High-throughput sequence analysis continues to revolutionize biological and clinical research and is now
revealing a large number of candidate structural variants suspected of contributing to human disease. To
model such structural variants in a mammalian whole animal context will require the ability to manipulate large
segments of the mouse genome in new and efficient ways. To that end, we have recently developed a
heretofore untested combination of technologies (that we call CRISPR/BAC/HR), bringing together — (1), the
long (up 300-kbp) physical extents of DNA manipulable in bacterial artificial chromosome (BAC) vectors; (2),
the introduction of such vectors, directed toward specific loci within the genomes of live mice, by zygotic
microinjection; and (3), the homologous recombination (HR) of such vectors at sites of frequent CRISPR/Cas9-
(CRISPR-) induced double strand breaks. Our preliminary data clearly show that this approach is eminently
feasible, as we were able to humanize an 18-kbp segment of the mouse Bcl2l11 gene with 25-kbp of DNA from
the orthologous leukemia-associated human locus, BCL2L11. This represents an order of magnitude increase
over the longest prior sequence addition (“knock-in”) employing plasmid vectors, CRISPR/Cas9 technology,
and homology directed repair.
Now, in the proposal outlined here, we will test our overarching hypothesis — that the extensive
homologies afforded by CRISPR/BAC/HR will drive efficient modification of the mouse genome over long
distances (10s to 100s of kbp), and can promote the modeling of human disease-associated intrachromosomal
and interchromosomal structural variants as well.
Our Aims are directed at optimizing, extending, and implementing our technology within the framework of
two compelling biological contexts (Gastric Cancer and Down Syndrome). First, we will assess the effects of
genomic insert size, homology arm length, and single-guide RNA placement on the efficiency of
CRISPR/BAC/HR while replacing large (10s to 100s of kbp) mouse genomic segments with orthologous
human genes borne on bacterial artificial chromosomes. Next, we will demonstrate that CRISPR/BAC/HR can
promote both precise tandem duplication along a single chromosome and reciprocal translocation between
chromosomes. Finally, we will use CRISPR/BAC/HR to create three important animal models — (1) Tandem
duplication of the mouse Bcl2l1 gene as a model of human gastric cancer, (2) Reciprocal translocation
between central mouse Chromosome 16 (Mmu16) and distal mouse Chromosome 17 (Mmu17) as a next
generation mouse model of Down Syndrome, and (3) Humanization of the mouse Bcl2l1 gene as a model to
assess the effect of pharmaceutical test agents on the activity of human BCL2L1.
