After the development of a primary vascular network, the developing embryo requires the formation of additional blood vessels or angiogen-esis. This process is largely driven by hypoxia, which serves as a stimulus for the release of angiogenic growth factors. One of the main classes of transcription factors that promote this process is the basic helix-loop-helix (bHLH) PAS domain family. A prototype member of this family is the arylhydrocarbon-receptor nuclear translocator (ARNT) (10). ARNT forms a heterodimeric complex with another PAS transcription factor, hypoxia-induced factor (HIF)-la (11). In response to oxygen deprivation, these transcription factors stimulate the expression of such angiogenic factors as vascular endothelial growth factor (VEGF) (12). Targeted disruption of the ARNT gene results in embryonic lethality by d 10.5 (13). Although a primary vascular network forms, the predominant defective angiogenesis occurs in the yolk sac and branchial arches, and overall growth of the embryos is stunted. These defects are similar to those observed in VEGF or tissue factor-deficient mice (14,15). Thus, although the primary vascular network develops, the angiogenic responses to hypoxia are severely impaired. Similar findings are observed in HIF-la knockout mice in which embryonic lethality occurs by d 10.5 as a result of cardiac and vascular malformations (16). Although neither of these transcription factors is expressed in a vascular-specific way, their roles in angiogenesis and vascular development are primarily related to their ability to stimulate the production of angiogenic factors such as VEGF in response to hypoxia. A third member of this family of transcription factors, endothelial PAS domain protein 1 (EPAS1), was recently identified (17). EPAS is predominantly expressed in endothelial cells and can also heterodimerize with ARNT. Targeted disruption of the EPAS gene has been evaluated by two different groups, resulting in two different phenotypes (18,19). Tian et al. (18) detected abnormalities in catecholamine homeostasis in EPAS-/- mice and no distinct abnormalities in blood vessel formation, whereas Peng et al. (19) identified vascular defects at later stages of embryogenesis during vascular remodeling in their EPAS-/- mice. The differences in the phenotype cannot be attributed to differences in targeting construct, since both groups disrupted the expression of the bHLH domain, but were more likely attributed to differences in the strain of the mice or subtle differences in the embryonic stem (ES) cells used. Although the formation of a primary vascular network or vasculogenesis occurs, later defects in vascular remodeling are observed during large vessel formation associated with hemorrhaging and the inability of the vessels to fuse properly. This suggests that all three of these PAS family members play a similar role in facilitating later stages of vascular remodeling and angiogenesis in the developing embryo.
Modulation of the function of HIF-1 a is also achieved by interaction with other proteins. The transcriptional adapter proteins p300 and CREB-binding protein (CBP) form a multiprotein/DNA complex together with HIF-1 a on the promoters of the VEGF and erythropoietin genes to promote expression of these genes in response to hypoxia (20). CBP-defi-cient mice exhibit abnormalities in both vasculogenesis and angiogenesis (21). In contrast, the von Hippel-Lindau tumor suppressor protein (pVHL) has been shown to promote proteolysis of HIF-1 a through ubiquitylation under normoxic conditions. Defective VHL function is associated with cancers that exhibit dysregulated angiogenesis and upregulation of hypoxia inducible genes (22).
The signaling mechanisms by which hypoxia activates HIF-1 a are beginning to be elucidated. The catalytic subunit of PI3-kinase, p110, plays a pivotal role in the induction of HIF-1 activity in response to hypoxia (23). Both induction of VEGF gene expression and HIF-1 a activity in response to hypoxia could be blocked by the addition of a PI3-kinase inhibitor. Further support of this concept comes from experiments in which VEGF gene expression and HIF-1 activity is induced by cotransfection of p110. Other studies have recently demonstrated that HIF-1 a activity may also be modulated by the mitogen-activated protein kinases p42 and p44 (24).
Was this article helpful?