Research Roundup: October 15, 2021
Prime editing efficiency improves, a new genomic resource for diabetes research, designing a better polygenic score report, and more
Welcome to the October 15, 2021 installment of Research Roundup, a recurring snapshot of recent studies published by scientists at the Broad Institute and their collaborators.
More efficient prime editing systems
Last week, core institute member and Merkin Institute for Transformative Technologies in Healthcare director David Liu and colleagues reported improvements to prime editing guide RNA that increase editing efficiency. This week, in Cell, Liu and Peter Chen, Jeffrey Hussmann and Jonathan Weissman (UCSF), Britt Adamson (Princeton), and others describe improvements to the protein component of prime editing. The team found that mismatch repair interferes with prime editing, decreasing editing efficiency and purity. They then developed new systems that include an engineered protein, MLH1dn, that temporarily inhibits mismatch repair. Using the improved systems, the researchers installed or corrected disease-associated mutations in cells with improved efficiency and fewer byproducts. Read more in a Broad news story, a tweetorial by Liu, and a STAT news story.
Directed evolution evolves
The Liu lab has worked for years to develop and refine directed evolution methods such as PACE for generating peptides, proteins, and enzymes with desired activities. In Nature Chemical Biology, Max Shen, Kevin Zhao, and Liu describe Evoracle, a machine learning approach that uses short-read DNA sequencing data from laboratory evolution experiments to reconstruct full-length evolving genes and identify gene variants with the highest desired activities. And in Nature Communications, Mary Morrison, Tina Wang, Aditya Raguram, Colin Hemez, and Liu unveil a new directed evolution method, pPACE, for the rapid continuous evolution of proteins containing disulfide bonds, such as antibodies in the space between E. coli's inner and outer membranes. Learn more in a pair of tweetorials by Liu.
TIGER eyes diabetes variants of every stripe
Genome-wide association studies have helped identify hundreds of signals associated with type 2 diabetes (T2D). To gain insight into the molecular mechanisms they affect, research scientist Josep Mercader of the Metabolism Program and collaborators created the Translational human pancreatic Islet Genotype tissue-Expression Resource (TIGER), a large expression regulatory variation resource aggregating >500 genomic datasets from human pancreatic islets, a tissue with a central pathogenic role in T2D. The data, made available through the TIGER portal, will facilitate translation of genetic signals into functional and molecular mechanisms, ultimately providing therapeutic insights. Read more in Cell Reports and this tweetorial from Mercader.
A better way to present polygenic scores to patients
Various companies and academic groups have started designing polygenic risk score reports for patients and caregivers, including a group led by associate member Amit Khera. Khera (MGH), associated researcher Deanna Brockman, Lia Petronio of Broad’s Pattern group, and colleagues reviewed nine existing reports and designed a mock two-page report for coronary artery disease risk. In interviews with ten people who reviewed the mock report, the team found that visual elements, simple descriptions of risk, and presentation of polygenic scores as percentiles were key to enabling clear understanding and interpretation. In response, the team modified their report, which is now in use at the MGH Preventive Genomics Clinic. Read more in BMC Medical Genomics.
New protocol for more accurate sequencing
Current methods to prepare DNA for sequencing via ‘End Repair/dA-Tailing’ may resynthesize large portions of each DNA duplex and lead to errors in sequencing, especially for techniques that rely on consensus reads from both strands. In Nucleic Acids Research, Kan Xiong, Douglas Shea, Justin Rhoades, Gerstner Center for Cancer Diagnostics associate director Viktor Adalsteinsson, and others in the Gerstner Center for Cancer Diagnostics confirmed that these methods can indeed introduce false mutations and presented a solution called Duplex-Repair, which limits resynthesis. Their method resulted in up to 8.9-fold higher accuracy in duplex sequencing of a cancer gene panel.
To learn more about research conducted at the Broad, visit broadinstitute.org/publications, and keep an eye on broadinstitute.org/news.
