Searching beyond the streetlight to uncover functional protein domains

From the Paddison Lab, Human Biology Division

The streetlight effect is a type of observational bias that occurs when people only search for something where it is easiest to look. Dr. Patrick Paddison, part of the Human Biology Division, explains that this streetlight effect extends to our understanding of the human proteome, specifically that we have a poor understanding of how protein activities are compartmentalized into distinct functional domains. This is due to a few reasons, including that about “75% of biomedical research is focused on 10% of the mammalian genome- we build on what we know” or what has already been ‘lit up’, but “we don’t know what we don’t know,” Dr. Paddison explains. Dr. Jake Herman, a postdoctoral researcher co-mentored by Dr. Paddison and Dr. Sue Biggins, wanted to use a CRISPR-based method to tackle this genome annotation problem. He also saw this as an opportunity to find “an exceedingly simple and inexpensive way to find separation of function mutations in human genes.” Dr. Herman goes on to describe, “my work was fundamentally motivated by the question: how are chromosomes accurately segregated during mitosis? And how is this process co-opted or compromised by specific oncogenic changes?” Despite scientists “knocking out or depleting mitotic factors for decades, […] to understand the molecular mechanisms more deeply we need better tools to interrogate proteins at the sequence level,” Dr. Herman notes. To develop a more systematic way of annotating protein coding genes, Dr. Herman and the Paddison and Biggins groups utilized a CRISPR-Cas9 screen approach to identify functional domains at the level of the gene’s DNA sequence, where they tested this approach on a set of mitotic genes. This work was recently published in Genes & Development.

Overview of CRISPR-Cas9 tiling screen
Schematic of CRISPR-Cas9 tiling screen to identify functional protein domains of cell proliferation genes. Image taken from original article.

CRISPR-Cas9 is a tool that enables editing of specific gene loci through using small sequence-specific guides called single guide RNAs (sgRNAs). The researchers used a CRISPR-Cas9 tiling screen to induce mutations across protein coding genes to identify functional regions at the sequence level. This approach results in Cas9-induced cutting at specific DNA sites, where these breaks are commonly repaired by error-prone nonhomologous end joining, leaving repair scars in the form of small frameshift or in-frame mutations. Pairing a large screen approach to induce mutations that “tile” targeted genes with next-generation sequencing and a functional cell outgrowth test, the researchers would be able to identify protein-coding regions of genes that signify functional domains essential for cell proliferation. As Dr. Herman was particularly interested in understanding chromosomal segregation during cell division (evident by his dividing cell tattoo), the researchers used their CRISPR tiling approach to test its ability to uncover functional domains of 48 mitotic genes, including ones involved in kinetochore assembly and microtubule dynamics. The researchers performed this screen in four diverse cell types and interestingly, found that many sgRNAs similarly affected proliferation across these unique backgrounds. Overall, Herman et al. identified hundreds of regions that were required for proliferation, including both previously characterized and novel domains.

One of the challenges the researchers faced was deciding what specifically defined a domain. To do so, they got help from Dr. Jun Zhu at the Icahn School of Medicine at Mount Sinai, to computationally identify DNA-sequences corresponding to structural protein domains. They identified 36 out of 48 genes where more than 9% of the sgRNAs targeting that gene were shown to be essential for cell proliferation in at least two different cell lines. Out of those “36 genes, we were able to define almost 200 new regions previously unstudied,” Dr. Paddison notes. The researchers then validated a set of 15 uncharacterized functional regions across six genes that had been identified by their tiling screen. The Paddison group generated cell lines harboring small mutations in the regions identified as being functional domains important for cell proliferation and asked if adding back the wildtype protein would be able to rescue proliferation defects. They found that in almost every case, the wildtype proteins were able to rescue the defect, suggesting that the CRISPR-Cas9 tiling screen reliably identified uncharacterized functional protein regions important for cell division. The researchers chose to carefully study one gene, Mad1 and identified previously uncharacterized domains that are essential for Mad1 kinetochore localization and chromosome segregation fidelity.

Dr. Herman exclaims that this work allowed them to create “an atlas of previously unstudied sequences that seem to be exceedingly important for their essential activities. We hope others in the field start digging into these motifs and figure out what exactly they do!” More broadly, due to the feasibility and cost-effectiveness of this approach, this tool allows “anyone studying an essential biological process [to] be able to implement this same approach to better understand proteins at the sequence level.” This work has also inspired a follow-up study in the Biggins group, which will investigate a new exciting functional domain identified by this study, in the gene chTOG, a kinetochore protein that the Biggins lab has studied for years. With support from Fred Hutch pilot funds and in collaboration with Dr. Lucas Sullivan in the Human Biology Division, Dr. Paddison will use this method to tile metabolic genes and identify “moonlighting” functions of metabolic enzymes, or additional roles for these proteins outside of their traditional metabolic functions. Collectively this method serves as a valuable tool to elucidate previously unknown protein functions and characterize their corresponding functional domains and will even be “useful for the most studied genes in very old fields, you’re always going to be able to find something new,” Dr. Paddison remarks.


This work was supported by the American Cancer Society, the National Institutes of Health, the Helen C. Kleberg Foundation and the Howard Hughes Medical Institute.

UW/Fred Hutch Cancer Consortium members Patrick Paddison and Sue Biggins contributed to this research.

Herman JA, Arora S, Carter L, Zhu J, Biggins S, Paddison PJ. Functional dissection of human mitotic genes using CRISPR-Cas9 tiling screens. Genes Dev. 2022 Apr 1;36(7-8):495-510. doi: 10.1101/gad.349319.121. Epub 2022 Apr 28. PMID: 35483740; PMCID: PMC9067404.