Dosage compensation in mammalian females is accomplished by the inactivation of one of the two X chromosomes in early development. The inactivation event encompasses most of the genes of the X chromosome, excluding but a small region involved in X-Y pairing and recombination. Once established, the inactive state is clonally inherited and quite stable, except in the female germline in which reactivation of the inactive X chromosome is an essential part of oogenesis. The molecular basis of initiation of X inactivation, spreading of inactivation to all but specific regions of the X, and reactivation of the X in oogenesis is poorly understood, but differences in chromatin conformation and DNA methylation have been implicated. It is now becoming apparent that perhaps ten percent of the mammalian genome functions differently depending on whether it was inherited maternally or paternally. The process of establishing this differential function based on parental origin is termed imprinting. A number of human genes are clearly subject to imprinting, and several diseases are phenotypically different in their expression due to parental origin. Studies of the human fragile X-linked mental retardation syndrome suggest that the mutation may involve imprinting in the female germline; an error in the normal X-chromosome inactivation-reactivation cycle is suspected. Imprinting has been demonstrated at the molecular level in autosomal genes and transgenes of mice, and appears to involve differential methylation in some cases. Imprinting of specific X-linked genes has not yet been described at a molecular level; this phenomenon must be demonstrated in order to understanding the imprinting that may be responsible for expression of the human fragile X-mental retardation syndrome. Our approach to the study of X-chromosome inactivation and imprinting utilizes transgenic mice having a foreign gene inserted on the X chromosome as a model system. Using this system, we hope to (1) determine which of the usual molecular correlates of X inactivation (methylation, DNase I resistance) are important to the process and which are merely correlated with inactivation, (2) to determine whether or not X-linked transgenes are differently expressed, organized in chromatin or methylated when inherited from the male or female parent, and (3) to determine whether or not the X chromosome which was inactivated in oogonial cells is differently expressed, organized in chromatin or methylated compared to the one that was on the active X.
