Medulloblastoma is the most frequent central nervous system (CNS) tumor of childhood and it is higly malignant with a dismal long term prognosis. This tumour is thought to result from neoplastic transformation of undifferentiated precursor cells in the developing cerebellum. The paired box-containing genes (PAX) are transcription factors that have been implicated in the regulation of morphogenesis of the CNS, e.g. Pax-5 has a key role in patterning of the midbrain. PAX-5 is specifically deregulated in medulloblastoma and its expression positively correlates with cell proliferation. This and other evidence strongly suggest that PAX genes may participate as dominant oncogenes in the formation of brain tumours such as medulloblastoma. We intend to test this hypothesis by using transgenic mice as an in vivo model system to study the role of PAX-5 in brain tumour formation. Two different strategies will be used to direct the PAX-5 expression to the cells of the external granular layer of the cerebellum from where medulloblastoma is thought to originate, i.e. a classical transgenic approach and a "knock-in" approach using embryonic stem cell technology.