H. Yamada et al., Cell injury unmasks a latent proangiogenic phenotype in mice with increased expression of FGF2 in the retina, J CELL PHYS, 185(1), 2000, pp. 135-142
Fibroblast growth factor-2 (FGF2) is a potent mitogen for vascular endothel
ial cells and exogenous administration of FGF2 stimulates angiogenesis. How
ever, increased expression of FGF2 in the retina does not cause angiogenesi
s. One possible explanation is that FGF2 may not be capable of initiating a
ngiogenesis unless it is administered in pharmacologic levels or there is c
oexpression of another angiogenic factor. Alternatively, there may be contr
ol mechanisms that sequester FGF2 in vivo, preventing it from manifesting i
ts in vitro angiogenic activity. We tested the first hypothesis by crossing
mice that express FGF2 in the retina with mice that express Vascular endot
helial growth factor (VEGF) in the retina. Surprisingly, despite comparable
levels of VEGF expression, mice that expressed both FGF2 and VEGF had sign
ificantly less neovascularization than mice that expressed VEGF alone. The
second hypothesis was tested by treating Rho/FGF2 transgenic mice with low-
intensity laser photocoagulation that disrupts photoreceptors, but does not
rupture Bruch's membrane, or intense laser that ruptures Bruch's membrane.
In Rho/FGF2 transgenics, but not wild type mice, choroidal neovascularizat
ion developed in areas of low-intensity laser. Both wild type and transgeni
c mice developed choroidal neovascularization in areas of intense laser tha
t ruptured Bruch's membrane, but the area of neovascularization was signifi
cantly greater in transgenics. These data suggest that increased retinal ex
pression of FGF2 is angiogenic only when it is accompanied by cell injury t
hat overcomes sequestration control mechanisms. (C) 2000 Wiley-Liss, Inc.