Role of ADTRP (Androgen-Dependent Tissue Factor Pathway Inhibitor Regulating Protein) in Vascular Development and Function

Abstract

Background The physiological function of ADTRP (androgen-dependent tissue factor pathway inhibitor regulating protein) is unknown. We previously identified ADTRP as coregulating with and supporting the anticoagulant activity of tissue factor pathway inhibitor in endothelial cells in vitro. Here, we studied the role of ADTRP in vivo, specifically related to vascular development, stability, and function. Methods and Results Genetic inhibition of Adtrp produced vascular malformations in the low-pressure vasculature of zebrafish embryos and newborn mice: dilation/tortuosity, perivascular inflammation, extravascular proteolysis, increased permeability, and microhemorrhages, which produced partially penetrant lethality. Vascular leakiness correlated with decreased endothelial cell junction components VE -cadherin and claudin-5. Changes in hemostasis in young adults comprised modest decrease of tissue factor pathway inhibitor antigen and activity and increased tail bleeding time and volume. Cell-based reporter assays revealed that ADTRP negatively regulates canonical Wnt signaling, affecting membrane events downstream of low-density lipoprotein receptor-related protein 6 ( LRP 6) and upstream of glycogen synthase kinase 3 beta. ADTRP deficiency increased aberrant/ectopic Wnt/β-catenin signaling in vivo in newborn mice and zebrafish embryos, and upregulated matrix metallopeptidase ( MMP )-9 in endothelial cells and mast cells ( MCs ). Vascular lesions in newborn Adtrp ^-/- pups displayed accumulation of MCs , decreased extracellular matrix content, and deficient perivascular cell coverage. Wnt-pathway inhibition reversed the increased mmp9 in zebrafish embryos, demonstrating that mmp9 expression induced by Adtrp deficiency was downstream of canonical Wnt signaling. Conclusions Our studies demonstrate that ADTRP plays a major role in vascular development and function, most likely through expression in endothelial cells and/or perivascular cells of Wnt-regulated genes that control vascular stability and integrity.

Keywords: vascular inflammation; vascular permeability; vascular stability, Wnt signaling, vascular function; vasculopathy.

Figure 1

Expression of adtrp in zebrafish embryos. A through E, Whole‐mount in situ hybridization for adtrp in zebrafish embryos using digoxigenin‐labeled riboprobes. A, Adtrp expression in anterior, medial, and posterior lateral plate mesoderm (lpm; arrows). B through D, Expression of adtrp is apparent in major blood vessels (arrows) at 16, 24, and 48 hours postfertilization (hpf). C, Right panel displays a cross‐section of 24‐hpf embryo confirming adtrp expression in dorsal aorta and cardinal vein. Bar: 50 μm. E, Negative control with sense riboprobe. F, mRNA expression of adtrp assayed by qPCR in zebrafish embryos through the embryonic development. n>30 biological replicates for each time point. cv indicates cardinal vein; da, dorsal aorta; nt, notochord.

Figure 2

Efficiency of adtrp knockdown in zebrafish embryos. A, Design of splice‐blocking morpholinos (MO) for adtrp1, the adtrp homologous gene on chromosome (Chr) 23; hereafter named “adtrp,” targeting exon2‐intron2 (e2i2) and exon3‐intron3 (e3i3) boundaries. B, mRNA expression of adtrp assayed by qPCR in 72 hours postfertilization (hpf) embryos shows 50% knockdown efficiency for both adtrp MOs used at 1 pmol MO/embryo. adtrp ^e2i2 MO, hereafter named “adtrp” MO, was used in all knockdown experiments unless otherwise stated. Statistics: 1‐way ANOVA with Bonferroni's multicomparison test. Values are mean±SEM fold‐change from control after correction for elongation factor 1a as internal control. ****P<0.0001. C, Effect of adtrp1 MO injection on mRNA expression of adtrp2 (adtrp homologous gene on Chr 13), assayed by qPCR in 72‐hpf embryos. Statistics: unpaired t test. Data are presented as for (B). NS indicates not significant, P>0.05. B and C, Experiments were repeated 3 times with qPCR technical duplicates. Each marker represents a group of at least 10 embryos.

Figure 3

Knockdown of adtrp causes vascular defects in zebrafish embryos. A, Bright‐field (left panels) and whole‐mount fluorescence confocal microscopy of 72 hours postfertilization (hpf) Tg(fli1:EGFP)y1 zebrafish embryos (right panels; EGFP, fluorescent vascular reporter) injected with scrambled (control) or adtrp morpholinos (MO)±human ADTRP mRNA. Human FLAG‐tagged ADTRP expression was verified by western blot (lower panel blot of 72‐hpf embryo lysates; protein loading: Ponceau Red staining). White arrows: intersegmental vessels (ISVs). White dashed line: caudal vein plexus (CVP) area. Bars: 100 μm. Quantification of vascular phenotype: scatter plots and statistics (1‐way ANOVA with Bonferroni's multicomparison test) of B the percentage of defective ISVs, and C the percentage area of CVP dilation normalized to tail area. Data are mean±SEM of biological replicates (n). ****P<0.0001 between all groups. D, Semithin sections at the CVP level (red dashed lines) of control (a) and adtrp MOs (b and c). White arrow: endothelial cells (ECs). Bars: a and b, 10 μm; c, 5 μm. E, O‐dianisidine staining for red blood cells in 72‐hpf embryos injected with adtrp or control MOs. Arrows: hemorrhage spots. F, Fluorescence confocal microscopy of 72‐hpf Tg(fli1‐EGFP)y1 embryos injected with adtrp ^e2i2+p53, adtrp ^e3i3, or p53 MOs shows nearly identical defective patterning of ISV (arrows) and dilated CVP (white dashed line areas) in all adtrp morphants, but not in the p53 MO–injected embryos. Bars: 100 μm. ADTRP indicates androgen‐dependent tissue factor pathway inhibitor regulating protein; EGFP, enhanced green fluorescent protein.

Figure 4

Expression of Adtrp in wild‐type (WT) mice. A, mRNA expression of Adtrp assayed by qPCR in WT mouse embryos is observed throughout embryonic development and increases with age. B, In situ hybridization for Adtrp in the lung (upper panels) and aorta (lower panel) of 2‐month‐old WT mice using digoxigenin‐labeled riboprobes shows Adtrp expression in endothelial cells (blue, left‐side panels). Right‐side panels: negative controls with sense riboprobe. Bars: 100 μm (lung), 50 μm (aorta). Adtrp indicates androgen‐dependent tissue factor pathway inhibitor regulating protein.

Figure 5

Generation and characterization of Adtrp global knockout mice. A, Bacterial artificial chromosome (BAC) gene targeting was used to simultaneously introduce loxP sites flanking both exons 2 and 5 of 9530008L14RIK gene (mouse Adtrp) to produce a frame‐shift event. The BAC vector also introduced FRT‐flanked Neo‐resistance cassette in intron 2 and Hyg‐cassette in intron 5. During the crossing of offspring positive for the mutant allele with ROSA26‐Flpe ⁺ mice to remove Hyg and Neo cassettes, Flpe recombinase also deleted the region between exons 3 and 5, thus generating Adtrp ^{Flpe‐Del/Flpe‐Del}, which was selected as the constitutive global knockout mouse, Adtrp ^−/−. B, Genomic DNA extracted from toe tips was subjected to PCR using primers pair #1, specific for LoxP1 site, to confirm the presence of lox allele. Homozygous Adtrp ^LoxP1/LoxP1 mice were further screened for the Flpe deletion by using primers pair #2, which amplifies a 511‐bp product only if the recombination occurred between exons 3 to 5. C, Agarose gel showing an example of genotyping (grayscale‐mode inverted image). D, mRNA expression of Adtrp assayed by qPCR in cephalic skin, lung, and heart of newborn (P₀) pups shows minimal expression in Adtrp ^−/− mice. Statistical analysis: unpaired t test. Values are fold‐change from wild‐type (WT; arbitrarily set to 1.0) after correction for β‐2 Microglobulin as internal control. Data are mean±SEM from experiments repeated at least 3 times. Markers represent means of qPCR technical duplicates performed on 5 biological replicates. E and F, Descriptive statistics derived from the breeding and backcrossing of Adtrp ^−/− mice shows partially penetrant lethality (35%) in newborn pups. E, Genotypic distribution among offspring of heterozygous (Het) pairs. F, Litter size and neo‐/perinatal death. Statistical analysis: 1‐way ANOVA with Bonferroni's multicomparison test. Data are mean±SEM. ****P<0.0001; NS, P>0.05. Adtrp indicates androgen‐dependent tissue factor pathway inhibitor regulating protein; bp, base pair; NS, not significant.

Figure 6

Newborn (P₀) Adtrp ^−/− mice display defective blood vessels. A, Gross morphology of cephalic skin in wild‐type (WT) and Adtrp ^−/− mice. B, Bright‐field microscopy showing vessel dilation and tortuosity, hemorrhage, and abnormal branching (arrows) in various organs. C, Immunofluorescence staining with anti‐CD31 (FITC, green; endothelial cell marker) IgG and whole‐mount confocal microscopy in WT and Adtrp ^−/− mice. Bars: 100 μm. D, Immunofluorescence whole‐mount confocal microscopy with anti–red blood cell (RBC) TER‐119 (FITC, green) IgG in the dura mater. Arrows: microhemorrhage. Bars, 50 μm. IgG indicates immunoglobulin G.

Figure 7

Newborn (P₀) Adtrp ^−/− mice display leaky vessels and defective endothelial cell junctions. A, Whole‐mount confocal fluorescence microscopy of P₀ pups cephalic skin after infusion of the permeability tracer, Sulfo‐NHS‐Biotin (SNB), and staining with Streptavidin‐Cy3 (red, left‐side panels). Red arrows: vascular leakage. Blue: TO‐PRO3 nuclear staining. Cy3‐channel presented in grayscale mode in the right‐hand panels for better visualization of the fluorescence. Bars: 100 μm. B, Double immunofluorescence with anti‐Fibrin (Cy3, red) and anti‐CD31 (FITC, green) IgGs, and whole‐mount confocal microscopy of skin from wild‐type (WT) and Adtrp ^−/− P₀ mice. Blue: nuclei. Red arrows in “Fibrin” panels: extravascular fibrin. Bars, 50 μm. C, Double immunofluorescence with anti‐VE‐Cadherin (VEC; Cy3, red) and anti‐CD31 (FITC, green) IgGs, and whole‐mount confocal microscopy (as for B) in WT and Adtrp ^−/− mice. Blue, nuclei. Bars, 10 μm. D, Double immunofluorescence whole‐mount confocal microscopy with anti‐Claudin‐5 (FITC, green) and anti‐CD31 (Cy3, red) IgGs in cephalic skin shows low levels of claudin‐5 (grayscale mode) in Adtrp ^−/− pups. Bars, 50 μm. E, Mean fluorescence intensity (MFI) and statistics (mean±SD by unpaired t test) of SNB/Cy3 (images in A). Each marker represents the mean of at least 10 regions of interest measured on 5 images/condition and expressed as arbitrary units (AU). ****P<0.0001. F MFI and statistics as for E of VEC immunostaining (images in C). **P<0.01. G mRNA expression of VEC (Cadherin5, Cdh5) and Claudin‐5 (Cldn5) in cephalic skin from P₀ pups. Values are fold‐change from WT (arbitrarily set to 1.0) after correction for β2‐Microglobulin as internal control. Statistics: mean±SEM by unpaired t test. **P<0.01. IgG indicates immunoglobulin G.

Figure 8

Newborn (P₀) Adtrp ^−/− mice display defective perivascular cell (PVC) coverage. A, Double immunofluorescence whole‐mount confocal microscopy with anti‐Desmin (Cy3, red) and anti‐CD31 (FITC, green) IgGs in the cephalic skin (upper row) and brain (lower row) of wild‐type (WT) and Adtrp ^−/− pups. Cy3‐channel in grayscale mode and red arrows in inserts indicate desmin disorganization and deficient mural cells coverage. Bars, 50 μm. B, Double immunofluorescence whole‐mount confocal microscopy with anti‐PDGFRβ (Cy3, red) and anti‐CD31 (FITC, green) IgGs in the skin illustrates decreased levels of red fluorescence intensity and disorganization around dilated vessels in Adtrp ^−/− pups (grayscale mode). Bars, 50 μm. IgG indicates immunoglobulin G; PDGFRβ, platelet‐derived growth factor receptor beta.

Figure 9

Adult Adtrp ^−/− mice display mild vascular dysfunction. A, Bright‐field microscopy of lung sections of young adult mice (2–3 months old) stained with hematoxylin and eosin (H&E) shows bleeding areas (red arrows) in Adtrp ^−/− mice. Black asterisks: blood vessels. Bars, 500 μm. B, Bright‐field microscopy of lung sections as in (A) stained with phosphotungstic acid (PTA) shows extravascular fibrin deposition (purple arrows) in Adtrp ^−/− mice. Bars, 100 μm. C, Activated partial thromboplastin time (aPTT), prothrombin time (PT), and thrombin‐antithrombin (TAT) complexes measured in plasma from young adult (2–3 months old) wild‐type (WT) and Adtrp ^−/− mice. Statistics: unpaired t test. Data are mean±SEM. NS, P>0.05. D, TFPI antigen (left panel) and specific activity (right panel) are both decreased in plasma from young adult (2–3 months old) WT and Adtrp ^−/− mice. TFPI activity assay used human coagulation factors and standard curve made with serial dilutions of pooled normal mouse plasma. E, TFPI antigen (left panel) and specific activity (right panel; as in D) measured in lung extracts from adult WT and Adtrp ^−/− mice. Statistical analysis: unpaired t test. Data are mean±SEM. ****P<0.0001; ***P<0.001; NS, P>0.05. C through E, Markers are mean values of technical duplicates performed on n biological replicates. F, Tail bleeding time (left panel) and volume of blood loss measured as hemoglobin content (right panel) are significantly increased in Adtrp ^−/− mice. Statistical analysis: unpaired t test. Data are mean±SEM. n, biological replicates. ****P<0.0001. NS indicates not significant; OD, optical density; TFPI, tissue factor pathway inhibitor.

Figure 10

Ultrastructural analysis of the defective blood vessels in Adtrp ^−/− newborn mice. Transmission electron microscopy of P₀ pups cephalic skin (A through E) and tail (G and H) of wild‐type (WT; A and G) and Adtrp ^−/− (B through F and H) mice. Red asterisks: severe extravascular edema (B, C, E, H). Black asterisks: degraded extracellular matrix (ECM), distorted and amorphous basement membrane (BM) with large intercellular spaces (C and D). Black arrows: defective perivascular cell coverage (B, C, E). Blue EC*: endothelial cells with heterogeneous shape and electron density (B, C, D, E). Mast cells (MC) accumulate in areas with vascular defects (F and H). Bars: (A) to (F) 5 μm, (G) 50 μm, and (H) 10 μm. BV indicates blood vessel; P, pericyte; RBC, red blood cells.

Figure 11

Evidence of decreased and/or degraded basement membrane in cephalic skin of P₀ Adtrp ^−/− mice. A, Double immunofluorescence with anti‐Laminin (Cy3, red) and anti‐CD31 (FITC, green) IgGs, and whole‐mount confocal microscopy in wild‐type (WT) and Adtrp ^−/− mice. Blue, nuclei. Bars, 50 μm. Histogram: mean fluorescence intensity (MFI) of laminin measured on 5 images/condition and expressed as mean±SD arbitrary units (AU). Statistics: unpaired t test. Each marker represents the mean of at least 10 regions of interest. ****P<0.0001. B, Total lysates of skin from WT and Adtrp ^−/− pups analyzed by SDS‐PAGE and immunoblotted with anti‐Laminin and anti‐β‐Actin antibodies (left panel, representative western blot from 3 biological replicates). Histogram: semiquantitative densitometry and statistics (mean±SEM by t test). Markers represent biological replicates. *P<0.05. C, Double immunofluorescence with anti‐Collagen‐IV (Cy5, blue; or white in grayscale mode) and anti‐erythroid cells TER‐119 (FITC, green) IgGs, and whole‐mount confocal microscopy in WT and Adtrp ^−/− mice. Arrows in inset: vessel dilation, microhemorrhage, and collagen disruption. Bars, 50 μm. For all panels, grayscale‐mode images are presented for better visualization of fluorescence intensity. IgG indicates immunoglobulin G; RBC, red blood cells.

Figure 12

Increased mast cell (MC) accumulation in newborn (P₀) Adtrp ^−/− mice. A, Transmission electron microscopy of the tail from wild‐type (WT) and Adtrp ^−/− pups. Yellow asterisks: MC accumulation. Black arrows: degranulated MC. Blue arrow: extravasated red blood cells. Red asterisks: decreased extracellular matrix density and enhanced edema. Bars, 10 μm. B, Bright‐field microscopy on whole‐mount diaphragms stained with Toluidine Blue showing MC (purple). Histogram: quantification and statistics (mean±SEM by unpaired t test) of MC numbers in the tail and diaphragm of P₀ pups (6 each). Markers are numbers of MCs counted in n fields of view (FOV) from at least 5 images/condition. ***P<0.001; ****P<0.0001. C, Double immunofluorescence with anti‐MC Tryptase (Cy3, red) and anti‐CD31 (FITC, green) IgGs, and whole‐mount confocal microscopy in cephalic skin of WT and Adtrp ^−/− newborn mice. Bars, 50 μm. BV indicates blood vessel; IgG, immunoglobulin G.

Figure 13

ADTRP deficiency increases canonical Wnt signaling. A, Total lysates of EA.hy926 ECs expressing nontargeting siRNA (siControl) or siADTRP analyzed by SDS‐PAGE and immunoblotted with anti‐ADTRP and anti‐GAPDH IgGs. Shown: representative western blot from 3 experiments. B, mRNA expression of LEF1 and AXIN2 assayed by qPCR in siControl or siADTRP‐expressing EA.hy926 ECs. Gray bars: treatment with Wnt3a conditioned medium for 3 hours. Values (mean±SEM by 1‐way ANOVA with Bonferroni's multicomparison test) are fold‐change from siControl (arbitrarily set to 1.0) after correction for β‐2 MICROGLOBULIN or 18S rRNA as internal controls. *P<0.05; **P<0.01; ****P<0.0001. C, Total lysates of siControl and siADTRP‐ EA.hy926 EC analyzed by SDS‐PAGE and immunoblotted with anti‐active β‐Catenin and anti‐GAPDH IgGs. Shown: representative western blot from 3 experiments. D, Immunofluorescence and confocal microscopy with anti‐active β‐Catenin (Cy3, red) IgG on siControl and siADTRP‐EA.hy926 ECs, and scatter plot and statistics of mean fluorescence intensity (MFI) expressed as mean arbitrary units (AU)±SD (unpaired t test) measured within nuclei (n cells/condition analyzed in at least 5 images recorded from 3 different experiments). White arrow: nuclear active β‐catenin. Blue, nuclei. Bars, 10 μm. ****P<0.0001. ADTRP indicates androgen‐dependent tissue factor pathway inhibitor regulating protein; ECs, endothelial cells; GAPDH, glyceraldehyde 3‐phosphate dehydrogenase; IgG, immunoglobulin G; LEF1, lymphoid enhancer binding factor 1; rRNA, ribosomal RNA; siRNA, small interfering RNA.

Figure 14

ADTRP negatively regulates Wnt/β‐catenin pathway. A, Western blot with anti‐FLAG and anti‐GAPDH IgGs in total lysates of HEK293T cells transfected with control or ADTRP‐FLAG plasmids. Shown: representative blot from 3 experiments. B, mRNA expression of LEF1 and AXIN2 in control and ADTRP‐expressing HEK293T cells. Gray bars: treatment with Wnt3a conditioned medium for 24 hours. Values (mean±SEM by 1‐way ANOVA with Bonferroni's multicomparison test) are fold‐change from control (arbitrarily set to 1.0) after correction for β‐2 MICROGLOBULIN as internal control. *P<0.05; **P<0.01. C, TOPFlash luciferase reporter activity and statistics (unpaired t test) in HEK293T cells expressing empty vector (control) or ADTRP‐FLAG, incubated with Wnt3a medium for 24 hours (gray bars). Data are mean±SEM. ****P<0.0001. D, Total lysates of HEK293T cells analyzed by SDS‐PAGE and immunoblotted with anti‐active β‐Catenin and anti‐GAPDH IgGs. E, Immunofluorescence and confocal microscopy with anti‐active β‐Catenin (Cy3, red) IgG and scatter plot and statistics of mean fluorescence intensity (MFI) expressed as mean arbitrary values (AU)±SD (unpaired t test) measured within nuclei (n cells/condition analyzed in at least 5 images recorded from 3 different experiments). White arrow: nuclear active β‐catenin. Blue, nuclei. Bars, 10 μm. ****P<0.0001. F, TOPFlash luciferase reporter activity and statistics in HEK293T cells incubated with 10 mmol/L of LiCl for 24 hours. Data are mean±SEM by unpaired t test. NS, P>0.05. G and H, TOPFlash luciferase reporter activity and statistics (1‐way ANOVA with Bonferroni's multicomparison test) in control and ADTRP‐expressing HEK293T cells at 48 hours after transfection with either full‐length β‐CATENIN or active β‐CATENIN–expressing plasmids (G, gray bars); or (H) constitutively active LRP6ΔN plasmid (gray bars). Data are mean±SEM. NS, P>0.05; ****P<0.0001. For all panels: Markers represent different experiments. ADTRP indicates androgen‐dependent tissue factor pathway inhibitor regulating protein; GAPDH, glyceraldehyde 3‐phosphate dehydrogenase; IgG, immunoglobulin G; LRP6, low‐density lipoprotein receptor‐related protein 6; LEF1, lymphoid enhancer binding factor 1; NS, not significant.

Figure 15

Ectopic and/or aberrant canonical Wnt signaling in newborn Adtrp ^−/− mice. A, Double immunofluorescence and whole‐mount confocal microscopy with anti‐CD31 (FITC, green) and anti‐β‐Galactosidase (β‐GAL; Cy3, red) IgGs in cephalic skin (upper panels) and brain (lower panels) of wild‐type (WT) BAT‐GAL ^z+ and Adtrp ^−/− BAT‐GAL ^z+ pups. Blue, nuclei. Pink arrows: β‐GAL in defective vessels. Bars, 50 μm. B, mRNA expression of Tcf4 in cephalic skin. Values are fold‐change from WT (arbitrarily set to 1.0) after correction for β‐2 Microglobulin as internal control. C, Representative western blot (upper panel) with anti‐active β‐Catenin and anti‐ β‐Actin IgGs in total lysates from WT and Adtrp ^−/− cephalic skin. Histogram (lower panel): semiquantitative densitometry. D, mRNA expression of Lef1 in primary fibroblasts cultured from WT and Adtrp ^−/− cephalic skin explants. Values are fold‐change from WT (arbitrarily set to 1.0) after correction for β‐2 Microglobulin as internal control. B through D, Statistics: unpaired t test. Data are mean±SEM. **P<0.01, ****P<0.0001. Markers represent biological replicates. Adtrp indicates androgen‐dependent tissue factor pathway inhibitor regulating protein; IgG indicates immunoglobulin G; Lef1, lymphoid enhancer binding factor 1; Tcf4, transcription factor 4.

Figure 16

Adtrp deficiency leads to increased expression of matrix metallopeptidase‐9 (Mmp9) mRNA, protein, and activity levels in zebrafish embryos and newborn mice. A, Whole‐mount in situ hybridization for mmp9 (blue) in 72 hours postfertilization (hpf) control and adtrp zebrafish morphants. Blue arrow: specific signal accumulation. Black arrow: caudal vascular plexus (CVP). n, biological replicates. Bars, 100 μm. B, Immunofluorescence staining with anti‐Mmp‐9 (Cy3, red) IgG on cross‐sectioned CVP (green) of 72‐hpf Tg(fli1:EGFP)y1 (EGFP, fluorescent vascular reporter) control and adtrp morphants±human ADTRP mRNA. Bars, 100 μm. C, Total lysates of 72‐hpf zebrafish embryos incubated ±100 μmol/L of pan‐MMP inhibitor GM6001 for 24 hours and analyzed by 10% gelatin gel electrophoresis. Left: representative zymogram (grayscale inverted image). Histogram: semiquantitative densitometry and statistics (mean±SEM by t test). *P<0.05. Each marker represents a group of at least 10 embryos. D, In situ zymography with MMP fluorescence substrate QXL™ 570‐KPLA‐Nva‐Dap(5‐TAMRA)‐AR‐NH2 (red, dequenched fluorescence) on cross‐sectioned CVP (green) of 48‐hpf Tg(fli1:EGFP)y1 embryos injected with Control or adtrp morpholinos (MOs). Arrows: MMP activity. E, Fluorescence confocal microscopy of 72‐hpf Tg(fli1:EGFP)y1 embryos injected with control or adtrp MOs (left panels; EGFP, fluorescent vascular reporter)±GM6001 as for C. Bars, 100 μm. Quantification of vascular defects (right panels): scatter plots and statistics (1‐way ANOVA with Bonferroni's multicomparison test) of the percentage of defective intersegmental vessels (ISVs) and percentage area of CVP dilation normalized to tail area. Data are mean±SEM of biological replicates (n). ****P<0.0001 between all groups. F, Double immunofluorescence and whole‐mount confocal microscopy with anti‐MMP‐9 (Cy3, red) and either anti‐CD31 or anti‐GR‐1 (FITC, green) in cephalic skin; or with anti‐MMP‐9 (Cy3, red) and anti‐mast cell (MC) Tryptase (FITC, green) in lung of wild‐type (WT) and Adtrp ^−/− P₀ pups. MMP‐9 associates with both CD31 and GR‐1–positive cells in the skin (lower panels, white arrow and inset), and with MC (lung panels, yellow arrows). Bars, 100 μm. G, Zymography (as in C) of conditioned media of primary fibroblasts cultured from P₀ WT and Adtrp ^−/− cephalic skin explants. Upper panel: representative zymogram (grayscale inverted image). Histogram: semiquantitative densitometry and statistics using t test. Data are mean±SEM of biological replicates. ****P<0.0001. Adtrp indicates androgen‐dependent tissue factor pathway inhibitor regulating protein; EGFP, enhanced green fluorescent protein; IgG indicates immunoglobulin G.

Figure 17

ADTRP deficiency‐induced matrix metallopeptidase‐9 (MMP‐9) is regulated by canonical Wnt signaling. A, Double immunofluorescence with anti‐MMP‐9 (FITC, green) and anti‐β‐Galactosidase (β‐GAL; Cy3, red), and confocal microscopy in whole‐mount P₀ pups lung of Adtrp ^−/− BAT‐GAL ^z+ and wild‐type (WT) BAT‐GAL ^z+. Arrows: partial expression of β‐GAL (active β‐catenin reporter) in MMP‐9–producing cells. Blue, nuclei. Bars, 50 μm. B, mRNA expression of Mmp9 in primary fibroblasts cultured from WT and Adtrp ^−/− P₀ pups cephalic skin explants. Gray bars: treatment with Wnt3a medium for 24 hours. Values are fold‐change from WT (arbitrarily set to 1.0) after correction for β‐2 Microglobulin as internal control. Data are mean±SEM by 1‐way ANOVA with Bonferroni's multicomparison test. Markers represent mean values of 3 replicate cell cultures isolated from 2 mice each. **P<0.01; ***P<0.001. C, mRNA expression of MMP9 in control and siADTRP‐EA.hy926 ECs incubated with Wnt3a medium for 24 hours. Markers represent different experiments and are fold‐change from siControl (arbitrarily set to 1.0) after correction for β‐2 MICROGLOBULIN as internal control. Data are mean±SEM by unpaired t test. ****P<0.0001. D, mRNA expression of mmp9 in control and adtrp morphants at 72 hours postfertilization after incubation with Wnt inhibitor IWR‐1 (20 μmol/L) for 8 hours. Values are fold‐change from control (arbitrarily set to 1.0) after correction for β‐actin or elongation factor 1a as internal controls. Each marker represents a group of at least 10 embryos. Data are mean±SEM by 1‐way ANOVA with Bonferroni's multicomparison test. **P<0.01; ****P<0.0001. ADTRP indicates androgen‐dependent tissue factor pathway inhibitor regulating protein; ECs, endothelial cells.