Autism Starts in Your Genes

Despite the recent resurgence of misinformation and debunked theories about what causes autism, scientific research has long shown that the risk of autism actually develops way before a child is born. In fact, a gene mutation that puts a child at risk for autism may occur soon after conception in the sperm or egg, or later as the fetus develops.

“At risk” doesn’t always guarantee autism. And in most cases, autism isn’t caused by a single gene, but a combination of small effects that are common in many genes. Plus, environmental factors combined with genetic factors may also play a role.

As you can see, there are still a lot of questions to be answered about autism spectrum disorders. One scientist examining the potential causes of autism is Daniel Geschwind, MD, PhD, the Gordon and Virginia MacDonald Distinguished Professor of Neurology, Psychiatry and Human Genetics at University of California, Los Angeles (UCLA). In his capacity as senior associate dean and associate vice chancellor of precision health, he also leads the Institute for Precision Health at UCLA.

Dr. Geschwind’s research lab studies how genes increase the risk for autism spectrum disorders. His laboratory has made major contributions to identifying genetic causes of autism spectrum disorder and defining the molecular pathology of autism and allied psychiatric disorders. He has also worked to increase diversity in autism research.

BrainWise recently interviewed Dr. Geschwind about his research. He will be a keynote speaker at this year’s annual conference of the National Academy of Neuropsychology—which produces BrainWise. His presentation will be “What Genetics Has Taught us about Autism Spectrum Disorder.”

This Q&A has been edited for length and clarity.

BrainWise: Tell me about your research of genes and autism.

Daniel Geschwind, MD, PhD: Genetics has really transformed our understanding of conditions that impact mental health broadly, autism being one of them. The recognition that autism had a clear genetic etiology was demonstrated by an early twin study in the 1970s. Then, studies over the last 30 years have shown that autism has an 80% heritability. That means that in the population, factors inherited from parents contribute the majority to somebody having autism. That’s very clear and very strong data. Undeniable.

The other interesting thing is that genetic factors can also cause a disorder in a non-heritable way. An example of that would be Down syndrome, a condition that is not inherited from the parents but caused by a new mutation in the egg or sperm. In autism, about 15% is contributed by these major forms of new mutations—where having one gene mutation is sufficient to cause autism, most likely. So, with 80% from heritable factors and 15% non-heritable, genetics contributes the vast majority of the chance of somebody having autism.

However, this is something people are not responsible for. It’s not like you, as a parent, gave kids autism because of what you ate, what you took, or your parenting style. None of that likely plays any role. It’s very clear that you don’t control heritable factors. It’s also not your actions that led to this. It’s nature.

Why is this important for research?

Take cancer, for example. We use genetics to identify mutations in tumors, and in about 10% of cancer cases, we identify a mutation that has a specific drug that can treat that specific cancer.

Especially for these rare forms of autism that have major gene mutations, that gives us a chance to understand the mechanism of disease. Researchers can genetically engineer mice or cells or other model systems that then they can study in the laboratory to understand how autism actually develops. And from that understanding, one can develop rational, targeted therapeutics.

I want to make it clear that we’re not saying that everybody with autism needs to get a drug or be treated with medication. In many of the cases, especially those with very high-functioning autism, it’s just a question of accommodating their needs. There’s a part of autism that’s on a normal continuum with the rest of behavior, social behavior, language, repetitive behavior, planning. We all have strengths and weaknesses, things we’re good at, things we’re not. There’s an aspect of autism where it’s not a disorder at all, but represents neurodivergence. These folks just have a different way of engaging with the world and thinking, and they need accommodation, rather than medical therapy or treatment.

But by identifying these genetic causes, it gives us an anchor that allows us to have a starting point for therapeutic development. The hope is that over the next decade, we’ll begin to have successful clinical trials and therapies for specific forms of autism caused by specific genes.

There are now a number of childhood neurologic disorders, especially neuromuscular disorders, where gene therapy has been shown to be very effective—almost a cure, which is kind of remarkable. We think that we can leverage the advances in gene therapy to focus on neurodevelopmental disorders like autism as well.

There are challenges though: When do we have to intervene? Do we have to diagnose somebody with a strong genetic form early during development, when they’re in utero? And do we have to get to in-utero gene therapy, which is another frontier in medicine? Or can we wait? All of those questions need biomedical research to answer them. That’s why the National Institutes of Health (NIH) funding is so important. Now that we have a lot of these genetic discoveries, we should allocate funding to use this information to develop therapies. We have hundreds of genes now that we know can cause autism and other neurodevelopmental disorders, and investigators should be funded to study those.

How many mutations are we talking about?

Of the 15% that’s caused by nonheritable genetic factors, none of them are very common. None account for more than 1% of autism—that means if I have 100 children, each one may have a different mutation.

There have now been close to 200 mutations that have been identified as autism-causing genes. But there might only be 40 or 50 patients in the world with one of those mutations. In some cases there are hundreds, and for some, there might just be a few dozen identified, so that collection of rare diseases makes it challenging.

At the same time, it allows us to begin to ask important questions. For example, a major theme in medicine that has kind of driven modern medicine for the last few 100 years, and that’s been refined over time, is a framing of diseases in what we call pathophysiology, which basically means how a medical condition came about.

An example would be heart attacks. Heart attacks are not caused by one thing. They can be caused by having an abnormal heart rhythm or by having cholesterol too high and having an artery blocked, or by having too high blood pressure that puts a stress on the heart. So, there are many different pathways, but it converges on essentially the same thing, which is low oxygenation to the heart muscle, which then stresses the heart muscle and it can’t work well.

So the question for disorders like autism is: What is the convergent pathology? What is the description of the disorder from a brain and behavior standpoint? Behaviorally, autism has been described. But how does that manifest in the brain during development?

By now having over 100 mutations, we can genetically engineer animals and cells—even human cells in a dish—to observe the effects of the mutations that cause autism. And we can begin to ask: Where and when do these mutations converge? What biological processes during brain development do they impact? And how does that impact brain function?

Everything we know points to all of that starting in utero during the development of the brain at mid-gestation. That’s a period of the highest risk. That helps us understand that you’re born with the predisposition—but it doesn’t mean it can’t be altered.

What are you excited about in your autism research right now?

One of the things we’ve been able to do—and others are working on this as well—is use the most advanced CRISPR engineering. Not to edit the genome—we don’t change the genome. But we use CRISPR to turn on genes that are muted or off. We have started to begin to explore that, and it shows some promise as a potential therapeutic for the rare and severe forms that usually are considered a medical genetic condition. And in that 15% of autism, most of those mutations are in one copy of the gene. So there’s one mutated copy that’s not expressing the protein properly, and there’s another that’s just fine but is not sufficiently expressed. So, we can actually use CRISPR to augment the gene expression on the remaining non-mutant copy, so that it actually becomes closer to normal and rescues the effects of the mutation. So, that’s looking very promising.

But what about the other 80% of causes of autism?

There still is a lot to learn. We have identified only the tip of the iceberg when it comes to the common genetic contributions to autism. This is an area of active investigation. However, what we have learned so far has given us important clues.

The past 30 years of NIH-funded research—as well as other foundation-funded research—has been extraordinary in moving from poorly supported theories from the 1960s about parenting to really understanding the biological origins of autism, which provides hope for treatment and for improving the lives of families and kids.