Related Topics
 

McKinnon: DNA Damage and Brain Tumors



In certain settings, notably in collaboration with p53 deficiency, we found that medulloblastoma brain tumors are a common outcome of DNA DSB repair deficiency. We inactivated key DNA double-strand break repair genes throughout the nervous system and found that despite neuraxis-wide gene inactivation tumors (medulloblastoma) formed primarily in the cerebellum, and likely reflects the high rate of neurogenesis in this tissue.  These mouse models have been important for understanding the molecular and cellular basis for this tumor type. For example, we found a restricted set of tumor-specific gene expression changes that were very similar to the profile of granule neuron precursors, thereby implicating this cerebellar progenitor type as a cell of origin of the medulloblastomas [Cancer Res. 63, 5428-37, 2003]. Additionally, we also identified a narrow range of recurring genomic events that are required for medulloblastoma arising from DNA repair deficiency [PNAS, 106:1880-5, 2009]. Foremost was the targeting of the Ptch1 gene, which was biallelically inactivated in all tumors examined indicating that Ptch1 loss underpins granule cell transformation.

Figure 5. Medulloblastoma in DNA repair deficient mice. Mice with inactivation of DNA repair factors (HR: Xrcc2 or NHEJ: Lig4) together with p53-/- deficiency develop medulloblastoma with complete penetrance by 3 months of age (A,B). In addition to inactivation of Ptch1 on chromosome 13, genomic analysis of tumors reveals a number of recurring genetic alterations including chromosomes 6, 12, and 19 in medulloblastomas. (C) Representative array comparative genomic hybridization of chromosomes 6, 12, and 19 in reapir-defcient medulloblastomas. (D) Representative spectral karyotype (SKY) analysis of Brca2Nes-Crep53-/- showing a large number of double minutes, indicated by a white box that contain amplified N-Myc as determined using fluorescence in situ hybridization analysis (Magnification: 1,000x). (F) Representative SKY of chr19 showing either deletion or translocation. (G) Genomic rearrangements of N-Myc, cyclin D2, Sufu, and Pten in the medulloblastomas. Each column represents a single tumor, and each line represents one chromosome. Each red square indicates that the chromosome exhibited genomic loss, whereas each green square indicates a region of genomic amplification.


We also uncovered a connection between chromosome 19 loss and disruption of HR (Xrcc2 and Brca2 inactivation) providing a direct connection between HR defects and chr19 in medulloblastoma.  In human medulloblastoma, a very similar region of chromosome 10q is often lost (this is syntenic with mouse chr19), although identification the specific gene(s) targeted by these deletions has been elusive. We also found an association between hemizygous deletion of PTEN (which is located on chr19) and loss of PTEN expression in a subset of tumor cells. Thus, disruption of DNA repair provides a genetic mechanism to generate mouse models containing somatic mutations that are critical to the tumorigenic process and which are also relevant to human disease. These findings also suggest that the generation of relevant mouse models for specific cancers by selective inactivation of DNA repair processes in defined brain regions will be particularly useful, and this is an approach we are currently pursuing. The conditional DNA repair models allow gene disruption to be targeted to specific tissues, and we are developing other brain tumor models in these mice. We are also determining the utility of DNA repair inhibitors as a therapeutic strategy for treating medulloblastoma, and eventually other brain tumor types.