < 0.001, (n = 3 independent experiments). Source numerical data and unprocessed blots are available in source data.
a, LIPTER was knocked out in hiPSCs using CRISPR /Cas-9. Dual gRNAs were designed to completely ablate LIPTER. Genotyping was carried out in individual hiPSC clones using PCR primer sets for detecting long deletions. b, Ratios of beating EBs during CM differentiation from WT and two LIPTERKO hiPSC lines. Dots represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. No significance detected. (n = 3 independent experiments). c, Representative FACS results of cTnT+ cells ratios at day 20 differentiation of WT and LIPTERKO hiPSC lines. d, Quantification of cTnT+ cell ratios at day 20 of differentiation. Bars represent mean values, (n = 2 independent experiments). e, RT-qPCR detecting expressions of CMs markers CTNT, MYH6 and MYH7 in WT and LIPTERKO hiPSC-EBs at day 20 of differentiation. Bars represent mean values, (n = 2 independent experiments). f, Representative FACS results of cTnT+ CM ratios in WT and LIPTERKO hiPSC-EBs at day 40 of differentiation. g, Nile Red and cTnT co-staining in WT, LIPTERKO and LIPTERKO/OE hiPSC-derived EB sections. h, Quantification of Nile Red positive areas in cTnT+ CM areas. Bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ***p < 0.001. (n = 4 independent experiments). i, RT-qPCR detection of LIPTER overexpression levels in LIPTERKO/OE hiPSC-CMs treated with various concentrations of doxycycline for 4 days. Bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ****p < 0.0001, **p < 0.01, n.s (no significance). (n = 3 independent experiments). Source numerical data are available in source data.
Extended Data Fig. 4 LIPTER deficiency compromises mitochondrial function and induces apoptosis of human iPSC-CMs.
a, Comparing the expression levels of CD36, TG synthesis/lipolysis related genes and PLIN5 in enriched WT and LIPTERKO hiPSC-CMs at day 40 of differentiation using RT-qPCR. Bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ****p < 0.0001, **p < 0.01. ns. no significance. (n = 4 independent experiments). b, Western blotting detection of GPAM, PLIN5 and ATGL1 protein expressions in enriched WT and LIPTERKO hiPSC-CMs at day 40 of differentiation. c, Statistical analysis of Western blotting results in b. Bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. **p < 0.01, *p < 0.05. (n = 3 times of independent experiments in the 1st and 3rd groups; n = 4 independent experiments in the 2nd group). d, WT and LIPTERKO hiPSC-derived embryoid bodies were treated with 200 µM palmitic acid for 4 days, followed by immunostaining for PLIN5 and Neil Red lipid staining on EB sections. Arrows indicate condensed PLIN5. The experiment was carried out 3 times with similar outcomes. e, Live cell images showing the fusion of Rhodamine B-palmitic acid (red) labeled-LDs with mitochondria (green) and subsequent diminishment in WT hiPSC-CMs over 45 min, while no apparent LD-mitochondria fusion was observed in LIPTERKO, and MYH10KO hiPSC-CMs. Images were from the continuous live cell imaging during 45 min of cultured hiPSC-CMs. The experiment was carried out 3 times with similar outcomes with statistical results in (f). f, Quantification of the ratios of LD fused with mitochondria in e. Bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. *p < 0.05. (n = 3 independent experiments). g, Analyses of the maximal oxygen consumption values with Palmitate:BSA between conditions without and with Etomoxir (Eto) to compare relative fatty acid oxidation (FAO) capabilities (yellow zone) in WT, LIPTER KO and LIPTER KO/OE hiPSC-CMs. The experiment was carried out 4 times with similar outcomes. h, Representative immunofluorescent images of TUNEL and cTnT co-staining on day 40 hiPSC-EB sections. The experiment was carried out 5 times with similar outcomes. i, Representative immunofluorescent images of Cleaved Caspase-3 and cTnT co-staining on day 40 hiPSC-EB sections. The experiment was carried out 4 times with similar outcomes. Source numerical data and unprocessed blots are available in source data.
Extended Data Fig. 5 NKX2.5 deficiency reduces LIPTER expression and phenocopies LIPTERKO hiPSC-CMs.
a, RT-qPCR analysis of cTnT, LIPTER, MYH10 and NKX2.5 expressions in WT hiPSC-CMs after infection with AAV9 virus carrying scramble control shRNA or two NKX2.5-shRNAs. (n = 5 independent experiments). b, Representative images of Oil Red O lipid staining for LD accumulation, and NKX2.5, cTnT, DAPI immunostaining in WT, LIPTER KO and LIPTER KO/OE hiPSC-CMs treated with control shRNA or NKX2.5-shRNA under low (5.5 mM) or high glucose (22 mM) conditions. The experiment was carried out 4 times with similar outcomes. c, Quantification of the ratios of Oil Red O positive areas in cTnT+ CM areas in b. (n = 4 independent experiments). d, Representative immunofluorescent images of TUNEL, NKX2.5, cTnT and DAPI staining in WT, LIPTER KO and LIPTER KO/OE hiPSC-CMs treated with control or NKX2.5 shRNA under high glucose conditions (22 mM). The experiment was carried out 4 times with similar outcomes. e, Quantification of the ratios of TUNEL+ CMs in cTnT+ CMs of (d). (n = 4 independent experiments). In a,c,e, bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05. Source numerical data are available in source data.
a, RT-qPCR analysis of RXRA and CEBPB expressions in WT hiPSC-CMs treated with high glucose for 2 weeks. (n = 4 independent experiments). b, Representative immunofluorescent images of RXRA and CEBPB staining in WT hiPSC-CMs treated with or without high glucose for 2 weeks. The experiment was carried out 3 times with similar outcomes. c, RT-qPCR detection of RXRA and CEBPB mRNA levels in human left ventricle tissues. NF (n = 9 samples), NF+T2DM (n = 4 samples), DCM (n = 12 samples) and DCM+T2DM (n = 7 samples). d, Representative immunofluorescent images of RXRA and CEBPB staining in human left ventricle tissues. The experiment was carried out with 3 samples per group with similar outcomes. In a,c, bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ****p < 0.0001, **p < 0.01, *p < 0.05. Source numerical data are available in source data.
a, RT-qPCR analysis of LIPTER, U6 (nuclear RNA) and GAPDH (cytosolic RNA) expression levels in fractionated nuclei and cytoplasm of hiPSC-CMs. Bars represent mean values ± s.e.m. (n = 3 independent experiments). b, Representative images showing no detectable interactions of giant lipid vesicles generated by TopFluor-PI with Alexa594-labeled LIPTER or Antisense-LIPTER (AS-LIPTER). The experiment was carried out 3 times with similar outcomes. c, A specific protein band only pulled down by LIPTER-MS2 in HEK293T cells and visualized in SDS PAGE gel by silver staining. MS2 binding protein was FLAG-tagged. The experiment was carried out 2 times with similar outcomes. d, Western blotting confirming MYH10 pulldown by LIPTER in HEK293T cells. The experiment was carried out 2 times with similar outcomes. e, 3D confocal immunofluorescent images showing colocalization of LIPTERYFP-MS2, MYH10, and LD in WT hiPSC-CMs (also see Supplementary Video 1). The lower right graph indicates the view of merged images. The experiment was carried out 3 times with similar outcomes. f, Immunofluorescent images showing MYH10-ACTIN cytoskeleton in WT hiPSC-CMs. The experiment was carried out 3 times with similar outcomes. Source numerical data and unprocessed blots are available in source data.
a, PCR detection of MYH10 null hiPSC clones confirms successful knockout of MYH10. b, Mitochondria stress assay results collected from a Seahorse XF96 Analyzer showing the differences of maximal oxygen consumption values with Palmitate:BSA between conditions without and with Etomoxir (Eto). The yellow zone in panels represent FAO capabilities of WT and MYH10KO hiPSC-CMs. c, Representative immunofluorescent images for TUNEL and cTnT co-staining in WT and MYH10 KO hiPSC-EB sections. d, Representative images showing LD accumulation and cytosolic distribution in WT and MYH10KO hiPSC-CMs treated with 200 μM palmitic acid for 6 h. The experiment was carried out 4 times with similar outcomes. d', Quantification of relative LD density in the 1/2 cytosolic areas of CMs in d. (n = 4 independent experiments). e, Quantification of fluorescence levels in mitochondria isolated from WT and MYH10KO hiPSC-CMs after Rhodamine B-palmitic acid treatment for 2 h. (n = 3 independent experiments). f, Representative images of Oil Red O lipid staining (first two columns); Nile Red and cTnT co-staining (3 rd column) in WT hiPSC-EBs treated with (S)-(-)-Blebbistatin or (R)-(-)-Blebbistatin for 10 days. g, Representative images of TUNEL and cTnT co-staining in WT hiPSC-EBs treated with (S)-(-)-Blebbistatin or (R)-(-)-Blebbistatin for 10 days. h, Quantification of TUNEL+ CM ratios in WT hiPSC-CMs treated with (S)-(-)-Blebbistatin or (R)-(-)-Blebbistatin for 10 days. (n = 5 independent experiments). i, Immunofluorescent images showing Myh10 deficiency in Myh10CKO mouse heart. In d', e, h, bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ***p < 0.001, **p < 0.01, *p < 0.05. Source numerical data are available in source data.
a, Representative FACS results showing cleaved-Caspase-3/cTnT double positive cell ratios in control and LIPTEROE hiPSC-CMs treated with or without 400 μM palmitic acid for 4 days. The experiment was carried out 3 times with similar outcomes. b, Representative Immunofluorescent images of TUNEL and cTnT co-staining in control and LIPTEROE hiPSC-EBs treated with 400 μM palmitic acid for 4 days. The experiment was carried out 3 times with similar outcomes. c, Schematic of LIPTER(Tg) mouse generation using CRISPR/Cas-9, with the Rosa26 locus targeted by a single gRNA. d, RT-qPCR examining LIPTER expression levels in WT and LIPTER(Tg) mouse hearts. (n = 5 mice per group). e, Representative images of WT and LIPTER (Tg) mouse hearts after 7 months of HFD feeding. The heart weight/tibia length ratio was calculated. (n = 5 mice in WT, and n = 6 in Tg). f, Representative immunofluorescent images of TUNEL and cTnT co-staining in mouse heart sections after 7 months of HFD feeding. The experiment was carried out 4 times with similar outcomes. g, Quantification of TUNEL+ CM ratios in (f). (n = 4 mice per group). h, RIP-seq results showing LIPTER pulldown in LIPTER (Tg) mouse heart using anti-Myh10 antibody, with the red arrow indicating enriched signals on exon 3 of transgenic LIPTER. Control Actb and Gapdh RNAs were not pulled down by anti-Myh10 antibody. The experiment was carried out in 2 mice each genotype with similar outcomes. In d, e, g, bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ***p < 0.001, *p < 0.05, (n.s), no significance. Source numerical data are available in source data.
Extended Data Fig. 10 CM-targeted LIPTER transgene exhibits cardiac protective effects in Leprdb/db mice.
a, Heart images of WT and db/db mice 6 weeks after injection of AAV9-cTnT-GFP, AAV9-cTnT-AS-LIPTER or AAV9-cTnT-LIPTER, with fluorescent images showing robust GFP expression post AAV9-cTnT-GFP injection. The experiment was carried out 4 times with similar outcomes. b, Representative images showing GFP expression in CMs of db/db mouse heart 6 weeks after AAV9-cTnT-GFP injection. The experiment was carried out 4 times with similar outcomes. c, RT-qPCR examining GFP expression levels in WT and db/db mouse hearts 6 weeks after AAV9 injection. (n = 5 mice per group, except n = 4 mice in the GFP group). d, Representative images of WGA staining in WT or db/db mouse hearts 6 weeks after AAV9 injection. The experiment was carried out 5 times with similar outcomes. e, Blood glucose levels of WT and db/db mice 6 weeks after AAV9 injection. WT mice without AAV9 injection also serve as control. (n = 5 mice per group, except n = 4 mice in the GFP group). f, EMSA results showing no direct interaction between LIPTER and TFG protein. The experiment was carried out 2 times with similar outcomes. In c and e, bars represent mean values ± s.e.m. Unpaired two-tailed t-test is used for comparison. ***p < 0.001, **p < 0.01. Source numerical data are available in source data.
Flow cytometry gating for cTnT positive. hiPSC-EBs were dissociated into single cells and then analysed by flow cytometry.
Supplementary Tables 1–6.
Stacked images from confocal microscopy showing co-localization of LIPTER, MYH10 and LDs in hiPSC-CMs. 3D video indicates LIPTER-MS2YFP (green), MYH10 (red), LDs (magenta) and DAPI (blue) staining in WT hiPSC-CMs.
LIPTER and LDs migrating together in hiPSC-CMs. Video showing LIPTER-MS2YFP (green) and Rhodamine-B-labeled-LDs (red) migrating together (yellow dots) only in cytosol of WT hiPSC-CMs. N, nucleus.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Statistical source data.
Unprocessed western blots and/or gels.
Unprocessed western blots and/or gels.
Unprocessed western blots and/or gels.
Unprocessed western blots and/or gels.
About this article
Cite this article
Han, L., Huang, D., Wu, S. et al. Lipid droplet-associated lncRNA LIPTER preserves cardiac lipid metabolism. Nat Cell Biol 25, 1033–1046 (2023). https://doi.org/10.1038/s41556-023-01162-4