Whole organism lineage tracing by combinatorial and cumulative genome editing

Abstract

Each of us begins life as a single fertilized cell, or zygote. The act of fertilization brings about an endeavored series of cell divisions, ultimately resulting in the forty trillion cells in each adult human. How are these cell divisions coordinated to produce the sophisticated organs and tissues of the body? This is a long-standing and open question in biology. Many fate-mapping and lineage tracing techniques have been developed to sample from the underlying developmental process, but we are still far from a holistic lineage map of vertebrates or mammals. Such a map would be transformative, placing physiology, cell biology, genomics, and anatomy onto a unified scaffold of development. Despite the enormous strides made over the last one hundred years, our understanding of lineage remains enormously fragmented for most organisms. Only one animal has had its lineage fully described, the thousand-cell worm Caenorhabditis elegans. There have been many challenges in expanding to larger, more complex animals: the orders of magnitude increase in cell numbers, the stochastic nature of development, and the non-transparency of the target animals. Many technological advances have been applied to further refine existing lineage maps, including microscopy, vital dyes, and genetic markers, but none of these technologies scale to the number of cells in even the simplest vertebrates. My goal in this dissertation is to harness technologies in the nascent field of genome engineering to encode information into individual cells, with the ultimate goal of recording whole organism lineage maps. In my first chapter, I describe methods and tools to harness the power of the CRISPR genome editing technology. I detail an approach for rapidly discovering target sites within arbitrary genome segments, aggregating potential off-target sequences, and ranking the results using a wide array of community developed scoring schemes. I then apply this method to characterize the enhancer region of the human gene MYC, as well as to aid in the design of a deletion scan for the regulatory region of the HPRT1 gene. The underlying tool for finding and characterizing CRISPR sites, FlashFry, has also been released to the community. I then leverage these computational tools in conjunction with published work in the field to generate a compact lineage recording technology. I combine multiple CRISPR targets into a compact barcode, and read this barcode out of individual cells in both RNA and DNA. This system is then shown to accumulate a diversity of alleles, a prerequisite for lineage tracing. Using a synthetic lineage created in cell culture we test the reconstructive power of our approach, and then apply this technology to recover the lineage of both embryonic and adult Danio rerio (zebrafish). Leveraging the large embryo size, I directly inject our marking reagents at the single cell stage, and observe diverse alleles through progressive stages of development. I then recover editing patterns consistent with marking of cells as early as the two cell stage. Using techniques borrowed from phylogenetics, I then reconstruct coarse lineage trees from many embryonic stages, recovering trees with upwards of 8 layers of cell division. In the adult zebrafish I successfully recover germ layer and organ relationships, and observe a lineage bottleneck for all the organ systems sampled.