Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The endothelial-enriched lncRNA LINC00607 mediates angiogenic function

The endothelial-enriched lncRNA LINC00607 mediates angiogenic function Long non-coding RNAs (lncRNAs) can act as regulatory RNAs which, by altering the expression of target genes, impact on the cellular phenotype and cardiovascular disease development. Endothelial lncRNAs and their vascular functions are largely undefined. Deep RNA-Seq and FANTOM5 CAGE analysis revealed the lncRNA LINC00607 to be highly enriched in human endothelial cells. LINC00607 was induced in response to hypoxia, arteriosclerosis regression in non-human primates, post- atherosclerotic cultured endothelial cells from patients and also in response to propranolol used to induce regression of human arteriovenous malformations. siRNA knockdown or CRISPR/Cas9 knockout of LINC00607 attenuated VEGF-A-induced angiogenic sprouting. LINC00607 knockout in endothelial cells also integrated less into newly formed vascular networks in an in vivo assay in SCID mice. Overexpression of LINC00607 in CRISPR knockout cells restored normal endothelial function. RNA- and ATAC-Seq after LINC00607 knockout revealed changes in the transcription of endothelial gene sets linked to the endothelial phenotype and in chromatin accessibility around ERG-binding sites. Mechanistically, LINC00607 interacted with the SWI/SNF chromatin remodeling protein BRG1. CRISPR/Cas9-mediated knockout of BRG1 in HUVEC followed by CUT&RUN revealed that BRG1 is required to secure a stable chromatin state, mainly on ERG-binding sites. In conclusion, LINC00607 is an endothelial-enriched lncRNA that maintains ERG target gene transcription by interacting with the chromatin remodeler BRG1 to ultimately mediate angiogenesis. Keywords Long non-coding RNA · BRG1 · Endothelial cell · Gene regulation · ERG · Hypoxia Introduction hypoxia [14, 67]. In response to growth factors and hypoxia, endothelial cells sprout from pre-existing vessels in the pro- Endothelial cells form the selectively permeable monolayer cess of angiogenesis [58]. This process is physiologically between the vessel and the blood. Resting endothelium pro- important and required for wound healing [29]. However, vides an anti-coagulant and anti-inflammatory surface and uncontrolled angiogenesis also contributes to pathological contributes to the control of local vascular tone. It also facili- conditions like macular degeneration and cancer [19]. tates the vascular response to inflammation, shear stress and Recent studies suggest that long non-coding RNAs (lncRNAs) are essential in the regulation of cardiovascular gene programs [2, 57, 65]. LncRNAs are RNA molecules Ralf P. Brandes and Matthias S. Leisegang share senior authorship. longer than 200 nucleotides in length, which may lack apparent protein-coding potential. They have independent * Ralf P. Brandes functions as RNAs, separate from potential peptide coding Brandes@vrc.uni-frankfurt.de abilities [57, 65]. Through different mechanisms lncRNAs * Matthias S. Leisegang impact on gene expression and therefore the cellular Leisegang@vrc.uni-frankfurt.de phenotype [65]. LncRNAs influence many aspects of cellular Extended author information available on the last page of the article Vol.:(0123456789) 1 3 5 Page 2 of 21 Basic Research in Cardiology (2023) 118:5 function among them nuclear architecture, transcription, Propranolol hydrochloride (P0884, Sigma-Aldrich, 100 µM), translation and mRNA stability [71]. TGF-β2 (100-35B, Peprotech, 10  ng/mL), Interleukin Transcriptional control can be exerted through interaction 1β (IL-1β, 200-01B, Peprotech, 1  ng/mL) and RNase A with or recruitment of chromatin remodeling complexes, (EN0531, Thermo Fisher). which subsequently alter the epigenetic landscape [65]. The following antibodies were used: β-Actin (A1978, Chromatin remodeling proteins regulate DNA accessibil- Sigma-Aldrich), BRG1 (ab110641, Abcam), VWF (ab6994, ity by restructuring, mobilizing, and ejecting nucleosomes Abcam), PECAM1 (sc-376764, Santa Cruz), FLT1 (#36110, [11] and thereby altering the binding of transcription factors ThermoFisher) and TGFB2 (sc374658, Santa Cruz). to their DNA targets [42]. One well-known multi-protein chromatin remodeling complex, the Switch/Sucrose Non- Cell culture and stimulation experiments Fermentable (SWI/SNF) complex, has Brahma related gene-1 (BRG1) as one of its core catalytic subunits, whose Pooled human umbilical vein endothelial cells (HUVEC) knockout is embryonic lethal in mice [10, 30]. Several lncR- were purchased from PromoCell (C-12203, Lot No. NAs are known to contribute to the function of BRG1, e.g. 405Z013, 408Z014, 416Z042, Heidelberg, Germany) and EVF2 directly inhibits the ATPase and chromatin remod- cultured at 37 °C with 5% C O in a humidified incubator. eling activity [12], MANTIS, which is now termed SMAN- Gelatin-coated dishes (356009, Corning Incorporated, USA) TIS according to the HUGO Gene Nomenclature Committee, were used to culture the cells. Endothelial growth medium stabilizes the interaction between BRG1 and BAF155 and (EGM), consisting of endothelial basal medium (EBM) recruits BRG1 to angiogenesis related genes [40] and Mhrt supplemented with human recombinant epidermal growth interacts with the helicase domain of BRG1 leading to the factor (EGF), EndoCGS-Heparin (PeloBiotech, Germany), inhibition of chromatin target recognition by BRG1 [23]. 8% fetal calf serum (FCS) (S0113, Biochrom, Germany), Xist binding inhibits BRG1 activity and functionally antag- penicillin (50 U/mL) and streptomycin (50 µg/mL) (15140- onizes the recruitment of associated SWI/SNF complexes 122, Gibco/Lifetechnologies, USA) was used. For each to the inactivated X chromosome [28]. MALAT1 forms a experiment, at least three different batches of HUVEC from complex with BRG1 and HDAC9, which inhibits the expres- passage 3 were used. sion of contractile proteins in aortic aneurysm [47]. These In hypoxia experiments, cells were incubated for 24 h in a examples highlight the fundamental importance of lncRNA- SciTive Workstation (Baker Ruskinn) at 1% O and 5% C O . 2 2 BRG1 interactions. For EndMT (endothelial to mesenchymal transition), In this study, we set out to identify endothelial-enriched HUVEC were stimulated for 5 d in differentiation medium lncRNAs that impact on angiogenic function and may (DM) consisting of endothelial basal medium (EBM) sup- therefore have disease or therapeutic relevance. This led plemented with 8% FCS, penicillin (50 U/mL), streptomycin to the identification of the lncRNA LINC00607, which is (50 µg/mL), l -glutamine, TGF-β2 (10 ng/mL) and IL-1β highly enriched in the endothelium. LINC00607 has been (1 ng/mL). previously described as a super enhancer-derived lncRNA induced by high glucose and TNFα levels [13]. Our study Human endothelial cells from plaque‑free revealed that LINC00607 is induced by hypoxia and sus- and plaque‑containing arteries tains endothelial gene transcription through interaction with the chromatin remodeling protein BRG1. Ultimately, Cells were isolated and cultured as previously described [6]. LINC00607 facilitates proper endothelial ERG-responsive Studies were approved from the scientific and ethic com - gene transcription and the maintenance of the angiogenic mittee of Hipokrateion University Hospital and the Goethe response. University (extension to SC55/22-2-2018). For patient char- acteristics, see Table 1. Cultured cells were stained using RNA-FISH. Materials and methods Experiments with Macaca fascicularis Materials Experiments on adult male Cynomolgus monkeys (Macaca The following chemicals and concentrations were used for fascicularis) were approved by the Institutional Care and stimulation: human recombinant VEGF-A 165 (R&D, 293- Use Committee of the University of Iowa [24] and vessels VE, 30 ng/mL), DMOG (400091, Merck, 1 mM), acrifla - were kindly provided by one of the co-authors (FJM). The vine (A8126, Sigma-Aldrich, 10 µM), Low Density Lipo- vessels originated from a previous study [24], in which protein from Human Plasma, oxidized (oxLDL, L34357, Macaca fascicularis were fed with three different diets, a Thermo Fisher, 10 µg/mL), DMSO (D2650, Sigma-Aldrich), normal diet, an atherosclerotic diet for 47 ± 10 (mean ± SE) 1 3 Basic Research in Cardiology (2023) 118:5 Page 3 of 21 5 Table 1 Clinical data from the human subjects RNA isolation, reverse transcription and RT‑qPCR Characteristics Plaque free subjects Plaque subjects Total RNA isolation was performed with the RNA Mini Kit Demographic data (Bio&Sell) according to the manufacturers protocol and  No 4 4 reverse transcription was performed with the SuperScript  Mean age (range) 68 ± 5 71 ± 4 III Reverse Transcriptase (Thermo Fisher) using a combina-  Male/female 4/0 4/0 tion of oligo(dT)23 and random hexamer primers (Sigma). Clinical data The resulting cDNA was amplified in an AriaMX cycler  Hypertension 0 4 (Agilent) with the ITaq Universal SYBR Green Supermix  Hyperlipidemia 0 4 and ROX as reference dye (Bio-Rad, 1725125). Rela-  Angiographic carotid steno- 0 4 tive expression of human target genes was normalized to sis < 87% β-Actin, or to UBC (for EndMT experiments), whereas for Macaca fascicularis genes GAPDH was used. Expression levels were analyzed by the delta-delta Ct method with the months, or an atherosclerotic diet with an additional recov- AriaMX qPCR software (Agilent). The following oligo- ery phase for 8 months. After isolation of RNA, RT-qPCR nucleotide sequences were used: human LINC00607, for- was performed for the orthologues of human GAPDH and ward 5′-CCA CCA CCA CCA TTA CTT TC-3′ and reverse LINC00607 with Macaca fascicularis (Mf) specific prim- 5′-AGG CTC TGT ATT CCC AAC TG-3′; human β-Actin, ers. The following oligonucleotide sequences were used: forward 5′-AAA GAC CTG TAC GCC AAC AC-3′ and Mf_LINC00607, forward 5′-CTG CAT GTC ACC GCA TAC reverse 5′-GTC ATA CTC CTG CTT GCT GAT-3′; human CC-3′ and reverse 5′-TGG CTC TGC TGG AGT AG-3′; Mf_ Calponin 1 (CNN1), forward 5′-CAT CGG CAA CTT CAT GAPDH, forward 5′-TGC ACC ACC AAC TGC TTA GC-3′ CAA GG-3′ and reverse 5′-CCT GCA GCC CAA TGA and reverse 5′-GGC GTG GAC TGT GGT CAT GAG-3′. TGT TC-3′; human Transgelin (TAGLN), forward 5′-TTC Additionally, tissue sections of cryopreserved carotid artery TGA GCA AGC TGG TGA AC-3′ and reverse 5′-AGT tissue from Macaca fascicularis [24] were subjected to TGG GAT CTC CAC GGT AG-3′; human Collagen, type RNAscope to visualize LINC00607. I, alpha 1 (COL1A1), forward 5′-TGC TGG TGC TCC TGG TAC TC-3′ and reverse 5′-GGG ACC ACG TTC Human brain arteriovenous malformation ACC ACT TG-3′; human Platelet endothelial cell adhe- under propranolol treatment sion molecule-1 (PECAM1), forward 5′-AAC AGT GTT GAC ATG AAG AGC C-3′ and reverse 5′-TGT AAA ACA Patients with arteriovenous malformation (AVM) evaluated GCA CGT CAT CCT T-3′; human Ubiquitin C (UBC), at University Hospital Frankfurt were entered into an ongo- forward 5′-TTG CCT TGA CAT TCT CGA TG-3′ and ing prospective registry. The study protocol was approved reverse 5′-ATC GCT GTG ATC GTC ACT TG-3′; human by the ethical committee of the Goethe University (approval R-spondin-3 (RSPO3), forward 5′-TGT GCA ACA TGC number UCT-63-2020, Frankfurt am Main, Germany). All TCA GAT TAC A-3′ and reverse 5′-TGC TTC ATG CCA patients with proved unruptured AVMs were included after ATT CTT TCC A-3′; human Transforming growth factor written informed consent. Patients with arteriovenous mal- beta-2 (TGFB2), forward 5′-GAG CTA TAT CAG ATT CTC formation (AVM) who underwent microsurgery and had AAG TC-3′ and reverse 5′-GCC ATC AAT ACC TGC AAA tissue available were further analyzed. We selected from TCT TG-3′; human Delta Like Canonical Notch Ligand 4 our tissue bank cases of unruptured brain AVMs in patients (DLL4), forward 5′-CAG CAC TCC CTG GCA ATG TA-3′ who did not undergo pre-surgical embolization. The patients and reverse 5′-CAC AGT AGG TGC CCG TGA AT-3′; did not undergo endovascular embolization before surgi- human Fms Related Receptor Tyrosine Kinase 1 (FLT1), cal resection, and medical records did not show previous forward 5′-AAA TGC CGA CGG AAG GAG AG-3′ and history of rupture. AVM tissue (pieces with a diameter of reverse 5′-AGG TTT CGC AGG AGG TAT GG-3′; human 0.5 cm) was cultured immediately after surgical resection in von-Willebrand-Faktor (VWF), forward 5′-CCT TGA CCT the presence of 100 μM propranolol or solvent DMSO for CGG ACC CTT ATG-3′ and reverse 5′-GAT GCC CGT 72 h. Afterwards, RNA was isolated and RT-qPCR was per- TCA CAC CAC T-3′; human Guanylate Cyclase 1 Soluble formed. Additionally, tissue sections of AVM were subjected Subunit Alpha 1 (GUCY1A1), forward 5′-AGA GCT GGA to RNAscope to visualize LINC00607. TGT CTA CAA GG-3′ and reverse 5′-CGC TAT CTG AAC AGC ATG AG-3′. 1 3 5 Page 4 of 21 Basic Research in Cardiology (2023) 118:5 GAC ACA CTG C-3′; BRG1: gRNA, 5′-CAC CGC ATG Knockdown with siRNAs CTC AGA CCA CCC AG-3′ and 5′-AAA CCT GGG TGG CTC TGA GCA TGC-3′. For LINC00607, gRNA-A was For small interfering RNA (siRNA) treatments, HUVEC (80–90% confluent) were transfected with GeneTrans II cloned into lentiCRISPRv2 with hygromycin resistance, gRNA-B was cloned into lentiCRISPRv2 with puromycin according to the instructions provided by MoBiTec (Göt- tingen, Germany). A Silencer® Select siRNA was used for resistance. For BRG1, lentiCRISPRv2 with puromycin resistance was used. After cloning, the gRNA-containing siRNA-mediated knockdown of LINC00607 (Thermo Fisher Scientific, s56342). As negative control, scrambled Stealth LentiCRISPRv2 vectors were sequenced and purie fi d. Lenti - virus was produced in Lenti-X 293 T cells (Takara, 632180) RNAi™ Med GC (Life technologies) was used. All siRNA experiments were performed for 48 h. using Polyethylenamine (Sigma-Aldrich, 408727), psPAX2 and pVSVG (pMD2.G). pMD2.G was a gift from Didier Protein isolation and western analyses Trono (Addgene plasmid #12259; http:// n2t. ne t/ addge ne: 12259; RRID:Addgene_12259). psPAX2 was a gift from For whole cell lysis, HUVEC were washed in Hanks solution Didier Trono (Addgene plasmid #12260; http:// n2t. ne t/ addg e ne: 12260; RRID:Addgene_12260). LentiCRISPRv2-pro- (Applichem) and lysed with RIPA buffer [1 × TBS, 1% Des- oxycholat, 1% Triton, 0.1% SDS, 2 mM Orthovanadat (OV), duced virus was transduced in HUVEC (p1) with polybrene transfection reagent (MerckMillipore, TR-1003-G) and for 10 nM Okadaic Acid (OA), protein-inhibitor mix (PIM), 40 µg/mL Phenylmethylsulfonyl fluoride (PMSF)]. After LINC00607 knockout selection was performed with puromy- cin (1 μg/mL) and hygromycin (100 µg/mL) for 6 d and for centrifugation (10 min, 16,000×g), protein concentrations of the supernatant were determined with the Bradford assay BRG1 only with puromycin (1 μg/mL) for 6 d. Validation of the CRISPR/Cas9 knockout of LINC00607 and the extract boiled in Laemmli buffer. Equal amounts of protein were separated with SDS-PAGE. Gels were blotted was performed from genomic DNA. Genomic DNA was isolated after selection. Cells were washed, collected and onto nitrocellulose membranes, which were blocked after- wards in Rotiblock (Carl Roth). After application of the first incubated with 500 µL lysis buffer (30 min, 56 °C, 800 rpm, 0.1 M Tris/HCl pH 8.5, 0.5 M NaCl, 0.2% SDS, 0.05 M antibody, an infrared-fluorescent-dye-conjugated secondary antibody (Licor) was used. Signals were detected with an EDTA, 22.2 mg/mL Proteinase K). After removing cell frag- ments (1 min, 13.000 rpm, 4 °C), DNA was precipitated infrared-based laser scanning detection system (Odyssey Classic, Licor). by adding the equal volume 100% isopropanol followed by centrifugation (10 min, 13.000 rpm, 4 °C). The DNA LentiCRISPRv2 was washed with 70% EtOH (10 min, 13.000 rpm, 4 °C), air-dried and dissolved in TE-Buffer (10 mM Tris/HCl pH Guide RNAs (gRNA) targeting LINC00607 were selected 8.0, 1 mM EDTA pH 8.0). CRISPR/Cas9 target sites were amplified by PCR with PCR Mastermix (ThermoFisher, using the publicly available CRISPOR algorithm (http:// crisp or. tefor. net/) [22]. A dual gRNA approach consisting K0171), containing forward and reverse primers (10 µM) and 100–500 ng DNA followed by agarose gel electropho- of gRNA-A and gRNA-B was used to facilitate the knock- out of LINC00607. gRNA-A targeted a region downstream resis and ethidium bromide staining. The following primers were used: LINC00607 CRISPR target site, 5′-CTT CAG of the transcriptional start site (TSS) and gRNA-B targeted a region upstream of the TSS. BRG1 knockout was per- CCC ACT GAG TCT TG-3′ and 5′-GAG GAA CCA GCC AGA ATA GC-3′; GAPDH, 5′-TGG TGT CAG GTT ATG formed using a single gRNA approach. The gRNAs were cloned into lentiCRISPRv2 vector backbone with Esp3I CTG GGC CAG-3′ and 5′-GTG GGA TGG GAG GGT GCT GAA CAC-3′. (Thermo Fisher, FD0454) according to the standard pro- tocol [61]. lentiCRISPRv2 was a gift from Feng Zhang Scratch‑wound migration assay (Addgene plasmid #52961; h tt p :/ / n 2t . n e t / a d dge n e : 5 29 6 1; RRID:Addgene_52961) [61]). The modification of the len- 30,000 HUVEC were seeded on ImageLock 96-well plates tiviral CRISPR/Cas9v2 plasmid with hygromycin resist- ance was provided by Frank Schnütgen (Dept. of Medicine, (Essen Bioscience). Once a monolayer had formed, this was scratched the following day with a 96-pin WoundMaker tool Hematology/Oncology, University Hospital Frankfurt, Goe- the University, Frankfurt, Germany). (Essen Bioscience). EGM was then refreshed to remove dead and scraped cells. Afterwards, the cells were imaged For annealing, the following oligonucleotides were used: LINC00607: gRNA-A, 5′-CAC CGC ATG TGC CCC CTT in an Incucyte imaging system for 11 h (one image every one hour, with the “phase” image channel and 10× magni- TGT TGA A-3′ and 5′-AAA CTT CAA CAA AGG GGG CAC ATG C-3′, gRNA-B, 5′-CAC CGC AGT GTG TCA fication). The Scratch Wound Cell Migration Module of the TGT TAT CTT G-3′ and 5′-AAA CCA AGA TAA CAT 1 3 Basic Research in Cardiology (2023) 118:5 Page 5 of 21 5 Incucyte S3 Live Cell Analysis System (Essen Bioscience) the cells on ice and irradiated with 0.150 J/cm 254 nm UV was used to monitor and analyze the cells. light. Cells were scraped twice in 500 µL Hanks buffer and centrifuged at 1000×g at 4 °C for 4 min. For nuclear pro- Spheroid outgrowth assay tein isolation, cells were resuspended in hypotonic buffer (10 mM HEPES pH 7.6, 10 mM KCl, 0.1 mM EDTA pH Spheroid outgrowth assays in HUVEC were performed as 8.0, 0.1 mM EGTA pH 8.0, 1 mM DTT, 40 µg/mL PMSF) described in [33]. Stimulation of spheroid outgrowth was and incubated on ice for 15 min. Nonidet was added to a performed with VEGF-A 165 (R&D, 293-VE, 30 ng/mL) for final concentration of 0.75% and cells were centrifuged 16 h. Spheroids were imaged with an Axiovert135 micro- (1 min, 4 °C, 16,000×g). The nuclear-pellet was washed scope (Zeiss). The cumulative sprout length and spheroid twice in hypotonic buffer, lysed in high salt buffer (20 mM diameter were quantified by analysis with the AxioVision HEPES pH 7.6, 400 mM NaCl, 1 mM EDTA pH 8.0, 1 mM software (Zeiss). EGTA pH 8.0, 1 mM DTT, 40 µg/mL PMSF) and centri- fuged (5 min, 4 °C, 16,000×g). 10% of the nuclear lysate Plasmid overexpression and Spheroid outgrowth was taken as input. 4 µg of antibody was pre-coupled to 30 assay µL protein G magnetic beads in bead wash buffer (20 mM HEPES pH 7.6, 200 mM NaCl, 1 mM EDTA pH 8.0, 1 mM Plasmid overexpression was performed using 700,000 EGTA pH 8.0, 1  mM DTT, 40  µg/mL PMSF) for 1  h at HUVEC and the Neon electroporation system (Invitrogen, RT, then washed once with high salt buffer and twice with 1400 V, 1 × 30 ms pulse) in E2 buffer for the following plas- bead wash buffer. The antibody-coupled beads were added mids (7 µg per transfection): pcDNA3.1 + LINC00607 and to the nuclear lysate and rotated for 1 h at 4 °C. Samples pcDNA3.1 + . Overexpression was performed 16 h prior to were placed on a magnetic bar and the lysate discarded. The conducting the spheroid outgrowth assay. beads were washed three times in high salt buffer (50 mM Tris–HCl, 1 M NaCl, 1 mM EDTA, 0.1% SDS, 0.5% Sodium Proliferation assay Deoxycholate, 1% NP-40, 1 mM DTT, 40 µg/mL PMSF) at 4 °C for 10 min per wash. Beads were then washed twice 5000 HUVEC (CTL or siLINC00607) were seeded on in bead wash buffer 2 (20  mM TrisHCl, 10  mM MgCl , 96-well plates, 16 h after siRNA transfection. Nuclei were 0.2% Tween, 1 mM DTT, 40 µg/mL PMSF). For RNase A stained with Incucyte® Nuclight Rapid Red Dye (Sartorius, treatment, beads were placed in a buffer containing 20 mM 4717) (1:2000 in EGM). The cells were imaged every 3 h in Tris–HCl, EDTA pH 8.0 and 2 µL of RNase A (10 mg/mL) an Incucyte imaging system, using “phase” and “red” image for 30 min at 37 °C and then washed again in bead wash channels and 10X magnification. The Proliferation Module buffer. For elution of RNA, the remaining wash buffer was of the Incucyte S3 Live Cell Analysis System (Essen Biosci- removed and 1 mL QIAzol (Qiagen) was added to the beads ence) was used to monitor and analyze the cells. and incubated at RT for 10 min. 400 µL chloroform was added to the samples and vortexed for 10 s followed by incu- Apoptosis assay bation for a further 10 min at RT. Samples were then centri- fuged at 12,000×g for 15 min at 4 °C. 500 µL of the upper 7500 HUVEC (CTL or siLINC00607) were seeded on aqueous phase was transferred to a new tube and 2 µL gly- 96-well plates, 16 h after siRNA transfection. Nuclei were cogen (GlycoBlue Coprecipitant, ThermoFisher, AM9515) stained with Incucyte® Annexin V Green Dye (Sartorius, and 500 µL isopropanol added. Samples were inverted mul- 4642) (1:500 in EGM). The cells were imaged every 3 h tiple times and incubated at RT for 10 min before being in an Incucyte imaging system, using “phase” and “green” centrifuged again at 12,000×g for 10 min. The supernatant image channels and 10X magnification. The Apoptosis Mod- was removed and the pellet washed with 1 mL 75% etha- ule of the Incucyte S3 Live Cell Analysis System (Essen nol by vortexing. The pellet was centrifuged at 7500×g for Bioscience) was used to monitor and analyze the cells. 5 min at 4 °C, dried and resuspended in 30 µL nuclease-free water. RNA samples were reverse transcribed for qPCR as RNA immunoprecipitation (RIP) described above. To identify RNAs bound to a protein of interest, specific In vivo matrigel plug assay antibodies and Pierce™ Protein G Magnetic Beads (88847, Thermo Fisher) were used to immunoprecipitate RNAs 150,000 HUVEC per plug were stained with Vybrant Dil bound to the target protein. Cells were grown to 80% conflu- (1:200 in 1 mL Basal Medium (EBM); Thermo Fisher, ence on a 10 cm plate (roughly 3 million cells) and washed V-22885). After incubation (45  min at 37  °C, 5  min at once with Hanks buffer. 6 mL Hanks buffer was added to 4 °C), cells were washed with EBM (Lonza), resuspended 1 3 5 Page 6 of 21 Basic Research in Cardiology (2023) 118:5 in EGM containing 20% methocel (Sigma-Aldrich) and version 1.26.0 [49]. Only genes with a minimum fold change cultured in hanging drops (25 µL/drop). Harvesting of of +/− 1.5 (log2 +/− 0.59), a maximum Benjamini–Hoch- spheroids and injection of matrigel containing sphe- berg corrected p-value of 0.05, and a minimum combined roids into SCID mice (Charles River Laboratories) was mean of 5 reads were deemed to be significantly differen- performed as described previously [36]. 21 d after injec- tially expressed. The Ensemble annotation was enriched tion, Isolectin GS-IB4 from Griffonia simplicifolia, Alexa with UniProt data (release 06.06.2014) based on Ensembl Fluor® 647 Conjugate (I32450, Thermo Fisher) was gene identifiers (Activities at the Universal Protein Resource administered intravenously and was allowed to circulate (UniProt) [1]). for 20 min. After transcardial perfusion of the animals, the plugs were dissected, cleaned, fixed in 4% Paraform- Assay for transposase‑accessible chromatin using aldehyde (PFA) and subsequently cleared following the sequencing (ATAC‑Seq) 3DISCO procedure [16]. Imaging was carried out with the Ultramicroscope II (UM-II, LaVision Biotec, Bielefeld) at 50,000 HUVEC were used for ATAC library preparation 16× magnification (10 Zoom body + 2 × Objective). Pic- using lllumina Tagment DNA Enzyme and Buffer Kit (Illu - tures were taken with a Neo 5.5 (3-tap) sCOMs Camera mina). The cell pellet was resuspended in 50 µL of the lysis/ (Andor, Mod. No.: DC-152q-C00-FI). The ImSpectorPro transposition reaction mix (25 µL TD-Buffer, 2.5 µL Nextera Version_3.1.8 was used. Quantification of 3D Images Tn5 Transposase, 0.5 µL 10% NP-40 and 32 µL H2O) and was performed with Imaris (Bitplane Version 9.6). The incubated at 37 °C for 30 min followed by immediate puri- surface function was used to manually delete auto fluo- fication of DNA fragments with the MinElute PCR Purifi- rescence signals and artefacts. Signal background was cation Kit (Qiagen). Amplification of Library and Indexing removed using baseline subtraction. Cells were detected was performed as described elsewhere [9]. Libraries were and counted with the Spots-Algorithm (estimated diam- mixed in equimolar ratios and sequenced on NextSeq500 eter = 10.0  μm; background subtraction = tr ue; “intensity platform using V2 chemistry. Trimmomatic version 0.39 center Ch = 3" above 395; Region Growing Type = Local was employed to trim raw reads after a quality drop below Contrast). Lower threshold was chosen depending to the a mean of Q20 in a window of 5 nt [7]. Only reads above background signal. Cells were considered incorporated in 15 nt were cleared for further analyses. These were mapped the vascular network with the threshold of the “intensity versus the hg38 version (emsambl release 101) of the human Max. channel = 2” above 575. genome with STAR 2.7.7a [15] using only unique align- ments to exclude reads with uncertain arrangement. Reads were further deduplicated using Picard 2.21.7 [8] to avoid RNA‑Seq PCR artefacts leading to multiple copies of the same origi- nal fragment. The Macs2 peak caller version 2.1.1 was 900  ng of total RNA was used as input for SMARTer employed to accommodate for the range of peak widths typi- Stranded Total RNA Sample Prep Kit—HI Mammalian cally expected for ATAC-Seq [73]. Minimum qvalue was set (Takara Bio). Sequencing was performed on the Next- to -4 and FDR was changed to 0.0001. Peaks overlapping Seq500 instrument (Illumina) using v2 chemistry, resulting ENCODE blacklisted regions (known misassemblies, satel- in average of 38 M reads per library with 1 × 75 bp single lite repeats) were excluded. In order to be able to compare end setup. The resulting raw reads were assessed for qual- peaks in different samples, the resulting lists of significant ity, adapter content and duplication rates with FastQC [4]. peaks were overlapped and unified to represent identical Trimmomatic version 0.39 was employed to trim reads after regions. The counts per unified peak per sample were com- a quality drop below a mean of Q20 in a window of 10 puted with BigWigAverageOverBed [32]. Raw counts for nucleotides [7]. Only reads between 30 and 150 nucleotides unified peaks were submitted to DESeq2 (version 1.20.0) were cleared for further analyses. Trimmed and filtered reads for normalization [49]. Peaks were annotated with the pro- were aligned versus the Ensembl human genome version moter of the nearest gene in range (TSS +/− 5000 nt) based hg38 (release 99) using STAR 2.7.3a with the parameter on reference data of GENCODE vM15. Peaks were deemed “–outFilterMismatchNoverLmax 0.1” to increase the maxi- to have significantly different counts between conditions at mum ratio of mismatches to mapped length to 10% [15]. The an average score of 20, and a log2 transformed fold change number of reads aligning to genes was counted with fea- of < − 0.59 or > 0.59. tureCounts 1.6.5 tool from the Subread package [46]. Only reads mapping at least partially inside exons were admit- RNA fluorescence in‑situ hybridization ted and aggregated per gene. Reads overlapping multiple genes or aligning to multiple regions were excluded. Dif- RNA-FISH was performed similar as described before ferentially expressed genes were identified using DESeq2 [56]. Briefly, cells grown on gelatin-coated 8-well µ-Slides 1 3 Basic Research in Cardiology (2023) 118:5 Page 7 of 21 5 (ibidi) were fixed in 4% paraformaldehyde (PFA) (in PBS, for 20 min using the Braun steamer. Treatment with protease 10 min, at RT) and washed 3 times with PBS. Cells were Plus for 30 min at 40 °C was performed using the HybEZ permeabilized in 0.5% Triton X-100 (in PBS, 5 mM vana- System. The probes (human LINC00607 (ACD #894351), dyl complex (VRC, NEB)) on ice for 10 min and washed Macaca fascicularis LINC00607 (ACD #1217651-C1)) and 3 times with PBS. Prior to hybridization, cells were rinsed amplification steps were carried out according to instructions once in 2xSSC. Hybridization was performed over night except that Amp5 was incubated for 45 min. Signal detection at 37  °C in hybridization buffer (10% dextran sulfate, was carried out by incubation with DAB for 20 min at RT. 50% formamide, 2xSSC, 400  µg E.coli tRNA, 0.02% Sections were mounted in EcoMount after dehydration and RNase-free bovine serum albumin, 2 mmol/L VRC) and analyzed by light microscopy. 10 nmol/L 5’TYE-665 labelled locked nucleic acid (LNA) detection probe (Qiagen). Custom LNA detection probes BRG1 CUT&RUN targeting LINC00607 were designed with the Qiagen GeneGlobe Custom LNA design tool and had the follow- BRG1 Cleavage Under Targets & Release Using Nuclease ing sequences: 5′-AGG AGC TGA GAT GCA CAT ACT- (CUT&RUN), a method established by Skene and Henikoff 3′. The cells were washed 4 times for 15  min in buffer in 2017 [64], was performed similarly as described in the containing 2xSSC and 50% formamide and were counter- EpiCypher CUT&RUN Protocol v2.0, but with minor modi- stained with DAPI (in PBS). Images were captured with a fications for the cell type and antibody used. Briefly, 500,000 laser confocal microscope LSM800 (Zeiss, Germany) and NTC or BRG1 knockout HUVEC were washed with wash analyzed with ZEN lite software (Zeiss, Germany). Fluo- buffer (20 mM HEPES pH 7.9, 150 mM NaCl, 500 nM sper - rescence intensities were analyzed with ImageJ software midine, 1X Roche Protein Inhibitor Cocktail) at RT. Cells (ImageJ, NIH). were resuspended in wash buffer and 10 µL BioMag®Plus Concanavalin A (ConA) beads (Polysciences, 86057-3) were added for 10 min at RT. Beads were separated on a magnetic Immunofluorescence rack and washed once before being resuspended in 100 µL antibody buffer (wash buffer, 0.25% Digitonin and 2 mM Cells were seeded on 8-well immunofluorescence plates EDTA) and 1 µL BRG1 antibody (Abcam, ab110641). Beads (Ibidi, Germany). After washing with PBS, the cells were were incubated with the antibody over night with gentle fixed with 4% PFA and permeabilized with 0.05% Triton shaking at 4 °C. The next day, beads were washed twice X-100. Cells were blocked in 3% BSA (bovine serum albu- with 200 µL 0.25% Digitonin wash buffer and resuspended min) for 30 min, followed by incubation at 4 °C overnight in Digitonin wash buffer containing 2 µL CUTANA™ pAG- with the primary antibody (1:200 dilution). After washing MNase (15–1016, EpiCypher, 15-1016) and incubated on with 0.3% Tween20 in PBS, the cells were incubated with ice for 30 min. Samples were washed twice and then resus- the secondary antibody (1:500 dilution; Alexa Fluor 647, pended in 100 µL Digitonin wash buffer containing 2 µL #A31573, Invitrogen, USA) for 30 min. The cells counter- CaCl at a final concentration of 100 mM and incubated stained with 4′,6-diamidino-2-phenylindole (DAPI). Images for 2 h at 4 °C with gentle shaking. 33 µL of 2X “stop solu- were captured with a laser confocal microscope LSM800 tion” (340 mM NaCl, 20 mM EDTA, 4 mM EGTA, 0.25% (Zeiss, Germany) and analyzed with ZEN lite software Digitonin, 100 µg/mL RNase A, 50 µg/mL Glycoblue) was (Zeiss, Germany). added to the beads and incubated at 37 °C for 10 min. Sam- ples were placed on a magnetic rack and the supernatant removed and kept for DNA purification. Briefly, 5X vol- RNAscope ume of binding buffer (20 mM HEPES pH 7.9, 20 mM KCl, 1 mM CaCl , 1 mM MnCl ) was added to the samples and 2 2 RNA in  situ hybridization was performed with the the pH adjusted with sodium acetate before being transferred RNAscope 2.5 HD Detection Reagents (322310), Advanced to a purification column (ActiveMotif, 58,002) and centri- Cell Diagnostics (ACD), Newark, CA, USA) similarly as fuged at 11,000×g for 30 s. The column was then washed described before [41]. The single-plex chromogenic brown with 750 µL wash buffer and dried by centrifugation for assay was used according to manufacturer’s instructions with 2 min. DNA was eluted with 25 µL elution buffer and the minor changes. OCT was removed from unfixed, cryopre- DNA concentration measured with a Qubit 3.0 Fluorometer served tissue-sections in PBS for 5 min, baked for 30 min at (Life Technologies). 60 °C and post-fixed in 4% PFA. Paraffin-embedded, forma- lin-fixed sections were deparaffinized prior to staining. Sec- tions were treated with hydrogen peroxide for 10 min. Target retrieval was performed with boiling in 1xRetrieval solution 1 3 5 Page 8 of 21 Basic Research in Cardiology (2023) 118:5 FANTOM5 CAGE and ENCODE expression data was Library preparation and sequencing of CUT&RUN samples obtained from the FANTOM5 website and was published elsewhere [21, 48, 55]. DNA libraries were prepared according to the manufac- Publicly available HUVEC ERG ChIP-sequencing data (GSE124891) was downloaded from the Gene Expression turer’s protocol (NEBNext® Ultra II, NEB) with some minor adjustments for CUT&RUN samples. Briefly, sam- Omnibus (GEO) [31]. ples were brought to 50 µL with 0.1X TE buffer and DNA end preparation performed as instructed but with incuba- ERG ChIP‑seq data analysis tion at 20 °C for 20 min and then 58 °C for 45 min. Adap- tor ligation was performed with a 1:10 dilution of adap- FASTQ files were trimmed with Trim Galore! [17] and tor (NEB, E6440S). For DNA purification, 0.9× Volume AMPure XP beads (Beckman Coulter, A63881) was added aligned to the Ensembl human genome version hg38 (ensembl release 104) using Bowtie2 [37, 38]. Duplicate to the samples and incubated for 5 min at RT. Beads were washed twice with 200 µL 80% ethanol and DNA eluted reads were removed with rmdup [43]. Peaks were called on the aligned data using MACS2 [18] and annotatePeaks with 17 µL 0.1X TE buffer for 2 min at RT. PCR amplifi- cation of the eluted DNA was performed as described in (HOMER) [25] used to identify the nearest genes to called peaks. the manufacturer’s protocol but with the addition of 2.5 µL Evagreen (20X) for visualization of the amplification Use of FANTOM5 CAGE ENCODE data for promoter curves on an AriaMx Real-time PCR system (Agilent). The denaturation and annealing/extension steps of the and expression analysis of LINC00607 PCR amplification were performed for around 12 cycles and stopped before the curves plateaued. A cleanup of The promoter of LINC00607 was defined as nucleotide sequence with a length of 1000 nt, starting from a promi- the PCR reaction was performed twice with 1.1× Ampure beads and eluted each time in 33 µL 0.1× TE buffer. DNA nent FANTOM5 CAGE region having multiple peaks in close vicinity (approx. 30 nt) going in upstream direction concentrations were measured with a Qubit (Thermo Fisher) and size distributions measured on a Bioanalyzer for 970 nt (hg38 chr2:215,848,858–215,849,857). Promoter analysis was performed with filters for the indicated tran- (Agilent). Sequencing was performed on the NextSeq1000/2000 scription factors with the MoLoTool (https://molo tool. aut os ome.or g/), an interactive web application suitable to identify (Illumina). The resulting raw reads were assessed for qual- ity, adapter content and duplication rates with FastQC DNA sequences for transcription factor binding sites (TFBS) with position weight matrices from the HOCOMOCO data- [4]. Trim Galore! [17] was used to trim reads before alignment to the Ensembl human genome version hg38 base [35]. To compare the individual lncRNA expression towards (ensembl release 104) using Bowtie2 [37, 38]. Duplicate reads were removed with rmdup [43] and coverage tracks all other cell types or tissues, each cell type-specific signal obtained with FANTOM5 CAGE (or ENCODE) [21, 48, generated with bamCoverage (deepTools Version 3.5.1) [59]. ComputeMatrix [59] and plotHeatmap (deepTools 55] was divided through the mean signal observed in all cell types or tissues and plotted. Version 3.5.1) were used on the coverage tracks to gener- ate heatmaps of BRG1 binding across the genome. Peaks Gene‑set enrichment analysis were called on the aligned data using MACS2 [18] and annotatePeaks (HOMER) [25] was used to identify the GSEA (Gene-Set Enrichment Analysis) [54, 66] was per- nearest genes to called peaks. formed based on the RNA-Seq data to identify gene sets that were significantly enriched from genes differently expressed Publicly available datasets between the NTC control and LINC00607 knockout. 1000 permutations were performed and gene sets were considered The following RNA-Seq datasets used in this study origi- statistically enriched with a nominal P < 0.05. nated from NCBI GEO: HUVEC treated with normoxia or hypoxia (GSE70330) [20], ACF treatments of HUVEC Differential ATAC‑sequencing analysis and intersection with gene‑linked regulatory under normoxia (GSE176555) or hypoxia (GSE186297)[63], EndMT treatments of HUVEC and PAEC (GSE118446) elements [53]. Alignment files arising from ATAC-sequencing data analysis detailed above were subjected to replicate-based differential 1 3 Basic Research in Cardiology (2023) 118:5 Page 9 of 21 5 peak calling using THOR (v0.13.1) [3], which employs a Expression)-ENCODE database revealed that LINC00607 hidden Markov model-based approach to identify differen- is one of the most endothelial-enriched lncRNAs (Fig. 1A). tially accessible regions of chromatin between conditions. Particularly high levels of LINC00607 were observed in aor- Differential peaks were those with reported adjusted p-val- tic, venous, lymphatic, thoracic and arterial ECs (Fig. 1B). ues less than 0.01. Differential peaks were subsequently Additionally, FANTOM5 CAGE-ENCODE cell-type expres- intersected with regulatory elements from EpiRegio (v1.0.0) sion data showed LINC00607 to be predominantly localized [5], a collection of regulatory elements and their associated in the nucleus (Fig.  1A), which was confirmed by RNA- genes. Genes whose expression is dependent on differen- fluorescence in situ hybridization (RNA-FISH) in HUVEC tially accessible regulatory elements were subjected to path- (Fig. 1C). RT-qPCR after reverse transcription with random way enrichment analysis using the ReactomePA (v1.36.0) or oligodT oligonucleotides revealed that LINC00607 has a [72] package for R. Subsequently, differential accessibility of poly-A tail (Fig. S1A). regulatory elements linked to genes differentially expressed Importantly, LINC00607 expression was altered in in RNA-Seq could be quantified for different gene sets, and various cardiovascular diseases. Endothelial cells isolated displayed graphically with ggplot2 (v3.3.5) [70]. Motif and cultured from plaque-containing arteries, regarded enrichment analysis of differential ATAC-sequencing peaks as post-atherosclerotic, showed significant upregulation was performed using HOMER (v4.11.1) [25] by providing of LINC00607 in atherosclerotic-derived ECs compared sequences underlying the peaks, and otherwise the default to plaque-free-derived ECs (Fig.  1D, E). LINC00607 parameters. expression was also increased in response to propranolol treatment of human arteriovenous malformation explants Data availability (Fig. 1F, S1B). Furthermore, the corresponding orthologue of LINC00607 (Fig. S1C) was strongly induced in Macaca The RNA-Seq and ATAC-Seq datasets have been deposited fascicularis samples undergoing atherosclerosis regression and are available at NCBI GEO with the accession number after a high fat diet (Fig. 1G, H). GSE199878: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. We next searched for potential gene regulatory mecha- cgi? acc= GSE19 9878 nisms responsible for controlling LINC00607 expression. BRG1 CUT&RUN datasets have been deposited and An analysis of the promoter region, defined here as the are available at NCBI GEO with the accession number FANTOM5 CAGE transcription start site signal to 1000 GSE201824: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. nucleotides (nt) upstream, revealed binding motifs for mul- cgi? acc= GSE20 1824 tiple transcription factors. In particular, ARNT (also known as Hypoxia Inducible Factor 1 Beta) and HIF1A (Hypoxia Statistics Inducible Factor 1 Alpha) were identified multiple times and in close proximity to the transcription start site, indicative Unless otherwise indicated, data are given as means ± stand- of transcriptional regulation by hypoxia (Fig. 1I). Indeed, ard deviation (SD). Calculations were performed with Prism LINC00607 expression was significantly increased when 8.0 or BiAS.10.12. The latter was also used to test for normal HUVEC were cultured under hypoxic conditions (1% oxy- distribution (Shapiro–Wilk) and similarity of variance. In gen) (Fig. 1J). A publicly available RNA-Seq dataset con- case of multiple testing, Bonferroni correction was applied. taining hypoxia-stimulated HUVEC [20] confirmed this For multiple group comparisons ANOVA followed by post finding (Fig.  1K); in fact, LINC00607 was among the top hoc testing was performed. Individual statistics of depend- upregulated lncRNAs in this dataset (Fig. 1L). Interestingly, ent samples were performed by unpaired t test, of unpaired stimulation of HUVEC with oxLDL and DMOG, the latter samples by unpaired t-test and if not normally distributed of which is known to stabilize HIF1α under both hypoxic by Mann–Whitney test. P values of < 0.05 was considered and normoxic conditions [27], increased LINC00607 expres- as significant. Unless otherwise indicated, n indicates the sion (Fig. 1M, S1D). Conversely, the DNA topoisomerase number of individual experiments. and HIF-inhibitor acriflavine (ACF) [63] led to a decrease in LINC00607 expression, which was exacerbated under hypoxia (Fig. 1N–P). Results In addition to HIF binding sites, the promoter analysis of LINC00607 yielded SMAD binding motifs. To test their LINC00607 is a highly endothelial‑enriched lncRNA relevance for LINC00607 expression, HUVEC were stimu- induced by hypoxia lated with TGF-ß2 and IL-1ß to induce endothelial to mes- enchymal transition (EndMT), a process in which SMADs A screen for the top-expressed endothelial lncR- play a central role [34]. Indeed, EndMT strongly increased NAs in the FANTOM5 CAGE (Cap Analysis of Gene the expression of LINC00607 (Fig. 1Q, S1E, F) and similar 1 3 5 Page 10 of 21 Basic Research in Cardiology (2023) 118:5 findings could be retrieved from publicly available RNA-Seq LINC00607 promotes sprouting, proliferation datasets [52] of HUVEC (Fig. 1R) and pulmonary arterial and vascularization endothelial cells (PAEC) (Fig. 1S). These data indicate that LINC00607 is an endothelial- In order to study the functional relevance of LINC00607 enriched lncRNA induced by transcription factors that are in endothelial cells, spheroid outgrowth assays were per- central in hypoxic and EndMT signalling. formed. In this assay, knockdown of LINC00607 with siRNA (Fig.  2A) suppressed sprouting in response to VEGF-A (Fig.  2B–D). Next, a LINC00607 knockout in HUVEC was achieved by CRISPR/Cas9-mediated 1 3 Basic Research in Cardiology (2023) 118:5 Page 11 of 21 5 ◂Fig. 1 LINC00607 is an EC-enriched lncRNA upregulated dur- normal angiogenic response to VEGF-A (Fig.  2M, N, ing hypoxia and EndMT. A FANTOM5 CAGE-ENCODE expres- S2D, E). This suggests that the RNA itself mediates the sion of the 9 highest endothelial expressed lncRNAs across dif- observed functional effects by acting in trans. ferent cell lines. Each cell type-specific signal was divided through Collectively, these data demonstrate that loss of the mean signal observed in all cell types. B FANTOM5 CAGE expression of the 9 highest expressed endothelial lncRNAs across LINC00607 limits endothelial angiogenic capacity. As different endothelial tissues. Calculation was performed as in A. C LINC00607 is not conserved in mice, its physiologi- RNA-FISH of LINC00607 in HUVEC. LINC00607 is labelled with cal importance was studied by assessing the capacity a 5’TYE-665 probe, DAPI is used to stain the nuclei. Scale bar indi- of HUVEC to integrate into the vascular network of cates 20 µm. D, E RNA-FISH of LINC00607 in cultured endothelial cells retrieved from human plaque-free (healthy, free) and plaque- matrigels when injected in SCID-mice. Importantly, in containing (atherosclerosis, cont.) arteries. Scale bar indicates this in vivo assay, knockout of LINC00607 significantly 20  µm. Fluorescence intensity per cell for TYE-665 (LINC00607) is decreased the capacity of HUVEC to be integrated into shown in E. n = 26–30, unpaired t test. F RT-qPCR of LINC00607 in the murine vascular network (Fig.  2O, P). These data human arteriovenous malformations (AVM) treated with and with- out the β-blocker propranolol for 72  h. n = 5, Mann–Whitney U test. demonstrate that LINC00607 acts in trans as a pro-angi- G RT-qPCR of the LINC00607 homologue in vessels originating ogenic lncRNA. from Macaca fascicularis treated either with a normal diet (CTL), a high fat diet (Ath) or with a high-fat diet and a subsequent recov- LINC00607 maintains transcription of genes ery phase (Reg). n = 3. One-way ANOVA with Bonferroni post hoc test. H RNA in  situ hybridization of carotid arteries  from Macaca involved in VEGF‑signalling fascicularis (Mf) with RNAscope. Arrows point to dots indicating LINC00607. Representative images are shown. Scale bar indicates To identify how LINC00607 impacts on angiogenic func- 10  µm. I Promoter analysis of LINC00607. A region starting from tion, gene expression was determined by RNA-Seq with and the LINC00607 transcriptional start site (TSS) and 1000 base pairs (bp) upstream was analyzed with MoLoTool and plotted according to without LentiCRISPR-mediated knockout of LINC00607 p value. J Relative LINC00607 expression in HUVEC treated with in HUVEC. Deletion of LINC00607 markedly impacted normoxia (NOX) or hypoxia (HOX), n = 7. Mann–Whitney t test. K endothelial gene expression (Fig.  3A–C, S3A–F), with LINC00607 gene read counts in HUVEC cultured under normoxic a greater tendency to decrease rather than increase the and hypoxic conditions, n = 3. Unpaired t test. L Volcano plot of log2 fold changes of lncRNAs expressed in hypoxia versus normoxia. M expression of protein-coding and non-coding RNAs (Fig. Relative expression of LINC00607 in HUVEC after stimulation with S3G, H). Due to the observed angiogenic defects, a Gene Set DMOG. DMSO served as control (CTL), n = 8, Mann–Whitney U Enrichment Analysis (GSEA) was performed for the VEGF- Test. N Relative expression of LINC00607 in HUVEC after stimula- signaling pathway. GSEA revealed a strong association of tion with acriflavine (ACF), n = 7, Mann–Whitney U Test. O, P  Vol- cano plot of log2 fold changes of lncRNAs in HUVEC treated with differentially expressed genes within the VEGF-signaling acriflavine (ACF) cultured under normoxia (O) or hypoxia (P). Q pathway after CRISPR/Cas9-mediated knockout of the Relative expression of LINC00607 in HUVEC under basal (CTL) or lncRNA (Fig. 3D). This GSEA result was associated with Endothelial-to-mesenchymal transition (EndMT) conditions. n = 3, numerous VEGF-signaling genes that were mainly down- Unpaired t test. R, S Volcano plot of log2 fold changes of lncRNAs after EndMT versus unstimulated control in HUVEC (R) or pulmo- regulated upon knockout of LINC00607 (Fig. 3E–G). nary arterial endothelial cells (PAEC) (S). Error bars are defined as mean ± SD. *p < 0.05 LINC00607 depletion reduces the accessibility of ETS transcription factor binding sites removal of the transcriptional start site of LINC00607 (Fig. S2A). Successful knockout was confirmed on the In order to determine whether the effects of LINC00607 levels of both the DNA (Fig.  2E) and RNA (Fig.  2F, loss of function and differential gene expression were a G). As with siRNA-mediated knockdown, knockout consequence of altered chromatin accessibility, an assay of LINC00607 inhibited VEGF-A-induced sprouting for transposase-accessible chromatin with sequencing (Fig.  2H–J). As additional functional assays, scratch (ATAC-Seq) was performed (Fig. S3I). Comparison of wound, proliferation and apoptosis assays were performed ATAC-Seq and RNA-Seq for the multiple VEGF-signaling to determine migratory capacity and proliferation. These genes revealed a similar effect of LINC00607 knockout on assays showed the negative effect on endothelial func- chromatin accessibility of gene-linked enhancers (as anno- tion as a consequence of LINC00607 loss (Fig.  2K, L, tated by EpiRegio [5]) and gene expression (Fig. 4A). This S2B, C). In order to study the mechanistic function of suggested that the lncRNA might directly influence the LINC00607, the RNA was overexpressed in knockout transcription of these genes by modulating the accessibil- cells. In the case of cis-action, i.e. local action of the ity of transcription factor binding sites. To investigate the RNA at the transcription site or a general transcriptional underlying mechanism of the profound changes in chro- importance of the gene locus, such a rescue experiment matin state and transcription, a transcription factor bind- should not restore function. However, transfection of ing analysis was performed using HOMER [25]. DNA- LINC00607 into LINC00607 knockout cells restored a motif enrichment analysis showed the basic region/leucine 1 3 5 Page 12 of 21 Basic Research in Cardiology (2023) 118:5 zipper motif (bZIP) to be more accessible under LINC00607 mediating LINC00607-dependent transcription (Fig. 4D). knockout (Fig. 4B). Interestingly, ERG (ETS Transcription Even though some of the ETS family members were dif- Factor ERG) and ETV2 (ETS Variant Transcription Fac- ferentially expressed in response to LINC00607 knockout, tor 2) motifs were identified as being less accessible after the expression of ERG remained unchanged (Fig. 4E). These LINC00607 knockout (Fig. 4C). ERG and ETV2 are both data indicate that LINC00607-dependent gene expression members of the ETS transcription factor family, recognizing is likely mediated through changes in ERG-induced gene the core consensus motif GGA(A/T) [69], and are highly expression, resulting from LINC00607-directed changes in important for endothelial gene expression in particular [51]. transcription factor binding site accessibility. Expression changes of a transcription factor might impact on the gene expression of its target gene. To exclude that LINC00607 maintains endothelial‑specific the differential gene expression in response to LINC00607 chromatin states through interaction with BRG1 loss of function was not caused through differential expres- sion of the transcription factors themselves, the expression The changes in chromatin accessibility and to ERG-binding of ETS family transcription factors was determined. As sites would naturally be caused by chromatin remodeling. determined from RNA-Seq, ERG was highly expressed in We have previously shown that an important chromatin normal HUVEC, whereas ETV2 expression was low. We remodeling protein interacting with lncRNAs in endothe- therefore selected ERG as a candidate transcription factor lial cells is the SWI/SNF member BRG1 [40]. Importantly, 1 3 Basic Research in Cardiology (2023) 118:5 Page 13 of 21 5 ◂Fig. 2 CRISPR/Cas9 KO and siRNA-knockdown reveal that and BRG1 knockout in HUVEC. BRG1 binding sites were LINC00607 is important for normal EC function. A RT-qPCR of located near the transcription start sites of many genes and, LINC00607 after siRNA-based knockdown for 48  h of LINC00607 upon BRG1 knockout, these sites were lost confirming the (607). Scrambled siRNA served as negative control (CTL). n = 6. specificity of BRG1 binding (Fig.  5D). To reveal the role Mann–Whitney test. B Spheroid outgrowth assay after siRNA-based knockdown of LINC00607 (607). Scrambled siRNA served as nega- of LINC00607 for BRG1 binding, differentially expressed tive control (CTL). Cells treated with or without VEGF-A (16 h) are genes identified by RNA-Seq were overlapped with differ - shown. C Quantification of cumulative sprout length from spheroid ential ATAC-Seq peaks having proximity to the transcrip- outgrowth assay shown in B, n = 28–30, two-way ANOVA. D Quan- tional start site and with genes BRG1 binding sites were tification of the ratio of cumulative outgrowth length and respective spheroid diameter from the spheroid outgrowth assay shown in B; identified by CUT&RUN. Surprisingly, there was a strong n = 28–30, two-way ANOVA. E PCR of Genomic DNA after lenti- overlap between LINC00607 differentially regulated genes viral CRISPR/Cas9-mediated knockout (KO) of LINC00607. Three and BRG1 target genes (Fig. 5E). BRG1-associated genes different batches (1–3) of HUVEC are shown. Non-targeting control exhibited a stronger and more significant decrease in expres- gRNAs (NTC) served as negative control. GAPDH served as house- keeping gene. F RT-qPCR of LINC00607 after CRISPR/Cas9-medi- sion after LINC00607 knockout compared to non-BRG1- ated knockout (KO) and control (NTC), n = 3. Unpaired t test. G IGV associated genes (Fig. 5F, G). Since the motif for the ERG genome tracks of RNA-Seq of the LINC00607 locus in HUVEC with transcription factor was strongly enriched in genes down- or without CRISPR/Cas9-mediated knockout of LINC00607 (KO) regulated after LINC00607 knockout, the described gene and control (NTC). H Spheroid outgrowth assay with LINC00607 knockout (KO) and control (CTL) in HUVEC. NTC served as nega- sets were further overlapped with a publicly available ERG tive control. Cells treated with or without VEGF-A are shown. I Chromatin immunoprecipitation-Seq (ChIP-Seq) [31] from Quantification of cumulative sprout length from spheroid outgrowth HUVEC. Importantly, almost all (1372 out of 1445) of the assay shown in H, n = 28–30, two-way ANOVA. J Quantification differentially accessible genes after LINC00607 knockout of the ratio of cumulative outgrowth length and respective spheroid diameter from spheroid outgrowth assay shown in H; n = 28–30, two- overlapping with BRG1 CUT&RUN binding sites were way ANOVA. K Scratch wound assay of LINC00607 KO and NTC shared with genes ERG binds close to (Fig. 5E). control cells (CTL). Representative images after 0  h and 4  h after To inspect these global associations in more detail, we scratch (blue line) are shown. L Quantification of relative wound clo- checked a handful of genes highly important in endothelial sure in LINC00607 KO and control (NTC) in HUVEC. n = 3, area under the curve (AUC) p < 0.0001 (two-tailed t test). M Spheroid out- cells manually: VWF (von Willebrand factor), SGK1 (Serum/ growth assay of HUVEC after CRISPR/Cas9-mediated LINC00607 Glucocorticoid Regulated Kinase 1), TSPAN12 (Tetraspanin KO or non target control (NTC). Cells treated under VEGF-A con- 12) and KDR (Kinase Insert Domain Receptor) (Fig. 5H). ditions for 16  h with/ without LINC00607 overexpression (OE) are SGK1, TSPAN12, VWF and KDR were among the strong- shown. Empty vector transfection served as control (CTL). N Quan- tification of the ratio of cumulative outgrowth length and respec- est differentially expressed genes after LINC00607 knock - tive spheroid diameter from spheroid outgrowth assay shown in M. out (Fig.  3A). Of these genes KDR, TSPAN12 and VWF n = 26–29, two-way ANOVA. O LINC00607 KO and control cells represented genes listed in the VEGF signaling pathway (NTC) after in vivo matrigel plug assay in SCID mice. HUVEC were (Fig. 3E), which we described in this study to be strongly embedded in matrigel, stained with Vybrant dil (red) and injected. Isolectin GS-b4 Alexa 647 conjugated stained vessels (green). Images affected by LINC00607 perturbation. Importantly, all these were taken by light sheet microscopy 21 days after injection. Scale genes contained a BRG1 and ERG signature at their tran- bar indicates 100  µm. Representative pictures are shown. P Quanti- scriptional start site (Fig.  5H) which indicates that many fication of cells per plug integrated into the newly formed vascular LINC00607-dependent genes are also BRG1 and ERG target network shown in 2O. n = 5–6. Mann Whitney U test. Error bars are defined as mean ± SD. *p < 0.05 genes. LINC00607 is required for the stable expression of these genes. RNA immunoprecipitation with antibodies against BRG1 yielded LINC00607 as an interaction partner of BRG1 Discussion (Fig. 5A). The interaction of the lncRNA with BRG1 was specific: in contrast to β-Actin mRNA, LINC00607 was In the present study, we identified LINC00607 to be spe- not pulled down by the non-primary antibody control IgG; cifically expressed in EC and to be important for vascular RNase A treatment was able to abolish the signal (Fig. 5A). sprouting and ERG-dependent gene expression through LINC00607 knockout did not affect BRG1 expression, which BRG1. Although already constitutively highly expressed indicates LINC00607 might influence BRG1 DNA binding in EC, LINC00607 itself was upregulated by hypoxia and activity. To test this, a lentiviral CRISPR/Cas9 knockout EndMT. Through RNA- and ATAC-Seq we identified of BRG1 in HUVEC was generated (Fig.  5B, C). Cleav- LINC00607 as a lncRNA important for central pathways age Under Targets & Release Using Nuclease (CUT&RUN) of endothelial cells, in particular for VEGF signaling. sequencing, a method to determine high-resolution map- After LentiCRISPR-mediated knockout of LINC00607, ping of DNA binding sites [64], was performed using endothelial cells exhibited an impaired response to VEGF- anti-BRG1 antibodies after both non-targeting control A in respect to vascular sprouting and a reduced ability to 1 3 5 Page 14 of 21 Basic Research in Cardiology (2023) 118:5 Fig. 3 RNA- and ATAC-Seq reveal that LINC00607 maintains for the Gene Ontology biological process (GOBP) Vascular Endothe- endothelial gene expression. A Heatmap of the top 50 differen- lial Growth Factor Signaling Pathway. E Heat map of VEGF-signal- tially expressed genes as determined by RNA-Seq with (KO) or ing pathway genes and their expression differences after RNA-Seq. without (NTC) CRISPR/Cas9-mediated knockout of LINC00607. Z-score represents up- (red, positive value) or down-regulated (blue, Three different batches of HUVEC are shown. Genes shown have a negative values) genes. F Examples of significantly downregulated padj < 0.05, and a log2 fold change greater than ± 0.585. Z-score rep- genes after LINC00607 knockout. IGV genome tracks of the FLT1 resents up- (red, positive value) or down-regulated (blue, negative and FLT4 locus. Shown are RNA-Seq reads in LINC00607 knock- values) genes. B Volcano plot of RNA-Seq showing the log2 fold out (KO, red) and control (NTC, blue). Tracks of three replicates changes (KO vs. NTC) of all genes expressed against their p-adjusted are overlaid. G Examples of significantly upregulated genes after value (p-adj). C Numbers of genes from different gene classes sig- LINC00607 knockout. IGV genome tracks of the VEGFC and SPRY2 nificantly altered by LINC00607 KO vs. NTC HUVEC determined by locus. Shown are RNA-Seq reads in LINC00607 knockout (KO, RNA-Seq. D Gene Set Enrichment Analysis (GSEA) of significantly red) and control (NTC, blue). Tracks of three replicates are overlaid. altered genes showing an enrichment score and signal to noise ratio Genomic coordinates correspond to hg38 1 3 Basic Research in Cardiology (2023) 118:5 Page 15 of 21 5 Fig. 4 ERG drives LINC00607-associated gene expression. A Over- enriched (B) or decreased-enriched (C) transcription factor motifs are lap of ATAC-Seq (enhancer accessibility) and RNA-Seq (gene shown. D Gene read counts of different transcription factors in NTC- expression) signals after knockout of LINC00607 in HUVEC. treated HUVEC (from RNA-Seq), n = 3. E Mean log2 fold change Indicated are genes involved in the VEGF-signaling pathway. B, (FC) of different transcription factors in the RNA-Seq comparing C  HOMER DNA-motif enrichment analysis of differential acces- LINC00607 KO and NTC control sible peaks (LINC00607 KO vs. NTC). Five most highly increased- integrate into the vascular network of SCID mice. Mecha- system and other tissues. For example, EVF2 has been nistically, the trans-acting lncRNA interacts with the chro- shown to inhibit the ATPase activity of BRG1 [12]. Addi- matin remodeling protein BRG1 in order to maintain chro- tionally, lncRNAs can stabilize or destabilize BRG1 inter- matin states for ERG-dependent transcription. Thereby, action with other proteins, as in the case of MALAT1 pro- LINC00607 preserves endothelial gene expression pat- moting BRG1 interaction with HDAC9 [47]. lncRNAs can terns, which are essential for angiogenesis. As exemplified also affect BRG1 gene targeting. For example, we have by lncRNA SAIL (scaffold attachment factor B interacting previously shown that the lncRNA SMANTIS guides BRG1 lncRNA), which has been reported to impair the transcrip- to specific genes related to endothelial lineage specifica- tion of fibrosis-related genes in human cardiac fibroblasts tion [40]. Our present results suggest that numerous target [50], such cardiovascular-specific gene expression control genes of LINC00607, BRG1 and ERG overlap arguing that mechanisms by lncRNAs could potentially be a common- the LINC00607 could potentially facilitate BRG1 binding type of mechanism. to genes linked to the endothelial phenotype. Recent stud- In terms of transcriptional control, lncRNAs can either ies highlight the importance of constant SWI/SNF remod- act in cis at nearby genes, or in trans genome wide [60]. eling to maintain a stable open chromatin state [26, 62]. Overexpression of LINC00607 restored endothelial function Our present observations suggest that LINC00607 provides after LINC00607 knockout, demonstrating that LINC00607 a specific link for ERG securing BRG1 binding to genes acts in trans rather than in cis, because the effect of locus- maintaining the endothelial phenotype. This specific context disruption by CRISPR/Cas9 gene editing was overruled by would explain why LINC00607 is so highly expressed in the plasmid-based overexpression of LINC00607. endothelial cells. Importantly, LINC00607 interacts with the chromatin Indeed, we uncovered a large overlap between genes remodeling protein BRG1. Several lncRNAs have been with altered chromatin state and differential expression linked to BRG1, inu fl encing its activity in the cardiovascular after LINC00607 knockout, and genes with binding sites 1 3 5 Page 16 of 21 Basic Research in Cardiology (2023) 118:5 Fig. 5 LINC00607 functions through interaction with the chroma- ChIP-Seq peak. F Median log2 fold change (FC) of differentially tin remodeler BRG1. A RNA-immunoprecipitation with antibod- expressed genes from LINC00607 knockout that are located near a ies against BRG1 with and without RNase A digestion, followed by differential ATAC-Seq peak of LINC00607 knockout and also found RT-qPCR of LINC00607 and β-Actin. IgG served as a non-primary near a BRG1 CUT&RUN peak (BRG1) or not (Non). G Median antibody control. n = 5. One-Way ANOVA with Bonferroni post p-adjusted value of differentially expressed genes from LINC00607 hoc test. B Western blot analysis with antibodies against BRG1 and knockout that are located near a differential ATAC-Seq peak of β-Actin of control (NTC) or BRG1 knockout HUVEC. C RT-qPCR LINC00607 knockout and also found near a BRG1 CUT&RUN of BRG1 after CRISPR/Cas9-mediated knockout, n = 3. Paired peak (BRG1) or not (Non). H Genome tracks of ATAC-Seq, t-test. D Chromatin accessibility heat map of differential peaks from RNA-Seq, BRG1 CUT&RUN and ERG ChIP-Seq. Loci of VWF, BRG1 CUT&RUN of control (NTC) and BRG1 knockout (KO) SGK1, TSPAN12 and KDR are shown. ATAC-Seq and RNA-Seq of HUVEC. Binding regions center-aligned to the transcription start LINC00607 KO (red) and NTC (blue) are shown. Tracks of replicates sites (TSS) ± 0.5 kb are shown. E Venn diagram showing the overlap were overlaid. CUT&RUN with anti-BRG1 antibodies of NTC or of genes located near a differential ATAC-Seq peak of LINC00607 BRG1 KO HUVEC are shown in black. ChIP-Seq of ERG is shown in knockout, genes near a BRG1 CUT&RUN peak and genes near ERG green. Error bars are defined as mean ± SD. *p < 0.05 for the ERG transcription factor. ERG belongs to the ETS role in the proposed mechanism of transcriptional control. transcription factor family, which act as key regulators of the Our findings advocate for LINC00607 as one link between majority of endothelial genes, as the ETS recognition motif BRG1-mediated stabilization of chromatin states and ERG can be found in promotors of many endothelial genes [68]. target gene expression in healthy endothelium. This shows the importance of LINC00607 for the expression The endothelial expression of LINC00607 was increased control of ERG-regulated genes through its interaction with by hypoxia, EndMT, and endothelial dysfunction as induced BRG1. In this context, it is interesting to note that SWI/ by TNFα and high glucose [13]. Under these stimuli, the SNF is required to maintain open chromatin [26, 62]. BRG1, upregulation of LINC00607 matches the transcription factor being the core member of SWI/SNF, could have a central binding motifs identified in the promotor analysis. Hypoxia 1 3 Basic Research in Cardiology (2023) 118:5 Page 17 of 21 5 Funding Open Access funding enabled and organized by Projekt signalling through VEGF is an important trigger for angio- DEAL. This work was supported by the Goethe University Frank- genic specification of endothelial cells [58] as well as a key furt am Main, the DFG excellence cluster “Cardiopulmonary Insti- mechanism contributing to chronic and acute cardiovascu- tute (CPI)” EXS2026. Furthermore, it was funded by the Deutsche lar diseases [39]. Through the upregulation of LINC00607 Forschungsgemeinschaft (DFG, German Research Foundation)— Project-ID 403584255—TRR 267 to TP A04, TP A06, TP B04, TP under hypoxic conditions, the pro-angiogenic endothelial B07, TP Z02, TP Z03. The project was also supported by the BHF/ phenotype could potentially be secured by tightening the DHF/DZHK grant “Exploiting endothelial long non-coding RNAs interaction of LINC00607 with BRG1. Specifically to HIF1- to promote regenerative angiogenesis in the damaged myocardium” controlled LINC00607 expression, we found that DMOG- (ReGenLnc) as well as the Dr. Rolf Schwiete-Stiftung. SIB is supported by the Emmy Noether Programme BI 2163/1-1 and the Johanna Quandt dependent upregulation and acriflavine-mediated HIF inhibi- Young Academy at Goethe University. FJM is supported in part by tion altered the expression of the lncRNA. This in particular Merit Review Award #BX001729 from the United States Department illustrates the close interaction of hypoxia-signaling, angio- of Veterans Affairs, Biomedical Laboratory Research and Develop- genesis and LINC00607. ment Service. The fact that EndMT induction by TGF-β2 and IL-1β also Declarations increased the expression of LINC00607 points towards a role for LINC00607 in expression control beyond the endothe- Conflict of interest The authors have declared that no conflict of inter - lial phenotype. Potentially LINC00607 guides BRG1 to est exists. genes involved in EndMT. In line with this, LINC00607 is Open Access This article is licensed under a Creative Commons Attri- also expressed in certain malignant cells. In this context, bution 4.0 International License, which permits use, sharing, adapta- LINC00607 was upregulated in doxorubicin-resistant thyroid tion, distribution and reproduction in any medium or format, as long cancer cells [44]. Furthermore, LINC00607 was described as you give appropriate credit to the original author(s) and the source, to be required for tumor proliferation of osteosarcoma cells provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are [75] and was downregulated in lung adenocarcinoma [74]. included in the article's Creative Commons licence, unless indicated Linking these findings to the present study, it could be specu- otherwise in a credit line to the material. If material is not included in lated that under basal conditions, LINC00607 guides BRG1 the article's Creative Commons licence and your intended use is not to ERG target genes and during endothelial dysfunction to permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a pro-angiogenic genes to maintain an open and accessible copy of this licence, visit http://cr eativ ecommons. or g/licen ses/ b y/4.0/ . chromatin state for ERG. Since genome-wide association studies identified many risk loci globally in non-coding regions, transcriptome-wide association studies could be References important in future to identify gene-trait associations [45] and thereby the importance of LINC00607 as a candidate 1. (2019) UniProt: a worldwide hub of protein knowledge. Nucleic gene in other disease-relevant tissues. Acids Res 47:D506-D515. https://doi. or g/10. 1093/ nar/ gky10 49 Taken together, these findings suggest that LINC00607 2. Abbas N, Perbellini F, Thum T (2020) Non-coding RNAs: emerging players in cardiomyocyte proliferation and cardiac is involved in securing endothelial BRG1 and ERG- regeneration. Basic Res Cardiol 115:52. https://d oi.o rg/1 0.1 007/ dependent target gene expression, as well as appropri- s00395- 020- 0816-0 ate responses to stress and cardiovascular diseases. The 3. Allhoff M, Seré K, Pires FJ, Zenke M, Costa GI (2016) Dif- function of LINC00607 in cardiovascular diseases with ferential peak calling of ChIP-seq signals with replicates with THOR. Nucleic Acids Res 44:e153. https://doi. or g/10. 1093/ nar/ hypoxia signaling needs more investigation as LINC00607 gkw680 might be a potential therapeutic target or clinical marker. 4. Andrews S. (2010) FastQC: a quality control tool for high throughput sequence data Supplementary Information The online version contains supplemen- 5. Baumgarten N, Hecker D, Karunanithi S, Schmidt F, List M, tary material available at https://doi. or g/10. 1007/ s00395- 023- 00978-3 . Schulz MH (2020) EpiRegio: analysis and retrieval of regula- tory elements linked to genes. Nucleic Acids Res 48:W193– Acknowledgements We thank Katalin Pálfi and Manuela Spaeth for W199. https:// doi. org/ 10. 1093/ nar/ gkaa3 82 excellent technical assistance and Dominik Fuhrmann for help with 6. Bibli S-I, Hu J, Looso M, Weigert A, Ratiu C, Wittig J, Dreko- the hypoxia chamber. lia MK, Tombor L, Randriamboavonjy V, Leisegang MS, Goy- mann P, Delgado Lagos F, Fisslthaler B, Zukunft S, Kyselova Author contributions FB, JAO, TW, SK, FJM, RPB and MSL designed A, Justo AFO, Heidler J, Tsilimigras D, Brandes RP, Dimme- the experiments. FB, JAO, TW, SG, JIP, ML, DR, GB, MDP, ZSM, ler S, Papapetropoulos A, Knapp S, Offermanns S, Wittig I, TL, SS, SH, SK, JK, FJM and MSL performed the experiments. FB, Nishimura SL, Sigala F, Fleming I (2021) Mapping the endothe- JAO, TW, SG, GB, ZSM, and MSL analyzed the data. FB, TW, SG and lial cell S-sulfhydrome highlights the crucial role of integrin MHS performed bioinformatics. KHP, RB, MM, JK, SSP, SIB, RH, sulfhydration in vascular function. Circulation 143:935–948. AHB, RAB and IW helped with research design, material accessibil- https:// doi. org/ 10. 1161/ CIRCU LATIO NAHA. 120. 051877 ity and advice. FB, RBP and MSL wrote the manuscript. All authors interpreted the data and approved the manuscript. 1 3 5 Page 18 of 21 Basic Research in Cardiology (2023) 118:5 7. Bolger AM, Lohse M, Usadel B (2014) Trimmomatic: a flexible H, Fagiolini M, Fairbairn L, Fang H, Farach-Carson MC, Faulkner trimmer for Illumina sequence data. Bioinformatics 30:2114– GJ, Favorov AV, Fisher ME, Frith MC, Fujita R, Fukuda S, Fur- 2120. https:// doi. org/ 10. 1093/ bioin forma tics/ btu170 lanello C, Furino M, Furusawa J-I, Geijtenbeek TB, Gibson AP, 8. Broad Institute (2019) Picard toolkit Gingeras T, Goldowitz D, Gough J, Guhl S, Guler R, Gustincich 9. Buenrostro JD, Wu B, Chang HY, Greenleaf WJ (2015) ATAC- S, Ha TJ, Hamaguchi M, Hara M, Harbers M, Harshbarger J, seq: a method for assaying chromatin accessibility genome- Hasegawa A, Hasegawa Y, Hashimoto T, Herlyn M, Hitchens KJ, wide. Curr Protoc Mol Biol. https:// doi. or g/ 10. 1002/ 04711 Ho Sui SJ, Hofmann OM, Hoof I, Hori F, Huminiecki L, Iida K, 42727. Mb212 9s109 Ikawa T, Jankovic BR, Jia H, Joshi A, Jurman G, Kaczkowski B, 10. Bultman S, Gebuhr T, Yee D, La Mantia C, Nicholson J, Gilliam Kai C, Kaida K, Kaiho A, Kajiyama K, Kanamori-Katayama M, A, Randazzo F, Metzger D, Chambon P, Crabtree G, Magnuson Kasianov AS, Kasukawa T, Katayama S, Kato S, Kawaguchi S, T (2000) A Brg1 null mutation in the mouse reveals functional Kawamoto H, Kawamura YI, Kawashima T, Kempfle JS, Kenna differences among mammalian SWI/SNF complexes. Mol Cell TJ, Kere J, Khachigian LM, Kitamura T, Klinken SP, Knox AJ, 6:1287–1295. https:// doi. org/ 10. 1016/ s1097- 2765(00) 00127-1 Kojima M, Kojima S, Kondo N, Koseki H, Koyasu S, Krampitz 11. Cairns BR (2007) Chromatin remodeling: insights and intrigue S, Kubosaki A, Kwon AT, Laros JFJ, Lee W, Lennartsson A, Li from single-molecule studies. Nat Struct Mol Biol 14:989–996. K, Lilje B, Lipovich L, Mackay-Sim A, Manabe R-I, Mar JC, https:// doi. org/ 10. 1038/ nsmb1 333 Marchand B, Mathelier A, Mejhert N, Meynert A, Mizuno Y, de 12. Cajigas I, Leib DE, Cochrane J, Luo H, Swyter KR, Chen Lima-Morais DA, Morikawa H, Morimoto M, Moro K, Mota- S, Clark BS, Thompson J, Yates JR, Kingston RE, Kohtz JD kis E, Motohashi H, Mummery CL, Murata M, Nagao-Sato S, (2015) Evf2 lncRNA/BRG1/DLX1 interactions reveal RNA- Nakachi Y, Nakahara F, Nakamura T, Nakamura Y, Nakazato K, dependent inhibition of chromatin remodeling. Development van Nimwegen E, Ninomiya N, Nishiyori H, Noma S, Noazaki 142:2641–2652. https:// doi. org/ 10. 1242/ dev. 126318 T, Ogishima S, Ohkura N, Ohimiya H, Ohno H, Ohshima M, 13. Calandrelli R, Xu L, Luo Y, Wu W, Fan X, Nguyen T, Chen C-J, Okada-Hatakeyama M, Okazaki Y, Orlando V, Ovchinnikov DA, Sriram K, Tang X, Burns AB, Natarajan R, Chen ZB, Zhong Pain A, Passier R, Patrikakis M, Persson H, Piazza S, Prendergast S (2020) Stress-induced RNA-chromatin interactions promote JGD, Rackham OJL, Ramilowski JA, Rashid M, Ravasi T, Rizzu endothelial dysfunction. Nat Commun 11:5211. https://doi. or g/ P, Roncador M, Roy S, Rye MB, Saijyo E, Sajantila A, Saka A, 10. 1038/ s41467- 020- 18957-w Sakaguchi S, Sakai M, Sato H, Savvi S, Saxena A, Schneider 14. Deanfield JE, Halcox JP, Rabelink TJ (2007) Endothelial func- C, Schultes EA, Schulze-Tanzil GG, Schwegmann A, Sengstag tion and dysfunction: testing and clinical relevance. Circulation T, Sheng G, Shimoji H, Shimoni Y, Shin JW, Simon C, Sugiy- 115:1285–1295. https://d oi.o rg/1 0.1 161/C IRCUL ATION AHA. ama D, Sugiyama T, Suzuki M, Suzuki N, Swoboda RK, Hoen 106. 652859 PAC, Tagami M, Takahashi N, Takai J, Tanaka H, Tatsukawa H, 15. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha Tatum Z, Thompson M, Toyodo H, Toyoda T, Valen E, van de S, Batut P, Chaisson M, Gingeras TR (2013) STAR: ultrafast Wetering M, van den Berg LM, Verado R, Vijayan D, Vorontsov universal RNA-seq aligner. Bioinformatics 29:15–21. https:// IE, Wasserman WW, Watanabe S, Wells CA, Winteringham LN, doi. org/ 10. 1093/ bioin forma tics/ bts635 Wolvetang E, Wood EJ, Yamaguchi Y, Yamamoto M, Yoneda M, 16. Ertürk A, Becker K, Jährling N, Mauch CP, Hojer CD, Egen JG, Yonekura Y, Yoshida S, Zabierowski SE, Zhang PG, Zhao X, Zuc- Hellal F, Bradke F, Sheng M, Dodt H-U (2012) Three-dimen- chelli S, Summers KM, Suzuki H, Daub CO, Kawai J, Heutink P, sional imaging of solvent-cleared organs using 3DISCO. Nat Hide W, Freeman TC, Lenhard B, Bajic VB, Taylor MS, Makeev Protoc 7:1983–1995. https:// doi. org/ 10. 1038/ nprot. 2012. 119 VJ, Sandelin A, Hume DA, Carninci P, Hayashizaki Y (2014) A 17. Krueger F, James F, Ewels P, Afyounian E, Schuster-Boeckler promoter-level mammalian expression atlas. Nature 507:462–470. B (2021) FelixKrueger/TrimGalore: v0.6.7 - DOI via Zenodo. https:// doi. org/ 10. 1038/ natur e13182 Zenodo 22. Haeussler M, Schönig K, Eckert H, Eschstruth A, Mianné J, 18. Feng J, Liu T, Qin B, Zhang Y, Liu XS (2012) Identifying ChIP- Renaud J-B, Schneider-Maunoury S, Shkumatava A, Teboul L, seq enrichment using MACS. Nat Protoc 7:1728–1740. https:// Kent J, Joly J-S, Concordet J-P (2016) Evaluation of o-t ff arget and doi. org/ 10. 1038/ nprot. 2012. 101 on-target scoring algorithms and integration into the guide RNA 19. Ferrara N, Mass RD, Campa C, Kim R (2007) Targeting VEGF- selection tool CRISPOR. Genome Biol 17:148. https://doi. or g/10. A to treat cancer and age-related macular degeneration. Annu 1186/ s13059- 016- 1012-2 Rev Med 58:491–504. https:// doi. org/ 10. 1146/ annur ev. med. 58. 23. Han P, Li W, Lin C-H, Yang J, Shang C, Nuernberg ST, Jin KK, 061705. 145635 Xu W, Lin C-Y, Lin C-J, Xiong Y, Chien H, Zhou B, Ashley E, 20. Fiedler J, Breckwoldt K, Remmele CW, Hartmann D, Dittrich M, Bernstein D, Chen P-S, Chen H-SV, Quertermous T, Chang C-P Pfanne A, Just A, Xiao K, Kunz M, Müller T, Hansen A, Geffers (2014) A long noncoding RNA protects the heart from patho- R, Dandekar T, Eschenhagen T, Thum T (2015) Development of logical hypertrophy. Nature 514:102–106. https://doi. or g/10. 1038/ long noncoding RNA-based strategies to modulate tissue vascu-natur e13596 larization. J Am Coll Cardiol 66:2005–2015. https:// doi. org/ 10. 24. Hathaway CA, Heistad DD, Piegors DJ, Miller FJ (2002) Regres- 1016/j. jacc. 2015. 07. 081 sion of atherosclerosis in monkeys reduces vascular superoxide 21. Forrest ARR, Kawaji H, Rehli M, Baillie JK, de Hoon MJL, levels. Circ Res 90:277–283. https:// doi. or g/ 10. 1161/ hh0302. Haberle V, Lassmann T, Kulakovskiy IV, Lizio M, Itoh M, 104724 Andersson R, Mungall CJ, Meehan TF, Schmeier S, Bertin N, 25. Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, Cheng Jørgensen M, Dimont E, Arner E, Schmidl C, Schaefer U, Med- JX, Murre C, Singh H, Glass CK (2010) Simple combinations vedeva YA, Plessy C, Vitezic M, Severin J, Semple CA, Ishizu Y, of lineage-determining transcription factors prime cis-regulatory Young RS, Francescatto M, Alam I, Albanese D, Altschuler GM, elements required for macrophage and B cell identities. Mol Cell Arakawa T, Archer JAC, Arner P, Babina M, Rennie S, Balwierz 38:576–589. https:// doi. org/ 10. 1016/j. molcel. 2010. 05. 004 PJ, Beckhouse AG, Pradhan-Bhatt S, Blake JA, Blumenthal A, 26. Iurlaro M, Stadler MB, Masoni F, Jagani Z, Galli GG, Schübeler Bodega B, Bonetti A, Briggs J, Brombacher F, Burroughs AM, D (2021) Mammalian SWI/SNF continuously restores local acces- Califano A, Cannistraci CV, Carbajo D, Chen Y, Chierici M, Ciani sibility to chromatin. Nat Genet 53:279–287. https:// doi. org/ 10. 1038/ s41588- 020- 00768-w Y, Clevers HC, Dalla E, Davis CA, Detmar M, Diehl AD, Dohi T, Drabløs F, Edge ASB, Edinger M, Ekwall K, Endoh M, Enomoto 1 3 Basic Research in Cardiology (2023) 118:5 Page 19 of 21 5 27. Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell function. Circulation 136:65–79. https:// doi. org/ 10. 1161/ cir cu SJ, von Kriegsheim A, Hebestreit HF, Mukherji M, Schofield latio naha. 116. 026991 CJ, Maxwell PH, Pugh CW, Ratcliffe PJ (2001) Targeting of 41. Leisegang MS, Bibli S-I, Günther S, Pflüger-Müller B, Oo JA, HIF-alpha to the von Hippel–Lindau ubiquitylation complex by Höper C, Seredinski S, Yekelchyk M, Schmitz-Rixen T, Schür- O2-regulated prolyl hydroxylation. Science 292:468–472. https:// mann C, Hu J, Looso M, Sigala F, Boon RA, Fleming I, Brandes doi. org/ 10. 1126/ scien ce. 10597 96 RP (2019) Pleiotropic effects of laminar flow and statins depend 28. Jégu T, Blum R, Cochrane JC, Yang L, Wang C-Y, Gilles M-E, on the Krüppel-like factor-induced lncRNA MANTIS. Eur Heart Colognori D, Szanto A, Marr SK, Kingston RE, Lee JT (2019) J 40:2523–2533. https:// doi. org/ 10. 1093/ eurhe artj/ ehz393 Xist RNA antagonizes the SWI/SNF chromatin remodeler BRG1 42. Li B, Carey M, Workman JL (2007) The role of chromatin during on the inactive X chromosome. Nat Struct Mol Biol 26:96–109. transcription. Cell 128:707–719. https:// doi. org/ 10. 1016/j. cell. https:// doi. org/ 10. 1038/ s41594- 018- 0176-82007. 01. 015 29. Johnson KE, Wilgus TA (2014) Vascular endothelial growth factor 43. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, and angiogenesis in the regulation of cutaneous wound repair. Adv Marth G, Abecasis G, Durbin R (2009) The sequence alignment/ Wound Care (New Rochelle) 3:647–661. https://doi. or g/10. 1089/ map format and SAMtools. Bioinformatics 25:2078–2079. https:// wound. 2013. 0517doi. org/ 10. 1093/ bioin forma tics/ btp352 30. Kadam S, Emerson BM (2003) Transcriptional specificity of 44. Li L, Gao Z, Zhao L, Ren P, Shen H (2021) Long non-coding human SWI/SNF BRG1 and BRM chromatin remodeling com- RNA LINC00607 silencing exerts antioncogenic effects on thy - plexes. Mol Cell 11:377–389. https:// doi. or g/ 10. 1016/ S1097- roid cancer through the CASP9 Promoter methylation. J Cell Mol 2765(03) 00034-0 Med 25:7608–7620. https:// doi. org/ 10. 1111/ jcmm. 16265 31. Kalna V, Yang Y, Peghaire CR, Frudd K, Hannah R, Shah AV, 45. Li L, Chen Z, von Scheidt M, Li S, Steiner A, Güldener U, Koplev Osuna Almagro L, Boyle JJ, Göttgens B, Ferrer J, Randi AM, S, Ma A, Hao K, Pan C, Lusis AJ, Pang S, Kessler T, Ermel R, Birdsey GM (2019) The transcription factor ERG regulates super- Sukhavasi K, Ruusalepp A, Gagneur J, Erdmann J, Kovacic JC, enhancers associated with an endothelial-specific gene expres- Björkegren JLM, Schunkert H (2022) Transcriptome-wide asso- sion program. Circ Res 124:1337–1349. https:// doi. org/ 10. 1161/ ciation study of coronary artery disease identifies novel suscep- CIRCR ESAHA. 118. 313788 tibility genes. Basic Res Cardiol 117:6. https:// doi. org/ 10. 1007/ 32. Kent WJ, Zweig AS, Barber G, Hinrichs AS, Karolchik D (2010) s00395- 022- 00917-8 BigWig and BigBed: enabling browsing of large distributed data- 46. Liao Y, Smyth GK, Shi W (2014) featureCounts: an efficient gen- sets. Bioinformatics 26:2204–2207. https://doi. or g/10. 1093/ bioin eral purpose program for assigning sequence reads to genomic forma tics/ btq351 features. Bioinformatics 30:923–930. https:// doi. or g/ 10. 1093/ 33. Korff T, Augustin HG (1998) Integration of endothelial cells in bioin forma tics/ btt656 multicellular spheroids prevents apoptosis and induces differen- 47. Lino Cardenas CL, Kessinger CW, Cheng Y, MacDonald C, tiation. J Cell Biol 143:1341–1352. https:// doi. org/ 10. 1083/ jcb. MacGillivray T, Ghoshhajra B, Huleihel L, Nuri S, Yeri AS, 143.5. 1341 Jaffer FA, Kaminski N, Ellinor P, Weintraub NL, Malhotra R, 34. Kovacic JC, Dimmeler S, Harvey RP, Finkel T, Aikawa E, Kren- Isselbacher EM, Lindsay ME (2018) An HDAC9-MALAT1- ning G, Baker AH (2019) Endothelial to mesenchymal transition BRG1 complex mediates smooth muscle dysfunction in thoracic in cardiovascular disease: JACC state-of-the-art review. J Am Coll aortic aneurysm. Nat Commun 9:1009. https:// doi. org/ 10. 1038/ Cardiol 73:190–209. https:// doi. org/ 10. 1016/j. jacc. 2018. 09. 089s41467- 018- 03394-7 35. Kulakovskiy IV, Vorontsov IE, Yevshin IS, Sharipov RN, Fedor- 48. Lizio M, Harshbarger J, Shimoji H, Severin J, Kasukawa T, Sahin ova AD, Rumynskiy EI, Medvedeva YA, Magana-Mora A, Bajic S, Abugessaisa I, Fukuda S, Hori F, Ishikawa-Kato S, Mungall CJ, VB, Papatsenko DA, Kolpakov FA, Makeev VJ (2018) HOCO- Arner E, Baillie JK, Bertin N, Bono H, de Hoon M, Diehl AD, MOCO: towards a complete collection of transcription factor Dimont E, Freeman TC, Fujieda K, Hide W, Kaliyaperumal R, binding models for human and mouse via large-scale ChIP-Seq Katayama T, Lassmann T, Meehan TF, Nishikata K, Ono H, Rehli analysis. Nucleic Acids Res 46:D252–D259. https:// doi. org/ 10. M, Sandelin A, Schultes EA, Hoen PAC, Tatum Z, Thompson M, 1093/ nar/ gkx11 06 Toyoda T, Wright DW, Daub CO, Itoh M, Carninci P, Hayashizaki 36. Laib AM, Bartol A, Alajati A, Korff T, Weber H, Augustin HG Y, Forrest ARR, Kawaji H (2015) Gateways to the FANTOM5 (2009) Spheroid-based human endothelial cell microvessel for- promoter level mammalian expression atlas. Genome Biol 16:22. mation in vivo. Nat Protoc 4:1202–1215. https://doi. or g/10. 1038/ https:// doi. org/ 10. 1186/ s13059- 014- 0560-6 nprot. 2009. 96 49. Love MI, Huber W, Anders S (2014) Moderated estimation of fold 37. Langmead B, Salzberg SL (2012) Fast gapped-read alignment change and dispersion for RNA-seq data with DESeq2. Genome with Bowtie 2. Nat Methods 9:357–359. https:// doi. org/ 10. 1038/ Biol 15:550. https:// doi. org/ 10. 1186/ s13059- 014- 0550-8 nmeth. 1923 50. Luo S, Zhang M, Wu H, Ding X, Li D, Dong X, Hu X, Su S, Shang 38. Langmead B, Trapnell C, Pop M, Salzberg SL (2009) Ultrafast W, Wu J, Xiao H, Yang W, Zhang Q, Zhang J, Lu Y, Pan Z (2021) and memory-efficient alignment of short DNA sequences to the SAIL: a new conserved anti-fibrotic lncRNA in the heart. Basic human genome. Genome Biol 10:R25. https:// doi. org/ 10. 1186/ Res Cardiol 116:15. https://doi. or g/10. 1007/ s00395- 021- 00854-y gb- 2009- 10-3- r25 51. Meadows SM, Myers CT, Krieg PA (2011) Regulation of endothe- 39. Lee JW, Ko J, Ju C, Eltzschig HK (2019) Hypoxia signaling in lial cell development by ETS transcription factors. Semin Cell human diseases and therapeutic targets. Exp Mol Med 51:1–13. Dev Biol 22:976–984. https://doi. or g/10. 1016/j. semcdb. 2011. 09. https:// doi. org/ 10. 1038/ s12276- 019- 0235-1 009 40. Leisegang MS, Fork C, Josipovic I, Richter FM, Preussner J, Hu 52. Monteiro JP, Bennett M, Rodor J, Caudrillier A, Ulitsky I, Baker J, Miller MJ, Epah J, Hofmann P, Günther S, Moll F, Valasarajan AH (2019) Endothelial function and dysfunction in the car- C, Heidler J, Ponomareva Y, Freiman TM, Maegdefessel L, Plate diovascular system: the long non-coding road. Cardiovasc Res KH, Mittelbronn M, Uchida S, Künne C, Stellos K, Schermuly 115:1692–1704. https:// doi. org/ 10. 1093/ cvr/ cvz154 RT, Weissmann N, Devraj K, Wittig I, Boon RA, Dimmeler S, 53. Monteiro JP, Rodor J, Caudrillier A, Scanlon JP, Spiroski A-M, Pullamsetti SS, Looso M, Miller FJ, Brandes RP (2017) Long Dudnakova T, Pflüger-Müller B, Shmakova A, von Kriegsheim noncoding RNA MANTIS facilitates endothelial angiogenic A, Deng L, Taylor RS, Wilson-Kanamori JR, Chen S-H, Stewart K, Thomson A, Mitić T, McClure JD, Iynikkel J, Hadoke PWF, 1 3 5 Page 20 of 21 Basic Research in Cardiology (2023) 118:5 Denby L, Bradshaw AC, Caruso P, Morrell NW, Kovacic JC, Ulit- generation web server for deep-sequencing data analysis. Nucleic sky I, Henderson NC, Caporali A, Leisegang MS, Brandes RP, Acids Res 44:W160–W165. https:// doi. org/ 10. 1093/ nar/ gkw257 Baker AH (2021) MIR503HG loss promotes endothelial-to-mes- 60. Rinn JL, Chang HY (2012) Genome regulation by long noncoding enchymal transition in vascular disease. Circ Res 128:1173–1190. RNAs. Annu Rev Biochem 81:145–166. https:// doi. org/ 10. 1146/ https:// doi. org/ 10. 1161/ CIRCR ESAHA. 120. 318124annur ev- bioch em- 051410- 092902 54. Mootha VK, Lindgren CM, Eriksson K-F, Subramanian A, Sihag 61. Sanjana NE, Shalem O, Zhang F (2014) Improved vectors and S, Lehar J, Puigserver P, Carlsson E, Ridderstråle M, Laurila E, genome-wide libraries for CRISPR screening. Nat Methods Houstis N, Daly MJ, Patterson N, Mesirov JP, Golub TR, Tamayo 11:783–784. https:// doi. org/ 10. 1038/ nmeth. 3047 P, Spiegelman B, Lander ES, Hirschhorn JN, Altshuler D, Groop 62. Schick S, Grosche S, Kohl KE, Drpic D, Jaeger MG, Marella NC, LC (2003) PGC-1alpha-responsive genes involved in oxidative Imrichova H, Lin J-MG, Hofstätter G, Schuster M, Rendeiro AF, phosphorylation are coordinately downregulated in human dia- Koren A, Petronczki M, Bock C, Müller AC, Winter GE, Kubicek betes. Nat Genet 34:267–273. https:// doi. org/ 10. 1038/ ng1180 S (2021) Acute BAF perturbation causes immediate changes in 55. Noguchi S, Arakawa T, Fukuda S, Furuno M, Hasegawa A, Hori chromatin accessibility. Nat Genet 53:269–278. https:// doi. org/ F, Ishikawa-Kato S, Kaida K, Kaiho A, Kanamori-Katayama M, 10. 1038/ s41588- 021- 00777-3 Kawashima T, Kojima M, Kubosaki A, Manabe R-I, Murata M, 63. Seredinski S, Boos F, Günther S, Oo JA, Warwick T, Izquierdo Nagao-Sato S, Nakazato K, Ninomiya N, Nishiyori-Sueki H, Ponce J, Lillich FF, Proschak E, Knapp S, Gilsbach R, Pflüger- Noma S, Saijyo E, Saka A, Sakai M, Simon C, Suzuki N, Tagami Müller B, Brandes RP, Leisegang MS (2022) DNA topoisomerase M, Watanabe S, Yoshida S, Arner P, Axton RA, Babina M, Baillie inhibition with the HIF inhibitor acriflavine promotes transcrip- JK, Barnett TC, Beckhouse AG, Blumenthal A, Bodega B, Bonetti tion of lncRNAs in endothelial cells. Mol Ther Nucleic Acids A, Briggs J, Brombacher F, Carlisle AJ, Clevers HC, Davis CA, 27:1023–1035. https:// doi. org/ 10. 1016/j. omtn. 2022. 01. 016 Detmar M, Dohi T, Edge ASB, Edinger M, Ehrlund A, Ekwall K, 64. Skene PJ, Henikoff S (2017) An efficient targeted nuclease strat- Endoh M, Enomoto H, Eslami A, Fagiolini M, Fairbairn L, Far- egy for high-resolution mapping of DNA binding sites. Elife. ach-Carson MC, Faulkner GJ, Ferrai C, Fisher ME, Forrester LM, https:// doi. org/ 10. 7554/ eLife. 21856 Fujita R, Furusawa J-I, Geijtenbeek TB, Gingeras T, Goldowitz 65. Statello L, Guo C-J, Chen L-L, Huarte M (2021) Gene regula- D, Guhl S, Guler R, Gustincich S, Ha TJ, Hamaguchi M, Hara M, tion by long non-coding RNAs and its biological functions. Hasegawa Y, Herlyn M, Heutink P, Hitchens KJ, Hume DA, Ikawa Nat Rev Mol Cell Biol 22:96–118. https:// doi. or g/ 10. 1038/ T, Ishizu Y, Kai C, Kawamoto H, Kawamura YI, Kempfle JS, s41580- 020- 00315-9 Kenna TJ, Kere J, Khachigian LM, Kitamura T, Klein S, Klinken 66. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, SP, Knox AJ, Kojima S, Koseki H, Koyasu S, Lee W, Lennartsson Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, A, Mackay-Sim A, Mejhert N, Mizuno Y, Morikawa H, Morimoto Mesirov JP (2005) Gene set enrichment analysis: a knowledge- M, Moro K, Morris KJ, Motohashi H, Mummery CL, Nakachi Y, based approach for interpreting genome-wide expression profiles. Nakahara F, Nakamura T, Nakamura Y, Nozaki T, Ogishima S, Proc Natl Acad Sci U S A 102:15545–15550. https:// doi. org/ 10. Ohkura N, Ohno H, Ohshima M, Okada-Hatakeyama M, Okazaki 1073/ pnas. 05065 80102 Y, Orlando V, Ovchinnikov DA, Passier R, Patrikakis M, Pombo 67. Tombor LS, Dimmeler S (2022) Why is endothelial resilience key A, Pradhan-Bhatt S, Qin X-Y, Rehli M, Rizzu P, Roy S, Sajantila to maintain cardiac health? Basic Res Cardiol 117:35. https://doi. A, Sakaguchi S, Sato H, Satoh H, Savvi S, Saxena A, Schmidl org/ 10. 1007/ s00395- 022- 00941-8 C, Schneider C, Schulze-Tanzil GG, Schwegmann A, Sheng G, 68. de Val S, Black BL (2009) Transcriptional control of endothelial Shin JW, Sugiyama D, Sugiyama T, Summers KM, Takahashi N, cell development. Dev Cell 16:180–195. https://doi. or g/10. 1016/j. Takai J, Tanaka H, Tatsukawa H, Tomoiu A, Toyoda H, van de devcel. 2009. 01. 014 Wetering M, van den Berg LM, Verardo R, Vijayan D, Wells CA, 69. Wei G-H, Badis G, Berger MF, Kivioja T, Palin K, Enge M, Bonke Winteringham LN, Wolvetang E, Yamaguchi Y, Yamamoto M, M, Jolma A, Varjosalo M, Gehrke AR, Yan J, Talukder S, Turunen Yanagi-Mizuochi C, Yoneda M, Yonekura Y, Zhang PG, Zucchelli M, Taipale M, Stunnenberg HG, Ukkonen E, Hughes TR, Bulyk S, Abugessaisa I, Arner E, Harshbarger J, Kondo A, Lassmann ML, Taipale J (2010) Genome-wide analysis of ETS-family DNA- T, Lizio M, Sahin S, Sengstag T, Severin J, Shimoji H, Suzuki binding in vitro and in vivo. EMBO J 29:2147–2160. https:// doi. M, Suzuki H, Kawai J, Kondo N, Itoh M, Daub CO, Kasukawa org/ 10. 1038/ emboj. 2010. 106 T, Kawaji H, Carninci P, Forrest ARR, Hayashizaki Y (2017) 70. Wickham H (2009) Ggplot2. Elegant graphics for data analysis. FANTOM5 CAGE profiles of human and mouse samples. Sci Use R! Springer, Dordrecht, New York Data 4:170112. https:// doi. org/ 10. 1038/ sdata. 2017. 112 71. Yao R-W, Wang Y, Chen L-L (2019) Cellular functions of long 56. Oo JA, Pálfi K, Warwick T, Wittig I, Prieto-Garcia C, Matkovic noncoding RNAs. Nat Cell Biol 21:542–551. https:// doi. org/ 10. V, Tomašković I, Boos F, Izquierdo Ponce J, Teichmann T, Petriu-1038/ s41556- 019- 0311-8 kov K, Haydar S, Maegdefessel L, Wu Z, Pham MD, Krishnan 72. Yu G, He Q-Y (2016) ReactomePA: an R/Bioconductor pack- J, Baker AH, Günther S, Ulrich HD, Dikic I, Leisegang MS, age for reactome pathway analysis and visualization. Mol Biosyst Brandes RP (2022) Long non-coding RNA PCAT19 safeguards 12:477–479. https:// doi. org/ 10. 1039/ C5MB0 0663E DNA in quiescent endothelial cells by preventing uncontrolled 73. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein phosphorylation of RPA2. Cell Rep 41:111670. https:// doi. org/ BE, Nusbaum C, Myers RM, Brown M, Li W, Liu XS (2008) 10. 1016/j. celrep. 2022. 111670 Model-based analysis of ChIP-Seq (MACS). Genome Biol 57. Oo JA, Brandes RP, Leisegang MS (2022) Long non-coding 9:R137. https:// doi. org/ 10. 1186/ gb- 2008-9- 9- r137 RNAs: novel regulators of cellular physiology and func- 74. Zhao B, Xu H, Ai X, Adalat Y, Tong Y, Zhang J, Yang S (2018) tion. Pflugers Arch 474:191–204. https:// doi. or g/ 10. 1007/ Expression profiles of long noncoding RNAs in lung adenocarci- s00424- 021- 02641-z noma. Onco Targets Ther 11:5383–5390. https://doi. or g/10. 2147/ 58. Potente M, Mäkinen T (2017) Vascular heterogeneity and spe-OTT. S1676 33 cialization in development and disease. Nat Rev Mol Cell Biol 75. Zheng Y, Chen Z, Zhou Z, Xu X, Yang H (2020) Silencing of 18:477–494. https:// doi. org/ 10. 1038/ nrm. 2017. 36 long non-coding RNA LINC00607 prevents tumor proliferation of 59. Ramírez F, Ryan DP, Grüning B, Bhardwaj V, Kilpert F, Rich- osteosarcoma by acting as a sponge of miR-607 to downregulate ter AS, Heyne S, Dündar F, Manke T (2016) deepTools2: a next E2F6. Front Oncol 10:584452. https://doi. or g/10. 3389/ f onc.2020. 1 3 Basic Research in Cardiology (2023) 118:5 Page 21 of 21 5 Authors and Affiliations 1,2 1,2 1,2 3 1 Frederike Boos  · James A. Oo  · Timothy Warwick  · Stefan Günther  · Judit Izquierdo Ponce  · 1,2 1,2 1,2 4,5 3,6 1,2 Melina Lopez  · Diba Rafii  · Giulia Buchmann  · Minh Duc Pham  · Zahraa S. Msheik  · Tianfu Li  · 1,2 1,2 7 8,9,10,11 12,13 Sandra Seredinski  · Shaza Haydar  · Sepide Kashefiolasl  · Karl H. Plate  · Rüdiger Behr  · 12,14 2,5,15,16 3,6,15 2,17 12,14,18 Matthias Mietsch  · Jaya Krishnan  · Soni S. Pullamsetti  · Sofia‑Iris Bibli  · Rabea Hinkel  · 19,20 2,5,21 2,5 1,2 22,23 Andrew H. Baker  · Reinier A. Boon  · Marcel H. Schulz  · Ilka Wittig  · Francis J. Miller Jr.  · 1,2 1,2 Ralf P. Brandes  · Matthias S. Leisegang 1 13 Institut für Kardiovaskuläre Physiologie, Fachbereich Platform Degenerative Diseases, German Primate Medizin der Goethe-Universität, Theodor-Stern-Kai 7, Center-Leibniz Institute for Primate Research, Göttingen, 60590 Frankfurt am Main, Germany Germany 2 14 German Center of Cardiovascular Research (DZHK), Partner Laboratory Animal Science Unit, German Primate Center, Site RheinMain, Frankfurt, Germany Leibniz Institute for Primate Research, Göttingen, Germany 3 15 Max-Planck-Institute for Heart and Lung Research, Cardio-Pulmonary Institute, Giessen, Germany Bad Nauheim, Germany Department of Medicine III, Genome Biologics, Frankfurt, Germany Cardiology/Angiology/Nephrology, Goethe University Hospital, Frankfurt am Main, Germany Institute for Cardiovascular Regeneration, Goethe University, Frankfurt, Germany Institute for Vascular Signalling, Goethe University, Frankfurt, Germany Department of Internal Medicine, Member of the DZL, Member of Cardio-Pulmonary Institute (CPI), Justus Liebig Institute for Animal Hygiene, Animal Welfare and Farm University, Giessen, Germany Animal Behavior, University of Veterinary Medicine, Hannover, Germany Department of Neurosurgery, University Hospital Frankfurt, Frankfurt, Germany Centre for Cardiovascular Science, The Queen’s Medical Research Institute, University of Edinburgh, Edinburgh, Institute of Neurology (Edinger Institute), Neuroscience Scotland Center, Goethe University, Frankfurt, Germany CARIM Institute, University of Maastricht, Maastricht, Frankfurt Cancer Institute, University Hospital, Goethe The Netherlands University, Frankfurt, Germany Department of Physiology, Amsterdam Cardiovascular German Cancer Consortium (DKTK), Partner Site Sciences, VU Medical Center, Amsterdam UMC, Frankfurt/Mainz, Frankfurt, Germany Amsterdam, The Netherlands German Cancer Research Centre (DKFZ), Heidelberg, Department of Medicine, Vanderbilt University Medical Germany Center, Nashville, USA DZHK (German Center for Cardiovascular Research), Veterans Affairs Medical Center, Nashville, TN, USA Partner Site Göttingen, Göttingen, Germany 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Basic Research in Cardiology Springer Journals

Loading next page...
 
/lp/springer-journals/the-endothelial-enriched-lncrna-linc00607-mediates-angiogenic-function-6ju14WHPaD
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2023
ISSN
0300-8428
eISSN
1435-1803
DOI
10.1007/s00395-023-00978-3
Publisher site
See Article on Publisher Site

Abstract

Long non-coding RNAs (lncRNAs) can act as regulatory RNAs which, by altering the expression of target genes, impact on the cellular phenotype and cardiovascular disease development. Endothelial lncRNAs and their vascular functions are largely undefined. Deep RNA-Seq and FANTOM5 CAGE analysis revealed the lncRNA LINC00607 to be highly enriched in human endothelial cells. LINC00607 was induced in response to hypoxia, arteriosclerosis regression in non-human primates, post- atherosclerotic cultured endothelial cells from patients and also in response to propranolol used to induce regression of human arteriovenous malformations. siRNA knockdown or CRISPR/Cas9 knockout of LINC00607 attenuated VEGF-A-induced angiogenic sprouting. LINC00607 knockout in endothelial cells also integrated less into newly formed vascular networks in an in vivo assay in SCID mice. Overexpression of LINC00607 in CRISPR knockout cells restored normal endothelial function. RNA- and ATAC-Seq after LINC00607 knockout revealed changes in the transcription of endothelial gene sets linked to the endothelial phenotype and in chromatin accessibility around ERG-binding sites. Mechanistically, LINC00607 interacted with the SWI/SNF chromatin remodeling protein BRG1. CRISPR/Cas9-mediated knockout of BRG1 in HUVEC followed by CUT&RUN revealed that BRG1 is required to secure a stable chromatin state, mainly on ERG-binding sites. In conclusion, LINC00607 is an endothelial-enriched lncRNA that maintains ERG target gene transcription by interacting with the chromatin remodeler BRG1 to ultimately mediate angiogenesis. Keywords Long non-coding RNA · BRG1 · Endothelial cell · Gene regulation · ERG · Hypoxia Introduction hypoxia [14, 67]. In response to growth factors and hypoxia, endothelial cells sprout from pre-existing vessels in the pro- Endothelial cells form the selectively permeable monolayer cess of angiogenesis [58]. This process is physiologically between the vessel and the blood. Resting endothelium pro- important and required for wound healing [29]. However, vides an anti-coagulant and anti-inflammatory surface and uncontrolled angiogenesis also contributes to pathological contributes to the control of local vascular tone. It also facili- conditions like macular degeneration and cancer [19]. tates the vascular response to inflammation, shear stress and Recent studies suggest that long non-coding RNAs (lncRNAs) are essential in the regulation of cardiovascular gene programs [2, 57, 65]. LncRNAs are RNA molecules Ralf P. Brandes and Matthias S. Leisegang share senior authorship. longer than 200 nucleotides in length, which may lack apparent protein-coding potential. They have independent * Ralf P. Brandes functions as RNAs, separate from potential peptide coding Brandes@vrc.uni-frankfurt.de abilities [57, 65]. Through different mechanisms lncRNAs * Matthias S. Leisegang impact on gene expression and therefore the cellular Leisegang@vrc.uni-frankfurt.de phenotype [65]. LncRNAs influence many aspects of cellular Extended author information available on the last page of the article Vol.:(0123456789) 1 3 5 Page 2 of 21 Basic Research in Cardiology (2023) 118:5 function among them nuclear architecture, transcription, Propranolol hydrochloride (P0884, Sigma-Aldrich, 100 µM), translation and mRNA stability [71]. TGF-β2 (100-35B, Peprotech, 10  ng/mL), Interleukin Transcriptional control can be exerted through interaction 1β (IL-1β, 200-01B, Peprotech, 1  ng/mL) and RNase A with or recruitment of chromatin remodeling complexes, (EN0531, Thermo Fisher). which subsequently alter the epigenetic landscape [65]. The following antibodies were used: β-Actin (A1978, Chromatin remodeling proteins regulate DNA accessibil- Sigma-Aldrich), BRG1 (ab110641, Abcam), VWF (ab6994, ity by restructuring, mobilizing, and ejecting nucleosomes Abcam), PECAM1 (sc-376764, Santa Cruz), FLT1 (#36110, [11] and thereby altering the binding of transcription factors ThermoFisher) and TGFB2 (sc374658, Santa Cruz). to their DNA targets [42]. One well-known multi-protein chromatin remodeling complex, the Switch/Sucrose Non- Cell culture and stimulation experiments Fermentable (SWI/SNF) complex, has Brahma related gene-1 (BRG1) as one of its core catalytic subunits, whose Pooled human umbilical vein endothelial cells (HUVEC) knockout is embryonic lethal in mice [10, 30]. Several lncR- were purchased from PromoCell (C-12203, Lot No. NAs are known to contribute to the function of BRG1, e.g. 405Z013, 408Z014, 416Z042, Heidelberg, Germany) and EVF2 directly inhibits the ATPase and chromatin remod- cultured at 37 °C with 5% C O in a humidified incubator. eling activity [12], MANTIS, which is now termed SMAN- Gelatin-coated dishes (356009, Corning Incorporated, USA) TIS according to the HUGO Gene Nomenclature Committee, were used to culture the cells. Endothelial growth medium stabilizes the interaction between BRG1 and BAF155 and (EGM), consisting of endothelial basal medium (EBM) recruits BRG1 to angiogenesis related genes [40] and Mhrt supplemented with human recombinant epidermal growth interacts with the helicase domain of BRG1 leading to the factor (EGF), EndoCGS-Heparin (PeloBiotech, Germany), inhibition of chromatin target recognition by BRG1 [23]. 8% fetal calf serum (FCS) (S0113, Biochrom, Germany), Xist binding inhibits BRG1 activity and functionally antag- penicillin (50 U/mL) and streptomycin (50 µg/mL) (15140- onizes the recruitment of associated SWI/SNF complexes 122, Gibco/Lifetechnologies, USA) was used. For each to the inactivated X chromosome [28]. MALAT1 forms a experiment, at least three different batches of HUVEC from complex with BRG1 and HDAC9, which inhibits the expres- passage 3 were used. sion of contractile proteins in aortic aneurysm [47]. These In hypoxia experiments, cells were incubated for 24 h in a examples highlight the fundamental importance of lncRNA- SciTive Workstation (Baker Ruskinn) at 1% O and 5% C O . 2 2 BRG1 interactions. For EndMT (endothelial to mesenchymal transition), In this study, we set out to identify endothelial-enriched HUVEC were stimulated for 5 d in differentiation medium lncRNAs that impact on angiogenic function and may (DM) consisting of endothelial basal medium (EBM) sup- therefore have disease or therapeutic relevance. This led plemented with 8% FCS, penicillin (50 U/mL), streptomycin to the identification of the lncRNA LINC00607, which is (50 µg/mL), l -glutamine, TGF-β2 (10 ng/mL) and IL-1β highly enriched in the endothelium. LINC00607 has been (1 ng/mL). previously described as a super enhancer-derived lncRNA induced by high glucose and TNFα levels [13]. Our study Human endothelial cells from plaque‑free revealed that LINC00607 is induced by hypoxia and sus- and plaque‑containing arteries tains endothelial gene transcription through interaction with the chromatin remodeling protein BRG1. Ultimately, Cells were isolated and cultured as previously described [6]. LINC00607 facilitates proper endothelial ERG-responsive Studies were approved from the scientific and ethic com - gene transcription and the maintenance of the angiogenic mittee of Hipokrateion University Hospital and the Goethe response. University (extension to SC55/22-2-2018). For patient char- acteristics, see Table 1. Cultured cells were stained using RNA-FISH. Materials and methods Experiments with Macaca fascicularis Materials Experiments on adult male Cynomolgus monkeys (Macaca The following chemicals and concentrations were used for fascicularis) were approved by the Institutional Care and stimulation: human recombinant VEGF-A 165 (R&D, 293- Use Committee of the University of Iowa [24] and vessels VE, 30 ng/mL), DMOG (400091, Merck, 1 mM), acrifla - were kindly provided by one of the co-authors (FJM). The vine (A8126, Sigma-Aldrich, 10 µM), Low Density Lipo- vessels originated from a previous study [24], in which protein from Human Plasma, oxidized (oxLDL, L34357, Macaca fascicularis were fed with three different diets, a Thermo Fisher, 10 µg/mL), DMSO (D2650, Sigma-Aldrich), normal diet, an atherosclerotic diet for 47 ± 10 (mean ± SE) 1 3 Basic Research in Cardiology (2023) 118:5 Page 3 of 21 5 Table 1 Clinical data from the human subjects RNA isolation, reverse transcription and RT‑qPCR Characteristics Plaque free subjects Plaque subjects Total RNA isolation was performed with the RNA Mini Kit Demographic data (Bio&Sell) according to the manufacturers protocol and  No 4 4 reverse transcription was performed with the SuperScript  Mean age (range) 68 ± 5 71 ± 4 III Reverse Transcriptase (Thermo Fisher) using a combina-  Male/female 4/0 4/0 tion of oligo(dT)23 and random hexamer primers (Sigma). Clinical data The resulting cDNA was amplified in an AriaMX cycler  Hypertension 0 4 (Agilent) with the ITaq Universal SYBR Green Supermix  Hyperlipidemia 0 4 and ROX as reference dye (Bio-Rad, 1725125). Rela-  Angiographic carotid steno- 0 4 tive expression of human target genes was normalized to sis < 87% β-Actin, or to UBC (for EndMT experiments), whereas for Macaca fascicularis genes GAPDH was used. Expression levels were analyzed by the delta-delta Ct method with the months, or an atherosclerotic diet with an additional recov- AriaMX qPCR software (Agilent). The following oligo- ery phase for 8 months. After isolation of RNA, RT-qPCR nucleotide sequences were used: human LINC00607, for- was performed for the orthologues of human GAPDH and ward 5′-CCA CCA CCA CCA TTA CTT TC-3′ and reverse LINC00607 with Macaca fascicularis (Mf) specific prim- 5′-AGG CTC TGT ATT CCC AAC TG-3′; human β-Actin, ers. The following oligonucleotide sequences were used: forward 5′-AAA GAC CTG TAC GCC AAC AC-3′ and Mf_LINC00607, forward 5′-CTG CAT GTC ACC GCA TAC reverse 5′-GTC ATA CTC CTG CTT GCT GAT-3′; human CC-3′ and reverse 5′-TGG CTC TGC TGG AGT AG-3′; Mf_ Calponin 1 (CNN1), forward 5′-CAT CGG CAA CTT CAT GAPDH, forward 5′-TGC ACC ACC AAC TGC TTA GC-3′ CAA GG-3′ and reverse 5′-CCT GCA GCC CAA TGA and reverse 5′-GGC GTG GAC TGT GGT CAT GAG-3′. TGT TC-3′; human Transgelin (TAGLN), forward 5′-TTC Additionally, tissue sections of cryopreserved carotid artery TGA GCA AGC TGG TGA AC-3′ and reverse 5′-AGT tissue from Macaca fascicularis [24] were subjected to TGG GAT CTC CAC GGT AG-3′; human Collagen, type RNAscope to visualize LINC00607. I, alpha 1 (COL1A1), forward 5′-TGC TGG TGC TCC TGG TAC TC-3′ and reverse 5′-GGG ACC ACG TTC Human brain arteriovenous malformation ACC ACT TG-3′; human Platelet endothelial cell adhe- under propranolol treatment sion molecule-1 (PECAM1), forward 5′-AAC AGT GTT GAC ATG AAG AGC C-3′ and reverse 5′-TGT AAA ACA Patients with arteriovenous malformation (AVM) evaluated GCA CGT CAT CCT T-3′; human Ubiquitin C (UBC), at University Hospital Frankfurt were entered into an ongo- forward 5′-TTG CCT TGA CAT TCT CGA TG-3′ and ing prospective registry. The study protocol was approved reverse 5′-ATC GCT GTG ATC GTC ACT TG-3′; human by the ethical committee of the Goethe University (approval R-spondin-3 (RSPO3), forward 5′-TGT GCA ACA TGC number UCT-63-2020, Frankfurt am Main, Germany). All TCA GAT TAC A-3′ and reverse 5′-TGC TTC ATG CCA patients with proved unruptured AVMs were included after ATT CTT TCC A-3′; human Transforming growth factor written informed consent. Patients with arteriovenous mal- beta-2 (TGFB2), forward 5′-GAG CTA TAT CAG ATT CTC formation (AVM) who underwent microsurgery and had AAG TC-3′ and reverse 5′-GCC ATC AAT ACC TGC AAA tissue available were further analyzed. We selected from TCT TG-3′; human Delta Like Canonical Notch Ligand 4 our tissue bank cases of unruptured brain AVMs in patients (DLL4), forward 5′-CAG CAC TCC CTG GCA ATG TA-3′ who did not undergo pre-surgical embolization. The patients and reverse 5′-CAC AGT AGG TGC CCG TGA AT-3′; did not undergo endovascular embolization before surgi- human Fms Related Receptor Tyrosine Kinase 1 (FLT1), cal resection, and medical records did not show previous forward 5′-AAA TGC CGA CGG AAG GAG AG-3′ and history of rupture. AVM tissue (pieces with a diameter of reverse 5′-AGG TTT CGC AGG AGG TAT GG-3′; human 0.5 cm) was cultured immediately after surgical resection in von-Willebrand-Faktor (VWF), forward 5′-CCT TGA CCT the presence of 100 μM propranolol or solvent DMSO for CGG ACC CTT ATG-3′ and reverse 5′-GAT GCC CGT 72 h. Afterwards, RNA was isolated and RT-qPCR was per- TCA CAC CAC T-3′; human Guanylate Cyclase 1 Soluble formed. Additionally, tissue sections of AVM were subjected Subunit Alpha 1 (GUCY1A1), forward 5′-AGA GCT GGA to RNAscope to visualize LINC00607. TGT CTA CAA GG-3′ and reverse 5′-CGC TAT CTG AAC AGC ATG AG-3′. 1 3 5 Page 4 of 21 Basic Research in Cardiology (2023) 118:5 GAC ACA CTG C-3′; BRG1: gRNA, 5′-CAC CGC ATG Knockdown with siRNAs CTC AGA CCA CCC AG-3′ and 5′-AAA CCT GGG TGG CTC TGA GCA TGC-3′. For LINC00607, gRNA-A was For small interfering RNA (siRNA) treatments, HUVEC (80–90% confluent) were transfected with GeneTrans II cloned into lentiCRISPRv2 with hygromycin resistance, gRNA-B was cloned into lentiCRISPRv2 with puromycin according to the instructions provided by MoBiTec (Göt- tingen, Germany). A Silencer® Select siRNA was used for resistance. For BRG1, lentiCRISPRv2 with puromycin resistance was used. After cloning, the gRNA-containing siRNA-mediated knockdown of LINC00607 (Thermo Fisher Scientific, s56342). As negative control, scrambled Stealth LentiCRISPRv2 vectors were sequenced and purie fi d. Lenti - virus was produced in Lenti-X 293 T cells (Takara, 632180) RNAi™ Med GC (Life technologies) was used. All siRNA experiments were performed for 48 h. using Polyethylenamine (Sigma-Aldrich, 408727), psPAX2 and pVSVG (pMD2.G). pMD2.G was a gift from Didier Protein isolation and western analyses Trono (Addgene plasmid #12259; http:// n2t. ne t/ addge ne: 12259; RRID:Addgene_12259). psPAX2 was a gift from For whole cell lysis, HUVEC were washed in Hanks solution Didier Trono (Addgene plasmid #12260; http:// n2t. ne t/ addg e ne: 12260; RRID:Addgene_12260). LentiCRISPRv2-pro- (Applichem) and lysed with RIPA buffer [1 × TBS, 1% Des- oxycholat, 1% Triton, 0.1% SDS, 2 mM Orthovanadat (OV), duced virus was transduced in HUVEC (p1) with polybrene transfection reagent (MerckMillipore, TR-1003-G) and for 10 nM Okadaic Acid (OA), protein-inhibitor mix (PIM), 40 µg/mL Phenylmethylsulfonyl fluoride (PMSF)]. After LINC00607 knockout selection was performed with puromy- cin (1 μg/mL) and hygromycin (100 µg/mL) for 6 d and for centrifugation (10 min, 16,000×g), protein concentrations of the supernatant were determined with the Bradford assay BRG1 only with puromycin (1 μg/mL) for 6 d. Validation of the CRISPR/Cas9 knockout of LINC00607 and the extract boiled in Laemmli buffer. Equal amounts of protein were separated with SDS-PAGE. Gels were blotted was performed from genomic DNA. Genomic DNA was isolated after selection. Cells were washed, collected and onto nitrocellulose membranes, which were blocked after- wards in Rotiblock (Carl Roth). After application of the first incubated with 500 µL lysis buffer (30 min, 56 °C, 800 rpm, 0.1 M Tris/HCl pH 8.5, 0.5 M NaCl, 0.2% SDS, 0.05 M antibody, an infrared-fluorescent-dye-conjugated secondary antibody (Licor) was used. Signals were detected with an EDTA, 22.2 mg/mL Proteinase K). After removing cell frag- ments (1 min, 13.000 rpm, 4 °C), DNA was precipitated infrared-based laser scanning detection system (Odyssey Classic, Licor). by adding the equal volume 100% isopropanol followed by centrifugation (10 min, 13.000 rpm, 4 °C). The DNA LentiCRISPRv2 was washed with 70% EtOH (10 min, 13.000 rpm, 4 °C), air-dried and dissolved in TE-Buffer (10 mM Tris/HCl pH Guide RNAs (gRNA) targeting LINC00607 were selected 8.0, 1 mM EDTA pH 8.0). CRISPR/Cas9 target sites were amplified by PCR with PCR Mastermix (ThermoFisher, using the publicly available CRISPOR algorithm (http:// crisp or. tefor. net/) [22]. A dual gRNA approach consisting K0171), containing forward and reverse primers (10 µM) and 100–500 ng DNA followed by agarose gel electropho- of gRNA-A and gRNA-B was used to facilitate the knock- out of LINC00607. gRNA-A targeted a region downstream resis and ethidium bromide staining. The following primers were used: LINC00607 CRISPR target site, 5′-CTT CAG of the transcriptional start site (TSS) and gRNA-B targeted a region upstream of the TSS. BRG1 knockout was per- CCC ACT GAG TCT TG-3′ and 5′-GAG GAA CCA GCC AGA ATA GC-3′; GAPDH, 5′-TGG TGT CAG GTT ATG formed using a single gRNA approach. The gRNAs were cloned into lentiCRISPRv2 vector backbone with Esp3I CTG GGC CAG-3′ and 5′-GTG GGA TGG GAG GGT GCT GAA CAC-3′. (Thermo Fisher, FD0454) according to the standard pro- tocol [61]. lentiCRISPRv2 was a gift from Feng Zhang Scratch‑wound migration assay (Addgene plasmid #52961; h tt p :/ / n 2t . n e t / a d dge n e : 5 29 6 1; RRID:Addgene_52961) [61]). The modification of the len- 30,000 HUVEC were seeded on ImageLock 96-well plates tiviral CRISPR/Cas9v2 plasmid with hygromycin resist- ance was provided by Frank Schnütgen (Dept. of Medicine, (Essen Bioscience). Once a monolayer had formed, this was scratched the following day with a 96-pin WoundMaker tool Hematology/Oncology, University Hospital Frankfurt, Goe- the University, Frankfurt, Germany). (Essen Bioscience). EGM was then refreshed to remove dead and scraped cells. Afterwards, the cells were imaged For annealing, the following oligonucleotides were used: LINC00607: gRNA-A, 5′-CAC CGC ATG TGC CCC CTT in an Incucyte imaging system for 11 h (one image every one hour, with the “phase” image channel and 10× magni- TGT TGA A-3′ and 5′-AAA CTT CAA CAA AGG GGG CAC ATG C-3′, gRNA-B, 5′-CAC CGC AGT GTG TCA fication). The Scratch Wound Cell Migration Module of the TGT TAT CTT G-3′ and 5′-AAA CCA AGA TAA CAT 1 3 Basic Research in Cardiology (2023) 118:5 Page 5 of 21 5 Incucyte S3 Live Cell Analysis System (Essen Bioscience) the cells on ice and irradiated with 0.150 J/cm 254 nm UV was used to monitor and analyze the cells. light. Cells were scraped twice in 500 µL Hanks buffer and centrifuged at 1000×g at 4 °C for 4 min. For nuclear pro- Spheroid outgrowth assay tein isolation, cells were resuspended in hypotonic buffer (10 mM HEPES pH 7.6, 10 mM KCl, 0.1 mM EDTA pH Spheroid outgrowth assays in HUVEC were performed as 8.0, 0.1 mM EGTA pH 8.0, 1 mM DTT, 40 µg/mL PMSF) described in [33]. Stimulation of spheroid outgrowth was and incubated on ice for 15 min. Nonidet was added to a performed with VEGF-A 165 (R&D, 293-VE, 30 ng/mL) for final concentration of 0.75% and cells were centrifuged 16 h. Spheroids were imaged with an Axiovert135 micro- (1 min, 4 °C, 16,000×g). The nuclear-pellet was washed scope (Zeiss). The cumulative sprout length and spheroid twice in hypotonic buffer, lysed in high salt buffer (20 mM diameter were quantified by analysis with the AxioVision HEPES pH 7.6, 400 mM NaCl, 1 mM EDTA pH 8.0, 1 mM software (Zeiss). EGTA pH 8.0, 1 mM DTT, 40 µg/mL PMSF) and centri- fuged (5 min, 4 °C, 16,000×g). 10% of the nuclear lysate Plasmid overexpression and Spheroid outgrowth was taken as input. 4 µg of antibody was pre-coupled to 30 assay µL protein G magnetic beads in bead wash buffer (20 mM HEPES pH 7.6, 200 mM NaCl, 1 mM EDTA pH 8.0, 1 mM Plasmid overexpression was performed using 700,000 EGTA pH 8.0, 1  mM DTT, 40  µg/mL PMSF) for 1  h at HUVEC and the Neon electroporation system (Invitrogen, RT, then washed once with high salt buffer and twice with 1400 V, 1 × 30 ms pulse) in E2 buffer for the following plas- bead wash buffer. The antibody-coupled beads were added mids (7 µg per transfection): pcDNA3.1 + LINC00607 and to the nuclear lysate and rotated for 1 h at 4 °C. Samples pcDNA3.1 + . Overexpression was performed 16 h prior to were placed on a magnetic bar and the lysate discarded. The conducting the spheroid outgrowth assay. beads were washed three times in high salt buffer (50 mM Tris–HCl, 1 M NaCl, 1 mM EDTA, 0.1% SDS, 0.5% Sodium Proliferation assay Deoxycholate, 1% NP-40, 1 mM DTT, 40 µg/mL PMSF) at 4 °C for 10 min per wash. Beads were then washed twice 5000 HUVEC (CTL or siLINC00607) were seeded on in bead wash buffer 2 (20  mM TrisHCl, 10  mM MgCl , 96-well plates, 16 h after siRNA transfection. Nuclei were 0.2% Tween, 1 mM DTT, 40 µg/mL PMSF). For RNase A stained with Incucyte® Nuclight Rapid Red Dye (Sartorius, treatment, beads were placed in a buffer containing 20 mM 4717) (1:2000 in EGM). The cells were imaged every 3 h in Tris–HCl, EDTA pH 8.0 and 2 µL of RNase A (10 mg/mL) an Incucyte imaging system, using “phase” and “red” image for 30 min at 37 °C and then washed again in bead wash channels and 10X magnification. The Proliferation Module buffer. For elution of RNA, the remaining wash buffer was of the Incucyte S3 Live Cell Analysis System (Essen Biosci- removed and 1 mL QIAzol (Qiagen) was added to the beads ence) was used to monitor and analyze the cells. and incubated at RT for 10 min. 400 µL chloroform was added to the samples and vortexed for 10 s followed by incu- Apoptosis assay bation for a further 10 min at RT. Samples were then centri- fuged at 12,000×g for 15 min at 4 °C. 500 µL of the upper 7500 HUVEC (CTL or siLINC00607) were seeded on aqueous phase was transferred to a new tube and 2 µL gly- 96-well plates, 16 h after siRNA transfection. Nuclei were cogen (GlycoBlue Coprecipitant, ThermoFisher, AM9515) stained with Incucyte® Annexin V Green Dye (Sartorius, and 500 µL isopropanol added. Samples were inverted mul- 4642) (1:500 in EGM). The cells were imaged every 3 h tiple times and incubated at RT for 10 min before being in an Incucyte imaging system, using “phase” and “green” centrifuged again at 12,000×g for 10 min. The supernatant image channels and 10X magnification. The Apoptosis Mod- was removed and the pellet washed with 1 mL 75% etha- ule of the Incucyte S3 Live Cell Analysis System (Essen nol by vortexing. The pellet was centrifuged at 7500×g for Bioscience) was used to monitor and analyze the cells. 5 min at 4 °C, dried and resuspended in 30 µL nuclease-free water. RNA samples were reverse transcribed for qPCR as RNA immunoprecipitation (RIP) described above. To identify RNAs bound to a protein of interest, specific In vivo matrigel plug assay antibodies and Pierce™ Protein G Magnetic Beads (88847, Thermo Fisher) were used to immunoprecipitate RNAs 150,000 HUVEC per plug were stained with Vybrant Dil bound to the target protein. Cells were grown to 80% conflu- (1:200 in 1 mL Basal Medium (EBM); Thermo Fisher, ence on a 10 cm plate (roughly 3 million cells) and washed V-22885). After incubation (45  min at 37  °C, 5  min at once with Hanks buffer. 6 mL Hanks buffer was added to 4 °C), cells were washed with EBM (Lonza), resuspended 1 3 5 Page 6 of 21 Basic Research in Cardiology (2023) 118:5 in EGM containing 20% methocel (Sigma-Aldrich) and version 1.26.0 [49]. Only genes with a minimum fold change cultured in hanging drops (25 µL/drop). Harvesting of of +/− 1.5 (log2 +/− 0.59), a maximum Benjamini–Hoch- spheroids and injection of matrigel containing sphe- berg corrected p-value of 0.05, and a minimum combined roids into SCID mice (Charles River Laboratories) was mean of 5 reads were deemed to be significantly differen- performed as described previously [36]. 21 d after injec- tially expressed. The Ensemble annotation was enriched tion, Isolectin GS-IB4 from Griffonia simplicifolia, Alexa with UniProt data (release 06.06.2014) based on Ensembl Fluor® 647 Conjugate (I32450, Thermo Fisher) was gene identifiers (Activities at the Universal Protein Resource administered intravenously and was allowed to circulate (UniProt) [1]). for 20 min. After transcardial perfusion of the animals, the plugs were dissected, cleaned, fixed in 4% Paraform- Assay for transposase‑accessible chromatin using aldehyde (PFA) and subsequently cleared following the sequencing (ATAC‑Seq) 3DISCO procedure [16]. Imaging was carried out with the Ultramicroscope II (UM-II, LaVision Biotec, Bielefeld) at 50,000 HUVEC were used for ATAC library preparation 16× magnification (10 Zoom body + 2 × Objective). Pic- using lllumina Tagment DNA Enzyme and Buffer Kit (Illu - tures were taken with a Neo 5.5 (3-tap) sCOMs Camera mina). The cell pellet was resuspended in 50 µL of the lysis/ (Andor, Mod. No.: DC-152q-C00-FI). The ImSpectorPro transposition reaction mix (25 µL TD-Buffer, 2.5 µL Nextera Version_3.1.8 was used. Quantification of 3D Images Tn5 Transposase, 0.5 µL 10% NP-40 and 32 µL H2O) and was performed with Imaris (Bitplane Version 9.6). The incubated at 37 °C for 30 min followed by immediate puri- surface function was used to manually delete auto fluo- fication of DNA fragments with the MinElute PCR Purifi- rescence signals and artefacts. Signal background was cation Kit (Qiagen). Amplification of Library and Indexing removed using baseline subtraction. Cells were detected was performed as described elsewhere [9]. Libraries were and counted with the Spots-Algorithm (estimated diam- mixed in equimolar ratios and sequenced on NextSeq500 eter = 10.0  μm; background subtraction = tr ue; “intensity platform using V2 chemistry. Trimmomatic version 0.39 center Ch = 3" above 395; Region Growing Type = Local was employed to trim raw reads after a quality drop below Contrast). Lower threshold was chosen depending to the a mean of Q20 in a window of 5 nt [7]. Only reads above background signal. Cells were considered incorporated in 15 nt were cleared for further analyses. These were mapped the vascular network with the threshold of the “intensity versus the hg38 version (emsambl release 101) of the human Max. channel = 2” above 575. genome with STAR 2.7.7a [15] using only unique align- ments to exclude reads with uncertain arrangement. Reads were further deduplicated using Picard 2.21.7 [8] to avoid RNA‑Seq PCR artefacts leading to multiple copies of the same origi- nal fragment. The Macs2 peak caller version 2.1.1 was 900  ng of total RNA was used as input for SMARTer employed to accommodate for the range of peak widths typi- Stranded Total RNA Sample Prep Kit—HI Mammalian cally expected for ATAC-Seq [73]. Minimum qvalue was set (Takara Bio). Sequencing was performed on the Next- to -4 and FDR was changed to 0.0001. Peaks overlapping Seq500 instrument (Illumina) using v2 chemistry, resulting ENCODE blacklisted regions (known misassemblies, satel- in average of 38 M reads per library with 1 × 75 bp single lite repeats) were excluded. In order to be able to compare end setup. The resulting raw reads were assessed for qual- peaks in different samples, the resulting lists of significant ity, adapter content and duplication rates with FastQC [4]. peaks were overlapped and unified to represent identical Trimmomatic version 0.39 was employed to trim reads after regions. The counts per unified peak per sample were com- a quality drop below a mean of Q20 in a window of 10 puted with BigWigAverageOverBed [32]. Raw counts for nucleotides [7]. Only reads between 30 and 150 nucleotides unified peaks were submitted to DESeq2 (version 1.20.0) were cleared for further analyses. Trimmed and filtered reads for normalization [49]. Peaks were annotated with the pro- were aligned versus the Ensembl human genome version moter of the nearest gene in range (TSS +/− 5000 nt) based hg38 (release 99) using STAR 2.7.3a with the parameter on reference data of GENCODE vM15. Peaks were deemed “–outFilterMismatchNoverLmax 0.1” to increase the maxi- to have significantly different counts between conditions at mum ratio of mismatches to mapped length to 10% [15]. The an average score of 20, and a log2 transformed fold change number of reads aligning to genes was counted with fea- of < − 0.59 or > 0.59. tureCounts 1.6.5 tool from the Subread package [46]. Only reads mapping at least partially inside exons were admit- RNA fluorescence in‑situ hybridization ted and aggregated per gene. Reads overlapping multiple genes or aligning to multiple regions were excluded. Dif- RNA-FISH was performed similar as described before ferentially expressed genes were identified using DESeq2 [56]. Briefly, cells grown on gelatin-coated 8-well µ-Slides 1 3 Basic Research in Cardiology (2023) 118:5 Page 7 of 21 5 (ibidi) were fixed in 4% paraformaldehyde (PFA) (in PBS, for 20 min using the Braun steamer. Treatment with protease 10 min, at RT) and washed 3 times with PBS. Cells were Plus for 30 min at 40 °C was performed using the HybEZ permeabilized in 0.5% Triton X-100 (in PBS, 5 mM vana- System. The probes (human LINC00607 (ACD #894351), dyl complex (VRC, NEB)) on ice for 10 min and washed Macaca fascicularis LINC00607 (ACD #1217651-C1)) and 3 times with PBS. Prior to hybridization, cells were rinsed amplification steps were carried out according to instructions once in 2xSSC. Hybridization was performed over night except that Amp5 was incubated for 45 min. Signal detection at 37  °C in hybridization buffer (10% dextran sulfate, was carried out by incubation with DAB for 20 min at RT. 50% formamide, 2xSSC, 400  µg E.coli tRNA, 0.02% Sections were mounted in EcoMount after dehydration and RNase-free bovine serum albumin, 2 mmol/L VRC) and analyzed by light microscopy. 10 nmol/L 5’TYE-665 labelled locked nucleic acid (LNA) detection probe (Qiagen). Custom LNA detection probes BRG1 CUT&RUN targeting LINC00607 were designed with the Qiagen GeneGlobe Custom LNA design tool and had the follow- BRG1 Cleavage Under Targets & Release Using Nuclease ing sequences: 5′-AGG AGC TGA GAT GCA CAT ACT- (CUT&RUN), a method established by Skene and Henikoff 3′. The cells were washed 4 times for 15  min in buffer in 2017 [64], was performed similarly as described in the containing 2xSSC and 50% formamide and were counter- EpiCypher CUT&RUN Protocol v2.0, but with minor modi- stained with DAPI (in PBS). Images were captured with a fications for the cell type and antibody used. Briefly, 500,000 laser confocal microscope LSM800 (Zeiss, Germany) and NTC or BRG1 knockout HUVEC were washed with wash analyzed with ZEN lite software (Zeiss, Germany). Fluo- buffer (20 mM HEPES pH 7.9, 150 mM NaCl, 500 nM sper - rescence intensities were analyzed with ImageJ software midine, 1X Roche Protein Inhibitor Cocktail) at RT. Cells (ImageJ, NIH). were resuspended in wash buffer and 10 µL BioMag®Plus Concanavalin A (ConA) beads (Polysciences, 86057-3) were added for 10 min at RT. Beads were separated on a magnetic Immunofluorescence rack and washed once before being resuspended in 100 µL antibody buffer (wash buffer, 0.25% Digitonin and 2 mM Cells were seeded on 8-well immunofluorescence plates EDTA) and 1 µL BRG1 antibody (Abcam, ab110641). Beads (Ibidi, Germany). After washing with PBS, the cells were were incubated with the antibody over night with gentle fixed with 4% PFA and permeabilized with 0.05% Triton shaking at 4 °C. The next day, beads were washed twice X-100. Cells were blocked in 3% BSA (bovine serum albu- with 200 µL 0.25% Digitonin wash buffer and resuspended min) for 30 min, followed by incubation at 4 °C overnight in Digitonin wash buffer containing 2 µL CUTANA™ pAG- with the primary antibody (1:200 dilution). After washing MNase (15–1016, EpiCypher, 15-1016) and incubated on with 0.3% Tween20 in PBS, the cells were incubated with ice for 30 min. Samples were washed twice and then resus- the secondary antibody (1:500 dilution; Alexa Fluor 647, pended in 100 µL Digitonin wash buffer containing 2 µL #A31573, Invitrogen, USA) for 30 min. The cells counter- CaCl at a final concentration of 100 mM and incubated stained with 4′,6-diamidino-2-phenylindole (DAPI). Images for 2 h at 4 °C with gentle shaking. 33 µL of 2X “stop solu- were captured with a laser confocal microscope LSM800 tion” (340 mM NaCl, 20 mM EDTA, 4 mM EGTA, 0.25% (Zeiss, Germany) and analyzed with ZEN lite software Digitonin, 100 µg/mL RNase A, 50 µg/mL Glycoblue) was (Zeiss, Germany). added to the beads and incubated at 37 °C for 10 min. Sam- ples were placed on a magnetic rack and the supernatant removed and kept for DNA purification. Briefly, 5X vol- RNAscope ume of binding buffer (20 mM HEPES pH 7.9, 20 mM KCl, 1 mM CaCl , 1 mM MnCl ) was added to the samples and 2 2 RNA in  situ hybridization was performed with the the pH adjusted with sodium acetate before being transferred RNAscope 2.5 HD Detection Reagents (322310), Advanced to a purification column (ActiveMotif, 58,002) and centri- Cell Diagnostics (ACD), Newark, CA, USA) similarly as fuged at 11,000×g for 30 s. The column was then washed described before [41]. The single-plex chromogenic brown with 750 µL wash buffer and dried by centrifugation for assay was used according to manufacturer’s instructions with 2 min. DNA was eluted with 25 µL elution buffer and the minor changes. OCT was removed from unfixed, cryopre- DNA concentration measured with a Qubit 3.0 Fluorometer served tissue-sections in PBS for 5 min, baked for 30 min at (Life Technologies). 60 °C and post-fixed in 4% PFA. Paraffin-embedded, forma- lin-fixed sections were deparaffinized prior to staining. Sec- tions were treated with hydrogen peroxide for 10 min. Target retrieval was performed with boiling in 1xRetrieval solution 1 3 5 Page 8 of 21 Basic Research in Cardiology (2023) 118:5 FANTOM5 CAGE and ENCODE expression data was Library preparation and sequencing of CUT&RUN samples obtained from the FANTOM5 website and was published elsewhere [21, 48, 55]. DNA libraries were prepared according to the manufac- Publicly available HUVEC ERG ChIP-sequencing data (GSE124891) was downloaded from the Gene Expression turer’s protocol (NEBNext® Ultra II, NEB) with some minor adjustments for CUT&RUN samples. Briefly, sam- Omnibus (GEO) [31]. ples were brought to 50 µL with 0.1X TE buffer and DNA end preparation performed as instructed but with incuba- ERG ChIP‑seq data analysis tion at 20 °C for 20 min and then 58 °C for 45 min. Adap- tor ligation was performed with a 1:10 dilution of adap- FASTQ files were trimmed with Trim Galore! [17] and tor (NEB, E6440S). For DNA purification, 0.9× Volume AMPure XP beads (Beckman Coulter, A63881) was added aligned to the Ensembl human genome version hg38 (ensembl release 104) using Bowtie2 [37, 38]. Duplicate to the samples and incubated for 5 min at RT. Beads were washed twice with 200 µL 80% ethanol and DNA eluted reads were removed with rmdup [43]. Peaks were called on the aligned data using MACS2 [18] and annotatePeaks with 17 µL 0.1X TE buffer for 2 min at RT. PCR amplifi- cation of the eluted DNA was performed as described in (HOMER) [25] used to identify the nearest genes to called peaks. the manufacturer’s protocol but with the addition of 2.5 µL Evagreen (20X) for visualization of the amplification Use of FANTOM5 CAGE ENCODE data for promoter curves on an AriaMx Real-time PCR system (Agilent). The denaturation and annealing/extension steps of the and expression analysis of LINC00607 PCR amplification were performed for around 12 cycles and stopped before the curves plateaued. A cleanup of The promoter of LINC00607 was defined as nucleotide sequence with a length of 1000 nt, starting from a promi- the PCR reaction was performed twice with 1.1× Ampure beads and eluted each time in 33 µL 0.1× TE buffer. DNA nent FANTOM5 CAGE region having multiple peaks in close vicinity (approx. 30 nt) going in upstream direction concentrations were measured with a Qubit (Thermo Fisher) and size distributions measured on a Bioanalyzer for 970 nt (hg38 chr2:215,848,858–215,849,857). Promoter analysis was performed with filters for the indicated tran- (Agilent). Sequencing was performed on the NextSeq1000/2000 scription factors with the MoLoTool (https://molo tool. aut os ome.or g/), an interactive web application suitable to identify (Illumina). The resulting raw reads were assessed for qual- ity, adapter content and duplication rates with FastQC DNA sequences for transcription factor binding sites (TFBS) with position weight matrices from the HOCOMOCO data- [4]. Trim Galore! [17] was used to trim reads before alignment to the Ensembl human genome version hg38 base [35]. To compare the individual lncRNA expression towards (ensembl release 104) using Bowtie2 [37, 38]. Duplicate reads were removed with rmdup [43] and coverage tracks all other cell types or tissues, each cell type-specific signal obtained with FANTOM5 CAGE (or ENCODE) [21, 48, generated with bamCoverage (deepTools Version 3.5.1) [59]. ComputeMatrix [59] and plotHeatmap (deepTools 55] was divided through the mean signal observed in all cell types or tissues and plotted. Version 3.5.1) were used on the coverage tracks to gener- ate heatmaps of BRG1 binding across the genome. Peaks Gene‑set enrichment analysis were called on the aligned data using MACS2 [18] and annotatePeaks (HOMER) [25] was used to identify the GSEA (Gene-Set Enrichment Analysis) [54, 66] was per- nearest genes to called peaks. formed based on the RNA-Seq data to identify gene sets that were significantly enriched from genes differently expressed Publicly available datasets between the NTC control and LINC00607 knockout. 1000 permutations were performed and gene sets were considered The following RNA-Seq datasets used in this study origi- statistically enriched with a nominal P < 0.05. nated from NCBI GEO: HUVEC treated with normoxia or hypoxia (GSE70330) [20], ACF treatments of HUVEC Differential ATAC‑sequencing analysis and intersection with gene‑linked regulatory under normoxia (GSE176555) or hypoxia (GSE186297)[63], EndMT treatments of HUVEC and PAEC (GSE118446) elements [53]. Alignment files arising from ATAC-sequencing data analysis detailed above were subjected to replicate-based differential 1 3 Basic Research in Cardiology (2023) 118:5 Page 9 of 21 5 peak calling using THOR (v0.13.1) [3], which employs a Expression)-ENCODE database revealed that LINC00607 hidden Markov model-based approach to identify differen- is one of the most endothelial-enriched lncRNAs (Fig. 1A). tially accessible regions of chromatin between conditions. Particularly high levels of LINC00607 were observed in aor- Differential peaks were those with reported adjusted p-val- tic, venous, lymphatic, thoracic and arterial ECs (Fig. 1B). ues less than 0.01. Differential peaks were subsequently Additionally, FANTOM5 CAGE-ENCODE cell-type expres- intersected with regulatory elements from EpiRegio (v1.0.0) sion data showed LINC00607 to be predominantly localized [5], a collection of regulatory elements and their associated in the nucleus (Fig.  1A), which was confirmed by RNA- genes. Genes whose expression is dependent on differen- fluorescence in situ hybridization (RNA-FISH) in HUVEC tially accessible regulatory elements were subjected to path- (Fig. 1C). RT-qPCR after reverse transcription with random way enrichment analysis using the ReactomePA (v1.36.0) or oligodT oligonucleotides revealed that LINC00607 has a [72] package for R. Subsequently, differential accessibility of poly-A tail (Fig. S1A). regulatory elements linked to genes differentially expressed Importantly, LINC00607 expression was altered in in RNA-Seq could be quantified for different gene sets, and various cardiovascular diseases. Endothelial cells isolated displayed graphically with ggplot2 (v3.3.5) [70]. Motif and cultured from plaque-containing arteries, regarded enrichment analysis of differential ATAC-sequencing peaks as post-atherosclerotic, showed significant upregulation was performed using HOMER (v4.11.1) [25] by providing of LINC00607 in atherosclerotic-derived ECs compared sequences underlying the peaks, and otherwise the default to plaque-free-derived ECs (Fig.  1D, E). LINC00607 parameters. expression was also increased in response to propranolol treatment of human arteriovenous malformation explants Data availability (Fig. 1F, S1B). Furthermore, the corresponding orthologue of LINC00607 (Fig. S1C) was strongly induced in Macaca The RNA-Seq and ATAC-Seq datasets have been deposited fascicularis samples undergoing atherosclerosis regression and are available at NCBI GEO with the accession number after a high fat diet (Fig. 1G, H). GSE199878: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. We next searched for potential gene regulatory mecha- cgi? acc= GSE19 9878 nisms responsible for controlling LINC00607 expression. BRG1 CUT&RUN datasets have been deposited and An analysis of the promoter region, defined here as the are available at NCBI GEO with the accession number FANTOM5 CAGE transcription start site signal to 1000 GSE201824: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. nucleotides (nt) upstream, revealed binding motifs for mul- cgi? acc= GSE20 1824 tiple transcription factors. In particular, ARNT (also known as Hypoxia Inducible Factor 1 Beta) and HIF1A (Hypoxia Statistics Inducible Factor 1 Alpha) were identified multiple times and in close proximity to the transcription start site, indicative Unless otherwise indicated, data are given as means ± stand- of transcriptional regulation by hypoxia (Fig. 1I). Indeed, ard deviation (SD). Calculations were performed with Prism LINC00607 expression was significantly increased when 8.0 or BiAS.10.12. The latter was also used to test for normal HUVEC were cultured under hypoxic conditions (1% oxy- distribution (Shapiro–Wilk) and similarity of variance. In gen) (Fig. 1J). A publicly available RNA-Seq dataset con- case of multiple testing, Bonferroni correction was applied. taining hypoxia-stimulated HUVEC [20] confirmed this For multiple group comparisons ANOVA followed by post finding (Fig.  1K); in fact, LINC00607 was among the top hoc testing was performed. Individual statistics of depend- upregulated lncRNAs in this dataset (Fig. 1L). Interestingly, ent samples were performed by unpaired t test, of unpaired stimulation of HUVEC with oxLDL and DMOG, the latter samples by unpaired t-test and if not normally distributed of which is known to stabilize HIF1α under both hypoxic by Mann–Whitney test. P values of < 0.05 was considered and normoxic conditions [27], increased LINC00607 expres- as significant. Unless otherwise indicated, n indicates the sion (Fig. 1M, S1D). Conversely, the DNA topoisomerase number of individual experiments. and HIF-inhibitor acriflavine (ACF) [63] led to a decrease in LINC00607 expression, which was exacerbated under hypoxia (Fig. 1N–P). Results In addition to HIF binding sites, the promoter analysis of LINC00607 yielded SMAD binding motifs. To test their LINC00607 is a highly endothelial‑enriched lncRNA relevance for LINC00607 expression, HUVEC were stimu- induced by hypoxia lated with TGF-ß2 and IL-1ß to induce endothelial to mes- enchymal transition (EndMT), a process in which SMADs A screen for the top-expressed endothelial lncR- play a central role [34]. Indeed, EndMT strongly increased NAs in the FANTOM5 CAGE (Cap Analysis of Gene the expression of LINC00607 (Fig. 1Q, S1E, F) and similar 1 3 5 Page 10 of 21 Basic Research in Cardiology (2023) 118:5 findings could be retrieved from publicly available RNA-Seq LINC00607 promotes sprouting, proliferation datasets [52] of HUVEC (Fig. 1R) and pulmonary arterial and vascularization endothelial cells (PAEC) (Fig. 1S). These data indicate that LINC00607 is an endothelial- In order to study the functional relevance of LINC00607 enriched lncRNA induced by transcription factors that are in endothelial cells, spheroid outgrowth assays were per- central in hypoxic and EndMT signalling. formed. In this assay, knockdown of LINC00607 with siRNA (Fig.  2A) suppressed sprouting in response to VEGF-A (Fig.  2B–D). Next, a LINC00607 knockout in HUVEC was achieved by CRISPR/Cas9-mediated 1 3 Basic Research in Cardiology (2023) 118:5 Page 11 of 21 5 ◂Fig. 1 LINC00607 is an EC-enriched lncRNA upregulated dur- normal angiogenic response to VEGF-A (Fig.  2M, N, ing hypoxia and EndMT. A FANTOM5 CAGE-ENCODE expres- S2D, E). This suggests that the RNA itself mediates the sion of the 9 highest endothelial expressed lncRNAs across dif- observed functional effects by acting in trans. ferent cell lines. Each cell type-specific signal was divided through Collectively, these data demonstrate that loss of the mean signal observed in all cell types. B FANTOM5 CAGE expression of the 9 highest expressed endothelial lncRNAs across LINC00607 limits endothelial angiogenic capacity. As different endothelial tissues. Calculation was performed as in A. C LINC00607 is not conserved in mice, its physiologi- RNA-FISH of LINC00607 in HUVEC. LINC00607 is labelled with cal importance was studied by assessing the capacity a 5’TYE-665 probe, DAPI is used to stain the nuclei. Scale bar indi- of HUVEC to integrate into the vascular network of cates 20 µm. D, E RNA-FISH of LINC00607 in cultured endothelial cells retrieved from human plaque-free (healthy, free) and plaque- matrigels when injected in SCID-mice. Importantly, in containing (atherosclerosis, cont.) arteries. Scale bar indicates this in vivo assay, knockout of LINC00607 significantly 20  µm. Fluorescence intensity per cell for TYE-665 (LINC00607) is decreased the capacity of HUVEC to be integrated into shown in E. n = 26–30, unpaired t test. F RT-qPCR of LINC00607 in the murine vascular network (Fig.  2O, P). These data human arteriovenous malformations (AVM) treated with and with- out the β-blocker propranolol for 72  h. n = 5, Mann–Whitney U test. demonstrate that LINC00607 acts in trans as a pro-angi- G RT-qPCR of the LINC00607 homologue in vessels originating ogenic lncRNA. from Macaca fascicularis treated either with a normal diet (CTL), a high fat diet (Ath) or with a high-fat diet and a subsequent recov- LINC00607 maintains transcription of genes ery phase (Reg). n = 3. One-way ANOVA with Bonferroni post hoc test. H RNA in  situ hybridization of carotid arteries  from Macaca involved in VEGF‑signalling fascicularis (Mf) with RNAscope. Arrows point to dots indicating LINC00607. Representative images are shown. Scale bar indicates To identify how LINC00607 impacts on angiogenic func- 10  µm. I Promoter analysis of LINC00607. A region starting from tion, gene expression was determined by RNA-Seq with and the LINC00607 transcriptional start site (TSS) and 1000 base pairs (bp) upstream was analyzed with MoLoTool and plotted according to without LentiCRISPR-mediated knockout of LINC00607 p value. J Relative LINC00607 expression in HUVEC treated with in HUVEC. Deletion of LINC00607 markedly impacted normoxia (NOX) or hypoxia (HOX), n = 7. Mann–Whitney t test. K endothelial gene expression (Fig.  3A–C, S3A–F), with LINC00607 gene read counts in HUVEC cultured under normoxic a greater tendency to decrease rather than increase the and hypoxic conditions, n = 3. Unpaired t test. L Volcano plot of log2 fold changes of lncRNAs expressed in hypoxia versus normoxia. M expression of protein-coding and non-coding RNAs (Fig. Relative expression of LINC00607 in HUVEC after stimulation with S3G, H). Due to the observed angiogenic defects, a Gene Set DMOG. DMSO served as control (CTL), n = 8, Mann–Whitney U Enrichment Analysis (GSEA) was performed for the VEGF- Test. N Relative expression of LINC00607 in HUVEC after stimula- signaling pathway. GSEA revealed a strong association of tion with acriflavine (ACF), n = 7, Mann–Whitney U Test. O, P  Vol- cano plot of log2 fold changes of lncRNAs in HUVEC treated with differentially expressed genes within the VEGF-signaling acriflavine (ACF) cultured under normoxia (O) or hypoxia (P). Q pathway after CRISPR/Cas9-mediated knockout of the Relative expression of LINC00607 in HUVEC under basal (CTL) or lncRNA (Fig. 3D). This GSEA result was associated with Endothelial-to-mesenchymal transition (EndMT) conditions. n = 3, numerous VEGF-signaling genes that were mainly down- Unpaired t test. R, S Volcano plot of log2 fold changes of lncRNAs after EndMT versus unstimulated control in HUVEC (R) or pulmo- regulated upon knockout of LINC00607 (Fig. 3E–G). nary arterial endothelial cells (PAEC) (S). Error bars are defined as mean ± SD. *p < 0.05 LINC00607 depletion reduces the accessibility of ETS transcription factor binding sites removal of the transcriptional start site of LINC00607 (Fig. S2A). Successful knockout was confirmed on the In order to determine whether the effects of LINC00607 levels of both the DNA (Fig.  2E) and RNA (Fig.  2F, loss of function and differential gene expression were a G). As with siRNA-mediated knockdown, knockout consequence of altered chromatin accessibility, an assay of LINC00607 inhibited VEGF-A-induced sprouting for transposase-accessible chromatin with sequencing (Fig.  2H–J). As additional functional assays, scratch (ATAC-Seq) was performed (Fig. S3I). Comparison of wound, proliferation and apoptosis assays were performed ATAC-Seq and RNA-Seq for the multiple VEGF-signaling to determine migratory capacity and proliferation. These genes revealed a similar effect of LINC00607 knockout on assays showed the negative effect on endothelial func- chromatin accessibility of gene-linked enhancers (as anno- tion as a consequence of LINC00607 loss (Fig.  2K, L, tated by EpiRegio [5]) and gene expression (Fig. 4A). This S2B, C). In order to study the mechanistic function of suggested that the lncRNA might directly influence the LINC00607, the RNA was overexpressed in knockout transcription of these genes by modulating the accessibil- cells. In the case of cis-action, i.e. local action of the ity of transcription factor binding sites. To investigate the RNA at the transcription site or a general transcriptional underlying mechanism of the profound changes in chro- importance of the gene locus, such a rescue experiment matin state and transcription, a transcription factor bind- should not restore function. However, transfection of ing analysis was performed using HOMER [25]. DNA- LINC00607 into LINC00607 knockout cells restored a motif enrichment analysis showed the basic region/leucine 1 3 5 Page 12 of 21 Basic Research in Cardiology (2023) 118:5 zipper motif (bZIP) to be more accessible under LINC00607 mediating LINC00607-dependent transcription (Fig. 4D). knockout (Fig. 4B). Interestingly, ERG (ETS Transcription Even though some of the ETS family members were dif- Factor ERG) and ETV2 (ETS Variant Transcription Fac- ferentially expressed in response to LINC00607 knockout, tor 2) motifs were identified as being less accessible after the expression of ERG remained unchanged (Fig. 4E). These LINC00607 knockout (Fig. 4C). ERG and ETV2 are both data indicate that LINC00607-dependent gene expression members of the ETS transcription factor family, recognizing is likely mediated through changes in ERG-induced gene the core consensus motif GGA(A/T) [69], and are highly expression, resulting from LINC00607-directed changes in important for endothelial gene expression in particular [51]. transcription factor binding site accessibility. Expression changes of a transcription factor might impact on the gene expression of its target gene. To exclude that LINC00607 maintains endothelial‑specific the differential gene expression in response to LINC00607 chromatin states through interaction with BRG1 loss of function was not caused through differential expres- sion of the transcription factors themselves, the expression The changes in chromatin accessibility and to ERG-binding of ETS family transcription factors was determined. As sites would naturally be caused by chromatin remodeling. determined from RNA-Seq, ERG was highly expressed in We have previously shown that an important chromatin normal HUVEC, whereas ETV2 expression was low. We remodeling protein interacting with lncRNAs in endothe- therefore selected ERG as a candidate transcription factor lial cells is the SWI/SNF member BRG1 [40]. Importantly, 1 3 Basic Research in Cardiology (2023) 118:5 Page 13 of 21 5 ◂Fig. 2 CRISPR/Cas9 KO and siRNA-knockdown reveal that and BRG1 knockout in HUVEC. BRG1 binding sites were LINC00607 is important for normal EC function. A RT-qPCR of located near the transcription start sites of many genes and, LINC00607 after siRNA-based knockdown for 48  h of LINC00607 upon BRG1 knockout, these sites were lost confirming the (607). Scrambled siRNA served as negative control (CTL). n = 6. specificity of BRG1 binding (Fig.  5D). To reveal the role Mann–Whitney test. B Spheroid outgrowth assay after siRNA-based knockdown of LINC00607 (607). Scrambled siRNA served as nega- of LINC00607 for BRG1 binding, differentially expressed tive control (CTL). Cells treated with or without VEGF-A (16 h) are genes identified by RNA-Seq were overlapped with differ - shown. C Quantification of cumulative sprout length from spheroid ential ATAC-Seq peaks having proximity to the transcrip- outgrowth assay shown in B, n = 28–30, two-way ANOVA. D Quan- tional start site and with genes BRG1 binding sites were tification of the ratio of cumulative outgrowth length and respective spheroid diameter from the spheroid outgrowth assay shown in B; identified by CUT&RUN. Surprisingly, there was a strong n = 28–30, two-way ANOVA. E PCR of Genomic DNA after lenti- overlap between LINC00607 differentially regulated genes viral CRISPR/Cas9-mediated knockout (KO) of LINC00607. Three and BRG1 target genes (Fig. 5E). BRG1-associated genes different batches (1–3) of HUVEC are shown. Non-targeting control exhibited a stronger and more significant decrease in expres- gRNAs (NTC) served as negative control. GAPDH served as house- keeping gene. F RT-qPCR of LINC00607 after CRISPR/Cas9-medi- sion after LINC00607 knockout compared to non-BRG1- ated knockout (KO) and control (NTC), n = 3. Unpaired t test. G IGV associated genes (Fig. 5F, G). Since the motif for the ERG genome tracks of RNA-Seq of the LINC00607 locus in HUVEC with transcription factor was strongly enriched in genes down- or without CRISPR/Cas9-mediated knockout of LINC00607 (KO) regulated after LINC00607 knockout, the described gene and control (NTC). H Spheroid outgrowth assay with LINC00607 knockout (KO) and control (CTL) in HUVEC. NTC served as nega- sets were further overlapped with a publicly available ERG tive control. Cells treated with or without VEGF-A are shown. I Chromatin immunoprecipitation-Seq (ChIP-Seq) [31] from Quantification of cumulative sprout length from spheroid outgrowth HUVEC. Importantly, almost all (1372 out of 1445) of the assay shown in H, n = 28–30, two-way ANOVA. J Quantification differentially accessible genes after LINC00607 knockout of the ratio of cumulative outgrowth length and respective spheroid diameter from spheroid outgrowth assay shown in H; n = 28–30, two- overlapping with BRG1 CUT&RUN binding sites were way ANOVA. K Scratch wound assay of LINC00607 KO and NTC shared with genes ERG binds close to (Fig. 5E). control cells (CTL). Representative images after 0  h and 4  h after To inspect these global associations in more detail, we scratch (blue line) are shown. L Quantification of relative wound clo- checked a handful of genes highly important in endothelial sure in LINC00607 KO and control (NTC) in HUVEC. n = 3, area under the curve (AUC) p < 0.0001 (two-tailed t test). M Spheroid out- cells manually: VWF (von Willebrand factor), SGK1 (Serum/ growth assay of HUVEC after CRISPR/Cas9-mediated LINC00607 Glucocorticoid Regulated Kinase 1), TSPAN12 (Tetraspanin KO or non target control (NTC). Cells treated under VEGF-A con- 12) and KDR (Kinase Insert Domain Receptor) (Fig. 5H). ditions for 16  h with/ without LINC00607 overexpression (OE) are SGK1, TSPAN12, VWF and KDR were among the strong- shown. Empty vector transfection served as control (CTL). N Quan- tification of the ratio of cumulative outgrowth length and respec- est differentially expressed genes after LINC00607 knock - tive spheroid diameter from spheroid outgrowth assay shown in M. out (Fig.  3A). Of these genes KDR, TSPAN12 and VWF n = 26–29, two-way ANOVA. O LINC00607 KO and control cells represented genes listed in the VEGF signaling pathway (NTC) after in vivo matrigel plug assay in SCID mice. HUVEC were (Fig. 3E), which we described in this study to be strongly embedded in matrigel, stained with Vybrant dil (red) and injected. Isolectin GS-b4 Alexa 647 conjugated stained vessels (green). Images affected by LINC00607 perturbation. Importantly, all these were taken by light sheet microscopy 21 days after injection. Scale genes contained a BRG1 and ERG signature at their tran- bar indicates 100  µm. Representative pictures are shown. P Quanti- scriptional start site (Fig.  5H) which indicates that many fication of cells per plug integrated into the newly formed vascular LINC00607-dependent genes are also BRG1 and ERG target network shown in 2O. n = 5–6. Mann Whitney U test. Error bars are defined as mean ± SD. *p < 0.05 genes. LINC00607 is required for the stable expression of these genes. RNA immunoprecipitation with antibodies against BRG1 yielded LINC00607 as an interaction partner of BRG1 Discussion (Fig. 5A). The interaction of the lncRNA with BRG1 was specific: in contrast to β-Actin mRNA, LINC00607 was In the present study, we identified LINC00607 to be spe- not pulled down by the non-primary antibody control IgG; cifically expressed in EC and to be important for vascular RNase A treatment was able to abolish the signal (Fig. 5A). sprouting and ERG-dependent gene expression through LINC00607 knockout did not affect BRG1 expression, which BRG1. Although already constitutively highly expressed indicates LINC00607 might influence BRG1 DNA binding in EC, LINC00607 itself was upregulated by hypoxia and activity. To test this, a lentiviral CRISPR/Cas9 knockout EndMT. Through RNA- and ATAC-Seq we identified of BRG1 in HUVEC was generated (Fig.  5B, C). Cleav- LINC00607 as a lncRNA important for central pathways age Under Targets & Release Using Nuclease (CUT&RUN) of endothelial cells, in particular for VEGF signaling. sequencing, a method to determine high-resolution map- After LentiCRISPR-mediated knockout of LINC00607, ping of DNA binding sites [64], was performed using endothelial cells exhibited an impaired response to VEGF- anti-BRG1 antibodies after both non-targeting control A in respect to vascular sprouting and a reduced ability to 1 3 5 Page 14 of 21 Basic Research in Cardiology (2023) 118:5 Fig. 3 RNA- and ATAC-Seq reveal that LINC00607 maintains for the Gene Ontology biological process (GOBP) Vascular Endothe- endothelial gene expression. A Heatmap of the top 50 differen- lial Growth Factor Signaling Pathway. E Heat map of VEGF-signal- tially expressed genes as determined by RNA-Seq with (KO) or ing pathway genes and their expression differences after RNA-Seq. without (NTC) CRISPR/Cas9-mediated knockout of LINC00607. Z-score represents up- (red, positive value) or down-regulated (blue, Three different batches of HUVEC are shown. Genes shown have a negative values) genes. F Examples of significantly downregulated padj < 0.05, and a log2 fold change greater than ± 0.585. Z-score rep- genes after LINC00607 knockout. IGV genome tracks of the FLT1 resents up- (red, positive value) or down-regulated (blue, negative and FLT4 locus. Shown are RNA-Seq reads in LINC00607 knock- values) genes. B Volcano plot of RNA-Seq showing the log2 fold out (KO, red) and control (NTC, blue). Tracks of three replicates changes (KO vs. NTC) of all genes expressed against their p-adjusted are overlaid. G Examples of significantly upregulated genes after value (p-adj). C Numbers of genes from different gene classes sig- LINC00607 knockout. IGV genome tracks of the VEGFC and SPRY2 nificantly altered by LINC00607 KO vs. NTC HUVEC determined by locus. Shown are RNA-Seq reads in LINC00607 knockout (KO, RNA-Seq. D Gene Set Enrichment Analysis (GSEA) of significantly red) and control (NTC, blue). Tracks of three replicates are overlaid. altered genes showing an enrichment score and signal to noise ratio Genomic coordinates correspond to hg38 1 3 Basic Research in Cardiology (2023) 118:5 Page 15 of 21 5 Fig. 4 ERG drives LINC00607-associated gene expression. A Over- enriched (B) or decreased-enriched (C) transcription factor motifs are lap of ATAC-Seq (enhancer accessibility) and RNA-Seq (gene shown. D Gene read counts of different transcription factors in NTC- expression) signals after knockout of LINC00607 in HUVEC. treated HUVEC (from RNA-Seq), n = 3. E Mean log2 fold change Indicated are genes involved in the VEGF-signaling pathway. B, (FC) of different transcription factors in the RNA-Seq comparing C  HOMER DNA-motif enrichment analysis of differential acces- LINC00607 KO and NTC control sible peaks (LINC00607 KO vs. NTC). Five most highly increased- integrate into the vascular network of SCID mice. Mecha- system and other tissues. For example, EVF2 has been nistically, the trans-acting lncRNA interacts with the chro- shown to inhibit the ATPase activity of BRG1 [12]. Addi- matin remodeling protein BRG1 in order to maintain chro- tionally, lncRNAs can stabilize or destabilize BRG1 inter- matin states for ERG-dependent transcription. Thereby, action with other proteins, as in the case of MALAT1 pro- LINC00607 preserves endothelial gene expression pat- moting BRG1 interaction with HDAC9 [47]. lncRNAs can terns, which are essential for angiogenesis. As exemplified also affect BRG1 gene targeting. For example, we have by lncRNA SAIL (scaffold attachment factor B interacting previously shown that the lncRNA SMANTIS guides BRG1 lncRNA), which has been reported to impair the transcrip- to specific genes related to endothelial lineage specifica- tion of fibrosis-related genes in human cardiac fibroblasts tion [40]. Our present results suggest that numerous target [50], such cardiovascular-specific gene expression control genes of LINC00607, BRG1 and ERG overlap arguing that mechanisms by lncRNAs could potentially be a common- the LINC00607 could potentially facilitate BRG1 binding type of mechanism. to genes linked to the endothelial phenotype. Recent stud- In terms of transcriptional control, lncRNAs can either ies highlight the importance of constant SWI/SNF remod- act in cis at nearby genes, or in trans genome wide [60]. eling to maintain a stable open chromatin state [26, 62]. Overexpression of LINC00607 restored endothelial function Our present observations suggest that LINC00607 provides after LINC00607 knockout, demonstrating that LINC00607 a specific link for ERG securing BRG1 binding to genes acts in trans rather than in cis, because the effect of locus- maintaining the endothelial phenotype. This specific context disruption by CRISPR/Cas9 gene editing was overruled by would explain why LINC00607 is so highly expressed in the plasmid-based overexpression of LINC00607. endothelial cells. Importantly, LINC00607 interacts with the chromatin Indeed, we uncovered a large overlap between genes remodeling protein BRG1. Several lncRNAs have been with altered chromatin state and differential expression linked to BRG1, inu fl encing its activity in the cardiovascular after LINC00607 knockout, and genes with binding sites 1 3 5 Page 16 of 21 Basic Research in Cardiology (2023) 118:5 Fig. 5 LINC00607 functions through interaction with the chroma- ChIP-Seq peak. F Median log2 fold change (FC) of differentially tin remodeler BRG1. A RNA-immunoprecipitation with antibod- expressed genes from LINC00607 knockout that are located near a ies against BRG1 with and without RNase A digestion, followed by differential ATAC-Seq peak of LINC00607 knockout and also found RT-qPCR of LINC00607 and β-Actin. IgG served as a non-primary near a BRG1 CUT&RUN peak (BRG1) or not (Non). G Median antibody control. n = 5. One-Way ANOVA with Bonferroni post p-adjusted value of differentially expressed genes from LINC00607 hoc test. B Western blot analysis with antibodies against BRG1 and knockout that are located near a differential ATAC-Seq peak of β-Actin of control (NTC) or BRG1 knockout HUVEC. C RT-qPCR LINC00607 knockout and also found near a BRG1 CUT&RUN of BRG1 after CRISPR/Cas9-mediated knockout, n = 3. Paired peak (BRG1) or not (Non). H Genome tracks of ATAC-Seq, t-test. D Chromatin accessibility heat map of differential peaks from RNA-Seq, BRG1 CUT&RUN and ERG ChIP-Seq. Loci of VWF, BRG1 CUT&RUN of control (NTC) and BRG1 knockout (KO) SGK1, TSPAN12 and KDR are shown. ATAC-Seq and RNA-Seq of HUVEC. Binding regions center-aligned to the transcription start LINC00607 KO (red) and NTC (blue) are shown. Tracks of replicates sites (TSS) ± 0.5 kb are shown. E Venn diagram showing the overlap were overlaid. CUT&RUN with anti-BRG1 antibodies of NTC or of genes located near a differential ATAC-Seq peak of LINC00607 BRG1 KO HUVEC are shown in black. ChIP-Seq of ERG is shown in knockout, genes near a BRG1 CUT&RUN peak and genes near ERG green. Error bars are defined as mean ± SD. *p < 0.05 for the ERG transcription factor. ERG belongs to the ETS role in the proposed mechanism of transcriptional control. transcription factor family, which act as key regulators of the Our findings advocate for LINC00607 as one link between majority of endothelial genes, as the ETS recognition motif BRG1-mediated stabilization of chromatin states and ERG can be found in promotors of many endothelial genes [68]. target gene expression in healthy endothelium. This shows the importance of LINC00607 for the expression The endothelial expression of LINC00607 was increased control of ERG-regulated genes through its interaction with by hypoxia, EndMT, and endothelial dysfunction as induced BRG1. In this context, it is interesting to note that SWI/ by TNFα and high glucose [13]. Under these stimuli, the SNF is required to maintain open chromatin [26, 62]. BRG1, upregulation of LINC00607 matches the transcription factor being the core member of SWI/SNF, could have a central binding motifs identified in the promotor analysis. Hypoxia 1 3 Basic Research in Cardiology (2023) 118:5 Page 17 of 21 5 Funding Open Access funding enabled and organized by Projekt signalling through VEGF is an important trigger for angio- DEAL. This work was supported by the Goethe University Frank- genic specification of endothelial cells [58] as well as a key furt am Main, the DFG excellence cluster “Cardiopulmonary Insti- mechanism contributing to chronic and acute cardiovascu- tute (CPI)” EXS2026. Furthermore, it was funded by the Deutsche lar diseases [39]. Through the upregulation of LINC00607 Forschungsgemeinschaft (DFG, German Research Foundation)— Project-ID 403584255—TRR 267 to TP A04, TP A06, TP B04, TP under hypoxic conditions, the pro-angiogenic endothelial B07, TP Z02, TP Z03. The project was also supported by the BHF/ phenotype could potentially be secured by tightening the DHF/DZHK grant “Exploiting endothelial long non-coding RNAs interaction of LINC00607 with BRG1. Specifically to HIF1- to promote regenerative angiogenesis in the damaged myocardium” controlled LINC00607 expression, we found that DMOG- (ReGenLnc) as well as the Dr. Rolf Schwiete-Stiftung. SIB is supported by the Emmy Noether Programme BI 2163/1-1 and the Johanna Quandt dependent upregulation and acriflavine-mediated HIF inhibi- Young Academy at Goethe University. FJM is supported in part by tion altered the expression of the lncRNA. This in particular Merit Review Award #BX001729 from the United States Department illustrates the close interaction of hypoxia-signaling, angio- of Veterans Affairs, Biomedical Laboratory Research and Develop- genesis and LINC00607. ment Service. The fact that EndMT induction by TGF-β2 and IL-1β also Declarations increased the expression of LINC00607 points towards a role for LINC00607 in expression control beyond the endothe- Conflict of interest The authors have declared that no conflict of inter - lial phenotype. Potentially LINC00607 guides BRG1 to est exists. genes involved in EndMT. In line with this, LINC00607 is Open Access This article is licensed under a Creative Commons Attri- also expressed in certain malignant cells. In this context, bution 4.0 International License, which permits use, sharing, adapta- LINC00607 was upregulated in doxorubicin-resistant thyroid tion, distribution and reproduction in any medium or format, as long cancer cells [44]. Furthermore, LINC00607 was described as you give appropriate credit to the original author(s) and the source, to be required for tumor proliferation of osteosarcoma cells provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are [75] and was downregulated in lung adenocarcinoma [74]. included in the article's Creative Commons licence, unless indicated Linking these findings to the present study, it could be specu- otherwise in a credit line to the material. If material is not included in lated that under basal conditions, LINC00607 guides BRG1 the article's Creative Commons licence and your intended use is not to ERG target genes and during endothelial dysfunction to permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a pro-angiogenic genes to maintain an open and accessible copy of this licence, visit http://cr eativ ecommons. or g/licen ses/ b y/4.0/ . chromatin state for ERG. Since genome-wide association studies identified many risk loci globally in non-coding regions, transcriptome-wide association studies could be References important in future to identify gene-trait associations [45] and thereby the importance of LINC00607 as a candidate 1. (2019) UniProt: a worldwide hub of protein knowledge. Nucleic gene in other disease-relevant tissues. Acids Res 47:D506-D515. https://doi. or g/10. 1093/ nar/ gky10 49 Taken together, these findings suggest that LINC00607 2. Abbas N, Perbellini F, Thum T (2020) Non-coding RNAs: emerging players in cardiomyocyte proliferation and cardiac is involved in securing endothelial BRG1 and ERG- regeneration. Basic Res Cardiol 115:52. https://d oi.o rg/1 0.1 007/ dependent target gene expression, as well as appropri- s00395- 020- 0816-0 ate responses to stress and cardiovascular diseases. The 3. Allhoff M, Seré K, Pires FJ, Zenke M, Costa GI (2016) Dif- function of LINC00607 in cardiovascular diseases with ferential peak calling of ChIP-seq signals with replicates with THOR. Nucleic Acids Res 44:e153. https://doi. or g/10. 1093/ nar/ hypoxia signaling needs more investigation as LINC00607 gkw680 might be a potential therapeutic target or clinical marker. 4. Andrews S. (2010) FastQC: a quality control tool for high throughput sequence data Supplementary Information The online version contains supplemen- 5. Baumgarten N, Hecker D, Karunanithi S, Schmidt F, List M, tary material available at https://doi. or g/10. 1007/ s00395- 023- 00978-3 . Schulz MH (2020) EpiRegio: analysis and retrieval of regula- tory elements linked to genes. Nucleic Acids Res 48:W193– Acknowledgements We thank Katalin Pálfi and Manuela Spaeth for W199. https:// doi. org/ 10. 1093/ nar/ gkaa3 82 excellent technical assistance and Dominik Fuhrmann for help with 6. Bibli S-I, Hu J, Looso M, Weigert A, Ratiu C, Wittig J, Dreko- the hypoxia chamber. lia MK, Tombor L, Randriamboavonjy V, Leisegang MS, Goy- mann P, Delgado Lagos F, Fisslthaler B, Zukunft S, Kyselova Author contributions FB, JAO, TW, SK, FJM, RPB and MSL designed A, Justo AFO, Heidler J, Tsilimigras D, Brandes RP, Dimme- the experiments. FB, JAO, TW, SG, JIP, ML, DR, GB, MDP, ZSM, ler S, Papapetropoulos A, Knapp S, Offermanns S, Wittig I, TL, SS, SH, SK, JK, FJM and MSL performed the experiments. FB, Nishimura SL, Sigala F, Fleming I (2021) Mapping the endothe- JAO, TW, SG, GB, ZSM, and MSL analyzed the data. FB, TW, SG and lial cell S-sulfhydrome highlights the crucial role of integrin MHS performed bioinformatics. KHP, RB, MM, JK, SSP, SIB, RH, sulfhydration in vascular function. Circulation 143:935–948. AHB, RAB and IW helped with research design, material accessibil- https:// doi. org/ 10. 1161/ CIRCU LATIO NAHA. 120. 051877 ity and advice. FB, RBP and MSL wrote the manuscript. All authors interpreted the data and approved the manuscript. 1 3 5 Page 18 of 21 Basic Research in Cardiology (2023) 118:5 7. Bolger AM, Lohse M, Usadel B (2014) Trimmomatic: a flexible H, Fagiolini M, Fairbairn L, Fang H, Farach-Carson MC, Faulkner trimmer for Illumina sequence data. Bioinformatics 30:2114– GJ, Favorov AV, Fisher ME, Frith MC, Fujita R, Fukuda S, Fur- 2120. https:// doi. org/ 10. 1093/ bioin forma tics/ btu170 lanello C, Furino M, Furusawa J-I, Geijtenbeek TB, Gibson AP, 8. Broad Institute (2019) Picard toolkit Gingeras T, Goldowitz D, Gough J, Guhl S, Guler R, Gustincich 9. Buenrostro JD, Wu B, Chang HY, Greenleaf WJ (2015) ATAC- S, Ha TJ, Hamaguchi M, Hara M, Harbers M, Harshbarger J, seq: a method for assaying chromatin accessibility genome- Hasegawa A, Hasegawa Y, Hashimoto T, Herlyn M, Hitchens KJ, wide. Curr Protoc Mol Biol. https:// doi. or g/ 10. 1002/ 04711 Ho Sui SJ, Hofmann OM, Hoof I, Hori F, Huminiecki L, Iida K, 42727. Mb212 9s109 Ikawa T, Jankovic BR, Jia H, Joshi A, Jurman G, Kaczkowski B, 10. Bultman S, Gebuhr T, Yee D, La Mantia C, Nicholson J, Gilliam Kai C, Kaida K, Kaiho A, Kajiyama K, Kanamori-Katayama M, A, Randazzo F, Metzger D, Chambon P, Crabtree G, Magnuson Kasianov AS, Kasukawa T, Katayama S, Kato S, Kawaguchi S, T (2000) A Brg1 null mutation in the mouse reveals functional Kawamoto H, Kawamura YI, Kawashima T, Kempfle JS, Kenna differences among mammalian SWI/SNF complexes. Mol Cell TJ, Kere J, Khachigian LM, Kitamura T, Klinken SP, Knox AJ, 6:1287–1295. https:// doi. org/ 10. 1016/ s1097- 2765(00) 00127-1 Kojima M, Kojima S, Kondo N, Koseki H, Koyasu S, Krampitz 11. Cairns BR (2007) Chromatin remodeling: insights and intrigue S, Kubosaki A, Kwon AT, Laros JFJ, Lee W, Lennartsson A, Li from single-molecule studies. Nat Struct Mol Biol 14:989–996. K, Lilje B, Lipovich L, Mackay-Sim A, Manabe R-I, Mar JC, https:// doi. org/ 10. 1038/ nsmb1 333 Marchand B, Mathelier A, Mejhert N, Meynert A, Mizuno Y, de 12. Cajigas I, Leib DE, Cochrane J, Luo H, Swyter KR, Chen Lima-Morais DA, Morikawa H, Morimoto M, Moro K, Mota- S, Clark BS, Thompson J, Yates JR, Kingston RE, Kohtz JD kis E, Motohashi H, Mummery CL, Murata M, Nagao-Sato S, (2015) Evf2 lncRNA/BRG1/DLX1 interactions reveal RNA- Nakachi Y, Nakahara F, Nakamura T, Nakamura Y, Nakazato K, dependent inhibition of chromatin remodeling. Development van Nimwegen E, Ninomiya N, Nishiyori H, Noma S, Noazaki 142:2641–2652. https:// doi. org/ 10. 1242/ dev. 126318 T, Ogishima S, Ohkura N, Ohimiya H, Ohno H, Ohshima M, 13. Calandrelli R, Xu L, Luo Y, Wu W, Fan X, Nguyen T, Chen C-J, Okada-Hatakeyama M, Okazaki Y, Orlando V, Ovchinnikov DA, Sriram K, Tang X, Burns AB, Natarajan R, Chen ZB, Zhong Pain A, Passier R, Patrikakis M, Persson H, Piazza S, Prendergast S (2020) Stress-induced RNA-chromatin interactions promote JGD, Rackham OJL, Ramilowski JA, Rashid M, Ravasi T, Rizzu endothelial dysfunction. Nat Commun 11:5211. https://doi. or g/ P, Roncador M, Roy S, Rye MB, Saijyo E, Sajantila A, Saka A, 10. 1038/ s41467- 020- 18957-w Sakaguchi S, Sakai M, Sato H, Savvi S, Saxena A, Schneider 14. Deanfield JE, Halcox JP, Rabelink TJ (2007) Endothelial func- C, Schultes EA, Schulze-Tanzil GG, Schwegmann A, Sengstag tion and dysfunction: testing and clinical relevance. Circulation T, Sheng G, Shimoji H, Shimoni Y, Shin JW, Simon C, Sugiy- 115:1285–1295. https://d oi.o rg/1 0.1 161/C IRCUL ATION AHA. ama D, Sugiyama T, Suzuki M, Suzuki N, Swoboda RK, Hoen 106. 652859 PAC, Tagami M, Takahashi N, Takai J, Tanaka H, Tatsukawa H, 15. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha Tatum Z, Thompson M, Toyodo H, Toyoda T, Valen E, van de S, Batut P, Chaisson M, Gingeras TR (2013) STAR: ultrafast Wetering M, van den Berg LM, Verado R, Vijayan D, Vorontsov universal RNA-seq aligner. Bioinformatics 29:15–21. https:// IE, Wasserman WW, Watanabe S, Wells CA, Winteringham LN, doi. org/ 10. 1093/ bioin forma tics/ bts635 Wolvetang E, Wood EJ, Yamaguchi Y, Yamamoto M, Yoneda M, 16. Ertürk A, Becker K, Jährling N, Mauch CP, Hojer CD, Egen JG, Yonekura Y, Yoshida S, Zabierowski SE, Zhang PG, Zhao X, Zuc- Hellal F, Bradke F, Sheng M, Dodt H-U (2012) Three-dimen- chelli S, Summers KM, Suzuki H, Daub CO, Kawai J, Heutink P, sional imaging of solvent-cleared organs using 3DISCO. Nat Hide W, Freeman TC, Lenhard B, Bajic VB, Taylor MS, Makeev Protoc 7:1983–1995. https:// doi. org/ 10. 1038/ nprot. 2012. 119 VJ, Sandelin A, Hume DA, Carninci P, Hayashizaki Y (2014) A 17. Krueger F, James F, Ewels P, Afyounian E, Schuster-Boeckler promoter-level mammalian expression atlas. Nature 507:462–470. B (2021) FelixKrueger/TrimGalore: v0.6.7 - DOI via Zenodo. https:// doi. org/ 10. 1038/ natur e13182 Zenodo 22. Haeussler M, Schönig K, Eckert H, Eschstruth A, Mianné J, 18. Feng J, Liu T, Qin B, Zhang Y, Liu XS (2012) Identifying ChIP- Renaud J-B, Schneider-Maunoury S, Shkumatava A, Teboul L, seq enrichment using MACS. Nat Protoc 7:1728–1740. https:// Kent J, Joly J-S, Concordet J-P (2016) Evaluation of o-t ff arget and doi. org/ 10. 1038/ nprot. 2012. 101 on-target scoring algorithms and integration into the guide RNA 19. Ferrara N, Mass RD, Campa C, Kim R (2007) Targeting VEGF- selection tool CRISPOR. Genome Biol 17:148. https://doi. or g/10. A to treat cancer and age-related macular degeneration. Annu 1186/ s13059- 016- 1012-2 Rev Med 58:491–504. https:// doi. org/ 10. 1146/ annur ev. med. 58. 23. Han P, Li W, Lin C-H, Yang J, Shang C, Nuernberg ST, Jin KK, 061705. 145635 Xu W, Lin C-Y, Lin C-J, Xiong Y, Chien H, Zhou B, Ashley E, 20. Fiedler J, Breckwoldt K, Remmele CW, Hartmann D, Dittrich M, Bernstein D, Chen P-S, Chen H-SV, Quertermous T, Chang C-P Pfanne A, Just A, Xiao K, Kunz M, Müller T, Hansen A, Geffers (2014) A long noncoding RNA protects the heart from patho- R, Dandekar T, Eschenhagen T, Thum T (2015) Development of logical hypertrophy. Nature 514:102–106. https://doi. or g/10. 1038/ long noncoding RNA-based strategies to modulate tissue vascu-natur e13596 larization. J Am Coll Cardiol 66:2005–2015. https:// doi. org/ 10. 24. Hathaway CA, Heistad DD, Piegors DJ, Miller FJ (2002) Regres- 1016/j. jacc. 2015. 07. 081 sion of atherosclerosis in monkeys reduces vascular superoxide 21. Forrest ARR, Kawaji H, Rehli M, Baillie JK, de Hoon MJL, levels. Circ Res 90:277–283. https:// doi. or g/ 10. 1161/ hh0302. Haberle V, Lassmann T, Kulakovskiy IV, Lizio M, Itoh M, 104724 Andersson R, Mungall CJ, Meehan TF, Schmeier S, Bertin N, 25. Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, Cheng Jørgensen M, Dimont E, Arner E, Schmidl C, Schaefer U, Med- JX, Murre C, Singh H, Glass CK (2010) Simple combinations vedeva YA, Plessy C, Vitezic M, Severin J, Semple CA, Ishizu Y, of lineage-determining transcription factors prime cis-regulatory Young RS, Francescatto M, Alam I, Albanese D, Altschuler GM, elements required for macrophage and B cell identities. Mol Cell Arakawa T, Archer JAC, Arner P, Babina M, Rennie S, Balwierz 38:576–589. https:// doi. org/ 10. 1016/j. molcel. 2010. 05. 004 PJ, Beckhouse AG, Pradhan-Bhatt S, Blake JA, Blumenthal A, 26. Iurlaro M, Stadler MB, Masoni F, Jagani Z, Galli GG, Schübeler Bodega B, Bonetti A, Briggs J, Brombacher F, Burroughs AM, D (2021) Mammalian SWI/SNF continuously restores local acces- Califano A, Cannistraci CV, Carbajo D, Chen Y, Chierici M, Ciani sibility to chromatin. Nat Genet 53:279–287. https:// doi. org/ 10. 1038/ s41588- 020- 00768-w Y, Clevers HC, Dalla E, Davis CA, Detmar M, Diehl AD, Dohi T, Drabløs F, Edge ASB, Edinger M, Ekwall K, Endoh M, Enomoto 1 3 Basic Research in Cardiology (2023) 118:5 Page 19 of 21 5 27. Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell function. Circulation 136:65–79. https:// doi. org/ 10. 1161/ cir cu SJ, von Kriegsheim A, Hebestreit HF, Mukherji M, Schofield latio naha. 116. 026991 CJ, Maxwell PH, Pugh CW, Ratcliffe PJ (2001) Targeting of 41. Leisegang MS, Bibli S-I, Günther S, Pflüger-Müller B, Oo JA, HIF-alpha to the von Hippel–Lindau ubiquitylation complex by Höper C, Seredinski S, Yekelchyk M, Schmitz-Rixen T, Schür- O2-regulated prolyl hydroxylation. Science 292:468–472. https:// mann C, Hu J, Looso M, Sigala F, Boon RA, Fleming I, Brandes doi. org/ 10. 1126/ scien ce. 10597 96 RP (2019) Pleiotropic effects of laminar flow and statins depend 28. Jégu T, Blum R, Cochrane JC, Yang L, Wang C-Y, Gilles M-E, on the Krüppel-like factor-induced lncRNA MANTIS. Eur Heart Colognori D, Szanto A, Marr SK, Kingston RE, Lee JT (2019) J 40:2523–2533. https:// doi. org/ 10. 1093/ eurhe artj/ ehz393 Xist RNA antagonizes the SWI/SNF chromatin remodeler BRG1 42. Li B, Carey M, Workman JL (2007) The role of chromatin during on the inactive X chromosome. Nat Struct Mol Biol 26:96–109. transcription. Cell 128:707–719. https:// doi. org/ 10. 1016/j. cell. https:// doi. org/ 10. 1038/ s41594- 018- 0176-82007. 01. 015 29. Johnson KE, Wilgus TA (2014) Vascular endothelial growth factor 43. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, and angiogenesis in the regulation of cutaneous wound repair. Adv Marth G, Abecasis G, Durbin R (2009) The sequence alignment/ Wound Care (New Rochelle) 3:647–661. https://doi. or g/10. 1089/ map format and SAMtools. Bioinformatics 25:2078–2079. https:// wound. 2013. 0517doi. org/ 10. 1093/ bioin forma tics/ btp352 30. Kadam S, Emerson BM (2003) Transcriptional specificity of 44. Li L, Gao Z, Zhao L, Ren P, Shen H (2021) Long non-coding human SWI/SNF BRG1 and BRM chromatin remodeling com- RNA LINC00607 silencing exerts antioncogenic effects on thy - plexes. Mol Cell 11:377–389. https:// doi. or g/ 10. 1016/ S1097- roid cancer through the CASP9 Promoter methylation. J Cell Mol 2765(03) 00034-0 Med 25:7608–7620. https:// doi. org/ 10. 1111/ jcmm. 16265 31. Kalna V, Yang Y, Peghaire CR, Frudd K, Hannah R, Shah AV, 45. Li L, Chen Z, von Scheidt M, Li S, Steiner A, Güldener U, Koplev Osuna Almagro L, Boyle JJ, Göttgens B, Ferrer J, Randi AM, S, Ma A, Hao K, Pan C, Lusis AJ, Pang S, Kessler T, Ermel R, Birdsey GM (2019) The transcription factor ERG regulates super- Sukhavasi K, Ruusalepp A, Gagneur J, Erdmann J, Kovacic JC, enhancers associated with an endothelial-specific gene expres- Björkegren JLM, Schunkert H (2022) Transcriptome-wide asso- sion program. Circ Res 124:1337–1349. https:// doi. org/ 10. 1161/ ciation study of coronary artery disease identifies novel suscep- CIRCR ESAHA. 118. 313788 tibility genes. Basic Res Cardiol 117:6. https:// doi. org/ 10. 1007/ 32. Kent WJ, Zweig AS, Barber G, Hinrichs AS, Karolchik D (2010) s00395- 022- 00917-8 BigWig and BigBed: enabling browsing of large distributed data- 46. Liao Y, Smyth GK, Shi W (2014) featureCounts: an efficient gen- sets. Bioinformatics 26:2204–2207. https://doi. or g/10. 1093/ bioin eral purpose program for assigning sequence reads to genomic forma tics/ btq351 features. Bioinformatics 30:923–930. https:// doi. or g/ 10. 1093/ 33. Korff T, Augustin HG (1998) Integration of endothelial cells in bioin forma tics/ btt656 multicellular spheroids prevents apoptosis and induces differen- 47. Lino Cardenas CL, Kessinger CW, Cheng Y, MacDonald C, tiation. J Cell Biol 143:1341–1352. https:// doi. org/ 10. 1083/ jcb. MacGillivray T, Ghoshhajra B, Huleihel L, Nuri S, Yeri AS, 143.5. 1341 Jaffer FA, Kaminski N, Ellinor P, Weintraub NL, Malhotra R, 34. Kovacic JC, Dimmeler S, Harvey RP, Finkel T, Aikawa E, Kren- Isselbacher EM, Lindsay ME (2018) An HDAC9-MALAT1- ning G, Baker AH (2019) Endothelial to mesenchymal transition BRG1 complex mediates smooth muscle dysfunction in thoracic in cardiovascular disease: JACC state-of-the-art review. J Am Coll aortic aneurysm. Nat Commun 9:1009. https:// doi. org/ 10. 1038/ Cardiol 73:190–209. https:// doi. org/ 10. 1016/j. jacc. 2018. 09. 089s41467- 018- 03394-7 35. Kulakovskiy IV, Vorontsov IE, Yevshin IS, Sharipov RN, Fedor- 48. Lizio M, Harshbarger J, Shimoji H, Severin J, Kasukawa T, Sahin ova AD, Rumynskiy EI, Medvedeva YA, Magana-Mora A, Bajic S, Abugessaisa I, Fukuda S, Hori F, Ishikawa-Kato S, Mungall CJ, VB, Papatsenko DA, Kolpakov FA, Makeev VJ (2018) HOCO- Arner E, Baillie JK, Bertin N, Bono H, de Hoon M, Diehl AD, MOCO: towards a complete collection of transcription factor Dimont E, Freeman TC, Fujieda K, Hide W, Kaliyaperumal R, binding models for human and mouse via large-scale ChIP-Seq Katayama T, Lassmann T, Meehan TF, Nishikata K, Ono H, Rehli analysis. Nucleic Acids Res 46:D252–D259. https:// doi. org/ 10. M, Sandelin A, Schultes EA, Hoen PAC, Tatum Z, Thompson M, 1093/ nar/ gkx11 06 Toyoda T, Wright DW, Daub CO, Itoh M, Carninci P, Hayashizaki 36. Laib AM, Bartol A, Alajati A, Korff T, Weber H, Augustin HG Y, Forrest ARR, Kawaji H (2015) Gateways to the FANTOM5 (2009) Spheroid-based human endothelial cell microvessel for- promoter level mammalian expression atlas. Genome Biol 16:22. mation in vivo. Nat Protoc 4:1202–1215. https://doi. or g/10. 1038/ https:// doi. org/ 10. 1186/ s13059- 014- 0560-6 nprot. 2009. 96 49. Love MI, Huber W, Anders S (2014) Moderated estimation of fold 37. Langmead B, Salzberg SL (2012) Fast gapped-read alignment change and dispersion for RNA-seq data with DESeq2. Genome with Bowtie 2. Nat Methods 9:357–359. https:// doi. org/ 10. 1038/ Biol 15:550. https:// doi. org/ 10. 1186/ s13059- 014- 0550-8 nmeth. 1923 50. Luo S, Zhang M, Wu H, Ding X, Li D, Dong X, Hu X, Su S, Shang 38. Langmead B, Trapnell C, Pop M, Salzberg SL (2009) Ultrafast W, Wu J, Xiao H, Yang W, Zhang Q, Zhang J, Lu Y, Pan Z (2021) and memory-efficient alignment of short DNA sequences to the SAIL: a new conserved anti-fibrotic lncRNA in the heart. Basic human genome. Genome Biol 10:R25. https:// doi. org/ 10. 1186/ Res Cardiol 116:15. https://doi. or g/10. 1007/ s00395- 021- 00854-y gb- 2009- 10-3- r25 51. Meadows SM, Myers CT, Krieg PA (2011) Regulation of endothe- 39. Lee JW, Ko J, Ju C, Eltzschig HK (2019) Hypoxia signaling in lial cell development by ETS transcription factors. Semin Cell human diseases and therapeutic targets. Exp Mol Med 51:1–13. Dev Biol 22:976–984. https://doi. or g/10. 1016/j. semcdb. 2011. 09. https:// doi. org/ 10. 1038/ s12276- 019- 0235-1 009 40. Leisegang MS, Fork C, Josipovic I, Richter FM, Preussner J, Hu 52. Monteiro JP, Bennett M, Rodor J, Caudrillier A, Ulitsky I, Baker J, Miller MJ, Epah J, Hofmann P, Günther S, Moll F, Valasarajan AH (2019) Endothelial function and dysfunction in the car- C, Heidler J, Ponomareva Y, Freiman TM, Maegdefessel L, Plate diovascular system: the long non-coding road. Cardiovasc Res KH, Mittelbronn M, Uchida S, Künne C, Stellos K, Schermuly 115:1692–1704. https:// doi. org/ 10. 1093/ cvr/ cvz154 RT, Weissmann N, Devraj K, Wittig I, Boon RA, Dimmeler S, 53. Monteiro JP, Rodor J, Caudrillier A, Scanlon JP, Spiroski A-M, Pullamsetti SS, Looso M, Miller FJ, Brandes RP (2017) Long Dudnakova T, Pflüger-Müller B, Shmakova A, von Kriegsheim noncoding RNA MANTIS facilitates endothelial angiogenic A, Deng L, Taylor RS, Wilson-Kanamori JR, Chen S-H, Stewart K, Thomson A, Mitić T, McClure JD, Iynikkel J, Hadoke PWF, 1 3 5 Page 20 of 21 Basic Research in Cardiology (2023) 118:5 Denby L, Bradshaw AC, Caruso P, Morrell NW, Kovacic JC, Ulit- generation web server for deep-sequencing data analysis. Nucleic sky I, Henderson NC, Caporali A, Leisegang MS, Brandes RP, Acids Res 44:W160–W165. https:// doi. org/ 10. 1093/ nar/ gkw257 Baker AH (2021) MIR503HG loss promotes endothelial-to-mes- 60. Rinn JL, Chang HY (2012) Genome regulation by long noncoding enchymal transition in vascular disease. Circ Res 128:1173–1190. RNAs. Annu Rev Biochem 81:145–166. https:// doi. org/ 10. 1146/ https:// doi. org/ 10. 1161/ CIRCR ESAHA. 120. 318124annur ev- bioch em- 051410- 092902 54. Mootha VK, Lindgren CM, Eriksson K-F, Subramanian A, Sihag 61. Sanjana NE, Shalem O, Zhang F (2014) Improved vectors and S, Lehar J, Puigserver P, Carlsson E, Ridderstråle M, Laurila E, genome-wide libraries for CRISPR screening. Nat Methods Houstis N, Daly MJ, Patterson N, Mesirov JP, Golub TR, Tamayo 11:783–784. https:// doi. org/ 10. 1038/ nmeth. 3047 P, Spiegelman B, Lander ES, Hirschhorn JN, Altshuler D, Groop 62. Schick S, Grosche S, Kohl KE, Drpic D, Jaeger MG, Marella NC, LC (2003) PGC-1alpha-responsive genes involved in oxidative Imrichova H, Lin J-MG, Hofstätter G, Schuster M, Rendeiro AF, phosphorylation are coordinately downregulated in human dia- Koren A, Petronczki M, Bock C, Müller AC, Winter GE, Kubicek betes. Nat Genet 34:267–273. https:// doi. org/ 10. 1038/ ng1180 S (2021) Acute BAF perturbation causes immediate changes in 55. Noguchi S, Arakawa T, Fukuda S, Furuno M, Hasegawa A, Hori chromatin accessibility. Nat Genet 53:269–278. https:// doi. org/ F, Ishikawa-Kato S, Kaida K, Kaiho A, Kanamori-Katayama M, 10. 1038/ s41588- 021- 00777-3 Kawashima T, Kojima M, Kubosaki A, Manabe R-I, Murata M, 63. Seredinski S, Boos F, Günther S, Oo JA, Warwick T, Izquierdo Nagao-Sato S, Nakazato K, Ninomiya N, Nishiyori-Sueki H, Ponce J, Lillich FF, Proschak E, Knapp S, Gilsbach R, Pflüger- Noma S, Saijyo E, Saka A, Sakai M, Simon C, Suzuki N, Tagami Müller B, Brandes RP, Leisegang MS (2022) DNA topoisomerase M, Watanabe S, Yoshida S, Arner P, Axton RA, Babina M, Baillie inhibition with the HIF inhibitor acriflavine promotes transcrip- JK, Barnett TC, Beckhouse AG, Blumenthal A, Bodega B, Bonetti tion of lncRNAs in endothelial cells. Mol Ther Nucleic Acids A, Briggs J, Brombacher F, Carlisle AJ, Clevers HC, Davis CA, 27:1023–1035. https:// doi. org/ 10. 1016/j. omtn. 2022. 01. 016 Detmar M, Dohi T, Edge ASB, Edinger M, Ehrlund A, Ekwall K, 64. Skene PJ, Henikoff S (2017) An efficient targeted nuclease strat- Endoh M, Enomoto H, Eslami A, Fagiolini M, Fairbairn L, Far- egy for high-resolution mapping of DNA binding sites. Elife. ach-Carson MC, Faulkner GJ, Ferrai C, Fisher ME, Forrester LM, https:// doi. org/ 10. 7554/ eLife. 21856 Fujita R, Furusawa J-I, Geijtenbeek TB, Gingeras T, Goldowitz 65. Statello L, Guo C-J, Chen L-L, Huarte M (2021) Gene regula- D, Guhl S, Guler R, Gustincich S, Ha TJ, Hamaguchi M, Hara M, tion by long non-coding RNAs and its biological functions. Hasegawa Y, Herlyn M, Heutink P, Hitchens KJ, Hume DA, Ikawa Nat Rev Mol Cell Biol 22:96–118. https:// doi. or g/ 10. 1038/ T, Ishizu Y, Kai C, Kawamoto H, Kawamura YI, Kempfle JS, s41580- 020- 00315-9 Kenna TJ, Kere J, Khachigian LM, Kitamura T, Klein S, Klinken 66. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, SP, Knox AJ, Kojima S, Koseki H, Koyasu S, Lee W, Lennartsson Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, A, Mackay-Sim A, Mejhert N, Mizuno Y, Morikawa H, Morimoto Mesirov JP (2005) Gene set enrichment analysis: a knowledge- M, Moro K, Morris KJ, Motohashi H, Mummery CL, Nakachi Y, based approach for interpreting genome-wide expression profiles. Nakahara F, Nakamura T, Nakamura Y, Nozaki T, Ogishima S, Proc Natl Acad Sci U S A 102:15545–15550. https:// doi. org/ 10. Ohkura N, Ohno H, Ohshima M, Okada-Hatakeyama M, Okazaki 1073/ pnas. 05065 80102 Y, Orlando V, Ovchinnikov DA, Passier R, Patrikakis M, Pombo 67. Tombor LS, Dimmeler S (2022) Why is endothelial resilience key A, Pradhan-Bhatt S, Qin X-Y, Rehli M, Rizzu P, Roy S, Sajantila to maintain cardiac health? Basic Res Cardiol 117:35. https://doi. A, Sakaguchi S, Sato H, Satoh H, Savvi S, Saxena A, Schmidl org/ 10. 1007/ s00395- 022- 00941-8 C, Schneider C, Schulze-Tanzil GG, Schwegmann A, Sheng G, 68. de Val S, Black BL (2009) Transcriptional control of endothelial Shin JW, Sugiyama D, Sugiyama T, Summers KM, Takahashi N, cell development. Dev Cell 16:180–195. https://doi. or g/10. 1016/j. Takai J, Tanaka H, Tatsukawa H, Tomoiu A, Toyoda H, van de devcel. 2009. 01. 014 Wetering M, van den Berg LM, Verardo R, Vijayan D, Wells CA, 69. Wei G-H, Badis G, Berger MF, Kivioja T, Palin K, Enge M, Bonke Winteringham LN, Wolvetang E, Yamaguchi Y, Yamamoto M, M, Jolma A, Varjosalo M, Gehrke AR, Yan J, Talukder S, Turunen Yanagi-Mizuochi C, Yoneda M, Yonekura Y, Zhang PG, Zucchelli M, Taipale M, Stunnenberg HG, Ukkonen E, Hughes TR, Bulyk S, Abugessaisa I, Arner E, Harshbarger J, Kondo A, Lassmann ML, Taipale J (2010) Genome-wide analysis of ETS-family DNA- T, Lizio M, Sahin S, Sengstag T, Severin J, Shimoji H, Suzuki binding in vitro and in vivo. EMBO J 29:2147–2160. https:// doi. M, Suzuki H, Kawai J, Kondo N, Itoh M, Daub CO, Kasukawa org/ 10. 1038/ emboj. 2010. 106 T, Kawaji H, Carninci P, Forrest ARR, Hayashizaki Y (2017) 70. Wickham H (2009) Ggplot2. Elegant graphics for data analysis. FANTOM5 CAGE profiles of human and mouse samples. Sci Use R! Springer, Dordrecht, New York Data 4:170112. https:// doi. org/ 10. 1038/ sdata. 2017. 112 71. Yao R-W, Wang Y, Chen L-L (2019) Cellular functions of long 56. Oo JA, Pálfi K, Warwick T, Wittig I, Prieto-Garcia C, Matkovic noncoding RNAs. Nat Cell Biol 21:542–551. https:// doi. org/ 10. V, Tomašković I, Boos F, Izquierdo Ponce J, Teichmann T, Petriu-1038/ s41556- 019- 0311-8 kov K, Haydar S, Maegdefessel L, Wu Z, Pham MD, Krishnan 72. Yu G, He Q-Y (2016) ReactomePA: an R/Bioconductor pack- J, Baker AH, Günther S, Ulrich HD, Dikic I, Leisegang MS, age for reactome pathway analysis and visualization. Mol Biosyst Brandes RP (2022) Long non-coding RNA PCAT19 safeguards 12:477–479. https:// doi. org/ 10. 1039/ C5MB0 0663E DNA in quiescent endothelial cells by preventing uncontrolled 73. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein phosphorylation of RPA2. Cell Rep 41:111670. https:// doi. org/ BE, Nusbaum C, Myers RM, Brown M, Li W, Liu XS (2008) 10. 1016/j. celrep. 2022. 111670 Model-based analysis of ChIP-Seq (MACS). Genome Biol 57. Oo JA, Brandes RP, Leisegang MS (2022) Long non-coding 9:R137. https:// doi. org/ 10. 1186/ gb- 2008-9- 9- r137 RNAs: novel regulators of cellular physiology and func- 74. Zhao B, Xu H, Ai X, Adalat Y, Tong Y, Zhang J, Yang S (2018) tion. Pflugers Arch 474:191–204. https:// doi. or g/ 10. 1007/ Expression profiles of long noncoding RNAs in lung adenocarci- s00424- 021- 02641-z noma. Onco Targets Ther 11:5383–5390. https://doi. or g/10. 2147/ 58. Potente M, Mäkinen T (2017) Vascular heterogeneity and spe-OTT. S1676 33 cialization in development and disease. Nat Rev Mol Cell Biol 75. Zheng Y, Chen Z, Zhou Z, Xu X, Yang H (2020) Silencing of 18:477–494. https:// doi. org/ 10. 1038/ nrm. 2017. 36 long non-coding RNA LINC00607 prevents tumor proliferation of 59. Ramírez F, Ryan DP, Grüning B, Bhardwaj V, Kilpert F, Rich- osteosarcoma by acting as a sponge of miR-607 to downregulate ter AS, Heyne S, Dündar F, Manke T (2016) deepTools2: a next E2F6. Front Oncol 10:584452. https://doi. or g/10. 3389/ f onc.2020. 1 3 Basic Research in Cardiology (2023) 118:5 Page 21 of 21 5 Authors and Affiliations 1,2 1,2 1,2 3 1 Frederike Boos  · James A. Oo  · Timothy Warwick  · Stefan Günther  · Judit Izquierdo Ponce  · 1,2 1,2 1,2 4,5 3,6 1,2 Melina Lopez  · Diba Rafii  · Giulia Buchmann  · Minh Duc Pham  · Zahraa S. Msheik  · Tianfu Li  · 1,2 1,2 7 8,9,10,11 12,13 Sandra Seredinski  · Shaza Haydar  · Sepide Kashefiolasl  · Karl H. Plate  · Rüdiger Behr  · 12,14 2,5,15,16 3,6,15 2,17 12,14,18 Matthias Mietsch  · Jaya Krishnan  · Soni S. Pullamsetti  · Sofia‑Iris Bibli  · Rabea Hinkel  · 19,20 2,5,21 2,5 1,2 22,23 Andrew H. Baker  · Reinier A. Boon  · Marcel H. Schulz  · Ilka Wittig  · Francis J. Miller Jr.  · 1,2 1,2 Ralf P. Brandes  · Matthias S. Leisegang 1 13 Institut für Kardiovaskuläre Physiologie, Fachbereich Platform Degenerative Diseases, German Primate Medizin der Goethe-Universität, Theodor-Stern-Kai 7, Center-Leibniz Institute for Primate Research, Göttingen, 60590 Frankfurt am Main, Germany Germany 2 14 German Center of Cardiovascular Research (DZHK), Partner Laboratory Animal Science Unit, German Primate Center, Site RheinMain, Frankfurt, Germany Leibniz Institute for Primate Research, Göttingen, Germany 3 15 Max-Planck-Institute for Heart and Lung Research, Cardio-Pulmonary Institute, Giessen, Germany Bad Nauheim, Germany Department of Medicine III, Genome Biologics, Frankfurt, Germany Cardiology/Angiology/Nephrology, Goethe University Hospital, Frankfurt am Main, Germany Institute for Cardiovascular Regeneration, Goethe University, Frankfurt, Germany Institute for Vascular Signalling, Goethe University, Frankfurt, Germany Department of Internal Medicine, Member of the DZL, Member of Cardio-Pulmonary Institute (CPI), Justus Liebig Institute for Animal Hygiene, Animal Welfare and Farm University, Giessen, Germany Animal Behavior, University of Veterinary Medicine, Hannover, Germany Department of Neurosurgery, University Hospital Frankfurt, Frankfurt, Germany Centre for Cardiovascular Science, The Queen’s Medical Research Institute, University of Edinburgh, Edinburgh, Institute of Neurology (Edinger Institute), Neuroscience Scotland Center, Goethe University, Frankfurt, Germany CARIM Institute, University of Maastricht, Maastricht, Frankfurt Cancer Institute, University Hospital, Goethe The Netherlands University, Frankfurt, Germany Department of Physiology, Amsterdam Cardiovascular German Cancer Consortium (DKTK), Partner Site Sciences, VU Medical Center, Amsterdam UMC, Frankfurt/Mainz, Frankfurt, Germany Amsterdam, The Netherlands German Cancer Research Centre (DKFZ), Heidelberg, Department of Medicine, Vanderbilt University Medical Germany Center, Nashville, USA DZHK (German Center for Cardiovascular Research), Veterans Affairs Medical Center, Nashville, TN, USA Partner Site Göttingen, Göttingen, Germany 1 3

Journal

Basic Research in CardiologySpringer Journals

Published: Jan 26, 2023

Keywords: Long non-coding RNA; BRG1; Endothelial cell; Gene regulation; ERG; Hypoxia

References