Research Roundup: July 9, 2021

The genomics of COVID, how the brain picks which synapses to sculpt, engineering yeast to treat IBD, and more

Susanna M. Hamilton
Credit: Susanna M. Hamilton

Welcome to the July 9, 2021 installment of Research Roundup, a recurring snapshot of recent studies published by scientists at the Broad Institute and their collaborators.

A COVID detector you can wear

By weaving biological circuits into clothing, researchers can make wearable sensors for monitoring a person's health or detecting pathogens or toxins. As a proof of concept, a team led by Peter Nguyen (Wyss Institute), Luis Soenksen  (Wyss), and Broad institute member James Collins used freeze-dried SHERLOCK assay components to create a face mask that, within 90 minutes and at room temperature, can tell whether a wearer has SARS-CoV-2 in their breath, with an accuracy rivaling gold-standard PCR testing. Their approach could also be used in a variety of public health and medical applications. Learn more in Nature Biotechnology and in news stories from the Wyss Institute and MIT.

Epitranscriptome drivers hit the road

It’s unclear how genetic variants, including those arising from genome-wide association studies, alter post-transcriptional mRNA modifications. Xushen Xiong, Lei Hou, and associate member Manolis Kellis of MIT and the Epigenomics Program describe 129 transcriptome-wide profiles of the most prevalent of these modifications, N6-methyladenosine (m6A), covering 91 individuals and four tissues (brain, heart, lung, and muscle) from the enhancing Genotype-Tissue Expression consortium. They discovered more than 9,000 quantitative trait loci targeting 1,270 m6A sites, with enrichment in disease-associated regions, and predicted new m6A-regulating RNA-binding proteins. Described in Nature Genetics, the study provides insights into epitranscriptomic gene-regulatory control and helps pave the way to new therapeutic targets. 

Take it to the (Bio)bank

New work by Alison Barton, Maxwell Sherman, Ronen Mukamel, and associate member Po-Ru Loh of Harvard Medical School and the Program in Medical and Population Genetics (MPG) uses imputation to link rare coding variants to diverse quantitative traits. The researchers used whole-exome sequencing data from 50,000 participants in the UK Biobank cohort, along with genotyping data on the full 500,000-person cohort, to impute rare variants across the exome. They analyzed 54 quantitative traits and uncovered more than 1,000 trait-variant associations, including rare variants with a large effect on height. The work foreshadows the insights that will be revealed as genetic biobank studies continue to grow. Read more in Nature Genetics and GenomeWeb.

How microglia help balance brain wiring 

Microglia are known to regulate synapses and wiring in the brain. Postdoctoral scholar Emilia Favuzzi and institute member Gordon Fishell of the Stanley Center for Psychiatric Research, and colleagues used techniques including MERFISH to show in mice that microglia physically and directly interact with inhibitory synapses during development. These microglia have a GABA receptor, making them sensitive to GABA-emitting inhibitory synapses. GABA appears to attract these microglia towards these synapses, where the cells engulf the synapses while leaving excitatory synapses alone. The team also found that deleting these GABA receptors from microglia in mice impaired this process and resulted in behavioral abnormalities. Learn more in Cell and a Harvard Medical School news story.

Circuit mechanism found for cognitive impairment

Some people with different neurodevelopmental disorders have similar symptoms, such as learning disabilities or attention deficits. Postdoctoral associate Dheeraj Roy and institute member Guoping Feng of the Stanley Center for Psychiatric Research, Ying Zhang (McGovern Institute, MIT), and colleagues have identified a mechanism in the brain’s thalamus that may underlie the cognitive impairment seen in some people with autism and schizophrenia. They found that key circuits in mice connecting the anterodorsal thalamus with the retrosplenial cortex (RSC), and the anteroventral thalamus with the RSC, are involved in memory formation and regulation. Deleting known autism and schizophrenia genes including PTCHD1 resulted in memory deficits and hyperexcitability of these circuits. Read more in Neuron and a story from MIT News.

Building better transversion base editors

Base editors that convert C•G-to-G•C are inefficient and can produce unpredictable results. Graduate student Luke Koblan, postdoctoral associate Mandana Arbab, associated researcher Max Shen, core institute member David Liu who is also director of the Merkin Institute for Transformative Technologies in Healthcare, and collaborators engineered new C•G-to-G•C base editors by testing different combinations of deaminase enzymes, Cas proteins, and DNA repair proteins identified from a CRISPR interference screen. The team selected the most promising editors and used them to train CGBE-Hive, a machine learning-based tool that predicts C•G-to-G•C editing efficiency and purity. The editors could correct 546 disease-related single-nucleotide variants with more than 90 percent precision and up to 70 percent efficiency. Read more about how these editors expand the base editing toolkit in Nature Biotechnology.

Engineered yeast robots to treat IBD

There is an unmet clinical need for localized and tunable inflammatory bowel disease (IBD) therapies. Using Saccharomyces cerevisiae (yeast used for making beer) and applying synthetic biology and directed evolution technology, associate member Francisco Quintana of the Immunology Program, Benjamin Scott (University of Toronto), Cristina Gutiérrez-Vázquez (Brigham and Women’s Hospital), and colleagues have developed a “designer” probiotic, an engineered, self-tunable yeast that can treat the inflammation, fibrosis and dysbiosis associated with IBD. In Nature Medicine, the researchers report that these probiotic yeasts called “Y-bots” (yeast robots) could suppress intestinal pathology in three pre-clinical models in mice and provide a new therapeutic platform for IBD and potentially other inflammatory disorders. Read more in a BWH news story.

Uncovering mechanisms of cancer drug resistance 

In a study published in the New England Journal of Medicine, Mark Awad (Dana Farber Cancer Institute), postdoctoral scholar Shengwu Liu, associate member Andrew Aguirre of the Cancer Program, colleagues at the Genetic Perturbation Platform including senior group leader Xiaoping Yang, research scientist Nicole Persky, and senior director David Root, and others provide a genomic landscape of tumors that have grown resistant to drugs targeting the abnormal KRASG12C protein (common cancer mutation). The mechanisms of acquired resistance to KRAS inhibitors were previously unknown. By analyzing tumor samples from 38 patients with cancers carrying KRASG12C mutations, the researchers confirm a variety of genomic and histologic mechanisms that lead to resistance to KRAS inhibitors and suggest that effective treatment for these cancers will likely require combinations of KRAS inhibitors and other targeted drugs. Read more in a DFCI news story.

Embryo selection pitfalls

In a New England Journal of Medicine report, associated scientists Patrick Turley and Alicia Martin, institute members Ben Neale and MPG co-director Heidi Rehm, core institute member Steve Hyman, and colleagues discuss the problems with using polygenic scores to choose among embryos produced through in vitro fertilization. In theory, this practice enables parents to select embryos with desirable genetics — but as the team warns, there are many limitations, risks, and potential unintended consequences. The researchers call for a society-wide conversation about embryo selection using polygenic scores, and, if this service continues to be offered, they provide some recommendations for responsible communication. Read more in a University of Southern California press release.

Shining a LAMP on COVID-19 diagnostics

A team led by Jonathan Schmid-Burgk, core institute member Feng Zhang, and colleagues presents a new COVID-19 diagnostic method called LAMP-Seq in Nature Biotechnology. This approach attaches molecular barcodes to individual samples, allowing them to be pooled for large-scale testing while retaining individual results. LAMP-Seq's streamlined protocol reduces supply chain issues and lowers the requirements for technical infrastructure in order to achieve accurate diagnoses quickly at low cost. The protocol could be "readily deployed for frequent testing as part of an integrated public health surveillance program," writes the team. Read more in a press release from University Hospital Bonn.

Large international effort uncovers host genetic factors in COVID-19

For the past year, thousands of researchers around the world have pooled data to understand the genetic factors underlying why some COVID-19 patients experience mild symptoms while others need hospitalization. The COVID-19 Host Genetics Initiative — including visiting researcher Andrea Ganna, Rachel Liao of the Data Sciences Platform, computational scientist Juha Karjalainen, and institute members Mark Daly and Ben Neale of the Program in Medical and Population Genetics — now reports findings from ~50,000 patients, revealing 13 loci associated with infection or severe disease. Published in Nature, the results indicate possible therapeutic targets and provide a model for large-scale collaborative genomic studies of infectious disease. Read more in a Broad news story, Nature News & Views, and coverage in STAT, The Boston Globe, The New York Times, The Los Angeles Times, USA Today, and Nature.

To learn more about research conducted at the Broad, visit, and keep an eye on