Fig 2: ATF4 Regulates JAG1 While JAG1 Regulates Angiocrine Factor Expression. (A-B) ChIP analysis with control rabbit IgG (ctrl IgG) and anti-ATF4 show JAG1 promoter region contains one CRE site approximately 0.6 kb upstream that was bound by ATF4 more in EC cells (MS1) exposed to ICN1 than in ECs exposed to WT bone marrow cells. (C-E) ATF4 over-expressing (C), ATF4 knockdown (D-E), or JAG1 knockdown (F-G) ECs were co-cultured with ICN1 cells (1.5 X106) for 24h. qRT-PCR of transcripts were standardized for beta-actin and expressed as fold changes relative to those in ECs transfected with control siRNA or pcDNA plasmid (n=6/group from 2 experiments) (D & F). siRNA knockdown efficiencies and down-regulation of JAG1 after ATF4 knockdown were shown by western blots (E & G). (H) Representative FACS profile of apoptotic ICN1 cells co-cultured with ECs expressing control or JAG1 siRNA from 3 similar experiments. Data shown in right are mean percentage of apoptotic WT or ICN1 cells co-cultured with ECs expressing control or JAG1 siRNA (n=4). Data shown in B-H are mean ± SD. Student t test was performed; *p<0.05; ** p<0.01.

Fig 3: STK11 ensures the ciliated cell fate commitment.a Immunofluorescence staining with antibodies against FOXJ1 and SCGB1A1 in the E16.5 lungs. White arrowheads indicate cells that do not express SCGB1A1 and FOXJ1. b The proportion of SCGB1A1−FOXJ1− cells (Left), and the proportion of FOXJ1+ cells (Right) in the intrapulmonary airways of E16.5 lungs (n = 3). c Immunofluorescence staining with antibodies against FOXJ1 and SCGB1A1 in the E18.5 lungs. White arrowheads indicate cells that do not express SCGB1A1 and FOXJ1. d The proportion of SCGB1A1−FOXJ1− cells (Left), and the proportion of FOXJ1+ cells (Right) in the intrapulmonary airways of E18.5 lungs (n = 3). e Immunofluorescence staining with antibodies against acetylated-α-tubulin and JAG1 (JAGGED1) in E16.5 lungs. Yellow arrowheads indicate the acetylated-α-tubulin− cells that express JAG1. f The proportion of JAG1+ cells in the intrapulmonary airways of E16.5 lungs (Control, n = 4; Shh-Stk11, n = 4). g Immunofluorescence staining with antibodies against acetylated-α-tubulin and JAG1 in E18.5 lungs. Yellow arrowheads indicate the acetylated-α-tubulin− cells that express JAG1. h The proportion of JAG1+ cells in the intrapulmonary airways of E18.5 lungs (Control, n = 4; Shh-Stk11, n = 5). ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001. Data shown in the graphs are means ± SEM. Student’s t-test. Scale bars: 11 μm

Fig 4: Repression of Gucy1a3 and Gucy1b3 correlates with changes in the Notch coactivators Fryl and Maml2. Panels A through D show correlations between Fryl and Maml2 and the sGC subunits Gucy1a3 and Gucy1b3 at the mRNA level. Data is from the microarray experiment. P-values and Spearman Rho coefficients are given in the respective panels. Correlations were also observed when AngII-treated mice were considered individually (Gucy1a3: PAngII = 0.005, Pvehicle = 0.06; Gucy1b3: PAngII = 0.015, Pvehicle = 0.24). Panel E shows western blots (n = 6/group) for aortic NOTCH2, JAG1, FRYL and MAML2 in vehicle and AngII-infused mice. Coomassie-stained proteins remaining on the gel after transfer was used as loading control for NOTCH2. HSP90 and GAPDH were used as loading controls for JAG1, FRYL and MAML2. Panels F and G show summarized western blot data for FRYL and MAML2, respectively (versus HSP90, *P < 0.05).

Fig 5: T-ALL Development Induced Up-regulation of JAG1 and Down-regulation of SCF and CXCL12 in ECs. (A-B) Representative FACS profile of bone marrow ECs (Lin-TER119-CD31+) and EC-expressing JAG1 (A). Frequency of bone marrow CD31+ ECs and mean MFI of JAG1 determined from 4 similar experiments (B). (C) Representative whole-mount imaging of CD31/CD144+ bone marrow vasculature in the control WT and ICN1 mice (left panel). Dilated small arteries and coalesced vessels were highlighted by arrows and dotted circles, respectively. Right panels plot the changes in CD31/CD144 volume (%) by whole-mount imaging. (D) Representative EC (CD45-TER119-CD31+) expression of CXCL12 by FACS analysis in control mice or in ICN1 leukemia mice. Mean MFIs were calculated from 4 similar experiments. (E) BMECs were co-cultured with control or ICN1 cells (1.5 X106) for 24h. qRT-PCR of EC expressions of transcripts were standardized for beta-actin and expressed as fold changes relative to those in EC co-cultured with control marrow cells (n=6/group from 2 experiments). (F) Representative western blot (left) with anti-JAG1 of BMECs co-cultured with control or ICN1 cells. Quantification of JAG1 expression normalized to β-actin from 3 similar experiments (right). Data shown in A, B, C (right), D-F are mean ± SD (n=3-6/group from 2-4 experiments). Student t test was performed; *p<0.05; ** p<0.01.