Functional genomic confounds of transgenic integration and methods for their delineation. J. C. Jacobsen1,2, C. Chiang2, C. Ernst2,3, A. J. Morton4, C. Hanscom2, S. J. Reid1, R. G. Snell1, M. E. MacDonald2, J. F. Gusella2, M. E. Talkowski2 1) Centre for Brain Research, School of Biological Sciences, The University of Auckland, Auckland, New Zealand; 2) Center for Human Genetic Research, Massachusetts General Hospital, Boston, MA, USA; 3) Department of Psychiatry, McGill University, Montreal, QC, Canada; 4) Department of Pharmacology, University of Cambridge, Cambridge, UK.

   The integration of exogenous DNA into a host genome has been an important route to generation of animal and cellular models for mechanistic exploration into human disease and development of therapeutics. In most such models, little is known concerning the structural integrity of the transgene, the precise site of integration, or the impact of integration on the host genome. Here, we provide methodological optimizations for the delineation of transgenic sequences using next-generation sequencing and illustrate the potential impact of transgenic integration in biological systems. We sequenced the genome and delineated the transgenic structure of seven Huntingtons disease (HD) transgenic animal models harboring the integration of two independent transgenes using targeted capture as well as whole-genome jumping libraries, followed by analyses of the integration site and assembly of the internal transgene structure. Our analyses revealed a transgenic architecture so complex as to be reminiscent of chromothripsis in some animals. A series of secondary experiments on the R6/2 mouse, the most widely used model system of HD, revealed significant structural rearrangement of the transgene, as well as insertion into a single location in chromosome 4, within intron 7 of the gene Gm12695, with coincident deletion of 5,444 bp of mouse DNA. Our analyses reveal that Gm12695 is normally expressed at negligible levels in mouse brain, but is expressed at dramatically increased levels in the brains of R6/2 mice compared to non-transgenic (wild-type) animals. This effect was consistent across multiple R6/2 litters and animals with varying CAG repeat lengths and disease severity. These data suggest transgenic integration can represent a legitimate confound in biological models of disease, and vouches for sequence-level resolution of transgene insertions prior to extensive investment in phenotypic characterization. The molecular and bioinformatics approaches we have developed argue strongly that this base-pair level approach is now both feasible and advisable.

You may contact the first author (during and after the meeting) at