Minor gene mutations can produce extensive disruption of protein interactions, study finds

For all but a few high-profile genes, the path by which genetic mutations lead to disease remains murky. In a new study, researchers at Dana-Farber Cancer Institute and allied institutions show that, in many cases, mutations disturb the activity of cell proteins in subtle rather than blunt ways, and that these disruptions can reveal which mutations are truly involved in disease and which are mere bystanders.

The study, published online today by the journal Cell, focused on protein interactions within cells – the connections established between one protein and another, or between a protein and a molecule such as DNA. Researchers mapped the interactions of thousands of mutant proteins, creating an interaction “profile” for each one. When they compared these profiles with the profiles for normal versions of the proteins, they discovered a range of differences: while some of the mutant proteins interacted much as their normal counterparts did, many either didn’t interact at all or interacted only somewhat as the normal proteins did.

The investigators further found that specific interaction profiles can point to distinct subtypes of disease. One profile might be associated with a particularly aggressive form of a disease while another profile indicates a more benign type. The finding could result in a new, highly accurate way of determining which subtype of a genetic disease a patient has and which therapy may work best.

The study is part of a larger effort at Dana-Farber to construct a map of the interactions of the various proteins in human cells. Proteins are the chief laborers of cell life: by interacting with other proteins, protein complexes, and molecules, they enable cells to survive, reproduce, and fulfill their role within the body. Genetic mutations can introduce errors in the makeup of specific proteins, potentially interfering with the proteins’ interactions and leading to disease.

It’s commonly thought that mutations yield proteins that are, essentially, duds – that fail to interact at all and are therefore useless to the cell. In the new study, researchers sought a detailed understanding of how mutations impact protein interactions.

“Sequencing studies have turned up more than 100,000 gene mutations and other variations associated with disease, ranging from cystic fibrosis to cancer,” says study co-author David E. Hill, PhD, of Dana-Farber. “Except in a small number of cases, the mechanism by which mutations give rise to disease has yet to be mapped out. We wanted to trace the line from mutations to disruptions in protein interactions more fully – because, ultimately, it is these disruptions that result in disease.”

Researchers began with 1,200 normal genes that can cause disease when inherited in an abnormal form. The researchers also collected up to four mutated versions of each of these proteins. The mutations were “missense” mutations, involving only a single letter change of the genetic code.

Nidhi Sahni, PhD, of Dana-Farber, Mikko Taipale, PhD, of the Whitehead Institute for Biomedical Research, and their colleagues first performed interaction tests between each of these 3,000-odd proteins and molecular “chaperones,” substances that help proteins fold properly so they can carry out their normal functions.

“We compared the normal and mutant proteins’ interactions with the chaperones,” says Sahni, a co-lead author of the study. “We found that in the majority of cases, chaperones that bound to normal proteins bound to the mutant ones as well.” A chaperone affects a mutant protein much as a truant officer affects a delinquent student – keeping it on the straight and narrow path and ensuring it behaves properly. “Chaperones serve a quality-control function on mutant proteins by ensuring they fold correctly, which stabilizes them,” Sahni remarks. This suggests that mildly mutant proteins aren’t always as disruptive to the cell as they might be thought to be.

In the next phase of the study, Sahni and her associates used the same group of 3,000 proteins in tests of their interactions with thousands of other proteins and sections of cell DNA. As before, they compared the results for normal proteins with those of mutant proteins.

They found that about two-thirds of the mutant proteins had a disruptive effect on protein-protein interactions. The extent of that disruption varied widely, however. In about half of the disrupted cases, mutant proteins didn’t react with any of the other proteins tested. In the other half, the mutants had some of the same interactions as the normal proteins, but not all of them. Among proteins whose interactions with DNA regulate gene expression, the majority of mutant proteins showed perturbed interactions with cellular DNA while maintaining interactions with other proteins.

The findings suggest that interaction profiles can help scientists distinguish between abnormal genes that are involved in disease versus those that aren’t. If a mutated gene gives rise to a protein with an abnormal interaction profile, that gene is a good candidate as a contributor to disease. Conversely, a mutated gene whose associated protein has a normal set of interactions may be less likely to be harmful.

The investigators also found that distinct subtypes of a single disease often have their own interaction profiles. “Different mutations in a single gene can give rise to specific interaction profiles,” Hill remarks. “These profiles, in turn, often correspond closely to particular subtypes of disease. This suggests that knowing the interaction profile for a particular patient can help determine which form of a disease a patient has, which, in turn, can help doctors choose the best treatment.”

Overall, “our results show that genetic mutations that interfere with certain protein activities – particularly their interactions with other proteins and with DNA – are common in genetic-based diseases,” Hill sums up. “By contrast, genetic variations that don’t result in disease generally preserve the normal pattern of interactions. Careful investigation of the network of protein interactions, therefore, offers a promising vehicle for understanding the connection between gene mutations and certain diseases.”

The first authors of the study are Nidhi Sahni, PhD, and Song Yi, PhD, of Dana-Farber, Mikko Taipale, PhD, of the Whitehead Institute for Biomedical Research, Juan Fuxman Bass, PhD, of the University of Massachusetts Medical School, and Jasmin Coulombe-Huntington, PhD, of McGill University. The senior authors are Frederick P. Roth, PhD, of Dana-Farber, the University of Toronto, and the Canadian Institute for Advanced Research, Yu Xia, PhD, of McGill University, Albertha J.M. Walhout, PhD, of the University of Massachusetts Medical School, Susan Lindquist, PhD, of the Whitehead Institute and the Massachusetts Institute of Technology, and Marc Vidal, PhD, of Dana-Farber. Co-authors are Yang Wang, PhD, Atanas Kamburov, PhD, Quan Zhong, PhD, Yun Shen, Alexandre Palagi, MSc, Adriana San-Miguel, PhD, Changyu Fan, Dawit Balcha, Amelie Dricot, Jennifer Walsh, Akash Shah, Xinping Yang, PhD, Ani Stoyanova, Michael A. Calderwood, PhD, Michael Cusick, PhD, Kourosh Salehi-Ashtiani, PhD, Benoit Charloteaux, PhD, David E. Hill, PhD, and Tong Hao, MA, of Dana-Farber; Fan Yang, PhD, Jochen Weile, MS, and Nozomu Yachie, PhD, of the University of Toronto; István Kovács, PhD, Amitabh Sharma, PhD, and Yang-Yu Liu, PhD, of Dana-Farber and Northeastern University; Albert-László Barabási, PhD, of Northeastern University, and Brigham and Women’s Hospital; Yves Jacob, MD, PhD, of Dana-Farber and Institut Pasteur, Centre National de la Recherche Scientifique, and Université Paris Diderot, Paris, France; Jian Peng, PhD, of MIT; Georgios Karras, PhD, Irina Krykbaeva, and Vikram Khurana, MBBS, PhD, of the Whitehead Institute; Luke Whitesell, MD, of the Whitehead and MIT; Mandy Lam, PhD, of the University of Toronto; George Tucker, PhD, Alex Leighton, and Bonnie Berger, PhD, of MIT; Daniel Jordan, MS, of Brigham and Women’s Hospital; and Shamil Sunyaev, PhD, of Brigham and Women’s and the Broad Institute of MIT and Harvard.

The research was supported by grants from the National Human Genome Research Institute, the National Science Foundation, the Natural Sciences and Engineering Research Council of Canada, the Krembil Foundation, a Canada Excellence Research Chair, the Ontario Research Fund, and the Canadian Institute for Advanced Research.