p16INK4A and p15INK4B in senescence, immortalization and cancer : gene transfer by adenovirus vectors

Abstract

The astrocytic gliomas are the most common form of brain tumors in humans. The malignant progression from low grade (astrocytoma, grade II) to high-grade (glioblastoma multiforme, grade, IV) astrocytomas is well documented and is accompanied by an increasing number of genetic aberrations. Many of these genetic alterations center round genes that regulate the G 1/S transition of the cell cycle. This is particularly the case with the genes encoding the cyclin dependent kinase inhibitors (CDKIs), p16 INK4A (p16) and p15 INK4B (p15). They are homozygously deleted in 40-50% of primary gliomas and in an even higher percentage in glioma cell lines. In addition, the p16 gene is frequently inactivated by point mutations and by DNA methylation, and mice that are deficient for the p16 gene spontaneously develop tumors within their first year of life. While the role of p16 in senescence and tumor suppression has been amply documented, the role of p 15 remained less clear.

Recombinant adenoviruses (Ad) expressing p15 (Adp15) or p16 (Adp16) were constructed and used to infect human glioma cells with different status of the p15, p16, and pRB genes. The results indicated that p15 is as potent as p16 in inducing cell cycle arrest and senescence-like changes in human glioma cells with an intact pRB. In addition, both p 15 and p 16 strongly inhibited telomerase activity in these cells. This shows the potential of p15 to function as a tumor suppressor and mediator of senescence. Together with the fact that p15 accumulates in T lymphocytes as they approach senescence and that p15 deficient mice display tumors of the hematopoietic cell system, these results suggest that p15 plays a role in regulating homeostasis in these (and perhaps other) cell systems.

The mechanism behind the synthesis of an alternative isoform of p 15 was also elucidated. This protein, termed p15.5, was found to be an N-terminally extended variant of p15, initiated from an upstream GUG codon. It was shown to separately, or in combination with p15, interact with both CDK4 and CDK6, and to induce cell cycle arrest and a senescent-like phenotype when transfected into human glioma cells.

The involvement of the p16/pRB pathway components in telomerase-mediated immortalization of bovine capillary endothelial (BCE) cells was determined. In these studies, transfection of the catalytic subunit of human telomerase (hTERT) alone was sufficient to immortalize BCE cells. Surprisingly, the telomere lengths in the hTERT-BCE cells were consistently shorter than in the senescent parental cells. pRB was hyperphosphorylated, and the expression of p16 (and p21CIP1) was repressed by promoter methylation. Reactivation of p16 by either Ad-mediated expression or by demethylation reversed the immortalized phenotype and induced senescence-like changes. These results suggest that the immortalization of BCE cells by hTERT is mediated through inactivation of crucial cellular senescence machineries including p161pR13 and p21CIP1, proteins that also by others have been implicated in inducing and maintaining a senescent phenotype in mammalian cells.

The ability of type C Ad to infect human cells is dependent on the expression of the coxsackie- and adenovirus receptor, hCAR. Since very little is known about the expression of hCAR in primary human gliomas and since hCAR has been implicated to function as a tumor suppressor, such expression studies were performed. A great variation in hCAR expression was detected, both in glioma cell lines and in the primary tumors, and the expression correlated well with Ad infectability. A significant decrease in the mean CAR expression levels was detected in the grade IV tumors as compared to the values for the grade II and grade III tumors. Moreover, a mean 12-fold higher expression of hCAR was detected in xenografts derived from glioblastomas, compared to the parental tumors. Interestingly, the two xenografts that did not show any upregulation of hCAR expression grew much faster than the hCAR-expressing cells. This points to an inverse correlation between growth rate and hCAR expression, and suggests that hCAR may contribute to suppression of tumor growth.