Background:The cyclin-dependent kinase inhibitor p27 is a central regulator of cell cycle progression, whose function is perturbed in many human cancers, either due to decreased expression or abnormal localization. p27 levels are negatively correlated with Skp2, the F-box protein of SCFSKP2, a E3 ubiquitin ligase targeting nuclear p27 for degradation. Skp2 has been shown to cooperate with mutant Ras in in vitro transformation assays, implicating Skp2 as a bona fide oncogene. In chronic myeloid leukemia cell lines, p27 is down-regulated in a Bcr-Abl dependent fashion, while cytoplasmic accumulation has been described in primary CML cells. We herefore hypothesized that Bcr-Abl may regulate p27 via Skp2.
Experimental approach and results:Mo7e-p210BCR-ABL treated with 2.5 μM imatinib arrested in G0/G1 in a time-dependent manner (53.6±2, 58.3±2, 71.9±1% at 4, 8 and 16h), correlated with reduced in vitro kinase activity of Cdk2 (32% of controls at 16h). Western blot analysis showed a marginal increase of cytoplasmic p27 and 2.5-fold accumulation of nuclear p27 that preceded G0/1 arrest. Despite the reduced Cdk2 activity, most p27 was phosphorylated on T187, a target of cdk2/cyclinE, suggesting reduced degradation of phospo-p27 (T187). Degradation of nuclear p27 is mediated by SCFSKP2 and degradation of cytoplasmic p27 by the recently discovered KPC complex. We therefore assayed expression SCFSKP2 components and KPC1/2 by immunoblot analysis of imatinib-treated cells. Skp2 expression was greatly reduced compared to controls while expression of other SCFSKP2 components and KPC1/2 was unchanged, consistent with up-regulation of nuclear but not cytoplasmic p27 and suggesting a central role of Skp2 in mediating p27 degradation in Bcr-Abl positive cells. To test whether Skp2 is crucial for Bcr-Abl-driven leukemogenesis, we infected bone marrow of Skp2+/+ and Skp2−/− mice with BCR-ABL retrovirus. No consistent difference was observed in B-cell transformation assay (Whitlock-Witte cultures). However, formation of myeloid colonies in semisolid media was reduced in Skp2−/− compared to Skp2+/+ marrow [46.4±10% of controls (p=0.002) without and 76.6±9% of controls (p=0.008) with cytokines, n=6]. Skp2+/+ mice transplanted with BCR-ABL infected Skp2−/− marrow had significantly longer median survival (19days, range 12–60days, n=8) compared to recipients of Skp2+/+ marrow (13days, range 12–22days, n=10) (p=0.0034) with significant reduction of spleen weight (0.42±0.07g vs. 0.28±0.09g, p=0.004) and white blood cell counts (median 59x103/μl, range 9.6–142x103/μl, vs. 7.9x103/μl, range 0.8–87x103/μl, p=0.02). Histology and immunophenotyping of tissues (blood, marrow, spleen) revealed no signinificant differences between Skp2+/+ and Skp2−/− mice.
Our data suggest that the primary cell cycle effect of Bcr-Abl kinase is up-regulation of Skp2. This leads to increased activity of SCFSKP2, inducing degradation of T187 phosphorylated p27 which in turn promotes cell cycle progression by relieving suppression of Cdk2.
Skp2 is required for Bcr-Abl to fully realize its potential to induce myeloproliferative disease, providing the first in vivo evidence that SKP2 is an oncogene.
Targeting Skp2-p27 interactions to prevent p27 degradation may be therapeutically useful in malignancies with a high Skp2/p27 ratio.
Disclosure: No relevant conflicts of interest to declare.