A long non-coding RNA LncSync regulates mouse cardiomyocyte homeostasis and cardiac hypertrophy through coordination of miRNA actions

Graphical abstract

Dear Editor,

Maintenance of cardiomyocyte (CM) homeostasis is essential for normal heart function. Long-term imbalance in heart homeostasis could elicit irreversible adaptive change in cell structure and function and tissue architecture, exemplified as cardiac hypertrophy and fibrosis, and eventually develop into heart failure (Shiojima et al., 2005). Therefore, identifying new genes and pathways regulating CM homeostasis may help better understand the cause of cardiac hypertrophy.

In recent years, researchers identified numerous long non-coding RNAs (lncRNAs) specifically expressed in the heart. However, many cardiac-specific lncRNAs seemed to be dispensable for heart development (Han et al., 2018; George et al., 2019). LncSync (C430049B03Rik) is a lncRNA located on mouse chrX:53053112–53057190 (⁻strand, mm10) and evolutionarily conserved in placental mammals (Fig. 1A). It has four exons, and exon 4 contains sequences encoding the miR-351 cluster: miR-351, miR-503, miR-322, and the target site of miR-181 (Fig. 1A). From E9.5 to E13.5, LncSync is significantly higher expressed in the heart than in other organs. In adult mice, it is only detected in the heart (Fig. 1B). RNA in situ hybridization revealed that LncSync is already detectable in the E9.5 heart and increased substantially in the ventricle at E10.5 and E11.5 (Fig. 1C). The expression pattern of LncSync can also be confirmed by the single-cell RNA-seq (scRNA-seq) analysis. It was expressed in Tnnt2⁺ embryonic CMs (Fig. S1A–D) (de Soysa et al., 2019; Pijuan-Sala et al., 2019). Similarly, MIR503HG, the ortholog of LncSync in humans, is enriched in the CMs of the human heart (Fig. S2A–D) (Haniffa et al., 2021). Subcellular fractionation assay showed that 75% of LncSync transcripts are present in the nucleoplasm and 23% in the cytoplasm, while Xist transcript was nearly 100% in the chromatin fraction (Fig. S3A). This result suggested that LncSync is a pri-microRNA transcript in the nucleoplasm. Accordingly, the expression levels of miR-351, miR-503, and miR-322 were parallel with that of LncSync during embryonic heart development (Fig. 1D). Intriguingly, LncSync also contained a sequence that is highly complementary to the seed sequence of the miR-181 family on exon 4 (Figs. 1E and S3B). We cloned the wild-type (WT) and mutated miR-181 targeting region of LncSync into a dual-luciferase reporter vector psiCheck-2 downstream of the Renilla luciferase (Fig. 1E). As expected, miR-181 significantly decreased the luciferase activity of the WT LncSync reporter compared to that of the mutant reporters (Fig. 1F). Transfection of miR-181 mimics to mESC overexpressing LncSync lead to a significant decrease in the levels of miR-351, miR-503, and miR-322 (Fig. S3C and 3D). Introducing miR-181 into purified E13.5 CMs also reduced LncSync and the miR-351 cluster levels (Fig. 1G). RNA-seq of E13.5 CMs transfected with miR-181 showed that 595 genes were significantly downregulated by miR-181. Gene Ontology (GO) terms associated with these genes include cardiac muscle growth, cardiac hypertrophy, and response to calcium ion (Figs. 1H and S3E), suggesting that miR-181 might impact CM development and function.

During mESC cardiac differentiation, LncSync expression was highly upregulated in plated embryonic bodies (EBs) on day 10 (D10) (Fig. S3F and S3G). Knocking out LncSync in mESCs caused a marked decrease in the contraction ratio of LncSyn^−/− EBs (Fig. 1I). In the meantime, induced overexpression of LncSync using doxycycline (dox) -inducible system in α-MHC-GFP reporter mESC significantly increased the GFP fluorescence and the ratio of beating EBs on D10 (Figs. 1J, S3F and S3H). These data demonstrated that LncSync could promote CM differentiation in vitro.

We generated LncSync^−/− mouse by CRISPR (Figs. 2A and S4A). LncSync^−/− mice were born with the expected Mendelian ratio based on the genotyping result of 114 pups (Fig. S4B) and viable and fertile. The heart morphology, weight, and structure of E13.5–E17.5 LncSync^−/− and WT embryos did not show obvious abnormality (Fig. S4C and S4D). Interestingly, in 24 weeks old mouse, LncSync^−/− hearts became grossly larger, and the left ventricular wall appeared substantially thicker on the cross-section (Fig. 2B). The heart weight/tibia length ratio (HW/TL) of WT hearts is about 0.1 g/cm, while the HW/TL of LncSync^−/− hearts increased to 0.15 g/cm, almost 50% heavier (Fig. 2C). Moreover, echocardiography results revealed that the left ventricular mass (LVmass) increased significantly in 24 weeks old LncSync^−/− mice (Fig. 2D and 2E). Wheat Germ Agglutinin (WGA) staining showed that the cross-sectional area of LncSync^−/− CMs was 70% larger, indicating that the size but not the number of CMs increased after LncSync deletion (Fig. 2F and 2G). Masson’s trichrome staining of cross sections of the heart revealed that areas with collagen deposition in LncSync^−/− hearts reached about 4%, significantly more extensive than that in the WT heart (about 1%) (Fig. 2H and 2I). In addition to the apparent cardiac hypertrophy and fibrosis, LncSync^−/− mice also developed hypertension. We measured the blood pressure of 24 weeks LncSync^−/− mice and found their blood pressure significantly higher than the WT mice (Fig. S5A). The morphology of the capillaries in the renal corpuscle and the lung appeared normal (Fig. S5B), suggesting that the LncSync^−/− mouse did not have pathological changes in the blood vessel. To prove that the cardiac hypertrophy of LncSync^−/− mice was not due to high blood pressure, we gave 13-weeks WT and LncSync^−/− mice angiotensin II (Ang II) for 28 days. The SBP and LV mass of LncSync^−/− mice were already significantly higher than that of WT mice at 13 weeks, but the difference narrowed over time, and we did not observe a significant difference between these two groups by 17 weeks (Fig. S5C, S5D, S5G and S5H). After 28 days, Ang II caused a significant increase in LV area in 17 weeks WT heart (Fig. S5E and S5F). On the other hand, untreated LncSync^−/− hearts displayed hypertrophy compared to WT hearts at 17 weeks, and there was little difference in LncSync^−/− hearts with or without Ang II treatment (Fig. S5E and Fig. S5E). These results strongly suggest that LncSync^−/− hearts became hypertrophic in a cell-autonomous manner and responded poorly to Ang II. Next, we purified E13.5 and 8 weeks old adult CMs and performed high-throughput sequencing of miRNAs and mRNAs (Figs. 2A and S6A). Both WT and LncSync^−/− CMs highly expressed CMs markers (Tnnt2, Tnni3), but lowly expressed EC markers (Kdr), smooth muscle cell (SMC) markers (Tagln), hematopoietic stem cell (HSC) markers (Ly6a, Kit), and pluripotent stem cell (PSC) markers (Nanog, Sox2, Pou5f1) (Fig. S6B). Eight hundred and fifty two and 324 genes were significantly up- and downregulated in the E13.5 LncSync^−/− CMs, respectively (Fig. 2J). GO and KEGG analysis revealed a significant enrichment of genes related to cardiac hypertrophy and ferroptosis (a form of regulated cell death caused by lipid hydroperoxides) in the LncSync^−/− CMs (Figs. 2K and S6C), suggesting that the deletion of LncSync induced a certain degree of ferroptosis in the embryonic heart. Similarly, 8-weeks adult LncSync^−/− CMs also significantly upregulated hypertrophic genes: α-skeleton actin 1 (Acta1), brain natriuretic peptide β (Nppb), and regulator of calcineurin 1 (Rcan1) and fibrotic genes: Collagen1a1 (Col1a1) and Collagen3a1 (Col3a1) (Fig. S6D).

We analyzed the miRNA profile of WT and LncSync^−/− CMs. As expected, LncSync hosted miR-351 cluster, including passenger strands (miR-503-3p, miR-351-3p, miR-322-3p), were the only microRNAs massively downregulated in both embryonic and adult LncSync^−/− CMs (Fig. S6E and S6F). To investigate the downstream target of LncSync-derived miRNAs in vivo, we performed miRNA target analysis using target prediction databases MirDB (Chen and Wang, 2020), Miranda (Betel et al., 2008), and TargetScan (Agarwal et al., 2015). Eighteen genes were significantly upregulated in the LncSync^−/− CMs and predicted targets of the miR-351 cluster (Fig. 2L). Hddc3 is a protein-coding gene whose expression level increased substantially in both embryonic and adult LncSync^−/− CMs (Fig. 2L). Sequence analysis and luciferase reporter assay confirmed that Hddc3 is a direct target of miR-351 (Fig. 2M and 2N). The RNA and protein levels of Hddc3 raised significantly in the LncSync^−/− heart (Figs. 2O–Q and S7A). These results strongly suggest that Hddc3 is upregulated in LncSync^−/− heart as a target of miR-351. A recent study reported that human HDDC3 (also named MESH1) is a cytosolic NADPH phosphatase, and the depletion of HDDC3 could protect cells from ferroptosis (Ding et al., 2020). To test whether Hddc3 can regulate ferroptosis in CMs, we treated E13.5 CMs with ferroptosis inducer erastin. Compared to the DMSO group, erastin treatment reduced CM viability to 0.3 (Fig. 2R). When Hddc3 was knocked down with siRNA (Fig. S7B), cell viability remained 0.7–0.9 with erastin treatment (Fig. 2R). This result indicated that reducing Hddc3 levels could indeed protect CMs from ferroptosis. Moreover, RNA-seq revealed that markers indicative of ferroptosis (Yang et al., 2014; Fang et al., 2019, 2020), such as Ptgs2, Tfrc (transferrin receptor), Slc7a11, and Fth1 (ferritin heavy polypeptide 1), were significantly upregulated in the LncSync^−/− CMs (Fig. S7B). The ferroptosis end-product malondialdehyde (MDA) elevated markedly in LncSync^−/− hearts (Fig. 2S). Transmission electron microscopy revealed that mitochondria were distorted and shrunken in the 24 weeks old LncSync^−/− adult CMs (Fig. S7C). The above results suggest that in the LncSync^−/− heart, the absence of the miR-351 cluster caused the derepression of Hddc3, which led to dysregulation of CMs metabolism and activation of the ferroptosis pathway.

Our study revealed the important physiological function of LncSync in mammalian hearts and uncovered a delicate cross-regulation pattern between lncRNA and microRNAs. LncSync and miR-351 control the levels of Hddc3. Without LncSync, HDDC3 protein levels increase, which might disrupt the metabolic balance in CMs and makes them susceptible to ferroptosis, and the KO mice eventually developed pathological cardiac hypertrophy. The regulatory relationship between LncSync and miR-181 is also interesting. There have been many reports about the versatile roles of miR-181 family members in embryo development and physiological and pathological processes, particularly in the cardiovascular system (Sun et al., 2014). We showed that miR-181 could target LncSync hence its derived miR-351 cluster. This finding expanded the scope of miR-181 targets. According to our microRNA profiling, the miR-181 level is significantly higher in adult CMs than in embryonic CMs (Fig. S6G and S6H), which may downregulate the LncSync post-transcriptionally and its derived miR-351 cluster in adult mice. The growth, differentiation, and function of the CMs need constant adjusting depending on the developmental stage and the physiological requirement. LncSync and the miR-351 cluster might serve as a rheostat to regulate CMs fate specification and homeostasis, for example, by reducing Hddc3 levels. Besides, LncSync could be under the control of miR-181 to fine-tune the CMs’ status in response to specific conditions. Deleting LncSync abrogates such regulation and eventually leads to dysregulation of CM homeostasis, slow onset of ferroptosis, and hypertrophic remodeling of the heart.

References

[1] Agarwal V, Bell GW, Nam JW et al. Predicting effective microRNA target sites in mammalian mRNAs. Elife 2015;4:e05005.

[2] Betel D, Wilson M, Gabow A et al. The microRNA.org resource: targets and expression. Nucl Acids Res 2008;36:D149–153.

[3] Chen Y, Wang X. miRDB: an online database for prediction of functional microRNA targets. Nucleic Acids Res 2020;48:D127–131.

[4] de Soysa TY, Ranade SS, Okawa S et al. Single-cell analysis of cardiogenesis reveals basis for organ-level developmental defects. Nature 2019;572:120–124.

[5] Ding CC, Rose J, Sun T et al. MESH1 is a cytosolic NADPH phosphatase that regulates ferroptosis. Nat Metab 2020;2:270–277.

[6] Fang X, Wang H, Han D et al. Ferroptosis as a target for protection against cardiomyopathy. Proc Natl Acad Sci USA 2019;116:2672–2680.

[7] Fang X, Cai Z, Wang H et al. Loss of cardiac ferritin H facilitates cardiomyopathy via Slc7a11-mediated ferroptosis. Circ Res 2020;127(4):486–501.

[8] Feng H, Schorpp K, Jin J et al. Transferrin receptor is a specific ferroptosis marker. Cell Rep 2020;30:3411–3423.

[9] George MR, Duan Q, Nagle A et al. Minimal in vivo requirements for developmentally regulated cardiac long intergenic non-coding RNAs. Development 2019;146(23):dev185314.

[10] Han X, Luo S, Peng G et al. Mouse knockout models reveal largely dispensable but context-dependent functions of lncRNAs during development. J Mol Cell Biol 2018;10:175–178.

[11] Haniffa M, Taylor D, Linnarsson S et al. A roadmap for the Human Developmental Cell Atlas. Nature 2021;597:196–205.

[12] Pijuan-Sala B, Griffiths JA, Guibentif C et al. A single-cell molecular map of mouse gastrulation and early organogenesis. Nature 2019;566:490–495.

[13] Shiojima I, Sato K, Izumiya Y et al. Disruption of coordinated cardiac hypertrophy and angiogenesis contributes to the transition to heart failure. J Clin Invest 2005;115:2108–2118.

[14] Sun X, Sit A, Feinberg MW. Role of miR-181 family in regulating vascular inflammation and immunity. Trends Cardiovasc Med 2014;24:105–112.

[15] Yang WS, SriRamaratnam R, Welsch ME et al. Regulation of ferroptotic cancer cell death by GPX4. Cell 2014;156:317–331.

RIGHTS & PERMISSIONS

©The Author(s) 2022. Published by Oxford University Press on behalf of Higher Education Press.