Our recent studies of recurrent amplicons in rhabdomyosarcoma (RMS) focused on the 12q13-q14 amplicon, which occurs preferentially in a subset of PAX3-FOXO1-positive RMS (25% of cases) and only in a smaller subset of PAX7-FOXO1-positive and fusion-negative cases. Our previous studies localized the minimal region of amplification to a 0.55 Mb region containing 28 genes, including the CDK4 proto-oncogene. Expression profiling studies in RMS tumors with and without this amplicon showed that 7 of these genes (including CDK4) were consistently overexpressed at the RNA level in amplified RMS tumors. We then focused on the CDK4 gene and assessed the relationship between CDK4 protein expression and amplification status through a collaboration with Drs. Svetlana Pack and Stephen Hewitt of the Laboratory of Pathology. Two tissue microarrays (TMA's) of RMS specimens were obtained from the Children's Oncology Group. Dr. Pack used fluorescence in situ hybridization probes to determine 12q13-q14 amplification status in each case on the TMA, and Dr. Hewitt used an immunohistochemical approach to assess CDK4 protein expression in these cases. The combined results from these TMA studies indicated that cases with 12q13-q14 amplification expressed high levels of CDK4 protein, cases with extra copies of chromosome 12 expressed intermediate protein levels and cases with normal 12q13-q14 copy number expressed low protein levels. To analyze the functional consequences of CDK4 amplification, IPTG-inducible shRNA constructs targeted against the CDK4 gene were transduced into Rh30 cells, a fusion-positive RMS line with 12q13-q14 amplification and associated high level expression of CDK4 mRNA and protein. Induction of the CDK4-shRNA constructs in Rh30 cells resulted in inhibition of CDK4 expression. In addition, this inhibition of CDK4 expression was associated with defects in cell growth, characterized as a G1 arrest by flow cytometry. In tumorigenesis studies in laboratory mice, CDK4 shRNA induction in Rh30 cells resulted in a delay in xenograft formation. Expression microarray analysis of Rh30 cells with inducible CDK4 shRNA demonstrated decreased expression of genes in the CDK4-RB pathway, such as E2F transcriptional targets and other cell cycle-related genes. In further experiments with the IPTG-inducible CDK4 shRNA constructs, induction of CDK4 shRNA in the Rh41 cell line, which does not have 12q13-q14 amplification, also demonstrated similar decreases in cell growth. This finding provides initial evidence that 12q13-q14 amplification or overexpression is not associated with increased sensitivity to CDK4 inhibition. To further analyze this issue, we compared the sensitivity of five fusion-positive RMS cell lines to growth inhibition by LEE011 (Novartis), a selective inhibitor of CDK4/CDK6. Treatment of these five lines with a range of LEE011 concentrations resulted in a dose-dependent decrease in cell numbers and a corresponding decrease in expression of CDK4-RB pathway targets. There was no relationship between sensitivity to growth inhibition and the CDK4 or CDK6 expression levels in these cell lines. In particular, the high CDK4 expression in Rh30 was not associated with higher susceptibility to growth inhibition.
