The lncRNA GATA6-AS epigenetically regulates endothelial gene expression via interaction with LOXL2 - PubMed (original) (raw)

doi: 10.1038/s41467-017-02431-1.

Nicolas Jaé 1 2, Andrea Knau 1, Simone F Glaser 1 2, Youssef Fouani 1, Oliver Rossbach 3, Marcus Krüger 4, David John 1 2, Albrecht Bindereif 3, Phillip Grote 1 2, Reinier A Boon 1 2, Stefanie Dimmeler 5 6

Affiliations

The lncRNA GATA6-AS epigenetically regulates endothelial gene expression via interaction with LOXL2

Philipp Neumann et al. Nat Commun. 2018.

Abstract

Impaired or excessive growth of endothelial cells contributes to several diseases. However, the functional involvement of regulatory long non-coding RNAs in these processes is not well defined. Here, we show that the long non-coding antisense transcript of GATA6 (GATA6-AS) interacts with the epigenetic regulator LOXL2 to regulate endothelial gene expression via changes in histone methylation. Using RNA deep sequencing, we find that GATA6-AS is upregulated in endothelial cells during hypoxia. Silencing of GATA6-AS diminishes TGF-β2-induced endothelial-mesenchymal transition in vitro and promotes formation of blood vessels in mice. We identify LOXL2, known to remove activating H3K4me3 chromatin marks, as a GATA6-AS-associated protein, and reveal a set of angiogenesis-related genes that are inversely regulated by LOXL2 and GATA6-AS silencing. As GATA6-AS silencing reduces H3K4me3 methylation of two of these genes, periostin and cyclooxygenase-2, we conclude that GATA6-AS acts as negative regulator of nuclear LOXL2 function.

PubMed Disclaimer

Conflict of interest statement

N.J. and S.D. applied for a patent on the use of long non-coding RNAs for the treatment of endothelial dysfunction. The remaining authors declare no competing financial interests.

Figures

Fig. 1

Fig. 1

The lncRNA GATA6-AS is hypoxia-induced and nuclear enriched in endothelial cells. a, b HUVECs were exposed to hypoxia (0.2% O2) for 12 and 24 h or kept under normoxic conditions and gene expression changes were assayed by deep sequencing of ribo-minus RNA. a Significantly regulated lncRNAs are highlighted in blue, GATA6-AS is highlighted in red (n = 2; statistical significance was assessed by Cuffdiff). b FPKM values were used to identify significantly regulated lncRNAs. VEGFA is shown as hypoxia-induced positive control (n = 2; SEM; *statistical significance was assessed by Cuffdiff). c HUVECs were exposed to hypoxia (0.2% O2) for 24 h or kept under normoxic conditions and regulation of GATA6-AS was confirmed by RT-qPCR using two primer sets targeting different regions of GATA6-AS. Relative expression was normalized to RPLP0 mRNA (n = 4; SEM; * t test p < 0.05). d RNA was isolated from nuclear and cytoplasmic fractions and analyzed by RT-qPCR using primers targeting the indicated transcripts (n = 4; SEM; * t test p < 0.05). e Akt phosphorylation was inhibited in HUVECs by addition of MK-2206 to the culture medium 1 h prior to hypoxic treatment (24 h, 0.2% O2). As controls, cells were incubated under normoxic conditions. Expression levels of GATA6-AS were determined by RT-qPCR and normalized to RPLP0 mRNA (n = 4; SEM; * t test p < 0.05)

Fig. 2

Fig. 2

GATA6-AS silencing reduces EndMT in vitro and augments vessel formation in a xenograft model. a HUVECs were transfected with GapmeRs targeting GATA6-AS or with control GapmeRs and relative expression of GATA6-AS was determined by RT-qPCR, normalized to RPLP0 mRNA (n = 4; SEM; * t test p < 0.05). b, c GATA6-AS-silenced cells or control cells were cultivated for 72 h in complete medium (−TGF-β2) or medium lacking endothelial growth factor and bovine brain extract, but containing 10 ng/µl TGF-β2 (+TGF-β2). b Relative expression levels of SM22 (left) and calponin (right) were determined by RT-qPCR, normalized to RPLP0 mRNA (n = 2–3; SEM; *t test p < 0.05). c Left: Cells were immunostained for SM22 (red) and VE-cadherin (green) and nuclei were counter stained using Hoechst (representative images are shown; scale bars are 50 µm). Right: For quantification, VE-cadherin gaps per mm2 were determined in 9–10 random fields per condition (n = 5; SEM; *t test p < 0.05). d HUVECs were transfected with GapmeRs targeting GATA6-AS or with control GapmeRs and used for in vitro spheroid sprouting assays under basal conditions and VEGFA (50 ng/ml) stimulation (n = 4; SEM; * t test p < 0.05; representative images are shown; scale bars are 50 µm). e GATA6-AS-silenced cells or control cells were used for in vitro transmigration assays under VEGFA (50 ng/ml) stimulation (n = 6; SEM; * t test p < 0.05). f–h Matrigel basement matrix plugs containing spheroids derived from GATA6-AS or control GapmeR treated HUVECs were injected subcutaneously in immunodeficient mice. Plugs were harvested 21 days later and used for immunohistochemistry detecting human endothelial cells (ulex rhodamine, red) and murine endothelial cells (isolectin-B4, white). Nuclei were stained with Hoechst (blue). f Implanted human endothelial cells were quantified as cells being ulex-posistive and Hoechst-positive (n = 10 plugs; SEM; *Mann–Whitney U test p < 0.05). g Murine endothelial cells were quantified as cells being isolectin B4-positive and Hoechst-positive (n = 10 plugs; SEM). h Representative images of GATA6-AS and control GapmeR matrigel plug sections. Scale bars are 50 µm

Fig. 3

Fig. 3

Antisense affinity selection of endogenous GATA6-AS–protein complexes. a Scheme for testing oligonucleotide accessibility of GATA6-AS. Binding of antisense DNA oligonucleotides to accessible sites within the GATA6-AS transcript results in the formation of DNA-RNA-heteroduplexes and subsequent RNA cleavage by RNase H. Cleaved sites cannot be amplified by flanking primers in subsequent RT-qPCR reactions. b DNA oligonucleotides (AS1–9; see Supplementary Fig. 1b for details) targeting GATA6-AS were incubated in HUVEC cellular extract and RNase H was added to degrade DNA-RNA-heteroduplexes. Following RNA preparation, RT-qPCR was used to assay for accessible sites. In control reactions, DNA oligonucleotides were omitted (n = 3; SEM; *t test p < 0.05). c Scheme for purifying endogenous GATA6-AS–protein complexes. An oligonucleotide-accessible site within GATA6-AS is targeted by a desthiobiotin-coupled 2′O-Me-RNA probe. Subsequently, the complex is recovered using streptavidin beads. d HeLa cell lysate was used to capture endogenous GATA6-AS–protein complexes by RNA affinity selection (AS1 probe). As control, scrambled 2′O-Me-RNA probes were used. Following elution with biotin, co-purified RNA fractions were analyzed for GATA6-AS enrichment by RT-qPCR (n = 4; SEM)

Fig. 4

Fig. 4

GATA6-AS is directly bound by LOXL2. a Total cell lysate was prepared from HeLa cells and 5% was taken as input. The remaining lysate was distributed between GATA6-AS or scramble control antisense affinity selections and co-purified proteins were assayed for LOXL2 by western blotting (n = 3; a representative image is shown). b HUVEC cell lysate was used for RNA immunoprecipitation using LOXL2 antibodies or isotype controls. Left: Co-purified RNA from LOXL2 and IgG IPs was assayed for GATA6-AS enrichment by RT-qPCR (n = 5; SEM; * t test p < 0.05). Right: IP specificity was controlled by LOXL2 western blotting (a representative image is shown). c In vitro transcribed full length GATA6-AS (1788nt) was used in His-LOXL2 pulldown assays. As control, mock pulldowns were performed without protein. Left: Co-purified GATAT6-AS was visualized by Northern blotting (a representative image is shown). Right: For quantification, intensities were determined by densitometry (n = 4; SEM; *t test p < 0.05). d Schematic outline of the LOXL2 iCLIP procedure in HUVECs (top) and identified LOXL2-binding sites within GATA6-AS (bottom). The crosslinked nucleotides of the LOXL2 iCLIP-tags are highlighted in red. e Full length GATA6-AS (1788nt) was in vitro transcribed and folded (5 min 65 °C → 10 min RT → 30 min 30 °C → 15 min 37 °C) or denatured (2 min 95 °C → ice) and used for His-tag pulldown using recombinant LOXL2. As controls, mock reactions were performed without protein. Left: Co-purified RNA was recovered and used for RT-qPCR (n = 4; SEM). Right: Integrity of GATA6-AS input after folding or denaturation was assayed by agarose gel electrophoresis (a representative image is shown). f Full length GATA6-AS and a 735nt truncated construct (nucleotides 643–1377) lacking both LOXL2 interaction sites were in vitro transcribed, folded, and used for His-LOXL2 pulldown. As controls, mock reactions were performed without protein. Co-purified RNA was recovered and used for RT-qPCR (n = 4; SEM; *t test p < 0.05)

Fig. 5

Fig. 5

GATA6-AS silencing does not influence intracellular and extracellular LOXL2 levels. a HUVECs were transfected with siRNAs targeting LOXL2 or with control siRNAs and relative expression of LOXL2 was determined by RT-qPCR, normalized to RPLP0 mRNA (n = 6; SEM; *t test p < 0.05). b LOXL2 protein levels were determined from the same cells by western blotting, using tubulin as loading control (n = 1). c LOXL2-silenced cells or control cells were used for in vitro spheroid sprouting assays under basal conditions (n = 6; SEM; * t test p < 0.05; representative images are shown; scale bars are 50 µm). d HUVECs were transfected with siRNAs or GapmeRs targeting LOXL2 or GATA6-AS, respectively. Left: Intracellular and extracellular LOXL2 levels were assayed by western blotting (representative images are shown). Right: For quantification, intracellular LOXL2 levels were normalized to tubulin (n = 5–6; SEM; *t test p < 0.05)

Fig. 6

Fig. 6

GATA6-AS acts as negative regulator of nuclear LOXL2 function in endothelial gene expression. a, b Left: HUVECs were transfected with GapmeRs or siRNAs targeting GATA6-AS or LOXL2, respectively and H3K4me3, H3 as well as tubulin levels were assayed by western blotting (representative images are shown). Right: For quantification, H3K4me3 levels were normalized to tubulin (n = 3–6; SEM; *t test p < 0.05). c Upon silencing of GATA6-AS, changes in global gene expression levels were determined by exon arrays (all genes; gray). Regulated genes (≥1.2 fold upregulated or downregulated) were analyzed in corresponding LOXL2-silenced samples (cut off; blue) and when found inversely regulated, highlighted in red. Percentages refer to the total of all cut off genes. Six data points are outside the axis limits (n = 2–4). d HUVECs were transfected with siRNAs or GapmeRs targeting LOXL2 or GATA6-AS, respectively and expression of LOXL2 and GATA6-AS, as well as of the indicated angiogenesis-related genes was determined by RT-qPCR, normalized to RPLP0 mRNA (n = 7–8; SEM; *t test p < 0.05). e HUVECs were transfected with GapmeRs targeting GATA6-AS or with control GapmeRs and used for H3K4me3 chromatin immunoprecipitations. Left: Co-precipitated DNA was assayed for PTGS2 and POSTN enrichment by qPCR. A genomic region lacking H3K4me3 signals was used as negative control (n = 4–10; SEM; *t test p < 0.05). Right: Specificity of the immunoprecipitation was controlled by isotype controls and H3K4me3 western blotting (representative images are shown). f Schematic model of GATA6-AS-regulated LOXL2 function. In the nucleus, LOXL2 catalyzes the oxidative deamination of activating H3K4me3 chromatin marks, a process which is negatively regulated by the hypoxia-induced nuclear lncRNA GATA6-AS

References

    1. Djebali S, et al. Landscape of transcription in human cells. Nature. 2012;489:101–108. doi: 10.1038/nature11233. - DOI - PMC - PubMed
    1. Esteller M. Non-coding RNAs in human disease. Nat. Rev. Genet. 2011;12:861–874. doi: 10.1038/nrg3074. - DOI - PubMed
    1. Dinger, M. E., Pang, K. C., Mercer, T. R. & Mattick, J. S. Differentiating protein-coding and noncoding RNA: challenges and ambiguities. PLoS Comput. Biol. 4, e1000176 (2008). - PMC - PubMed
    1. Hu W, Alvarez-Dominguez JR, Lodish HF. Regulation of mammalian cell differentiation by long non-coding RNAs. EMBO Rep. 2012;13:971–983. doi: 10.1038/embor.2012.145. - DOI - PMC - PubMed
    1. Fatica A, Bozzoni I. Long non-coding RNAs: new players in cell differentiation and development. Nat. Rev. Genet. 2014;15:7–21. doi: 10.1038/nrg3606. - DOI - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources