ANXA2 supports EC barrier function in vitro and vascular integrity in the hypoxic lung. (A) Permeability of confluent Anxa2^+/+ and Anxa2^−/− CMEC monolayers to 2,000-kD FITC–dextran assessed over 24 h (n = 3 replicates, Student’s t test). (B) TEER measured across confluent CMEC monolayers on days 2–8 (n = 3–9, Student’s t test). (C and D) CMEC monolayers were transduced with empty or ANX2-encoding adenovirus, and permeability to FITC–dextran (n = 6 replicates, ANOVA) was assessed over a 2-h period and TEER (n = 10 replicates, ANOVA) over 60 h. (E and F) Permeability of confluent Anxa2^+/+ and Anxa2^−/− LMEC monolayers to 40-kD Texas red–dextran (E) or 500-kD FITC–dextran (F) over 24 h (n = 3 replicates, Student’s t test). (G) EB was extracted from lung and brain from highly perfused Anxa2^+/+ and Anxa2^−/− normoxic or posthypoxic mice (n = 5–6 replicates, ANOVA). (H) Corresponding wet/dry weight organ ratios (n = 6 replicates, ANOVA), plasma VEGF levels (I; n = 6 replicates, ANOVA), and hemoglobin saturation (J; n = 7–16; replicates, ANOVA) were assayed. (K) Image analysis of sections (M) stained for extravascular albumin in five random images from each of three mice per group. Insets show no primary antibody control. Bars, 20 µm. (L) Densitometric analysis of immunoblots of whole-lung protein (N) from highly perfused lung tissue (n = 3 mice per group, ANOVA). Data in A, B, and C–N are representative of nine, six, and three experiments, respectively. Black and white squares or bars indicate Anxa2^+/+ and Anxa2^−/− cells or mice, respectively. Data are expressed as mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001.