Discovery of hundreds of genes potentially associated with ALS may steer scientists toward treatments

Using machine learning, Stanford Medicine scientists and their colleagues have found hundreds of genes that could play a role in amyotrophic lateral sclerosis.

- By Hanae Armitage

Using a computer algorithm based in machine learning, scientists have identified nearly 700 genes that may be implicated in amyotrophic lateral sclerosis, also known as Lou Gehrig's disease.
Giovanni Cancemi

Thanks to a super-powered genetic sleuthing method, Stanford School of Medicine scientists have discovered almost 700 genes potentially associated with ALS, creating new avenues for drug discovery and a better understanding of the debilitating neurological disease.

Fifteen genes have been shown previously to contribute to the onset of amyotrophic lateral sclerosis, but only a small proportion of people with non-inherited ALS harbor a mutation in one or more of those genes. Scientists suspect many more genes are involved.

“There’s this big gap between the genes known to be involved in ALS and the number of genes we suspect to be involved,” said Michael Snyder, PhD, professor and chair of genetics at Stanford. “Our study opens up a lot of new possible genes to explore in connection with ALS.”

Globally, more than 200,000 people are living with ALS, also known as Lou Gehrig’s disease. It erodes patients’ voluntary muscle movement, crippling their ability to walk, talk, eat and, eventually, breathe. Patients survive an average of two to four years after onset of the disease.

Finding ALS-associated genes among masses of genomic data has proved difficult — like finding a few needles in a field of haystacks that stretches for miles. But Sai Zhang, PhD, a genetics instructor and one of the lead authors of a paper describing the research, devised a way to scour massive datasets and find genes that are relevant to disease, in this case ALS.

“We developed a computer algorithm, called RefMap, which is rooted in machine learning, a powerful data analysis methodology that automatically identifies patterns from massive complex data.” Zhang said. “We want to let the data lead us.”

Michael Snyder

In addition to finding many genes that could contribute to the ALS, the researchers believe the study has settled a few important questions about the disease.

“There’s a long-standing debate about where ALS originates in the cell,” said Johnathan Cooper-Knock, a Stanford visiting scholar and lecturer at the University of Sheffield in the United Kingdom. “This new technique has surfaced genetic evidence that really pins down the axon of motor neurons as the place of disease origin.” (The axon of the neuron is a long cord that helps transmit electrical signals from one neuron to another.)

A paper describing the study was published Jan. 18 in Neuron. Snyder, the Stanford W. Ascherman, MD, FACS, Professor of Genetics, is the senior author. Zhang and Cooper-Knock are co-lead authors.

Hundreds of new targets

Typically, ALS researchers investigate one gene at a time, performing in-depth analyses to tease out if and how that gene might contribute to the onset of the disease. The Stanford team’s approach was to cast a net far and wide for genes that may play a role in ALS. Zhang trained the algorithm to sift through millions of data points from studies known as genome-wide associated screens, which contain anonymized genetic information from thousands of patients with and without ALS. The strategy was to look for genetic mutations that often occur in people who have ALS.

The team narrowed the search further: Sorting through ALS patients’ data, the algorithm looked for mutations only in genes that support motor neuron function. “Searching only in motor neurons allowed our approach to discover more risk genes compared with previous methods,” Zhang said. The analysis spit out 690 candidate genes, some that were already known to be implicated in ALS.

“We can use this information to learn more about how and why motor neurons fail in ALS,” Cooper-Knock said. As an example, he added, “Many of the genes we uncovered pointed to the disease originating in the axon of the cell, rather than the cell body.”

“Previously it was not clear if axon defects were an effect of the disease, but our results indicate these defects are likely causative,” added Snyder.

KANK1

One gene, which repeatedly showed up in the data analysis, caught the researchers’ attention: KANK1, which is involved in functions at the very end of the axon. Through a series of experiments using stem cells and gene editing, the team showed that mutations in this gene lead to loss of a protein called TDP-43 from the nucleus of motor neurons, a hallmark of ALS.

“If you were to look in the brains of 100 people with ALS and analyze the motor neurons, you’d see this loss of TDP in something like 98,” Cooper-Knock said. “It’s almost the definition of ALS. If this phenomenon isn’t occurring, you probably don’t have ALS.”

“The finding is an exciting discovery, but it’s too early to consider KANK1 a drug target.” Zhang said. “More research will be needed to determine if reversing the effects of a mutated KANK1 gene can help treat the disease.”

The team also plans to do some experimental work on other “hits” from their dataset to determine whether any of the other hundreds of genes identified in the analysis could lead to ALS pathology.

Other Stanford co-authors of the study are life science researchers Minyi Shi, PhD, and Annika Weimer, PhD.

Researchers from the University of Sheffield; UC San Francisco; the Montreal Neurological Institute; Lund University; the Weizmann Institute of Science; and the University Medical Center in Utrecht, the Netherlands, contributed to this study.

This study was funded by the European Research Council, Health~Holland, the ALS Foundation Netherlands, the National Lottery of Belgium, the KU Leuven Opening the Future Fund, the Kingsland Fellowship, the My Name’5 Doddie Foundation, the Wellcome Trust and the National Institute for Health Research.

About Stanford Medicine

Stanford Medicine is an integrated academic health system comprising the Stanford School of Medicine and adult and pediatric health care delivery systems. Together, they harness the full potential of biomedicine through collaborative research, education and clinical care for patients. For more information, please visit med.stanford.edu.

2023 ISSUE 3

Exploring ways AI is applied to health care