Inefficient vascular supply and the resultant reduction in tissue oxygen tension often lead to neovascularization in order to satisfy the needs of the tissue. Examples include the compensatory development of collateral blood vessels in ischaemic tissues that are otherwise quiescent for angiogenesis and angiogenesis associated with the healing of hypoxic wounds. But the presumptive hypoxia-induced angiogenic factors that mediate this feedback response have not been identified. Here we show that vascular endothelial growth factor (VEGF; also known as vascular permeability factor) probably functions as a hypoxia-inducible angiogenic factor. VEGF messenger RNA levels are dramatically increased within a few hours of exposing different cell cultures to hypoxia and return to background when normal oxygen supply is resumed. In situ analysis of tumour specimens undergoing neovascularization show that the production of VEGF is specifically induced in a subset of glioblastoma cells distinguished by their immediate proximity to necrotic foci (presumably hypoxic regions) and the clustering of capillaries alongside VEGF-producing cells.

Vascular endothelial growth factor (VEGF) is a secreted endothelial cell-specific mitogen. To evaluate whether VEGF may play a role in angiogenesis, we have determined the spatial and temporal patterns of expression of VEGF and VEGF receptors during natural angiogenic processes taking place within the female reproductive system. Four angiogenic processes were ana-

Expression of vascular endothelial growth factor (VEGF), an endothelial cell-specific mitogen and a potent angiogenic factor, is upregulated in response to a hypoxic or hypoglycemic stress. Here we show that the increase in steady-state levels of VEGF mRNA is partly due to transcriptional activation but mostly due to increase in mRNA stability. Both oxygen and glucose deficiencies result in extension of the VEGF mRNA half-life in a protein synthesis-dependent manner. Viewing VEGF as a stress-induced gene, we compared its mode of regulation with that of other stress-induced genes. Results showed that under nonstressed conditions, VEGF shares with the glucose transporter GLUT-1 a relatively short half-life (0.64 and 0.52 h, respectively), which is extended fourfold and more than eightfold, respectively, when cells are deprived of either oxygen or glucose. In contrast, the mRNAs of another hypoxia-inducible and hypoglycemia-inducible gene, grp78, as well as that of HSP70, were not stabilized by these metabolic insults. To show that VEGF and GLUT-1 are coinduced in differentially stressed microenvironments, multicell spheroids representing a clonal population of glioma cells in which each cell layer is differentially stressed were analyzed by in situ hybridization. Cellular microenvironments conducive to induction of VEGF and GLUT-1 were completely coincidental. These findings show that two different consequences of tissue ischemia, namely, hypoxia and glucose deprivation, induce VEGF and GLUT-1 expression by similar mechanisms. These proteins function, in turn, to satisfy the tissue needs through expanding its vasculature and improving its glucose utilization, respectively.

Perfusion insufficiency, and the resultant hypoxia, often induces a compensatory neovascularization to satisfy the needs of the tissue. We have used multicellular tumor spheroids, simulating avascular microenvironments within a clonal population of glioma tumor cells, in conjunction with in situ analysis of gene expression, to study stress inducibility of candidate angiogenic factors. We show that expression of vascular endothelial growth factor (VEGF) is upregulated in chronically hypoxic niches (inner layers) of the spheroid and that expression is reversed when hypoxia is relieved by hyperoxygenation. Acute glucose deprivation-

VEGF production in the myocardium is significantly upregulated by hypoxia in vitro and by ischaemia in vivo. These results suggest that VEGF is a likely mediator in the natural process of ischaemia induced myocardial neovascularisation.

No abstract

SNP-based research strongly affects our biomedical and clinically associated knowledge. Nonunique and false-positive SNP existence in commonly used datasets may thus lead to biased, inaccurate clinically associated conclusions. We designed a computational study to reveal the degree of nonunique/false-positive SNPs in the HapMap dataset. Two sets of SNP flanking sequences were used as queries for BLAT analysis against the human genome. 4.2% and 11.9% of HapMap SNPs align to the genome nonuniquely (long and short, respectively). Furthermore, an average of 7.9% nonunique SNPs are included in common commercial genotyping arrays (according to our designed probes). Nonunique SNPs identified in this study are represented to various degrees in clinically associated databases, stressing the consequence of inaccurate SNP annotation and hence SNP utilization. Unfortunately, our results question some disease-related genotyping analyses, raising a worrisome concern on their validity.

No abstract