Double minute amplification of mutant PDGF receptor α in a mouse glioma model

2015-02-16 00:01:22

Scientific Reports; 16 February 2015 ; DOI: 10.1038/srep08468

Hongyan Zou, Rui Feng, Yong Huang, Joseph Tripodi, Vesna Najfeld, Nadejda M. Tsankova, Maryam Jahanshahi, Lorin E. Olson, Philippe Soriano, Roland H. Friedel


In primary brain tumors, oncogenes are frequently amplified and maintained on extrachromosomal DNA as double minutes (DM), but the underlying mechanisms remain poorly understood. We have generated a mouse model of malignant glioma based on knock-in of a mutant PDGF receptor α (PDGFRα) that is expressed in oligodendrocyte precursor cells (OPCs) after activation by a Cre recombinase. In the tumor suppressor INK4/Arf-/- background, mutant animals frequently developed brain tumors resembling anaplastic human gliomas (WHO grade III). Besides brain tumors, most animals also developed aggressive fibrosarcomas, likely triggered by Cre activation of mutant PDGFRα in fibroblastic cell lineages. Importantly, in the brain tumors and cell lines derived from brain tumor tissues, we identified a high prevalence of DM Pdgfra gene amplification, suggesting its occurrence as an early mutational event contributing to the malignant transformation of OPCs. Amplicons extended beyond the Pdgfra locus and included in some cases neighboring genes Kit and Kdr. Our genetically defined mouse brain tumor model therefore supports OPC as a cell of origin for malignant glioma and offers an example of a defined temporal sequence of mutational events, thus providing an entry point for a mechanistic understanding of DM gene amplification and its functionality in gliomagenesis.


Focal amplification of genomic DNA is a pathological hallmark in many solid tumors, including glioblastoma, the most frequent and deadly form of primary brain cancer. Amplicons typically comprise 0.5-10 MB of DNA and exist either as circular extrachromosomal structures termed double minutes (DM), or as intrachromosomal concatenated repeats termed homogenously staining regions (HSR). These two forms of gene amplification are interrelated and possibly interconvertable. Several models have been proposed on how DM arise during chromosome replication, including replication fork stalling, fork collision, and fragmentation of larger chromosomal segments. However, mechanistic details of DM formation and its functionality in cancer development remain poorly understood. For instance, despite high prevalence in glioblastoma, it is unclear whether DM gene amplification occurs as an early mutational event contributing to malignant transformation, or at a later stage in advanced glioblastomas as a result of global genomic instability.

Increased activity of the receptor tyrosine kinase (RTK) PDGF receptor α (PDGFRα) is frequently encountered in glioblastoma and commonly associated with amplification of the PDGFRA gene. Additionally, PDGFRA amplification is often associated with point mutations or structural variants that are thought to increase the intensity of PDGFRα signaling. A number of mouse glioma models have been generated by overexpressing secreted PDGF ligands in the brain, but since secreted PDGF ligands may act through both autocrine and paracrine mechanisms, the cell of origin cannot be conclusively defined in these glioma models. These models also do not address the prevalence of PDGFRA gene amplification and its oncogenic role in gliomagenesis. Providing an adequate animal model to investigate these fundamental questions is central to advancing glioma research and therapeutic options.

We have developed a new glioma mouse model based on cell-autonomous activation of PDGFRα in oligodendrocyte precursor cells (OPCs). Specifically, we utilized a conditional knock-in of a PDGFRα point mutation that reduces auto-inhibition of the kinase domain. The knock-in design ensures that the expression of mutant PDGFRα is under the control of the endogenous Pdgfra promoter, which, in the CNS, is active in OPCs19. When bred on INK4/Arf-/- background, a common tumor suppressor deletion in glioblastoma, mice developed spontaneous primary brain tumors between 15-30 weeks of age. Brain tumors displayed pathological characteristics of human high grade gliomas (WHO grade III), thus providing direct evidence for OPC as a cell of origin for malignant glioma. Importantly, we identified a high prevalence of amplification of the Pdgfra mutant allele as DM in not only advanced, but also early stage grade III gliomas, thus supporting RTK amplification as an important early event in the malignant transformation of OPCs. In summary, our study describes a novel glioma model that provides an example of a defined temporal sequence of mutational events in the malignant transformation of OPCs, starting from an activating RTK mutation and loss of a tumor suppressor, followed by amplification of the mutant RTK in the form of DM.

To Read Full Article, Click Here

Key Words

Cancer genetics | Cancer models