An international research effort headed by a team at the University of Michigan (UM) has linked more than 140 genetic variants and 150 protein-coding genes with an increased risk for developing atrial fibrillation (AF), a heart rhythm disorder that affects an estimated 2.7 to 6.1 million people in the United States and more than 30 million world-wide.
The scientists carried out a genome-wide association study (GWAS) that analyzed genetic data from more than 1 million people, including patients with AF and healthy controls, to find which variants might play a contributing role. They hope that the findings will offer up new genetic insights into the biological processes that cause AF, which could support screening tests, and potential new targets for drug discovery and development.
As a result of their work, the researchers have used the new data to build a statistical tool that can predict an individual’s risk for developing the disorder based on which of the genetic variants are present. “This polygenic risk score could be used in a clinical setting to help identify high-risk individuals, so that preventive measures can be taken before they develop AF, which can have fatal consequences,” commented research co-lead Cristen Willer, Ph.D., associate professor at UM Department of Internal Medicine, and head of the university’s Willer lab. The GWAS data might help to increase our understanding of why and how AF is triggered and maintained, so that genetic screening of patients who already have AF could help clinicians decide on the best course of therapy.
The protein-coding genes and regulatory sequences identified could also represent targets both for developing new treatments for AF, and for repurposing existing medicines that are used to treat unrelated disorders. “Each one of the genetic variants and genes that we have identified represents a clue to the underlying pathways and processes that cause AF,” said Willer in an interview with Clinical OMICs. “Further research will be needed to identify which could feasibly be used to develop new drugs for treating or preventing the disorder.”
The UM team and colleagues in the U.S., Japan, and Scandinavia report their studies in Nature Genetics, in a paper entitled, “Biobank-driven genomic discovery yields new insight into atrial fibrillation biology.”
Potentially Thousands of AF-Related Variants
Although AF is more prevalent in older people, the disorder still affects an estimated 2% of people under-65. And as a major risk factor for stroke and heart failure, AF contributes to roughly 130,000 deaths every year in the U.S. and represents an increasingly significant healthcare burden as the global population ages. “More than a quarter of people over the age of 80 years suffer from atrial fibrillation, but treatments are limited, and we don’t really understand the biological or genetic basis of the condition,” Willer acknowledged. Current treatments for AF include drugs that can either help to control the heart’s rhythm, or that reduce the risk of stroke. In some cases surgical procedures can block the abnormal electrical signals that cause the upper chambers of the heart to beat irregularly.
“Atrial fibrillation is a complex disorder, and there isn’t a single causative genetic factor,” Willer continued. “While the risk of AF in any one individual may be increased by environmental factors such as heart surgery, it’s generally believed that there could be hundreds or even thousands of genetic variants that impact on AF risk.” And although each of these may only have a very small standalone effect on risk, “… carrying many of these variants in combination has an additive effect that can increase your overall risk of AF significantly.”
Developing an AF Risk Score
Large biobanks containing tissue samples and data on the health, lifestyle, and demographics of many thousands of healthy people and individuals with different diseases have made it possible to study how genetic variation across our genomes impacts on health and disease. “Being able to tap into these large biobanks means that we can now study sample sizes that are big enough to identify genetic risk factors for complex disorders such as AF,” Willer commented. “GWAS is a very powerful approach for finding common variants that exert very small individual effects on disease risk. But when we started this study two years ago there had been relatively few GWAS focused on AF.”
For their study, the UM team and colleagues in the U.S., Denmark, Norway, Iceland, Spain, and Japan drew on data from six different large-scale biobanks in the U.K, U.S., and Scandinavia: UM’s Michigan Genomics Initiative (MGI), UK Biobank, Norway’s HUNT study, DiscovEHR, Iceland’s deCODE Genetics, and AFGen Consortium.
They tested 34,740,186 individual genetic variants for an association with AF, by comparing the prevalence of each variant in 60,620 individuals with AF and 970,216 healthy controls. Analyzing the results identified 142 independent genetic risk variants sited within 111 different regions of the genome—80 of which had not previously been associated with AF—and linked 151 functional genes directly with an increased risk for AF.
The researchers then applied mathematical tools to construct a polygenic risk score that could predict an individual’s risk of AF, dependent upon which of the genetic variants they carried. “The risk score takes into account an aggregation of all the genetic variants,” Willer told Clinical OMICs. “The higher the score, the greater the predicted risk of AF.” When the researchers tested the score on the UK Biobank data, they found that it was highly predictive, but also very specific for AF. Interestingly, the results showed that individuals who developed AF at a younger age were more likely to have a greater number of the genetic variants than those who developed AF later in life.
AF Variants Implicated in Other Heart Conditions
Many of the identified protein-coding genes were likely involved in cardiac muscle structure and function, ion channel function and calcium signalling, and hormone signalling. “This is what we expected,” Willer commented. Other genetic risk variants were also close to genes in which mutations are already known to cause serious heart problems in humans.
More unexpectedly, a number of the genetic variants are related to genes that are involved in fetal heart development, but which are normally switched off in the adult heart.
“For example, there was a strong association between AF and the gene for fetal myosin heavy chain,” Willer said. Myosin heavy chain is a key protein involved in muscle contraction. During fetal heart development, cells switch on the fetal myosin heavy chain gene to produce the fetal form of the protein. The fetal gene is subsequently switched off, and the adult gene switched on to produce the adult form of the myosin heavy chain protein.
The identified link between fetal heart genes and AF hints at two potential mechanisms, Willer explained. “It is possible that variations in fetal cardiac protein-coding genes could impact directly on heart structure or function during cardiac development, which then predisposes to AF later in life. Alternatively, genetic variation occurring in regulatory elements might trigger the undesirable reactivation of fetal genes in the adult heart, which might alter electrical signalling, for example, and lead to AF.”
This second hypothesis was supported by studies in a rabbit model of induced AF, which showed that at least one of the fetal genes identified through the GWAS was reactivated in rabbit hearts in parallel with the induction of AF.
Effectively, one of the genetic variants acted as a molecular switch that turned off production of adult myosin heavy chain and switched production of the fetal form of myosin heavy chain back on. This switch altered the contractile properties of the heart muscle, which can trigger AF. “If we can demonstrate that turning off the fetal genes in the stressed adult heart can relieve AF, this may represent a potential future therapeutic approach,” Willer suggested.
AF-Linked Genes Key to Future Treatments
The team separately identified known drug compounds that interact with 32 of the 151 AF-linked functional genes. Some of these drugs are already known to be able to either control or trigger cardiac arrhythmias, but other compounds were originally developed to treat completely unrelated diseases, including neuropsychiatric or inflammatory disorders. “Whether the highlighted drugs can be used to treat or prevent atrial fibrillation requires further evaluation, but the findings can be used as a foundation for directing future functional experiments and clinical trials,” the authors stated in their published paper. “These findings need confirmation but provide a foundation for directing future functional experiments to better understand the biology underlying atrial fibrillation.”
Just a couple of weeks before the UM team’s paper was released in Nature Genetics, the journal published a meta-analysis of existing AF GWAS. This study, carried out by a large international research group headed by Patrick Ellinor, Ph.D., and scientists at the Broad Institute of MIT and Harvard, included 65,446 cases with AF. The potential now exists to pool data from both studies, to generate an even more detailed list of genetic variants, and a more accurate polygenic risk score, Willer explained to Clinical OMICS. “The Ellinor study gives us a major opportunity to collaborate and increase scientific knowledge even further. One of our next steps will be to see if we can work with the group to put all the data together. Combining the two datasets will increase the sample size significantly,”—to more than 93,000 AF cases—“and will almost certainly lead to the identification of additional genes as potential targets, as well as providing even greater detailed insights into the biological basis of AF. We should also be able to generate a much more informative polygenic risk score for identifying individuals at high risk of AF. How well that risk score performs in the clinic will need to be verified in randomized clinical trials.”
Ultimately, the ability to identify who is most at risk of developing AF should allow the development of preventive approaches that improve patient outcomes and quality of life. “Discovery of novel genetic variants and genes important for atrial fibrillation was only possible because we combined information from multiple biobanks from around the world in a large collaborative effort,” concluded first author Jonas Bille Nielsen, M.D., Ph.D., a cardiovascular researcher at UM. “Combining the advantages of each of the data sources helped us to better understand the biology underlying atrial fibrillation.”
These findings need confirmation but provide a foundation for directing future functional experiments to better understand the biology underlying AF.
Besides leading to palpitations and reduced fitness, AF is a major risk factor for stroke, heart failure, and death, Nielsen said. “If you just once have seen a person disabled, bound to a wheel chair, and unable to speak as a result of a major stroke, likely caused by atrial fibrillation, you know it is an important disease to treat and prevent. However, current treatment options are limited, can cause serious side effects, and are very rarely curative.”
A better understanding of the biological processes underlying AF could lead to better treatment and prevention, he noted. The team’s identification of more than 150 genes that are likely to be involved in the initiation and progression of AF should prove valuable in the search for novel therapies. “An ongoing, randomized trial is currently testing whether continuous monitoring of the heart rhythm using a small implantable device can led to earlier detection of atrial fibrillation and prevent strokes,” Nielsen continued. “If this or similar studies come out positive, it is likely that high-risk persons, such as those with a very high genetic burden or risk score for developing atrial fibrillation, will benefit the most from a device and the implantation could eventually be guided by genetics in combination with other risk factors. This or similar approaches of personalized medicine using genetic information from hundreds or even thousands of variants still needs to be proven to be effective in randomized trials, but is more likely to be successful as our understanding of the genetic mechanisms underlying atrial fibrillation and other common diseases is expanding rapidly due to large collaborative efforts like the ones we just presented.”
This article was originally published in the September/October 2018 issue of Clinical OMICs. For more content like this and details on how to get a free subscription, go to www.clinicalomics.com.