Integrating lineage across individuals with phylogenetic tree comparison

The cell lineage tree describes every cell division from the fertilized egg to the adult. Tracing cell lineage trees informs our understanding of the processes that sculpt development and disease. We established a new method for tracing lineage trees that uses CRISPR to incrementally add genetic “scars” to cells and their descendants. These scars are stored in the genome of every cell, and can be recovered using DNA or RNA sequencing from single cells alongside their molecular identity. We use the relatedness of scars between cells to compute estimates of an individual’s cell lineage tree. However, we lack methods to quantify the accuracy of each branch in the tree, and how to compare and integrate trees across individuals.

Here we propose an experimental and mathematical framework for computing CRISPR lineage trees. This project brings together expertise in genome editing, sequencing, and phylogenetics. First, we will experimentally recover genetic scars from hundreds of thousands of cells across many individual animals. Second, we will adapt phylogenetic methods for tree reconstruction and comparison to determine the most likely lineage tree for each individual. Third, we will merge lineage trees from all individuals to generate a “meta tree,” and quantify tree variance across all branches of the meta tree. This project will generate a platform for lineage comparison between individuals, opening a new door into our understanding of healthy and diseased development.

Current Status

2021-09-15
Abstract:
The cell lineage tree describes every cell division from the fertilized egg to the adult individual, conceptually similar to the family trees that relate us to our parents and grandparents. We have developed a new method for tracing lineage trees by using CRISPR to incrementally add genetic “scars” to cells and their descendants. We use phylogenetic methods to compute estimates of an individual’s cell lineage tree by using the scars collected from many cells. However, these methods deliver a limited number of scars, and we lack methods to quantify the accuracy of each branch in the tree. The goal of our project was to develop better scarring methods using a combined experimental and computational approach. We developed new experimental methods for CRISPR mutagenesis in zebrafish embryos. Next, we integrated these CRISPR tools into a new system for CRISPR lineage tracing which permits scarring at three timepoints, in a sequential order. Finally, we built a new computational method that scores the likelihood of all possible CRISPR scars. Our methods are superior to the current state-of-the-art for lineage tracing applications. Together, we anticipate that these methods will inform our understanding of the processes that sculpt animal development, and provide insight into how developmental errors can cause disease