Depletion of endogenous c-Kit prevents in vivo EPC recruitment to neoformed vessels. (A) TECs transfected with the empty vector (left panel) or with the mbKitL construct (right panel) were analyzed by FACS. Light lines indicate preimmune mouse IgG used as negative control; dark lines, mbKitL expression. (B) Fluorescent EPCs were let to adhere to TECs, transfected with the empty vector or with the mbKitL construct. EPCs or transfected TECs were pretreated with K44, K45 (EPC), or anti-KitL (TEC) as indicated. Adherent EPCs were counted for statistical analysis (data are the mean ± SD; *P < .05 control vs experimental groups). (C) Sections of Matrigel plugs from SCID mice injected with untreated EPCs (i-ii), EPCs depleted of c-Kit (iii-iv), or EPCs treated with imatinib mesylate (v-vi) were analyzed by immunohistochemistry. To detect CSFE-positive EPCs, anti–Fluorescein/Oregon Green polyclonal Abs and horseradish peroxidase–labeled antirabbit were used as described in “Materials and methods.” Black arrows indicate vessels; red arrows indicate some positive EPCs. Magnification × 200 (i,iii,v); × 400 (ii,iv,vi). (D) The quantification of neovascularization was performed on hematoxylin-eosin–stained sections, and the results were expressed as percentage ± SD of the vessel area to the total Matrigel area. Five different sections from each individual experimental group (4 mice per group) were analyzed. (E) Matrigel plugs from different experimental groups (lane 1: untreated EPCs; lane 2: c-Kit–depleted EPCs; lane 3: imatinib mesylate–treated EPCs) were subjected to RNA extraction, and RQ-PCR was performed to assess the expression of the mbKitL on injected TECs. Fibroblasts were used as positive control (+). Three groups of mice (each of 4 mice) were used for the in vivo experiments.