Functional analysis of wild-type and malignant glioma derived CDKN2Aβ alleles: Evidence for an RB-independent growth suppressive pathway

Abstract

The tumor suppressor gene CDKN2A (p16/MTS1/INK4A), which encodes the cyclin-dependent kinase inhibitor p16^INK4a, is a target of 9p21 deletions during the malignant progression of human gliomas. This gene also encodes a second protein product (human p16β, murine p19^ARF), which originates from an unrelated exon of CDKN2A (exon 1β) spliced onto exon 2 in an alternate reading frame. Cell cycle arrest by p16β is caused by an as yet unidentified pathway. In order to test the candidacy of p16β as a glioma suppressor, we replaced p16^INK4a, p15^INK4b and p16β wild-type as well as a series of seven glioma-derived p16β alleles (R87H, A112V, R120H, A121V, G125R, A128A and A128V), into glioma cell lines that had either CDKN2A⁻/RB⁺ (U-87MG and U-251MG) or CDKN2A⁺/RB⁻ (LN-319) endogenous backgrounds and demonstrated that p16β can act as a functional glioma cell growth suppressor. Moreover, p16β, but not p16^INK4a or p15^INK4b inhibited the growth of RB-negative LN-319 cells, indicating that p16β likely exerts its effects through an RB-independent pathway. In vitro and in vivo assays of pRB phosphorylation were consistent with this interpretation. Since none of the glioma-derived p16β mutations inactivated their growth suppressive activities, it appears that mutations in CDKN2A exon 2 (which is shared in the coding sequences of p16^INK4a and p16β) likely exclusively target p16^INK4a.