NO is required for proper MT1-MMP homeostasis and function in migratory ECs. (A) Subconfluent MLECs from wild-type and eNOS-null mice were analyzed for MT1-MMP membrane expression by flow cytometry. The means plus or minus SD of the percent of MT1-MMP–positive cells from 5 independent experiments are represented. *P < .01 (versus eNOS+/+) (top). The Western blot shows MT1-MMP expression in total cell lysates from wild-type and eNOS-deficient MLECs (bottom). MT1-MMP expression in HUVECs is shown for comparison. (B) MT1-MMP subcellular localization was visualized by immunofluorescence staining in MLECs from wild-type and eNOS-deficient mice; arrowheads indicate MT1-MMP staining at membrane protrusions (top). The chart shows quantification of the percentage of cells showing MT1-MMP clusters at lamellipodia/filopodia versus pointlike staining (n = 80 cells counted from 3 independent experiments) (bottom). (C) MT1-MMP activity was detected as the area of COL I degradation (dark) by wild-type and eNOS-deficient and MT1-MMP–deficient MLECs (top). Arrowheads indicate individual endothelial cells. One representative experiment is shown. The chart shows quantification of COL I degradation areas from wt (+/+) and MT1-MMP– and eNOS-deficient cultures (-/-). The effect of stimulation with the eNOS activator bradykinin on COL I degradation is shown. Data are the means plus or minus SD (n = 5 for MT1-MMP and n = 4 for eNOS; bottom). Absolute value corresponding to FI = 1 is 94.5 μm². #P < .01 (versus +/+); *P < .01 (versus NS). (D) Top panel: Migration of wild-type and eNOS-null MLECs was analyzed in Transwell chambers after exposure of cells for 16 hours to the neutralizing anti–MT1-MMP mAb LEM-2/63 or control anti-CD31 mAb. Treatments with CCL2 and CXCL12 (10 nM) were included as controls of MT1-MMP–dependent and MT1-MMP–independent migration, respectively. Data are the means plus or minus SD of the FI above untreated wild-type cells from 3 independent experiments. The absolute value corresponding to FI = 1 was 18 migrated cells/field. *P = .05; **P < .02 (versus None); +P < .02 (versus No Ab); #P < .01 (versus eNOS+/+). Bottom panel: Cord formation by wild-type and eNOS-null MLECs was analyzed by seeding onto Matrigel in the presence of the anti–MT1-MMP or control anti-CD31 mAbs. Treatments with CCL2 and CXCL12 (10 nM) were included as controls of MT1-MMP dependence and independence, respectively. Cord formation was quantified after 6 hours. Data are the means plus or minus SD of the FI above tube formation by untreated wild-type cells from 5 (eNOS+/+) and 4 (eNOS-/-) independent experiments. The absolute value corresponding to FI = 1 was 48.21 cords/field. *P < .04 (versus None); +P < .01 (versus No Ab). eNOS-/- MLECs were significantly less efficient at forming cords than eNOS+/+ MLECs; #P < .04 (versus eNOS+/+None). All experiments were run in triplicate.