Labs discover new mechanisms that cause irregular heartbeat microbiologystudy

Medicine often takes a one-size-fits-all approach, but a disorder’s root cause can vary. Vanderbilt researchers have found that for people with long QT syndrome, a heart condition that causes an irregular heartbeat, a more tailored approach could be beneficial.

The research is published in PNAS. It was led by Vanderbilt University Professor of Biochemistry and Vice Dean of Basic Sciences Charles Sanders, and Northwestern University Professor and Chair of Pharmacology Dr. Al George.

Congenital LQTS is caused by genetic mutations that affect KCNQ1 potassium channels in the heart’s muscle cells. These channels act as pathways for electrical currents that help reset the heart’s electrical activity after each contraction. For patients with LQTS, this process is prolonged, leading to an irregular heartbeat. An inadequate understanding of what causes these potassium channels to malfunction makes it difficult to create effective treatments for LQTS patients, especially considering that over 2,500 mutations in KCNQ1 have been documented in people, most of them with unclear pathogenicity.

To address this knowledge gap, the Sanders lab, in collaboration with the George lab at Northwestern University, characterized 61 KCNQ1 gene variants to determine whether and how they might lead to LQTS. The chosen variants represented amino acid changes located across the full span of the protein sequence.

The researchers characterized channel variants using a variety of methods. Using flow cytometry, they recorded whether the potassium channel coded for by each variant made it to the cell’s surface—a critical step in heart contraction. They also tested how resilient the channels were by heating them. Some remained as heat-resistant as the wild type channel, while others proved much less stable; unstable channels are more likely to misfold and less likely to successfully be transported to the cell surface. Third, the researchers used electrophysiology experiments to measure electrical currents in the potassium channels, thereby measuring how well each variant functioned.

Although LQTS is traditionally thought to be caused by loss-of-function mutations in KCNQ1, the exhaustive testing conducted by first author and Sanders lab Ph.D. student Kathryn Butcher (Brewer) indicated that the tested KCNQ1 variants can be sorted into five classes that all display loss of function, but through different mechanisms.

Some variants were stable and trafficked correctly but exhibited reduced channel activity; some were normal except they were regulated differently by a change in transmembrane voltage; others were seen to be unstable and mistrafficked; and others were stable but still mistrafficked.

Of note, one group of tested variants demonstrated wild-type-like properties, which would suggest that those mutations are “benign” and don’t lead to disease.

“Our results indicate that if you suffer from LQTS and find that you are carrying a KCNQ1 variant, you can’t assume that a particular molecular mechanism underlies your condition,” Sanders said. “If they haven’t been done yet, someone needs to do the necessary experiments to establish the exact disease mechanism of a particular patient, which in the future might help determine their treatment.”

Several disease-prediction algorithms exist that purport to predict whether a particular protein variant is pathogenic or benign. To see if these algorithms could potentially help clinicians determine into which class a particular patient’s KCNQ1 mutation falls—and through that knowledge, how to treat that patient—the researchers tested eight such algorithms against their experimental data.

Generally, the algorithms successfully predicted disease in which the variants improperly folded, but they often could not predict disease in variants with a modified response to electrical signals. Stanford University’s REVEL predicted disease incidence best, with a prediction accuracy of 85%. AlphaMissense, a Google DeepMind tool related to AlphaFold2, was not among the best-performing algorithms. The researchers hope that future work will focus on improving these prediction algorithms.

According to Sanders, this paper was the result of a hugely collaborative “roll up your sleeves and do the experiments” effort in which each lab was responsible for a different set of experiments. Notably, Professor of Cell and Developmental Biology Dylan Burnette and his graduate student James Hayes dropped everything to help Sanders and George when reviewers requested additional experiments, providing data that was pivotal in the final acceptance of the paper.

“Kathryn and Carlos Vanoye from the George lab were exceedingly patient and methodical in their work, though they were supported by other lab members and the crucial work by James Hayes and Dylan Burnette.” Sanders said. “It takes a lot of fortitude to complete a project like this. I am so proud of all of them.”

The study is a step toward treatments tailored to a patient’s genetic makeup.

“This work is part of the growing field of personalized medicine, often also referred to as precision medicine,” Sanders said. “Our results could be used by physicians treating patients who carry the mutations we have looked at.”

One LQTS patient with a particular variant, for example, might require a drug that enhances potassium channel functioning while another LQTS patient might require the exact opposite: a drug that calms down their overactive potassium channels.

“Our results complement fantastic pioneering work on the personalized medicine of ion channel-based heart disease being carried out by a number of labs in the medical center, including those of Dan Roden, Andrew Glazer, Jerod Denton, Brett Kroncke, and Bjorn Knollmann, among others,” Sanders said.

“Some of these labs are now using deep mutational scanning to analyze huge numbers of channel variants. Our work may provide a benchmark data set for testing and may further illuminate results from their high-throughput testing on KCNQ1.”