Accelerated tumor growth and immature tumor blood vessels in Lyl1-deficient mice. (A) Accelerated growth of syngeneic tumors in Δ^Lyl1/Δ^Lyl1 mice. (Left) Palpable tumors were measured every day using a caliper. The tumor volumes (in cubic millimeters) were calculated according to the formula: volume = Pi/6 × (width)²  × (length). Shown is the volume of LLC tumors implanted in Δ^Lyl1/Δ^Lyl1 and in their WT and WT/Δ^Lyl1 littermates as a function of time. Error bars represent means ± SEM, WT (n = 7); WT/Δ^Lyl1 (n = 7); Δ^Lyl1/Δ^Lyl1 (n = 8). **P < .01. Data are representative of 4 separate experiments. (Right) Shown is the volume of B16-F10 tumors implanted in Δ^Lyl1/Δ^Lyl1 and in their WT littermates as a function of time. Error bars represent means ± SEM, WT (n = 9); Δ^Lyl1/Δ^Lyl1 (n = 6). *P < .05. (B) Enlargement of tumor blood vessels in Δ^Lyl1/Δ^Lyl1 mice. The entire surfaces of CD31-stained sections of LLC tumor from Δ^Lyl1/Δ^Lyl1 and from their WT and WT/Δ^Lyl1 littermates were visualized with a nanozoomer slide scanner controlled by the NDP.View software. (Left) Representative microscopy images of CD31 immunostaining of tumor sections showing enlargement of tumor blood vessels in Lyl1-deficient mice. Scale bar: 100 μm. (Right) Distribution of tumor blood vessels according to their Ferret diameter in WT, WT/Δ^Lyl1 and Δ^Lyl1/Δ^Lyl1 mice. 80 to 200 open vessels were selected per tumor section stained with anti-CD31 and their Ferret diameter was calculated using ImageJ software (National Institutes of Health). For each section, selected vessels were ranked according to their diameter and the following distribution was established: < 30 μm, 30-70 μm, and > 70 μm. Analyzed tumors: WT (n = 7), WT/Δ^Lyl1 (n = 6), Δ^Lyl1/Δ^Lyl1 (n = 9). **P < .01, using the χ² test. (C) Reduced pericyte coverage of tumor blood vessels in Δ^Lyl1/Δ^Lyl1 mice. (Left) Microscopy images illustrating the severe reduction of pericyte coverage of tumor blood vessels from Δ^Lyl1/Δ^Lyl1 mice. LLC tumor sections were double-stained for the endothelial cell marker CD31 (green) and for the pericyte marker NG2 (red). Scale bar: 100 μm. (Right) The extent of vessel coverage by pericytes was determined on 4 to 6 random fields by measuring, with ImageJ software, the proportion of CD31-positive vessels covered by NG2- or α-SMA–immunoreactive cells. NG2 coverage was calculated as the percentage of NG2-positive vessels compared with the number of CD31-positive vessels. The data are presented as the mean ± SEM. Analyzed individual tumors: WT (n = 7), WT/Δ^Lyl1 (n = 8); Δ^Lyl1/Δ^Lyl1 (n = 7). ***P < .001. Tumor sections were double-stained for the endothelial cell marker CD31 and for the pericyte marker α-SMA (see supplemental Figures 2-3). α-SMA coverage was calculated as the percentage of α-SMA–positive vessels compared with the number of CD31-positive vessels. The data are presented as the mean ± SEM. Analyzed individual tumors: WT (n = 4), Δ^Lyl1/Δ^Lyl1 (n = 6). ***P < .001. (D) Increased vascular permeability of tumors in Δ^Lyl1/Δ^Lyl1 mice Δ^Lyl1/Δ^Lyl1 and WT/Δ^Lyl1 mice with LLC tumor sizes between 200 and 500 mm³ were used to measure tumor vascular permeability by Evans blue extravasation as an index of albumin leakage. Evans blue diffused within tumor was quantified at 630 nm. The amount of EBD (micrograms per gram of dry weight) was calculated from standard curve to determine the concentration of extravasated dye in the tumor. The mean is represented as a black bar on this scatter plot. Each point represents data for a separate mouse: WT/Δ^Lyl1 (n = 7), Δ^Lyl1/Δ^Lyl1 (n = 6). *P < .05.