Transgene- and locus-dependent imprinting reveals allele-specific chromosome conformations

A Mouse Model for Imprinting of the Human Retinoblastoma Gene

The human RB1 gene is imprinted due to integration of the PPP1R26P1 pseudogene into intron 2. PPP1R26P1 harbors the gametic differentially methylated region of the RB1 gene, CpG85, which is methylated in the female germ line. The paternally unmethylated CpG85 acts as promoter for the alternative transcript 2B of RB1, which interferes with expression of full-length RB1 in cis. In mice, PPP1R26P1 is not present in the Rb1 gene and Rb1 is not imprinted. Assuming that the mechanisms responsible for genomic imprinting are conserved, we investigated if imprinting of mouse Rb1 can be induced by transferring human PPP1R26P1 into mouse Rb1. We generated humanized Rb1_PPP1R26P1 knock-in mice that pass human PPP1R26P1 through the mouse germ line. We found that the function of unmethylated CpG85 as promoter for an alternative Rb1 transcript and as cis-repressor of the main Rb1 transcript is maintained in mouse tissues. However, CpG85 is not recognized as a gametic differentially methylated region in the mouse germ line. DNA methylation at CpG85 is acquired only in tissues of neuroectodermal origin, independent of parental transmission of PPP1R26P1. Absence of CpG85 methylation in oocytes and sperm implies a failure of imprint methylation establishment in the germ line. Our results indicate that site-specific integration of a proven human gametic differentially methylated region is not sufficient for acquisition of DNA methylation in the mouse germ line, even if promoter function of the element is maintained. This suggests a considerable dependency of DNA methylation induction on the surrounding sequence. However, our model is suited to determine the cellular function of the alternative Rb1 transcript.

Implications of COMT long-range interactions on the phenotypic variability of 22q11.2 deletion syndrome

22q11.2 deletion syndrome (22q11DS) results from a hemizygous microdeletion on chromosome 22 and is characterized by extensive phenotypic variability. Penetrance of signs, including congenital heart, craniofacial, and neurobehavioral abnormalities, varies widely and is not well correlated with genotype. The three-dimensional structure of the genome may help explain some of this variability. The physical interaction profile of a given gene locus with other genetic elements, such as enhancers and co-regulated genes, contributes to its regulation. Thus, it is possible that regulatory interactions with elements outside the deletion region are disrupted in the disease state and modulate the resulting spectrum of symptoms. COMT, a gene within the commonly deleted ~3 Mb region has been implicated as a contributor to the neurological features frequently found in 22q11DS patients. We used this locus as bait in a 4C-seq experiment to investigate genome-wide interaction profiles in B lymphocyte and fibroblast cell lines derived from both 22q11DS and unaffected individuals. All normal B lymphocyte lines displayed local, conserved chromatin looping interactions with regions that are lost in atypical and distal deletions, which may mediate similarities between typical, atypical, and distal 22q11 deletion phenotypes. There are also distinct clusterings of cis interactions based on disease state. We identified regions of differential trans interactions present in normal, and lost in deletion-carrying, B lymphocyte cell lines. This data suggests that hemizygous chromosomal deletions such as 22q11DS can have widespread effects on chromatin organization, and may contribute to the inherent phenotypic variability.