Targeted inactivation of the tyrosine kinase domain of Flt-1 in ES cells and mice. (A) Predicted structures of wild-type, targeted, and endogenously expressed soluble Flt-1 (Upper), as well as the restriction maps of flt-1 wild-type allele, replacement-type targeting vector, and the targeted allele after homologous recombination (Lower). The targeting vector contains a PGK-neo cassette (pPGKneopA) and the diphtheria toxin A gene in the upstream region of the construct. (B and C) Genotype of a litter from flt-1^TK+/− and flt-1^TK+/− interbreeding. Tail DNA was isolated and analyzed by PCR (B) and Southern blotting (C) by using the probe shown in A after digestion with EcoRI. Primers used also are shown in A as FP (forward primer), RP (reverse primer), and RN (reverse neo primer). (D) Northern blot analysis of poly(A) RNAs (3 μg/lane) extracted from wild-type, flt-1^TK+/−, and flt-1^TK−/− mice of whole embryos. Probes used were the 3′ half of mouse flt-1 cDNA (nucleotide residue 2643–4008) containing the tyrosine kinase domain (Upper Left) and the 5′ half of flt-1 cDNA (69–1919) (Lower Left) and neo sequence (Right). The arrow in the upper left indicates the full-length mouse flt-1 mRNA of 7.0 kb long. (E) Immunoblot analysis of proteins obtained from mouse lung with various antibodies (Ab). pre, preimmune rabbit serum; aS, Ab against the soluble Flt-1 specific for the carboxyl terminal 31 amino acids (arrowhead, Upper); aKI, Ab against the kinase insert that is encoded from exon 21 (arrow); aN, Ab against the amino terminal region of the Flt-1 extracellular domain; aTr, Ab specific to the carboxyl terminal sequence of truncated Flt-1 (arrow, Lower). Arrowheads show nonspecific bands (Lower).

Endothelial cell growth and vascular permeability in flt-1^TK−/− homozygous mice. (A) Growth stimulatory activity of 100 ng/ml of mouse VEGF on primary sinusoidal endothelial cells derived from wild-type and flt-1^TK−/− mice by using the MTS assay. The assay was performed in triplicate for one mouse, and four or five mice for each type were used. The relative mean absorbances at 490 nm after stimulation with VEGF were shown in comparison with the mean absorbances without VEGF as control (100%). The average value for wild-type mice without VEGF was 0.31 OD₄₉₀, and that for flt-1^TK−/− homozygous mice was 0.45 OD₄₉₀. This slight difference in the basal levels of two types of mice appears mostly because of the technical difficulty for the preparation of mouse liver sinusoidal endothelial cells. (B) VEGF-dependent increase in vascular permeability of flt-1^TK−/− mice. −VEGF, with 2 ml of 1% BSA-PBS solution; +VEGF, with 200 ng of mouse VEGF in 2 ml of 1% BSA-PBS solution. The absorbances at 605 nm after stimulation with VEGF or with BSA alone (control) were indicated (see Materials and Methods).

No detectable abnormalities in the endothelial cells and blood vessels in vivo and in vitro of flt-1^TK−/− mice. Wild-type (A, C, E, G, and I) and flt-1^TK−/− (B, D, F, H , and J) mice were used. Arrowheads show the structure of dosal aorta at E9.0 (A and B) and endothelial cell layers of the dosal aorta at E10.5 (C and D) in embryos. Arrows show blood vessels in brain (E and F) and lung (G and H) from adult mice. Embryos and organs were fixed with 10% formaldehyde, and sections were stained with hematoxylin-eosin. (E and F, Upper) Endothelial cells stained with antibody against von Willebrand factor in frozen sections of brain. (I and J) Spindle-like sinusoidal endothelial cells of liver were cultured with 100 ng/ml of VEGF.

Loss of VEGF- and PlGF-dependent macrophage migration in flt-1^TK−/− homozygous mice. (A) Reverse transcription–PCR used for the detection of flt-1 mRNA in mouse macrophages. Primers: #9F (exon 9 forward primer) and #11R (exon 11 reverse primer) for the extracellular domain (Top), #17F (exon 17 forward primer) and #18R (exon 18 reverse primer) for kinase domain (Middle), and mouse β actin primers (Bottom). (B) VEGF₁₆₅ or PlGF-2 at different concentrations, as well as Zymosan-activated serum (ZAS), were added to the lower chamber of the Boyden chamber apparatus. The results are mean percentages (±SD) of migrated cells in 10 high-power fields. For each experiment, 4–5 wild-type or flt-1^TK−/− homozygous mice were used, and each experiment was carried out in triplicate. The mean numbers of migrated macrophages without chemoattractant were shown as 100 and compared with those with chemoattractants. (The real mean numbers of migrated macrophages without chemoattractant were 120 in the wild-type and 112 in the flt-1^TK−/− homozygous mice.)