Minimal in vivo requirements for developmentally regulated cardiac long intergenic non-coding RNAs

Abstract

Long intergenic non-coding RNAs (lincRNAs) have been implicated in gene regulation, but their requirement for development needs empirical interrogation. We computationally identified nine murine lincRNAs that have developmentally regulated transcriptional and epigenomic profiles specific to early heart differentiation. Six of the nine lincRNAs had in vivo expression patterns supporting a potential function in heart development, including a transcript downstream of the cardiac transcription factor Hand2, which we named Handlr (Hand2-associated lincRNA), Rubie and Atcayos We genetically ablated these six lincRNAs in mouse, which suggested genomic regulatory roles for four of the cohort. However, none of the lincRNA deletions led to severe cardiac phenotypes. Thus, we stressed the hearts of adult Handlr and Atcayos mutant mice by transverse aortic banding and found that absence of these lincRNAs did not affect cardiac hypertrophy or left ventricular function post-stress. Our results support roles for lincRNA transcripts and/or transcription in the regulation of topologically associated genes. However, the individual importance of developmentally specific lincRNAs is yet to be established. Their status as either gene-like entities or epigenetic components of the nucleus should be further considered.

Keywords: Gene regulation; Heart development; Long non-coding RNA.

Fig. 1.

Epigenetically regulated cardiac lincRNAs. (A) Differentiation progression of mESCs into cardiomyocytes used for lincRNA candidate selection. ESC, embryonic stem cell; MES, mesoderm; CP, cardiac progenitor; CM, cardiomyocyte. (B) Criteria for lincRNA identification and the resulting nine candidates. Asterisks indicate names assigned by the Bruneau lab.

Fig. 2.

Genomic characterization of Rubie and Handlr in vitro. (A) UCSC Genome Browser tracks of Rubie RNA-seq and overlaid histone H3 ChIP-seq at ESC, MES, CP and CM stages of in vitro differentiation; RefSeq annotation in blue. (B) Quantified expression of Rubie and Bmp4 at each differentiation stage. (C) 3D Genome Browser Hi-C heatmap of chromosome interactions around Bmp4 and Rubie loci. (D) UCSC Genome Browser tracks of Handlr RNA-seq and overlaid histone H3 ChIP-seq at ESC, MES, CP and CM stages of in vitro differentiation. Ensembl annotation in red; observed exon structure of predominant Handlr transcript in black with blue asterisks. (E) Quantified expression of Handlr and Hand2 at each differentiation stage. (F) 3D Genome Browser Hi-C heatmap of chromosome interactions around Handlr and Hand2 loci. ESC, embryonic stem cell; MES, mesoderm; CP, cardiac progenitor; CM, cardiomyocyte; blue, ESC; green, MES; orange, CP; red, CM; K4me3, histone H3 lysine 4 trimethylation; K27me3, histone H3 lysine 27 trimethylation; K27Ac, histone H3 lysine 27 acetylation; TAD, topologically associated domain. Data are mean±s.e.m. in B and E.

Fig. 3.

lincRNA expression patterns in vivo. (A) In situ hybridization staining for Rubie from E7.5 to E9.5. (B) In situ hybridization staining for Handlr at E8.5 and E9.5. (C) In situ hybridization staining for Atcayos at E9.5 and E10.5. (D) In situ hybridization staining for HrtLincR4 at E8.25, E8.5 and E9.5. (E) In situ hybridization staining for HrtLincR5 at E8.0 and E9.5. (F) In situ hybridization staining for HrtLincRX from E7.25 through E10.5. A, anterior; P, posterior; Em, embryonic region; Ex, extra-embryonic region; CC, cardiac crescent; Ht, heart tube; H, heart; OV, otic vesicle. Scale bars: 100 µm; 500 µm for E10.5 in C and F. All images are white balanced for clarity.

Fig. 4.

Cas9 ablation of cardiac lincRNAs in vivo and effects on local gene expression. (A) lincRNA TSS/promoter ablation strategy. TSS, transcriptional start site; tru-sgRNA, truncated single guide RNA. (B) Schematic for RT-qPCR on the anterior half of an E8.25 embryo and gDNA genotyping on the posterior half of the embryo. A, anterior; P, posterior; dotted red line indicates a bisection point; colored arrows indicate the orientation of genotyping primers; scissors, Cas9 cut sites. (C) Left: electrophoresed gDNA PCR genotyping products of Rubie alleles and quantification of the resulting Rubie and Bmp4 expression in anterior E8.25 embryos. Right: correlation between Rubie expression and Bmp4 expression for all genotypes or Rubie^+/− only. (D) Electrophoresed gDNA PCR genotyping products of Handlr alleles and quantification of the resulting Handlr and Hand2 expression in anterior E8.25 embryos. (E) Electrophoresed gDNA PCR genotyping products of Atcayos alleles and quantification of the resulting Atcayos and Nmrk2 expression in anterior E8.25 embryo. (F) Electrophoresed gDNA PCR genotyping products of HrtLincR4 alleles and quantification of the resulting HrtLincR5 and Mn1 expression in anterior E8.25 embryos. (G) Electrophoresed gDNA PCR genotyping products of HrtLincR5 alleles and quantification of the resulting HrtLincR4 and Trabd2b expression in anterior E8.25 embryos. (H) Left: electrophoresed gDNA PCR genotyping products of HrtLincRX alleles and quantification of the resulting HrtLincRX and Plac1 expression in anterior E8.25 embryos. (I) IntaRNA 2.0 binding prediction between HrtLincRX 3′ miRNAs and Plac1. *P<0.05; **P<0.01; ***P<0.005; n.s., not significant; Student's two-tailed t-test. Data are mean±s.e.m. in C-H.

Fig. 5.

Viability and phenotypic effects after lincRNA knockout. (A) Offspring recovered at weaning from Rubie^+/−×Rubie^+/− cross versus expected Mendelian ratios. (B) Representative sporadic circling behavior in Rubie^−/− offspring; #, only observed in Rubie^−/− genotype over 2+ years of observation (left-handedness not exclusive in all observed cases). (C) Offspring recovered at weaning from Handlr^+/−×Handlr^+/− cross versus expected Mendelian ratios. (D) Offspring recovered at weaning from Atcayos^+/−×Atcayos^+/− cross versus expected Mendelian ratios; P-values in A, C and D derived from χ² test. (E) LV mass calculated by echocardiographic measurements in Handlr wild-type versus null and in Atcayos wild-type versus null litter-matched adult males. (F) LV contractility measured by echocardiography in Handlr wild-type versus null and in Atcayos wild-type versus null litter-matched adult males. (G) Time course of cardiac contractility after TAC in Handlr wild-type versus null litter-matched adult males. (H) Time course of cardiac contractility after TAC in Atcayos wild-type versus null litter-matched adult males. (I) Week 8 post-TAC LV and lung weights in Handlr wild-type versus null litter-matched adult males. (J) Week 8 post-TAC LV and lung weights in Atcayos wild-type versus null litter-matched adult males. *P<0.05; **P<0.01; ***P<0.005; Student's two-tailed t-test, except I, Z-test. LV, left ventricle; FAC, fractional area shortening; TAC, transverse aortic constriction. Phenotypic data are mean±s.e.m. n=4-9 for all measurements.

Fig. 6.

Effect of Rubie ablation on heart development at E15.5. (A-C,E) Oblique transverse Hematoxylin and Eosin histological sections of cardiac ventricular and OFT morphogenesis at E15.5. (A) Representative wild-type morphology. (B) Representative Rubie^−/− morphology. (C) Representative Bmp4^+/− (Actb-Cre⁺) morphology. (D) Left: electrophoresed gDNA PCR genotyping products of Rubie and Actb-Cre transgene alleles. Right: offspring recovered at weaning from Rubie^+/−; Actb-Cre⁺×Bmp4^fl/fl mating versus expected Mendelian ratios. (E) Representative Bmp4^+/−; Rubie^+/− (Actb-Cre⁺) morphology in two separate individuals. RV, right ventricle; LV, left ventricle; OFT, outflow tract; WT, wild type. Scale bars: 300 µm. Arrows indicate distorted OFT orientation. P-values were obtained using a χ² test.