Genes and pathways driving glioblastomas in humans and murine disease models.

Human malignant gliomas arise from neural progenitor cells and/or dedifferentiated astrocytes. By now, they are genetically so well characterized that several murine glioma models have emerged that faithfully reiterate the typical histological features of the disease. In experimental animals, only one or two elements of the growth factor/Ras, PI3K/PTEN/PKB, p53/ARF/HDM2, and p16/Rb/cyclinD/CDK4 pathways are targeted. In human gliomas, many additional genes and pathways are targeted due to a most severe mutator phenotype that leads to the accumulation of countless epigenetic and genetic alterations. Changes that convey a growth advantage are selected for, leading to overgrowth of precursor cell populations with increasingly malignant tumor cell clones. While murine models represent a powerful tool for elucidating the role of genetic pathways, mechanisms of response and resistance to new therapeutic agents might be fundamentally different due to the high degree of genomic instability in the human disease. In fact, little is known about the molecular causes of genomic instability involved in gliomas, except for the rare Turcot's syndrome, O(6)-methylguanine-DNA methyltransferase, and the apurinic/apyrimidinic endonuclease Ape-1. Novel approaches that selectively exploit fundamental metabolic differences between tumor and normal cells have to consider these fundamental differences between human disease and presently available, highly sophisticated animal models.