Precise and Expansive Genome Positioning for CRISPR Edits
New insights on microbial adaptive immune systems, known as CRISPR, have brought about genome-wide design across all domains of life. Common to all CRISPR systems are associated (Cas) proteins that are guided by RNA molecules to memorize, search, and destroy invading genetic threats. Per the simplicity of its RNA-programmable nature, CRISPR has facilitated numerous productive efforts to cure or diagnose disease, engineer ecosystems, and enhance agriculture, livestock, and biomanufacturing.
Whereas an ideal DNA-editing platform would achieve perfect accuracy on desired cellular and genomic targets, the evolved defense function of CRISPR did not directly select for such ability. Its natural error-prone targeting fidelity is of particular concern at the scale of editing plants and animals, which, unlike unicellular microbes that only carry a mega-base in a single chromosome, instead contain 1000x larger genomes in as many as trillions of cells. This thesis overcomes three critical challenges for scaling CRISPR to be a precise biologically-universal gene-editing tool through: 1) engineering specificity for the type of cells to edit, 2) improving DNA target accuracy, and 3) broadening the editable portion of the genome.
My proposed solutions integrate computational prediction and biological evaluation to address said challenges: 1) To target within multicellular heterogeneity, I optimize new oligonucleotide-sensing structural motifs and embed them into guides that can control CRISPR nuclease activity based on cell-type transcriptome patterns; 2) To discern among increased similarity between a target and other sequences in larger genomes, I apply thermodynamic principles for changing biochemical compositions in guides that can evade subtly mismatched off-target sites; 3) To access the narrow genomic window for positioning single-base-resolution edits, I develop new bioinformatics workflows that discover genomic target ranges of uncharacterized Cas proteins. We demonstrate the effectiveness of our solutions in the context of in vitro, bacterial, and human cell culture assays, which contribute advancements in the precision and generality for CRISPR gene-editing.
Joseph Jacobson, Professor of Media Arts and Sciences, MIT
Ed Boyden, Professor of Media Arts and Sciences, MIT
George Church, Professor of Genetics, Harvard Medical School