High Mitogenic Stimulation Arrests Angiogenesis Supplementary Information Pontes-Quero et al. High mitogenic stimulation arrests angiogenesis Samuel Pontes-Quero, Macarena Fernández-Chacón, Wen Luo, Federica Lunella, Verónica Casquero-Garcia, Irene Garcia-Gonzalez, Ana Hermoso, Susana F. Rocha, Mayank Bansal, Rui Benedito. It includes; Supplementary Figures 1-8 and Supplementary Table 1. 0.0 0.5 1.0 Supplementary Fig. 1 e d b c a f Control DAPT (24h) Anti-Dll4 (24h) Rbpj 0.0 0.5 1.0 R el at iv e R bp j g en e ex pr es si on R el at iv e ge ne e xp re ss io n Rbpj iKO-EC Rbpj iKO-EC Rbpj Hey1 NS NS Efnb2f/f Rbpj f/f iKO-EC Tomato TomatoControl Rbpj ox Retina ECs Ladder Tg(iSuRe-Cre) Construct Retina ECs + + LoxP MbTomato LoxP CAG Prom. N-PhiM 2A Int-Cre WPREpA pA DBZ (24h)Vehicle (24h)DBZ (24h)Vehicle (24h) Venus 9% 46%53%90% Mouse ES cells and other lineages Mouse ECs IC A M 2 Venus Mouse ES cells and other lineages Mouse ECs IC A M 2 Venus 0 10 20 30 H U VE C s KI 67 +/ D AP I+ (% ) DMSO DBZ KI67 DAPI KI67 DAPI ** ** ** ** ** ** ** * Supplementary Fig. 1 Proliferative eect of Notch inhibition in vitro and validation of dierent pharmacological and genetic approaches used to interfere with Notch signalling. a) Inhibition of Notch signalling with DBZ for 24h increases the frequency of KI67+ HUVECs. n=4 independent experi- ments. b) Inhibition of Notch signalling with DBZ for 24h in ECs derived from embryonic stem cells carrying a Rosa26knock-in FUCCI allele, increases the frequency of ECs (ICAM2+) in S/G2/M (Venus+). c) Diagram representing the iSuRe-Cre allele. After inducible CreERT2 recombination, cells express MbTomato and a intron containing Cre. d) Detection of the Rbpj oxed allele by PCR shows that deletion of the allele is complete in MbTomato+ cells collected by FACS from RbpjiKO-EC mutants. e) qRT-PCR of RNA collected from CD31-APC+ Retina ECs of Rbpjf/f (n=3) or RbpjiKO-EC mutants (n=3), shows the pronounced reduction in Rbpj mRNA levels in MbTomato+ cells. f) Comparison by qRT-PCR of canonical endothelial Notch targets (Hey1 and Efnb2) expression in retina ECs collected by FACS from animals receiving DAPT (for 24h, n=3) or Anti-Dll4 (for 24h, n=3) or with induced Rbpj deletion for 5 days (n=3). Error bars indicate StDev; NS, non-signicant; * p< 0.05; ** p<0.005 relative to control. Two-tailed unpaired T-test (a, e). One-way ANOVA with Tukey’s post hoc test (f ). Source data are provided as a Source Data le. Pontes-Quero et al. 26% Supplementary Fig. 2 R el at iv e ge ne e xp re ss io n R el at iv e ge ne e xp re ss io n ES cells initial recombination ratio ES cells initial recombination ratio e f g d iChr-Notch-v2-Mosaic iChr-Notch-v2-Mosaic x Tie2-Cre iChr-Notch-Mosaic Lox3Lox3 Lox1Lox1 Lox2 Lox2 HA-H2B-Cherry DN-Rbpj2A2A pA pA pACAG Rosa26Rosa26 N-Phi-M Retina (P20) HA-Cherry V5-GFP ERG Total= 1277 cells 20% 5% 75% Liver (P20) EGFP Cerulean Total= 47607 cells 28% 1% 70% EGFP Cherry Heart (P20) Total= 19523 cells 26% 6% 68% EGFP Cerulean EGFP Cherry NICD-PEST H2B-EGFP-V5 His-H2B-Cerulean P20E8.5 E9.5E8.5 1,2kb 5kb 8,2kb 36% 52% 12% Gated on Pecam1+ events Gated on Pecam1+ events 0.0 0.5 1.0 Control DN-Maml1 DN-Rbpj Rbpj h 0 2 4 6 8 Control NICD-PEST Hey1 Hey1 Hey2Efnb2 iKO-EC **** ** ** ** **** * iChr-Notch-Mosaic x Tie2-Cre a b c Lox3Lox3 Lox1Lox1 Lox2 Lox2 DN-Maml12AHis-H2B-Cherry H2B-EGFP-V5pA pA pACAG Rosa26Rosa26 2AHA-H2B-Cerulean NICD-PESTN-Phi-M 1,2kb 2,8kb 5kb Cerulean Cherry Cherry EGFP E9.5 Tie2-Cre+ Ratio 57% n= 3161cells n= 1232 cells Cherry EGFP HA-Cerulean 50 µm 50 µm n= 6335 cells FACS E9.5 Embryos 32% 44% 24% 17% 829 1800 1800 532 Pontes-Quero et al. Supplementary Fig. 2 Supplementary Fig. 2 iChr-Notch-Mosaic constructs, mosaic ratios and canonical Notch target genes proling. a, b) Diagram of iChr-Notch-Mosaic allele, with the distances between loxP sites and obtained frequencies in mouse ES cells after Cre plasmid transfection. c) Mice carrying the iChr-Notch-Mosaic and Tie2-Cre alleles have recombination of the construct between E8.0 and E8.5. FACS analysis of embryos at E9.5 (n=4) reveals the proportion of ECs expressing the dierent markers at that timepoint. Note the dierence between ES cells (a) and E9.5 embryos (c). d) The iChr-Notch-V2-Mosaic allele contains DN-Rbpj instead of DN-Maml1, and a change in the position of the control cassette (Cerulean+), resulting in the lower occurrence of these cells. e, f) Mice carrying the iChr-Notch-V2-Mosaic and Tie2-Cre alleles will have recombination of the construct between E8.0 and E8.5. FACS analysis of animals at P20 reveals the proportion of ECs expressing the dierent markers at that timepoint. Note the dierence between the initial recombination ratio in ES cells (d) and in the vasculature of the dierent organs (e, f ). g, h) Comparison by qRT-PCR of canonical Notch targets expression in ECs collected by FACS from iChr-Notch-Mosaic (n=3) and RbpjiKO-EC animals (n=3). Error bars indicate StDev; NS, non-signicant; * p< 0.05; ** p<0.005. One-way ANOVA with Tukey’s post hoc test (g) and Two-tailed unpaired T-test (h). Source data are provided as a Source Data le. Pontes-Quero et al. 50 µm 50 µm a b g h dc e f ERG EdU Double+ 50 mm Control IgG 72h Anti-Dll4 72h Control IgG 12h Anti-Dll4 12h Angiogenic front Mature area Angiogenic front Mature area 0 20 40 10 30 0 100 200 300 Angiogenic front EC n um be r / m m 2 IgG 72h a-Dll4 72h IgG 72h a-Dll4 72h IgG 72h a-Dll4 72h IgG 72h a-Dll4 72h Angiogenic front Angiogenic front Angiogenic front EC S -p ha se ra tio (% ) 0 10 20 30 40 Mature Area EC S -p ha se ra tio (% ) 0 50 100 150 200 250 Mature Area EC n um be r / m m 2 0 100 200 250 150 50EC n um be r / m m 2 IgG 12h a-Dll4 12h An gio ge nic fr on t 200 µm 200 µm 50 µm 50 µm 50 µm 50 µm *** *** *** ** N.S. IgG 12h a-Dll4 12h 0 20 40 60 EC S -p ha se ra tio (% ) * ER G E dU Do ub le+ ERG EdU Double+ ERG EdU Double+ Supplementary Fig. 3 Supplementary Figure 3. Temporal analysis of the proliferative eect of Notch inhibition in angiogenic and mature vessels in vivo. a, b) Confocal micrographs of retinal vasculature from animals treated with IgG (control) and anti-Dll4 for 72h and immunos- tained with anti-Erg (red signal, labels EC nuclei) and with EdU labelling in the nuclei of all cells in S-phase in the 4 h before dissection. Blue signal is EdU in the nuclei of non-endothelial (ERG-) cells, and green signal indicates ERG+/EdU+ endothelial nuclei (ERG+ Red and EdU+ Blue signals when colocalize result in pink color, that was pseudocolored to green to better highlight ECs in S-phase. Higher magnication pictures are provided to show the distinct eects of Dll4/Notch inhibition in the proliferative angiogenic front and in the mature/quiescent area. c-f) Charts showing that Dll4/Notch signaling inhibition with anti-Dll4 for 72h leads to premature cell cycle exit at the angio- genic front and cell cycle entry of quiescent/mature ECs, indicated by the changes in the frequency of Erg+ cells in S-phase (EdU+, green) observed in these two distinct areas. This results in an increased EC density in the mature vascular area. g, h) Similar analysis in animals treated with anti-Dll4 for only 12h indicates that at this stage there is not a signicant dier- ence in EC density or the frequency of ECs in S-phase. Charts show comparative analysis of large microscopy elds taken from 4 retinas per group. Error bars indicate SEM; NS, non-signicant; * p< 0.05; ** p<0.005; *** p<0.0005. Two-tailed unpaired T-test. Source data are provided as a Source Data le. Pontes-Quero et al. bc Control IgG 24h Control IgG 48hAnti-Dll4 24h Anti-Dll4 48h Control IgG 24h Control IgG 48hAnti-Dll4 24h Anti-Dll4 48h G FP /C er ule an K i67 G FP /C er ule an K i67 d f g a 0 50 100 150 % K i6 7+ E C s IgG 24h a-Dll4 24h IgG 48h a-Dll4 48h % K i6 7+ E C s 0 20 40 60 IgG 24h a-Dll4 24h IgG 48h a-Dll4 48h ***** * ** e Lox3Lox3 Lox1Lox1 Lox2 Lox2 H2B-EGFP-V5pA pA pACAG Rosa26Rosa26 H2B-Kate2-HisHA-H2B-CeruleanPGK-Neo STOP Gt(Rosa)26Sor tm (iChr-Cerulean/GFP/Kate2 Mosaic) Supplementary Fig. 4 Angiogenic front Angiogenic front Angiogenic front Mature Area Mature Area Mature Area Pontes-Quero et al. Supplementary Figure 4. Eect of Notch inhibition on the frequency of endothelial cells in cycle (Ki67+) in angiogenic and mature vessels. a) Diagram illustrating the mouse iChr-Cerulean/GFP allele used to specically label the nuclei of the endothelium with GFP and Cerulean proteins, in order to be possible to perform the co-immunostaining with rabbit anti-Ki67 and goat anti-GFP to detect the nuclei of ECs (we could not use the incompatible rabbit anti-ERG, that was combined with EdU in Sup. Fig. 3). b-e) Confocal micrographs of retina vessels from animals carrying the Tie2-Cre and iChr-Cerulean/GFP reporter allele. In these mice, most ECs will express Cerulean or GFP in the nucleus, both recognized by the anti-GFP antibody. This allows detection in the same immunostaining of Ki67, present in all cycling cells. Yellow nuclei correspond to GFP+/Ki67+ cycling ECs and green nuclei to non-cycling ECs (GFP+/KI67-). Scale bars, 50µm. f, g) Quantication of several microscopic elds as represented in b-e. Error bars indicate SEM; NS, non-signicant; * p< 0.05; ** p<0.005; *** p<0.0005. One-way ANOVA with Tukey’s post hoc test. Source data are provided as a Source Data le. M at ur e ar ea a Control IgG 24h Anti-Dll4 24h IsolectinB4 P-ERK IsolectinB4 P-ERK 50 µm P-ERK P-ERK Supplementary Fig. 5 b Control Anti-Dll4 0 100 200 300 400 500 P -E R K A .U ./m m 2 va sc ul at ur e * 0 25 50 75 100 125 P -E R K s ig na l A .U . MbTomato- MbTomato+ IsolectinB4 MbTomato-2A-VEGFR2Ac. P-ERK P-ERKIsolectinB4 P-ERK * MbTomato-2A-Vegfr2 Strong MbTomato Pontes-Quero et al. Supplementary Figure 5. ERK activity in mature ECs after Dll4/Notch inhibition or VEGFR2 activation. a) Confocal micrographs of the mature retinal vasculature of control (n=3) and anti-Dll4 (n=4) treated animals, showing very low basal P-ERK levels in mature vessels (compare with Fig. 3a), that increase after Dll4/Notch signaling blockade for only 24h. However, P-ERK levels do not reach the very high level detected at the angiogenic front (compare with Fig. 3a right). b) Confocal micrographs of P6 retina vessels from animals (n=3) carrying the iMb-Vegfr2-Mosaic and Cdh5-CreERT2 alleles, three days after tamoxifen injection. Expression of MbTomato-2A-Vegfr2Ac (yellow arrows) induces a marked upregulation of P-ERK levels in endothelial cells. Each dot in the chart represents the average relative P-ERK signal in a single cell segmented area, in relation to the background signal in the IsolectinB4 negative area. Error bars indicate StDev; * p< 0.05. Two-tailed unpaired T-test. Source data are provided as a Source Data le. An gio ge nic F ro nt An gio ge nic F ro nt a b c d Angiogenic Front P7 a-Dll4 Tam. P1 P3 P5 IgG Control 48h IgG Control 48h Anti-Dll4 48h Anti-Dll4 48h MbYFP-2A-VEGFR2TKMut ERG ERG MbYFP-2A-VEGFR2TKMut Mature Area Mature Area * ** * * * * ** * ** * ** ** Control IgG a-Dll4 48h 0 20 40 60 80 100 *** Control IgG (24h) Anti-Dll4 (24h) Pe ca m He y1 Efn b2 Dll 4 An gp t2 Es m1 Ap ln Cd k2 Cd k4 Cd kn 1a Cd kn 1b Trp 53 Ve gfr 2 Ve gfr 1 Ve gfr 3 -5 -4 -3 -2 -1 1 2 3 4 Anti-Dll4 NSNS NS NS NS NS NS NSNS NS NS * * * * ** * * * *** ** ** ** ** ** * ** Notch targets Tip cell-enriched Cell cycle Anti-VEGF ** Supplementary Fig. 6 1 1 2 2 KI 67 + E Cs (% ) MbYFP-2A-VEGFR2TKMut KI67 MbYFP-2A-VEGFR2TKMut MbYFP-2A-VEGFR2TKMut KI67 Pontes-Quero et al. Supplementary Fig. 6 Supplementary Fig. 6. Notch strongly represses the expression of tip cell enriched genes and Cdkn1a (p21) but not Vegfrs. a) Confocal micrographs of the control and anti-Dll4 treated retinal vasculature stained with IsolectinB4. b) Evaluation by qRT-PCR (n=3 independent experiments) of genes dierentially expressed by angiogenic retina vessels after anti-Dll4 (n=6 animals) or anti-VEGF (n=6 animals) treatment for 24h. Tip cells enriched genes and Cdkn1a (p21) are signicant- ly regulated by Notch and VEGF in a opposite manner. c, d) Confocal micrographs of P7 retinas from iMb-Vegfr2-Mosaic Cdh5-CreERT2 mice, 4 days after inducing with tamoxifen the expression of MbYFP-2A-VEGFR2TkMut in some ECs, and 2 days after receiving control IgG or Anti-Dll4 for 48h. When Dll4/Notch signaling is inhibited, VEGFR2TKMut expressing mature ECs enter the cell cycle (KI67+, asterisk) and are able to proliferate. Note that there is some unspecic KI67 signals in the membrane of some cells. In higher magnication panels shown in 1 and 2 is possible to see a dierence in the morphology of MbYFP-2A-VEGFR2TkMut expressing cells induced by anti-Dll4 treatment. Error bars indicate StDev in b and SEM in d; NS, non-signicant; * p< 0.05; ** p<0.005; *** p<0.0005. Two-tailed unpaired T-test. Source data are provided as a Source Data le. Pontes-Quero et al. Gt(Esm1) Targeting vector tm(HA-H2B-Cerulean-2A-CreERT2) CRISPR/Cas9 Guide Sequence: (PAM 10bp upstream of Esm1 ATG) ATG ATG ATG iCreERT2 pA Esm1Esm1 2AHA-H2B-Cerulean E1 E2 E3 UTR E1 P3 P1 P2P4 E2 E3 UTR Neo iCreERT2 pA2AHA-H2B-Cerulean Neo FRTFRT FRT 413bp 454bp Wild-type Esm1 Targeted Esm1 (after FlpO) UTR E1 Southern Blot ES cells DNA digested with EcoRV NEO Probe PCR Screening Adult mice DNA Primers P1+P2 Primers P3+P4 KI /K I KI /K I W t W t EcoRV EcoRV 10kb 579bp409bp 500 100 400 100 GGCTCTTCATGTCTCGCAGC TGG a b c Supplementary Fig. 7 Esm1-HA-H2B-Cerulean-2A-CreERT2 x R26-LoxP-STOP-LoxP-Tdtomato IsolectinB4 TdTomato CD1 CD2 BF2 BF1 BF3 BE1 BE2 BE3 CD1 EC surface IsolectinB4 Mb2-HA-Tfp1 HA-H2B-Cerulean α-GFP α-Dsred α-HA Mb2-YFP H2B-EGFP HA-H2B-Cerulean Mb2-Tomato H2B-Cherry 4 signals 1 1 2 2 3 33 iMb2-Control-Mosaic iChr2-Control-Mosaic Tamoxifen P3 -> Analysis P6 Cdh5-CreERT2 88µm 88µm 88µm 88µm 66 µm 66 µm 28 µm 28 µm 28 µm 28 µm Mb2-YFP Mb2-Tomato Mb2-HA-Tfp1 H2B-Cherry H2B-EGFP HA-H2B-Cerulean N-PhiM L2 L2 CAGRosa26 Rosa26 L1 L1L3 L3 L2 L2L1 L1L3 L3 X X CAG E F DA B C Rosa26Rosa26 N-PhiM Pontes-Quero et al. Supplementary Fig. 7 Pontes-Quero et al. Supplementary Figure 7. A mouse line to label and genetically pulse ECs with high Esm1 expression. a) Targeting vector and Esm1 allele genetic maps. To label the nuclei and induce genetic recombination in individual endothelial tip cells, we introduced the HA-H2B-Cerulean-2A-CreERT2-Sv40pA-FRT-Neo-FRT cassette in the Esm1 starting codon (ATG) by CRISPR/Cas9-induced double-strand break and homologous recombination-dependent repair. The donor vector contained two homology arms with 413 and 454bp, respectively, and was electroporated into ES cells. After selec- tion in G418 (neomycin), the resistant clones were screened for correct vector integration, rst by PCR with the indicated primers and after by Southern blot with a labeled probe against the Neo gene. The FRT-Neo-FRT cassette was later ipped out by transient transfection of FlpO, and clones without Neo resistant cells were used to generate mice. b) Confocal micrographs from P7 retinas of mice with the above mentioned alleles and pulsed with tamoxifen at P3. In most retinas, there were either no labelled cells, or only 1 clone with labelled cells per ank (each dissected and at-mounted retina has 4 anks). Clones were of dierent sizes (see also Fig. 4c chart). Given the very low frequency of Tomato+ cells and their clonal and nearby distribution in the tissue, it is statistically highly improbable that clones with more than 2 cells arose from more than one independent stochastic tip-cell recombination event. c) Representative confocal micrographs showing the 4 signals detected after immunostaining with the indicated antibod- ies, of the postnatal day 6 mouse retina of Dual ifgMosaic mice, carrying the Cdh5-CreERT2 allele, and pulsed once with tamoxifen at P3. Magnied boxed areas (1 to 3) and two-letter codes show selected double-recombined cell clones and the observed combination of FPs A to F, at higher magnication. The chance of a given dual recombination event is lower than single recombination event, enabling the identication and quantication of single-cell derived clones (see also Pontes-Quero et al., 2017). This method was used for the quantications shown in Fig. 4d chart. Cherry p21IsolecB4 Cherry GFP p21 Cherry GFP KI67 GFP Cherry p21GFP p21GFP Cherry p21 Fr eq ue nc y of p 21 + E C s (% ) Fr eq ue nc y of K I6 7 + EC s (% ) n= 22 1 n= 69 5 Angiogenic Area Angiogenic Area Leading Edge Angiogenic Area Leading Edge 0 20 40 60 80 Cherry GFP Strong GFP n= 22 n= 17 5 n= 49 0 n=1546 n=421 n= 43 5 n= 13 n= 15 5 n= 11 3 n= 18 n= 93 n= 22 0 20 40 60 80 100 Cherry GFP Strong GFP Leading Edge Cherry KI67IsolecB4 GFP Cherry KI67GFP KI67GFP Cherry p21 74.09% Cherry+ 23.56% GFP+ 2.35% Strong GFP+ 60.34% Cherry+ 32.07% GFP+ 7.59% Strong GFP+ a c d e b Supplementary Figure 8 iChr-Notch-Mosaic: iChr-Notch-Mosaic Lox3Lox3 Lox1Lox1 Lox2 Lox2 DN-Maml12AHis-H2B-Cherry H2B-EGFP-V5pA pA pACAG Rosa26Rosa26 2AHA-H2B-Cerulean NICD-PESTN-Phi-M Pontes-Quero et al. Supplementary Figure 8 Pontes-Quero et al. Supplementary Figure 8. Single cells with a decrease in Notch signalling exit cell cycle and sprout more frequently. a, b) Confocal micrographs of the angiogenic front of P6 retinas from animals expressing the iChr-Notch-Mosaic allele in endothelial cells (IsolectinB4+). Individual cells with lower Notch signalling (GFP+) are more frequently p21+, especially if they express GFP strongly and are at the leading edge of the vessels. Blue arrowheads indicate GFP+/p21+ cells. Pink arrowheads indicate Cherry+/p21+ cells. c, d) Confocal micrographs of the angiogenic front of P6 retinas from animals expressing the iChr-Notch-Mosaic allele in endothelial cells (IsolectinB4+). Individual cells with lower Notch signalling (GFP+) are less frequently KI67+, especially if they express GFP strongly and are at the leading edge of the vessels. e) Cells with a decrease in Notch signalling (GFP+), are more frequently found at the leading edge of the vessels. Primer Name Primer Sequence PCR Ta Band Size Purpose hVEGFR2 seq F GCGGCACGAAATATCCTCT hVEGFR2 seq R ATTTCCCACAGCAAAACACC CTP PHI F ACGTGAAGCTGAGCAAGGAT CTP H2B R CTTAGTCACCGCCTTCTTGG LoxP3 seq F TCAATGTATCTTAAGGCGTGACT mTFP1 R TACTTGGTGAAGGCCCTGTT Esm1 Trans F CTCCGTGCTAAGGGACTCTG H2B PCR REV CCTTAGTCACCGCCTTCTTG Esm1 WT F AACAAGAGAGGCTGGCAAGA Esm1 WT R TCCATGCCTGAGACTGTACG RR711 GCACTTGCTCTCCCAAAGTC RR713 CTTTAAGCCTGCCCAGAAGA CAG PCR F CGGGGTCATTAGTTCATAGCC CAG PCR R CACCTCGACCATGGTAATAGC Cdh5 Trans F GGAGGCTGGAAAGTAGAGCA CreM R TCCCTGAACATGTCCATCAG RbpjLox F ATAATTTGCCAAGCCAAAGC RbpjLox R GCTCCCCACTGTTGTGAACT Dll4 lox (Duarte) GTGCTGGGACTGTAGCCACT Dll4 lox (Duarte) R TGTTAGGGATGTCGCTCTCC p21f_2B9 ACCCAGCAAAGCCTTGATTCT NeoF_8B6 CCTTCTATCGCCTTCTTGACGA p21r_3B2 CAGGTCGGACATCACCAGGAT 60C 206bp To detect the iMb-VEGFR2-Mosaic allele 60C 550bp To detect the iChr-Mosaic allele 60C 600bp To detect the iMb-Mosaic allele 60C 350bp Mutant Band 300 bp Wt band To detect Esm1 Knock-In 60C 350bp Mutant Band 250 bp Wt band To detect Rosa26 Knock-in 60C 550bp To detect the VE-cadherin(Pac)-CreERT2 allele 60C 600bp KO band 760bp Wt band To detect the p21 KO allele 60C 350bp Floxed band 200 bp Wt band To detect the Rbpj floxed allele 60C 500bp Floxed band 400 bp Wt band To detect the Dll4floxed allele Supplementary Table 1 Primers used to genotype mice Pontes-Quero et al.