Genomic Imprinting in Development and Disease

Basic Sciences

Investigators

Linked publications & trials

Abstract

Unlike the majority of genes expressed in mammals, imprinted genes are expressed from only one parental allele- which allele depends on the particular gene. Thus the Insulin like growth factor 2 (Igf2) is expressed almost exclusively from the paternal allele, whereas p57Kip2, a CDK inhibitor is expressed from the maternal allele. This form of gene regulation is, among vertebrates, unique to mammals. Why it exists is still unclear. Much evidence has suggested that imprinted genes are involved in regulating cell proliferation and viability. Androgenetic embryos, in which the entire genome is paternal in origin, exhibit overgrowth of the extraembryonic membranes and in increase in fetal size at mid gestation. Parthenogenetic embryos, where the entire genome is maternal in origin show retarded embryonic growth. To facilitate an understanding of imprinting and to identify novel imprinted genes, we established fibroblast lines which are either exclusively androgenetic or parthenogenetic in origin. The lines show diametrically opposite patterns of growth with the androgenetic cells having a shorter cell cycle time, reaching a higher saturation density and forming tumors, whereas the parthenotes senesced and died. Using mouse lines deficient for imprinted genes such as Igf2, one of its receptors, the Igf2r, and p57Kip2 revealed that Igf2 was a major determinant regulating proliferation and viability of these cells. In addition to these growth studies, we have used these lines to identify novel imprinted genes. Using a suppressive subtractive screen we identified the nuclear receptor cofactor repressor/activator Zac1 and epsilon sarcoglycan as being imprinted genes expressed from the paternal allele, as well as an Est that is strongly expressed in the brain, and is transcribed from the maternal allele. Current studies are centered on determining the function of these genes in development and growth regulation. We have developed several mouse models for a human congenital disease associated with a defect in imprinting called Prader-Willi syndrome. In this condition, newborns are featured by hypotonia, have disrupted breathing patterns and often fail to thrive past the first year of their lives. Those that survive develop an eating disorder and hyperphagia (excessive food intake) frequently resulting in obesity. The disease is associated with loss of part of the paternal chromosome 15. A homologous region is found on mouse chromosome 7. We recently described the derivation of mice lacking the paternal allele of a gene Necdin located in the Prader-Willi region. Such mice die shortly after birth due to respiratory problems and have mimicked one aspect of Prader-Willi syndrome. Their respiratory physiology is being studied in greater detail to determine the function of Necdin. We have also analyzed in gene targeting mouse model the function of another closely linked gene to Necdin, Magel2, that is also expressed from the paternal allele specifically in several key areas in the brain known to regulate the circadina rhythm and the appetite. We were able to demonstrate that the heterozygous animals bearing a disrupted Magel2 allele inherited paternally show severe disruption of circadian rhythmicity. All KO mice became severely arrhythmic within a short time period when subject to entraining in the constant darkness, whereas the control wild-type littermates displayed the robust circadian rhythm after several weeks of the same regime. We moreover were able to identify disrupted feeding behavior in Magel2 null mice, although contrary to teh human PWS individuals Magel2 KO animals appeared to consume on the average less food than their wild-type littermates, and displayed the collinear supression in the appetite-regulating hormone ghrelin. On the neuro-anatomical level we found that in Magel2 null mice the count of neurons expressing arousal and appetite regulating neuropeptide orexin is almost twice reduced as compared to the wild-type animals. In summary, we conclude that our Magel2 deficient mouse line provides a satisfactory model for the second stage of PWS pathology. Overall, a molecular analysis of imprinting will provide insights into the epigenetic control of gene expression, an aspect that is of increasing relevance to understanding the regulation of certain tumor suppressor genes in cancer formation