Nuclear S-nitrosylation impacts tissue regeneration in zebrafish

Ethical approval

This work complied with all relevant ethical regulations for animal testing and research. Animals were housed and all experiments were carried out in accordance with the recommendations of the Institutional Animal Care and Use Committee at the Houston Methodist Research Institute, and with the United Kingdom Animals (Scientific Procedures).

Act 1986 at the Queens Medical Research Institute research facilities.

Zebrafish aquaculture and husbandry

Adult zebrafish—wild-type Wik and Tg(nfkb:EGFP)^nc1 strains—were maintained according to standard procedures. Fish were kept at 28 °C under a 14/10 h light/dark cycle and fed with dry meal (Gemma Micro, Westbrook, ME) twice per day. Embryos were obtained by natural mating and kept in E3 embryo medium at 28.5 °C. Surgical procedures were performed under anesthesia with Tricaine (also named MS-222, Sigma-Aldrich, St Louis, MO, cat. E10521) 0.02 mg/mL on embryos and 0.05 mg/mL in adult zebrafish.

Zebrafish tailfin amputation and regeneration

Caudal tailfin amputation surgeries were performed as previously described³⁹. Briefly, fish were anaesthetized and amputations were made by using a sterile razor blade, removing half of the tailfin. At day 3, 5, and 10 post-amputation (dpa) (uninjured tailfin was used as control) the regrown tissue was carefully resected and immediately processed for nuclear protein extraction. Total regeneration was measured as previously described⁴⁰. Briefly, fin images were collected before amputation and time points after amputation. The new tissue area (in pixels) of the caudal fin from the new distal fin edge to the amputation plane was quantified in each fish using Image J software. The percentage of regeneration for each fin at each time point was defined as percentage of regeneration = 100 × (regenerated tissue area/original fin area amputated). The collection of the distal (blastema) and proximal (regenerating) regions of the tailfin tissue for PCR analysis was performed under a fluorescence stereomicroscope Leica M205. The Tg(fli1:EGFP) zebrafish was used and the distal ends of the newly forming (GFP⁺) vessel branches was used as boundary to separate by dissection the two regions. Then, we placed the tissues in an Eppendorf tube and immediately extracted RNA that was used for PCR analysis.

Blood vessel density

At 5 dpa, the adult Tg(fli1:EGFP) zebrafish were anesthetized. Then, fish were transferred on a wet sponge, previously soaked in tank water, to keep the zebrafish skin moist during imaging. Images of the caudal fin including the regenerated tissue were captured under a fluorescence stereomicroscope Leica M205 equipped with a camera. The images were collected with a Leica LAS X software and analysed by Image J software. The second and third rays from the dorsal edge of the fin were used for measurement of vessel area and reported as mm².

Protein extraction

Extraction of nuclear proteins

Nuclear proteins were extracted from the caudal fin tissue using the NE-PER Extraction Reagents kit (Thermo Fisher Scientific, San Jose, CA, cat. 78833) according to the manufacturer’s instructions, supplemented with protease inhibitor cocktail. In brief, regenerating tailfin tissue was resected, cut into small pieces and placed in a microcentrifuge tube. Then, tissue was washed three times with chilled PBS, centrifuged at 500 × g for 1 min and supernatant discarded. Using a motor-driven pestle (Sigma-Aldrich, St Louis, MO, cat. Z359971), tissue was homogenized in solution CER I, that breaks plasma membrane but not nuclei, added with protease/phosphatase inhibitors cocktail (1:100, Thermo Fisher Scientific, San Jose, CA, cat. 78442). The tube was vigorously vortexed for 30 s and put on ice for 10 min. Then, chilled CER II was added to the tube, vortexed for 10 s and incubated on ice for 1 min. The tube was centrifuged for 5 min at 16,000 × g and the supernatant, containing cytoplasmic proteins, was transferred to a clean pre-chilled tube and stored at −80 °C. The pellet, containing nuclei, was resuspended in chilled NER solution, vortexed on the highest setting for 15 s. The sample was placed on ice for 45 min, vortexed for 20 s every 10 min. Then, the tube was centrifuged at 16,000 × g for 10 min and the supernatant, containing nuclear extract, immediately transferred to a clean pre-chilled tube.

Extraction of total proteins

Zebrafish embryos were euthanised with an overdose of tricaine, washed three times in PBS and homogenized with a motor-driven pestle (Sigma-Aldrich, St Louis, MO, cat. Z359971) in 100 mL RIPA buffer (25 mmol/L Tris-HCl pH 7.6, 150 mmol/L NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS), supplemented with protease/phosphatase inhibitors. The lysate was kept on ice for 40 min. Then, the tube was centrifuged at 3000 × g for 5 min and the supernatant transferred to a clean pre-chilled tube.

In both nuclear and total proteins extraction, bicinchoninic acid (BCA) protein assay (Thermo Fisher Scientific, San Jose, CA, cat. 23225) was used to measure protein concentration.

Labeling of protein S-nitrosothiols

Labeling of S-nitrosothiols in nuclear proteins was achieved using Iodoacetyl Tandem Mass Tag (iodoTMT) kit (Thermo Fisher Scientific, San Jose, CA, cat. 90103). First, nuclear protein extracts were acetone-precipitated at −20 °C for 2 h, then centrifuged at 15,000 × g for 10 min and the pellet solubilized in 500 mL HENS buffer (Thermo Fisher Scientific, San Jose, CA, cat. 90106) at a protein concentration of 1 mg/mL. Equal amounts of nuclear protein from each sample were iodoTMT-labeled. To generate a positive control sample, an aliquot of protein from control was added with 200 μM S-nitrosoglutathione (GSNO, Sigma-Aldrich, St Louis, MO, cat. N4148) for 30 min at room temperature (RT). Experimental samples were incubated for 30 min at RT after adding MMTS (10 μL of 1 M) to block-free cysteine thiols. Then, proteins were precipitated with pre-chilled acetone (1 mL per sample) at −20 °C for 2 h to remove MMTS. Samples were centrifuged at 10,000 × g for 10 min at 4 °C, the pellet resuspended in 500 mL of HENS buffer and to each was added 5 mL of iodoTMT reagent, previously dissolved in liquid chromatography/mass spectrometry (LC/MS)-grade methanol, and 10 μL of 1M sodium ascorbate (Sigma-Aldrich, St Louis, MO, cat. A4034). As a negative control, 10 μL of ultrapure water instead of sodium ascorbate was added in a protein sample. All samples were incubated for 1 h at 37 °C, protected from light. The reaction was quenched by adding 20 μL of 0.5M DTT and incubated for 15 min at 37 °C, protected from light. All experimental samples labeled with iodoTMT sixplex were combined, added with six volumes of pre-chilled acetone and incubated at −20 °C overnight. The sample was centrifuged at 10,000 × g for 10 min at 4 °C and the pellet dissolved in 3 mL HENS buffer. Then, 100 μL of 0.5 M iodoacetamide were added and the sample incubated at 37 °C for 1 h protected from light. Sample was precipitated with pre-chilled acetone and the pellet allowed to dry for 10 min.

Protein digestion for mass-spec analysis

The pellet was dissolved in 50 mM ammonium bicarbonate (Sigma-Aldrich, St Louis, MO, cat. 09830) and digested using trypsin enzyme (GenDepot, cat. T9600) at 37 °C overnight. The peptide mixture was acidified using 10% formic acid and dried using a vacuum concentrator (Thermo Fisher Scientific, San Jose, CA, cat. SPD120).

Enrichment of iodoTMT-labeled S-nitrosopeptides

The anti-TMT Antibody Resin (Thermo Fisher Scientific, San Jose, CA, cat. 90076) was washed three times with Tris Buffered Saline (TBS) (Thermo Fisher Scientific, San Jose, CA, cat. 28358). Previous labeled and lyophilized peptides were resuspended in TBS (a small portion of unfractionated sample was stored for direct analysis of the non-enriched samples). Then, peptides were added to the anti-TMT resin (100 μL of settled resin for every 1 mg of iodoTMT Reagent-labeled peptides) and incubated at RT for 4 h. Finally, the resin was washed three times (5 min per wash) with TBS and then three times with water. The sample was eluted with TMT Elution Buffer (Thermo Fisher Scientific, San Jose, CA, cat. 90104). The eluate was frozen and lyophilized, using a vacuum concentrator and then the sample resuspended in a solution of 5% methanol/0.1% formic acid. Then, 1–5 μL of sample were injected directly onto an LC-MS/MS system.

LC/MS-MS

The mass spectrometry analysis of S-nitrosopeptides was carried out on a nano-LC 1200 system (Thermo Fisher Scientific, San Jose, CA) coupled to Orbitrap Fusion™ Lumos ETD (Thermo Fisher Scientific, San Jose, CA) mass spectrometer. The peptides were loaded onto a Reprosil-Pur Basic C18 (1.9 µm, Dr. Maisch GmbH, Germany) pre-column of 2 cm × 100 µm and in-lined an in-housed 5 cm x 150 µm analytical column packed with Reprosil-Pur Basic C18 beads. The peptides were separated using a 75 min discontinuous gradient of 5-28% acetonitrile/0.1% formic acid at a flow rate of 750 nL/min. The eluted peptides were directly electro-sprayed into the mass spectrometer. The instrument used the multi-notch MS3-based TMT method. The full MS scan was performed in Orbitrap in the range of 375–1500 m/z at 120,000 resolution followed by ion trap CID-MS2 fragmentation at precursor isolation width of 0.7 m/z, AGC of 1 × 10⁴, maximum ion accumulation time of 50 ms. The top ten fragment ions from MS2 was selected for HCD-MS3 with isolation width of 2 m/z, AGC 5 × 10⁴, collision energy 65%, maximum injection time of 54 ms. The RAW file from mass spectrometer was processed with Proteome Discoverer 2.1 (Thermo Fisher Scientific, San Jose, CA) using Mascot 2.4 (Matrix Science) with percolator against Zebrafish Uniprot database. The precursor ion tolerance and product ion tolerance were set to 20 ppm and 0.5 Da, respectively. Variable modifications of oxidation on Methionine (+15.995 Da) and iodoTMT tag (+329.2266 Da) on cysteine residues was used. The general quantification in consensus workflow used unique and razor peptides with top three peptides for area calculation, while reporter quantification used co-isolation threshold of 50 and average reporter S/N threshold of 10. The assigned peptides and PSMs were filtered at 1% FDR.

Bioinformatic analysis

The computational detection strategy identifies peptides exhibiting iodoTMT tag modifications. Proteome Discoverer software (Thermo Fisher Scientific, San Jose, CA) was used to search MS/MS spectra against the Zebrafish UniProt database (Danio rerio; UP000000437) using Mascot 2.3 search engine. The iodoTMTsixplex quantification method within Proteome Discoverer software was used to calculate the reporter ratios with a mass tolerance ±10 ppm. Search algorithms, including MS-Fragger was also used in the analysis. Hierarchical clustering of S-nitrosylated protein expression heatmap was conducted using MEV based on Pearson correlation distance metric and the average linkage method. Ingenuity pathway analysis (IPA, Ingenuity systems Qiagen, Redwood City, CA) was used to assess Gene Ontology (GO) and IPA analysis to explore the function of differential S-nitrosylated proteins. Enrichment Q values of S-nitrosylated protein pathways were defined based on EdgeR FDR cutoff 1e−5. The GO category was classified by Fisher’s exact test, and the p-value was corrected by the false discovery rate (FDR) calculation.

The scRNA-seq of the regenerative caudal fin was analysed following the methods described in the manuscript²³. The matrix count from cell ranger were obtained on GEO database accession GSE137971. Downstream analysis was performed on R using the package Seurat. Clustering analysis was performed on the integrated dataset and we found six clusters. These clusters were annotated as Superficial/intermediate/basal epithelial, mucosal-like, hematopoietic and mesenchymal based on markers described in Hou et al.²³. Differential gene expression between the different time points compared to pre-injury was done using the function FindMarkers from Seurat (based on Wilcoxon test followed by Bonferroni correction). Violin Plot of kdm1a expression was obtained using the Vln Plot function of Seurat.

Chromatin immunoprecipitation (ChIP)-PCR assay

ChIP assay was performed following the manufacturer’s instructions (Cell Signaling Technology, Beverly, MA). Briefly, 15 tailfins of adult zebrafish per group were disaggregated in single cells as described above. DNA and protein were crosslinked by 1% formaldehyde. Chromatin was isolated and digested with Micrococcal Nuclease. Then, the DNA-protein complex was precipitated with control IgG or antibody against H3K4me2 (rabbit polyclonal, ChIP grade) overnight at 4 °C and protein A/G conjugated magnetic beads for 1 h. Cross-links were reversed. The extracted DNA was used as template for PCR amplification of the targeted promoter region. The extracted DNA from unprecipitated DNA-protein complex was used as input. The promoter regions of ten genes known to be involved in neoangiogenesis (kdr, vegfaa, cdh5, tek, tie2, tbx20, fgf2, angpt2, mmp2, and cd31) were identified in ENSEMBL. The gene sequence up to 600 bp upstream of the TSS was validated this sequence on genome.ucsc.edu to confirm it was upstream of our gene of interest. We also looked for the presence of CpG islands and TATA box. Hence, we designed four couples of primers using Primer Blast for each gene that matched in this region and around the TSS, and that could generate amplicons which size was no more than 120–130 bp to allow both primers to find their target on one fragment of ChIP DNA, if present. In silico PCR (UCSC) was used to confirm that primers matched our region of interest and the amplicon size, and then primers were further validated by PCR using genomic DNA.

Quantification of Kdm1a demethylase activity

Kdm1 Activity Colorimetric Kit (Abcam, Cambridge, UK, cat. ab113459) was used to quantify Kdm1 activity. Nuclear proteins were extracted from the regenerating caudal fin as shown above and an input of 10 μg per sample was used for the enzymatic analysis. The experiment was run in triplicate. A standard curve was prepared with Kdm1a assay buffer and assay standard solution, containing demethylated histone H3K4, diluted at concentration between 0.2 and 5 ng/μL. Sample wells were added with Kdm1a assay buffer, Kdm1a substrate (containing di-methylated histone H3K4) and 10μg of nuclear extract. No nuclear extract was added in blank wells. The strip-well microplate was covered with adhesive film to avoid evaporation, and incubate at 37 °C for 2 h. At this stage, active Kdm1a binds to the substrate and removes methyl groups from the substrate. Then, the reaction solution was removed and each well washed three times with wash buffer. Capture antibody, that recognizes Kdm1a-demethylated products, was added to each well, the strip-well was covered with aluminum foil to protect from light and incubated at RT for 60 min. Antibody solution was removed and each well washed three times with wash buffer. Then, detection antibody was added to each well, covered again with aluminum foil and incubated at RT for 30 min. Detection antibody solution was removed and each well was washed four times with wash buffer. Developer solution was added and the microplate incubated at RT for 10 min protected from light. In presence of methylated DNA, the solution will turn blue. At this point, stop solution was added to each well to quench the enzymatic reaction. Absorbance was read on a microplate reader Infinite M1000 (Tecan, Männedorf, Switzerland) at a wavelength at 450 nm with an optional reference wavelength of 655 nm. The activity of Kdm1a enzyme is proportional to the optical density (OD) intensity measured. Accordingly, Kdm1a activity was calculated using the following formula:

Kdm1a suppression

The knockdown (KD) of kdm1a gene (NM_001242995.1) in zebrafish was achieved by injection of antisense morpholino (Mo) (Gene Tools, Philomath, Oregon) oligo targeting the mRNA AUG translational start site (sequence 5′-TTGGACAACATCACAGATGACAGAG-3′). A 5-base pair mismatch Mo (sequence 5′-TTGcAgAACATgACAcATcACAGAG-3′) was used as control to detect possible off-target effects. A second antisense oligo targeting i3e4 splice junction of kdm1a, sequence 5′- CTACACCTGAGAAACCCAACATTTC-3′ was used to corroborate data obtained with MOs that block translation.

Using a standard microinjector (IM300 Microinjector; Narishige, Tokyo, Japan), an optimized dose of 0.4 ng (0.5 nL bolus) of morpholino placed in a pulled glass capillary was injected in each embryo at 1–2 cell stage, just beneath the blastoderm.

For KD of kdm1a in adult zebrafish, a vivo-Mo version was used, where the standard Mo is bound to a synthetic scaffold containing guanidinium groups as a delivery moiety in adult tissues. An antisense vivo-Mo that targets human b-globin intron mutation 5′-CCTCTTACCTCATTACAATTTATA-3′ was used as negative control (Gene Tools LLC, Philomath, Oregon). Adult zebrafish were anesthetized in tricaine 0.05 mg/mL, and 2 μL of 0.1 mmol/L vivo-Mo solution, previously loaded in a glass capillary, was injected into the retro-orbital vein, as previously described⁴¹, on days 10, 8, 6, 4, 2 and 0 before tailfin amputation.

kdm1a construct

A construct with kdm1a gene of Danio rerio (NM_001242995.1) was prepared to generate kdm1a modified (m) mRNA and was assembled from synthetic oligonucleotides. Modifications in the triplet code (n = 7 silent mutations) were inserted in the sequence corresponding to the mRNA AUG translational start site (i.e., Mo binding site) to prevent Mo recognition in rescue experiments. The fragment was inserted in the pMA (GeneArt, Invitrogen, Carlsbad, CA) cloning vector and cloned in transformed Escherichia coli bacteria (strain K12/DH10B, Invitrogen, Carlsbad, CA) and then purified. The final construct was verified by sequencing and the sequence identity within the insertion site was 100%.

Site-directed mutagenesis

Site-directed mutagenesis of Kdm1a was carried out with Q5 Site-Directed Mutagenesis kit (New England Biolabs, Ipswich, MA, cat. E0554) according to manufacturers’ instructions. Standard primers for kdm1a were used for exponential amplification of the plasmid DNA (F 5′-GACAGCCAGTCGAGGAGAAC-3′ and R 5′-TGCGACGTACGAGTATGAGC-3′), and mutagenic primers to create substitution of Cys334-to-Ala (C334A) in the plasmid were designed with the software NEBase Changer (F 5′-AAAACAGAAGGCTCCCCTCTATGA GGC-3′ and R 5′-ATCTTAGCCAGCTCCATATTG-3′).

In vitro transcription of kdm1a

Wild type and mutated kdm1a mRNA (C334A), with 7-methyl guanosine cap structure at the 5′end and poly(A) tail at the 3′end, was transcribed from the constructs using HiScribe T7 ARCA mRNA Kit (New England Biolabs, Ipswich, MA, cat. E2060) following manufacturers’ instructions.

Rescue of kdm1a knockdown by Kdm1a mRNA

To determine whether the effects of the kdm1a KD on zebrafish embryos phenotype and tailfin regeneration were specifically due to loss of kdm1a, we co-injected kdm1a-Mo with kdm1a mRNA wild-type as a rescue. A bolus of 1 nL of solution containing 0.5 ng of kdm1a-Mo a and 1 ng of Kdm1a RNA wild-type was injected into each egg.

Rescue of kdm1a knockdown by kdm1a mRNA C334A

Co-injection of kdm1a -Mo and kdm1a mRNA C334A was performed to assess whether the absence of the S-nitrosylated cysteine affected the ability of the mRNA to rescue phenotype and tailfin regeneration associated with kdm1a KD. A bolus of 1 nL of solution containing 0.5 ng of kdm1a-Mo and 1 ng of kdm1a mRNA C334A was injected into each egg.

Pharmacological modulation of S-nitrosylation

Adult zebrafish were anesthetized in tricaine 0.05 mg/mL. A solution of 2 μL of iNos inhibitor N(ω)-nitro-l-arginine methyl ester (L-NAME, Sigma-Aldrich, St Louis, MO, cat. N5751) 10 or 50 mM diluted in sterile PBS (from stock solution of 250 mM), or of nitric oxide (NO) scavenger Phenyl-4,4,5,5-tetramethyl imidazoline-1-oxyl 3-oxide (PTIO, Sigma-Aldrich, St Louis, MO, cat. P5084) 3 or 10 mM diluted in sterile PBS (from stock solution of 100 mM), or of NO donor S-Nitroso-N-acetyl-dl-penicillamine (SNAP, Sigma-Aldrich, St Louis, MO, cat. N3398) 10 or 30 mM diluted in DMSO (from stock solution of 100 mM), or PBS as control, was loaded on a glass capillary, prepared in advance with a micropipette puller (Narishige, Inc., PC-10) and connected to a microinjector (IM300 Microinjector; Narishige, Tokyo, Japan) and was injected into the retro-orbital vein as previously described⁴¹ on days 6, 4, 2, and 0 before tailfin amputation.

The concentration of the Nos inhibitors, L-NAME and 1400W, were adopted after a pilot study with doses up to 250 mM. The survival was recorded and fish were monitored for any physical or behavioral abnormalities at a range of doses. For both compounds at a dose of 250 nM, survival after four injections was approximately 40%, and lethargy and reduced swim were observed. At 150 mM the survival increased to 65% with no evident abnormal behavior; whereas at 50 mM, the dose adopted for the study, the survival was 85% and no evident defects were observed (Supplementary Fig. S2C).

Optimization of the injection procedure

In our pilot studies, injection of physiological solutions every other day and alternating the eye at each injection reduced fish mortality. In this way, the effect of a drug on fish survival/mortality and phenotype can be better evaluated. Therefore, all the solutions of drugs were injected in the retro-orbital veins every other day and alternating the eyes, so that each eye was injected only twice at a distance of four days.

Defining the zebrafish phenotype

Whole embryo phenotype following Mo and mRNA treatments were described on the basis of the following morphologic features observed under bright-field microscopy: reduced body length, curved body, reduced swimming, chorionated larvae at 4 dpf, edema. The phenotype was assessed using a simple six points scoring system, according to the severity of that feature and where one point was normal. At least four different clutches of larvae were assessed under each of the treatment groups. Data were reported graphically as divided in two groups: normal, i.e., embryos not showing any abnormal features, and abnormal, i.e., embryos showing one or more of the features described above.

Cardinal vein blood velocity

Blood velocity was estimated in the posterior cardinal vein⁴² by assessing frame by frame motion of single blood cells determined from video images captured in the zebrafish tail at the level of the cloaca. Four erythrocytes per fish (at least five embryos per group) over ten frames at video frame-rate of 30 frames per second (fps) were analyzed using ImageJ to determine mean blood cell velocity (μm s⁻¹).

Kaplan–Meier analysis of survival

Kaplan–Meier analysis was used to measure the survival of adult zebrafish or larvae following each defined treatment, using PRISM 7 software.

Immunoprecipitation

Immunoprecipitation experiments were performed using the Pierce Classic Magnetic IP/Co-IP Kit (Thermo Fisher Scientific, San Jose, CA, cat. MAN0011737), according to manufacturer’s instructions.

RNA extraction and quantitative PCR

mRNA was extracted from embryos using column purification (RNeasy Mini Kit, Qiagen, Hilden, Germany, cat. 74104) according to the manufacturer’s instructions. Working surfaces were cleaned with RNase Zap (Thermo Fisher Scientific, San Jose, CA, cat. AM9780) to deactivate environmental RNase. Efficient disruption and homogenization of tissue was done using sterile RNase-free disposable pestles (Thermo Fisher Scientific, San Jose, CA, cat. 12-141-368) mounted on a cordless motor for 30 s and then passing the lysate 5–10 times through the needle (18–21 gauge) mounted on a RNAse free syringe. RNA integrity was assessed on basis of 18S and 28S ribosomal RNA (rRNA) bands. mRNA was reverse transcribed into cDNA using qScript cDNA Synthesis Kit (Quanta Bio, Beverly, MA, cat. 95047), Primers (IDT Technologies, Coralville, Iowa) targeting genes of interest (see Table S2) and SYBR Green PCR kit (Invitrogen, Carlsbad, CA) were used for real-time qPCR, that was performed with the QuantStudio 12 k Flex system (Applied Biosystems, Foster City, CA) following the manufacturer’s instructions. Gene expression was expressed as relative fold changes using the ΔCt method of analysis and normalized to β-actin.

Western blotting

Lysates containing 20 μg of protein each were added with Laemmli buffer (4×) and deionised water to reach a final volume of 20 μL. A sample containing pre-stained protein standard (BioRad, Hertfordshire, UK, cat. 1610375) was used to assess molecular mass of protein bands. Samples were heated at 95 °C for 5 min and loaded on a polyacrylamide gel (4–15% gradient) (BioRad, Hertfordshire, UK, cat. 4561083). Electrophoresis was performed for 30 min at a voltage of 100 V and then for 60 min at 150 V. Gels were transferred on PVDF membranes (Amersham Hybond, Sigma-Aldrich, St Louis, MO, cat. GE10600023) for 2 h at 100V. Membranes were blocked with non-fat milk 5% in PBST (PBS+0.1% Tween) for 1 h at RT and probed with primary antibody overnight at 4 °C. Antibodies used were: Kdm1a rabbit polyclonal (1:200, Thermo Fisher Scientific, San Jose, CA, cat. PA1-41697); anti-iNos mouse monoclonal (1:200, BD Transduction Laboratories, San Jose, CA, cat. 610432); anti-β-tubulin rabbit polyclonal (1:500, Abcam, Cambridge, UK, cat. ab6046), Anti-Histone H3 nuclear marker, rabbit polyclonal (1:500, Abcam, Cambridge, UK, cat. ab1791); anti-Histone H3 (1:500, unmodified Lys4), mouse monoclonal (1:500, Merck Millipore, Massachusetts, USA, cat. 05-1341); anti-monomethyl-Histone H3 (Lys4), rabbit polyclonal (1:500, Merck Millipore, Massachusetts, USA, cat. 07-436); anti-dimethyl-Histone H3 (Lys4), rabbit monoclonal (1:500, Merck Millipore, Massachusetts, USA, cat. 04-790); anti-monomethyl-Histone H3 (Lys9), rabbit polyclonal (1:500, Merck Millipore, Massachusetts, USA, cat. ABE101); anti-dimethyl-Histone H3 (Lys9) rabbit polyclonal (1:500, Merck Millipore, Massachusetts, USA, cat. 07-212); anti-Rcor1 rabbit polyclonal (1:200, Invitrogen, Carlsbad, CA, cat. PA5-41564); anti-Hdac1 rabbit polyclonal (1:200, Abcam, Cambridge, UK, cat. ab33278); anti-Rbbp4 rabbit polyclonal (1:200, Biorbyt, Cat. orb583248); anti-Chd4 rabbit polyclonal (1:200, Biorbyt, Cambridge, UK, cat. orb575051); anti-TMT mouse monoclonal (1:200, Thermo Fisher Scientific, San Jose, CA, cat. 90075). Membranes were washed three times (5 min per wash) with PBS and incubated with HRP-conjugated goat anti-mouse (1:2000, Santa Cruz Biotechnology, Dallas, USA, SC-2005) or anti-rabbit (1:2000, Santa Cruz Biotechnology, Dallas, USA, SC-2004) antibodies for 1 h at RT. Then, membranes were washed again three times with PBS (5 min per wash). Antigen-antibody complexes were detected by incubation for 5 min to the enhanced chemiluminescence solution (ECL, Amersham) followed by exposure to a photographic film (BioMax XAR Film Kodak, Sigma-Aldrich, St Louis, MO). The film was developed and band density was quantified by densitometry using ImageJ. β-tubulin and Histone H3 were used as loading control for cytoplasmic and nuclear protein, respectively.

Sequence and structural analysis of Kdm1a protein

Similarities of zebrafish and human Kdm1a proteins were assessed using Protein Blast (https://blast.ncbi.nlm.nih.gov/Blast.cgi). Kdm1a crystal structure was obtained from the Protein Data bank (https://www.ebi.ac.uk/pdbe/entry/pdb/6nqm).

Enzymatic isolation of endothelial cells from zebrafish tailfin

Cells were isolated according to⁴³ with some modifications. In brief, amputated tailfin from adult Tg(fli1:EGFP)^y1 zebrafish were placed in chilled PBS, washed with calcium-free Ringer solution (116 mM NaCl, 2.6 mM KCl, and 5 mM Hepes, pH 7.0), and replaced with 1 mL solution of trypsin 0.25% (Gibco) added with 50 μg collagenase P (Roche) and 1 mM EDTA. Tissue was disaggregated first using fine scissors and then by pipetting the solution with a 200 μL pipette tip every 5 min for about 30 min. Cell suspensions were filtered through a 40-μm cell strainer (BD Biosciences) into FACS tubes.

Flow cytometry characterization of fli1
⁺ cells from zebrafish

Cell samples were run on a BD FACS Aria (BD Biosciences). FSC-H and FSC-A were used to select cell singlets; 4′,6-diamidino-2-phenylindole (DAPI) to select viable single cells; wild-type (Wik) zebrafish were used to set the gate between GFP⁻ (i.e., fli1⁻) and GFP⁺ (i.e., fli1⁺) cells. At least 10,000 of fli1⁻ and fli1⁺ cells (excitation [Ex]: 488 nm; emission [Em]: 530 nm) were sorted into chilled PBS and 10% fetal bovine serum for further analysis. FlowJo 10 (Becton and Dickinson) was used to analyse data.

Statistical analysis

Results were expressed as the mean ± SEM. Each experiment was performed three times (biological replicates). The Shapiro–Wilk test was used to confirm the null hypothesis that the data follow a normal distribution. Statistical comparisons between two groups or multiple groups were then performed, respectively, via Student t-test or ANOVA test using PRISM 7 software followed by Bonferroni post hoc test. Log-rank test and Gehan–Breslow–Wilcoxon test were used for statistical analysis of the Kaplan–Meier curves. A p value < 0.05 was considered significant.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.