Tissue-specific patterns of imprinting revealed by analysis of monoallelic expression in human populations. T. Lappalainen1,2, Y. Baran3, E. Tsang4, T. Tukiainen5, M. A. Rivas6, M. Pirinen7, M. Gutierrez-Arcelus8, The. GTEx Consortium9, D. G. MacArthur5,9, S. B. Montgomery4, N. A. Zaitlen10 1) New York Genome Center, New York, NY; 2) Department of Systems Biology, Columbia University, New York, NY; 3) The Blavatnik School of Computer Science, Tel-Aviv University, Israel; 4) Departments of Pathology and Genetics, Stanford University, CA; 5) Analytic and Translational Genetics Unit, Massachusetts General Hospital, Boston, MA; 6) Wellcome Trust Center for Human Genetics, Oxford, UK; 7) Institute for Molecular Medicine Finland, University of Helsinki, Finland; 8) Department of Genetic Medicine and Development, University of Geneva, Switzerland; 9) The Broad Institute, Boston, MA; 10) Department of Medicine, University of California San Francisco, CA.
Imprinting is an epigenetic mechanism that leads to parent-of-origin effects via imbalanced expression of the maternally and paternally inherited alleles, and it has been shown to play a role in Mendelian and common disease and cancer. This study is the first systematic characterization of imprinting in multiple primary tissues from adult humans, using allele-specific expression data of the GTEx project with 1652 RNA-seq samples from 178 unrelated individuals. Our novel filtering and likelihood-based approach based on probabilistic generative models distinguishes imprinting from other sources of monoallelic expression. We identify 47 imprinted genes, of which 29 have been identified before. However, 34% of known imprinted genes are biallelic in our data set, demonstrating how poorly previous catalogs capture imprinting in primary tissues of adults. Methylation array data shows low correlation with imprinting status in expression data, supporting recent reports on complex epigenetic mechanisms underlying parental monoallelic expression. We show widespread tissue-specificity of imprinting, with 37/47 imprinted genes being biallelic in at least one tissue. The direction of imprinting can also change between tissues: the IGF2 gene with previously known maternal imprinting is in fact paternally imprinted in the brain. We also observe instances in which some individuals are imprinted while others exhibit biallelic expression. In muscle, females have less imprinting than males (p=0.0053) and imprinted genes have more sex-specific expression (p=0.012), pointing to gender-specific parental effects. While testis has less imprinting than other tissues (p=3.8 x 10-5), in general imprinted genes are highly expressed in tissues with endocrinological functions, consistent with their role in growth regulation. We also putatively characterize how monoallelic expression could modify the phenotypic impact and penetrance of functional coding variants in imprinted genes. Altogether, our results demonstrate that imprinting is not a stable property of a gene but can vary substantially between tissues and individuals. This adds to our understanding of the mechanisms of imprinting and its role in the function of the human genome.
You may contact the first author (during and after the meeting) at