Research Roundup: June 7, 2019

Welcome to the June 7, 2019 installment of Research Roundup, a recurring snapshot of recent studies published by scientists at the Broad Institute and their collaborators.

A peek behind cells’ otherwise “normal” facade

Mutations often crop up in our cells as we age — generally without ill effects, but on rare occasions contributing to disease. To see how widespread these mutated cells (called “somatic clones”) can be, Keren Yizhak, Cancer Program institute member Gad Getz, and colleagues analyzed RNA sequencing data from the GTEx project, which studied normal tissues donated by individuals who died from causes other than cancer. They found that 95 percent of the nearly 500 people they studied had somatic clones present in at least one part of the body, including some clones harboring mutations in cancer-associated genes. Learn more in Science and a Broad news story, and check out coverage from NPR, STAT, Nature, and Science News.

A new gene-editing CAST member

In Science, a team led by Jonathan Strecker, Alim Ladha, and core institute member Feng Zhang reports a new gene-editing approach that can precisely and efficiently insert large DNA sequences into a genome. The system, called CRISPR-associated transposase (CAST), is a completely new platform to integrate genetic sequences into cellular DNA, addressing a long-sought goal for precision gene editing. The team molecularly characterized and harnessed the natural CAST system from cyanobacteria, also unveiling a new way that some CRISPR-associated systems perform in nature: not to protect bacteria from viruses, but to facilitate the spread of transposon DNA. Check out more in coverage from STAT and New Scientist.

Some brain cells play favorites

Oligodendrocytes are brain cells that generate myelin to insulate neurons, but it’s been unclear if they do so discriminately. New work from Marzieh Zonouzi and Jeff Lichtman from Harvard and institute member Paola Arlotta in the Stanley Center for Psychiatric Research provides evidence that oligodendrocytes recognize the class identity of individual types of interneurons that they target — some prefer inhibitory interneurons, while others target excitatory interneurons or show no bias. Appearing in Cell Reports, the study suggests that the interactions between oligodendrocytes and neurons are very specific, but more work remains to fully understand the myelination preferences of oligodendrocytes.

Of the same mind

A team led by Silvia Velasco and Paola Arlotta has developed a new method for growing human brain organoids — 3D tissue cultures that mimic the human brain — that consistently grows the same types of cells, in the same order, as the developing human cerebral cortex. Reporting in Nature, the researchers performed single-cell RNA sequencing analysis of 21 individual organoids and observed remarkable consistency. The advance could change the way researchers study neuropsychiatric diseases and test the effectiveness of drugs. Read more in the Harvard Gazette.

A long look at linear loss landscapes

In the Proceedings of Machine Learning Research, Daniel Kunin from Stanford, institute scientists Jonathan Bloom and Cotton Seed, and Stanley Center postdoctoral fellow Aleksandrina Goeva use a topological lens to fill out and tie together research on linear representations over the last forty years, deriving from first principles a unified characterization of the loss landscapes of regularized linear autoencoders. At next week’s International Conference on Machine Learning, they will present this work, which suggests new learning algorithms for PCA and, just maybe, the brain.

BETting on a promising therapy

MYC-amplified medulloblastoma is an aggressive pediatric brain tumor. Previous research has shown BET-bromodomain inhibition (BETi) to be a potential therapeutic strategy in targeting medulloblastomas. However, the underlying mechanisms for its action is not completely understood. A team led by associate member Mimi Bandopadhayay, Broad alum Cory Johannessen, and associate member Rameen Beroukhim from the Cancer Program used gene expression profiling and various screening methods to uncover key genes which mediate responses to BETi, such as bHLH/homeobox transcription factors and cell cycle regulators. Reporting in Nature Communications, the team’s findings provide insights on BETi’s mechanisms and how combining cell-cycle inhibitors with BETi can be a potential therapeutic approach for targeting MYC-amplified medulloblastomas.

Analyzing ancestry in Africa and the Arctic

Associate member David Reich in the Broad Program in Medical and Population Genetics and colleagues examine ancient DNA to trace humans’ advances into new frontiers in two recent papers. One, published in Nature, clarifies the relationship between the first settlers of the American Arctic, the Palaeo-Eskimos, and later Arctic peoples. Based on DNA from 48 ancient Arctic dwellers and data from five populations of modern Arctic natives, Reich's team suggest that Na-Dene and Eskimo-Aleut-speaking groups have significant Palaeo-Eskimo ancestry, and that several Arctic people groups stemmed from a Palaeo-Eskimo-related Siberian source. In the second paper, appearing in Science, Reich and colleagues shed light on how farming and herding were brought to eastern Africa. Their analysis of DNA from 41 ancient individuals from Kenya and Tanzania supports the theory that herders and farmers arrived and mixed with foragers in several distinct phases.

A LIGER bred not for magic, but for science

Biologists are creating ever-increasing numbers of single-cell datasets, and combining these diverse data to compare similar cells across individuals, tissues, and species is key to understanding the distinct roles of cell types. In a paper published in Cell, Joshua Welch, Stanley Center associate member Evan Macosko, and colleagues took on this integration challenge with a tool they developed: Linked Inference of Genomic Experimental Relationships (LIGER). This tool groups similar cells from different datasets together and maps the relationships of groups to one another. LIGER could help locate cell types within tissues, identify meaningful differences between healthy and disease states, and aid discovery of cell-type-specific genetic regulatory elements. Learn more in a Broad news story.

Synapse scholarship synthesized

The synapse is the fundamental information processing unit of the brain, and lies at the root of a range of brain disorders. To speed progress in understanding synapse biology, members of the SYNGO Consortium, established four years ago by institute member Guoping Feng and core institute memeber and Stanley Center director Steven Hyman, have released SYNGO 1.0: a comprehensive, expert-curated, ontology-guided knowledge base — integrated into the widely-used GO bioinformatic resource — describing the neuroscience community’s current insights into the function and localization of more than 1,100 genes active at the synapse. Learn more about SYNGO 1.0 in Neuron and a Broad news story.

To learn more about research conducted at the Broad, visit broadinstitute.org/publications, and keep an eye on broadinstitute.org/news.