Human Genome Meeting 2003
Cancun Convention Center, Cancun, Mexico
April 28, 2003
336 (15. Comparative genomics)
Poster presentation

Mouse Impact represents a solitary imprinted gene independent of regulation as a chromosomal domain
Kohji Okamura1, Yuriko Hagiwara-Takeuchi1, Yoshiyuki Sakaki1,2, and Takashi Ito3
1Hum. Genome Ctr., Inst. Med. Sci., Univ. Tokyo, Japan, 2Genomic Sci. Ctr., RIKEN, Japan, 3Cancer Res. Inst., Kanazawa Univ., Japan
Clustering into chromosomal domains is one of the characteristic features of imprinted genes. Indeed, among 50 imprinted genes identified so far, almost half were mapped to mouse chromosome 7. While this may well illustrate a prevailing mechanism of parent-of-origin-specific regulation, we provide an exceptional case for 'solitary' imprinted genes. Mouse Impact, which was identified by a systematic screening method making good use of differential display technique, is exclusively expressed from the paternal allele. This gene is the first and sole imprinted gene mapped to chromosome 18, which was suggested to bear some. Intriguingly, its human homologue IMPACT is not imprinted, although allele-specific expression is usually conserved between mouse and human. In this study, we performed allelic expression and DNA methylation analyses of the genes lying next to Impact and elucidated a unique feature of Impact as a solitary imprinted gene. First, we examined human OSBPL1A/OSBPL1B and HRH4 which reside upstream and downstream of IMPACT, respectively. OSBPL1B encodes an alternative transcript of OSBPL1A produced from different starting point. These genes are expressed from both parental alleles like IMPACT and no differentially methylated region was detected in this region. Next, we revealed a conserved genomic organization of the mouse Impact locus and successfully identified mouse homologues for the neighbors. Similar to human, Osbpl1a and Hrh4 are expressed biallelically. We, however, found that the longer transcript, Osbpl1b, whose promoter is located only 20-kb apart from that of Impact, is preferentially expressed from the paternal allele. Since DNA methylation analysis revealed the Osbpl1b CpG island is hypomethylated on both alleles, the biased expression may be due to the allele-specific chromatin structure of the promoter region of imprinted Impact. Apart from this biased transcription, mouse Impact locus fails to show any evidence for imprinted cluster. U2af1-rs1 and NNAT were also demonstrated as solitary imprinted genes. Because both genes occur in introns of nonimprinted genes and have few introns, they are conceivably generated as retrotransposons, which are often subject to imprinting. In contrast, Impact consists of 11 exons and is not embedded in an intron of another gene. It is thus unlikely that Impact was generated by a retrotranspositional event. Hence, the structural element unique to mouse Impact, such as the differentially-methylated intronic CpG island containing characteristic tandem repeats, may represent the basis for the species-specific imprinting. Thus, Impact would serve as an ideal model for the investigation of imprinted expression achieved by a unique mechanism independent of the regulation as a chromosomal domain.