NTera2 and SH-SY5Y: Cell culture models of regulation of brain-specific imprinting in the Prader-Willi chromosome 15q11--q13 region

Date of Completion

January 2004


Biology, Genetics




Genomic imprinting is a complex ‘epigenetic process’, which results in the transcriptional silencing of one of the two parental alleles early in development. ‘Epigenetic’ implies that ‘gene expression’ is inherited from generation to generation, rather than the primary sequence of DNA. Genomic imprinting is, therefore, a paradox to the central dogma of Mendelian Genetics. The chromosome 15q11–q13 gene locus is subject to genomic imprinting. Prader-Willi Syndrome (PWS) and Angelman Syndrome (AS) are distinct epigenetic disorders, in which the brain is the major organ affected. PWS results from loss of expression of multiple paternally expressed genes on 15q11–q13. AS results from loss of a single maternally expressed, brain specific gene, UBE3A. Adding to the complexity of the region is UBE3A antisense (UBE3A ATS), a paternally expressed brain specific gene. Understanding the regulation of imprinted genes on chromosome 15 has proven to be complex. ^ Recently, Runte et al., (2001) reported that UBE3A ATS is one long processed transcription unit, consisting of 148 alternatively spliced exons spanning 460Kb, in which all paternally expressed genes along chromosome 15 are regulated by the SNURF/SNRPN imprinting center (IC). Two human cell lines, which have the potential to differentiate into post-mitotic neurons, were used to study the development and regulation of brain specific transcripts on chromosome 15. NTera2 is a human embryonic cell line; SH-SY5Y is a human neuroblastoma cell line. ^ Regional differences in the distribution of non-brain specific and brain specific transcripts were identified. Transcripts in the centromeric region of chromosome 15 were expressed in undifferentiated NTera2 cells and throughout the course of differentiation into neurons. However, transcripts in the telomeric region of chromosome 15, were expressed in neuronal cells only. A transition region separated the two expression sub-regions. While SH-SY5Y cells showed a similar pattern of expression in the centromeric region, neuronal specific antisense transcripts were not detected in SH-SY5Y neurons. ^ The finding of distinct regions of expression within 15q11–q13 imposed a conundrum, given the Runte et al., (2001) data. It is likely that distinct expression sub-regions, identified in this thesis, come about via brain specific alternative splicing of the primary transcript. ^