If you’ve been told you have your mother’s chin or your father’s eyes, you can thank your genes — the biological instruction manual passed down from each parent.
But tiny “typos,” or genetic mutations, can sneak into segments of DNA. Many of these are harmless, but some can cause health problems. Geneticists search for these single blemishes to understand how diseases are inherited.
It’s not always so simple, however.
Two or more genes can team up and change the outcome of a physical or molecular trait. This phenomenon, known as epistasis, occurs through complex interactions between genes that are functionally related — such as those that support protein creation.
Identifying these group dynamics provides crucial clues to how genetic diseases manifest and should be treated. But they’re not easily detected and often fly under the radar. To help root out these connections, Euan Ashley, MB ChB, DPhil, the Arthur L. Bloomfield Professor of Medicine and the Roger and Joelle Burnell Professor of Genomics and Precision Health, and a team of scientists, including co-corresponding author Bin Yu, PhD, a professor of statistics and of electrical engineering and computer sciences at the University of California, Berkeley, have developed computational techniques to identify and understand the hidden ways epistasis influences inherited diseases.
In a demonstration of the new technology’s success, Ashely last year published a study in Nature Cardiovascular Research that established epistasis as a key driver of cardiac hypertrophy, a condition in which the heart thickens, resulting in a restriction of the heart’s ability to pump blood. For years, scientists believed the disease had a simpler genetic origin, focusing on single mutations as the root cause. But now scientists are recognizing that interactions among multiple genes can influence who develops it and the severity of the condition.
In adding to that understanding, the team found three additional gene pairs underlying the illness, which may redefine how cardiac hypertrophy can be inherited, diagnosed and treated.
Epistasis research could not only lead to promising new therapies for heart diseases but may also open a door to deeper understanding of a wide variety of complex genetic disorders in the future.
- Intro
- 1
- 2
- 3
- 4
- End
How epistasis works
Genes dictate how physical and molecular traits appear in the body. Take dimples — they’re determined by a single gene. However, in more complex traits, such as hair loss, epistasis occurs, relying on two or more genes. These genes operate like teammates, working together to achieve a single outcome. The interactions between these gene pairs can modify, amplify or mute the expression of a biological characteristic.
A few examples:
- Yellow Labradors: Labrador retrievers with yellow coats have a pair of genes that collaborate to override the genes encoding for black- or brown-colored fur. Either of the genes functioning alone wouldn’t have the genetic sway to change the color.
Albinism: People and animals with albinism, which results in a lack of pigment in skin, hair and eyes, have the condition when several genes combine forces to mask the genes that normally code for pigmentation.
The exact mechanism of how these genes interact is intricate and not well understood, said Qianru Wang, PhD, a postdoctoral scholar in Ashley’s lab and the lead author of the paper. But understanding a gene’s function can provide insights into this molecular mystery.
For example, if two of the genes encode two proteins involved in the same pathway, a protein-protein interaction may be the culprit behind the physical or molecular outcome.
Revealing hidden genetic teamwork
Epistasis is relatively overlooked in human genetic diseases. Only a handful of studies have investigated its role in heart disease, and they have yet to turn up much that’s helpful in understanding how heart disease manifests or should be treated. The findings from these early studies were based on a concept called statistical epistasis, which depends on how often a gene pair interaction is detected in people with a disease compared to those without it.
But searching for epistasis in human biology’s large library of genetic information is computationally challenging, and identifying interacting gene pairs is only the first step. Ashley and his team needed a way to scour large datasets and measure the effect of multigene interactions on physical characteristics.
In the case of cardiac hypertrophy, the team wanted to understand how epistasis could change the severity of the heart ventricular wall’s thickness — better yet, if they could show that these gene interactions were behind the disease. “Can we establish causality?” asked Ashley, the Roger and Joelle Burnell Professor in Genomics and Precision Health. “The only way we can show that is to go into cells and actually manipulate those genes.”
Ashley’s lab and its collaborators collected nearly a decade’s worth of data, bringing together genetic analyses from more than 300 human hearts and nearly 30,000 magnetic resonance images of hearts, provided by the data repository UK Biobank.
What the new tools do
In the past, scientists used an approach known as a genome-wide association study — which, Wang notes, was considered the gold standard. But because the approach highlighted one genetic variant at a time, it missed the chance to study whether or how genetic variants might work together.
Finding those linkages required sifting through large databases and identifying subtle connections between gene networks. But the team designed tools that could pick up on multiple mutations at once.
To speed that time-consuming (and detailed) process along, the team turned to artificial intelligence. They filtered through 15 million genetic variations associated with heart thickening, whittling it to around 1,400 mutations most likely to form partnerships. They then, created a machine learning model to identify epistasis candidates and rank the significance of these potential collaborations. The researchers trained the AI model to identify key gene partnerships predicted to impact the shape, size and weight of the heart muscle cells and tissue.
From that analysis, mutations in three genes surfaced: CCDC141 (a new gene not yet extensively characterized), TTN (a well-known gene associated with heart muscle disease) and IGF1R (a gene that regulates cardiac development and metabolism). Mutations in only one of these genes left heart cells relatively unaffected, compared to the effect of multiple mutations (though single mutations can still impact heart health). But in collaboration, they seemed problematic. To prove the mutations banded together to damage or rescue the heart, the team needed to test their effects in real human hearts.
To probe cellular effects, the researchers used small segments of RNA to switch off pairs of genes in different combinations within lab-grown heart muscle cells. Then, they ran hundreds of thousands of these cells through a new tool Wang developed — rapidly pumping them through narrow microfluidic channels to form a single line like a tiny highway. A high-speed camera snapped photos of the cells as they whizzed by, providing images of each cell’s morphology. The team found that no matter the combination of pairs between TTN, CCDC141 and/or IGF1R, the heart cells’ size enlargement (a hallmark of hypertrophic cardiomyopathy) was significantly reduced when both genes in the couple were turned off.
“We are excited because this could be a new genetic story for the cause behind cardiac hypertrophy,” Wang said. “This tool may open some new frontiers on how we diagnose and treat these types of diseases.”
What’s next for epistasis research
Ashley’s team plans to refine its epistasis detection algorithm to identify more genetic interactions and investigate other complex diseases such as cancer. They are also exploring how genetic mutations and environmental factors interact to shape complex human traits that may be involved in disease development. “In time, it is likely we will target more than one gene at a time,” Ashley said, “and these kinds of epistatic relationships will be the first place we start to look for those bigger effects.”
Wang hopes epistasis can inform new therapies. Often, when a single gene takes the brunt of the blame for a disease, the treatment also targets these one-off mutations. But epistatic gene interactions can help researchers better understand the network of disruption that needs a remedy, informing a more complete treatment plan.
“This can change the way we think about complex diseases and can lead to a promising new frontier in developing therapies,” Wang said.
- 8
- 9
- 2
