Parental origin-specific expression of Mash2 is established at the time of implantation with its imprinting mechanism highly resistant to genome-wide demethylation

The Mash2 gene encodes a basic helix-loop-helix transcription factor, which is highly expressed in diploid trophoblast cells of the postimplantation mouse embryo and is required for development of the spongiotrophoblast in order to form a functional placenta. Genomic imprinting of Mash2 has been previously reported; transcriptional inactivation of the paternal wild-type allele in heterozygotes carrying a maternal null allele results in a null-equivalent embryonic lethal phenotype. In order to study the Mash2 imprinting mechanism, we have created a new allele at this locus carrying a targeted insertion of an IRES (internal ribosome entry site)-lacZ cassette within the 3' untranslated region of the gene (referred to as 'Mash2-lacZ'). This new allele has made it feasible to monitor paternal Mash2 expression in a wild-type-equivalent background. Our data suggest that parental origin-specific expression of Mash2 begins in the early postimplantation conceptus (5.5 dpc) at the time when trophoblast-specific expression is observed. We also show that the paternal allele is continuously repressed up to 9.5 dpc in the developing ectoplacental cone (EPC) and early chorio-allantoic placenta, with some cells escaping paternal repression. When maternally inherited, lacZ expression from this allele reflects the expression pattern of endogenous Mash2 transcripts up to 8.5 dpc. Furthermore, we have addressed the question of a requirement for DNA methylation for the Mash2 imprinting mechanism by crossing our Mash2-lacZ mice with mice mutant for Dnmt1 (DNA-methyltransferase1). Our results show a partial loss of transcriptional repression of the paternal allele in Dnmt1 deficient background. Interestingly, however, this is not sufficient to eliminate the highly biased parental allele-specific expression of Mash2. Thus, the preferential maternal expression of the gene is still maintained in Dnmt1 null mutant embryos, although methylation analyses demonstrate that the Mash2 locus is highly demethylated in Dnmt1 null mutant embryos. The locus is also highly demythyled in wild-type EPCs. Our results suggest the possibility that a mechanism other than DNA methylation, such as allele-specific chromatin conformation, may be involved in maintenance of parental origin-specific expression of Mash2. Copyright (C) 1999 Elsevier Science Ireland Ltd.