LSD1 inhibitor ameliorates autism-associated traits in new mouse model. We need to talk.

In findings published March 26 in Molecular Psychiatry, a team of researchers from Tokyo University, Juntendo University, and RIKEN report that they have successfully created a new mouse model of autism spectrum disorder in what they have named Kmt2c^+/fs mice. Then things get weird.

They engineered the animals to have only one functioning copy of the lysine methyltransferase 2c (KMT2C) gene, which encodes one of the catalytic units of histone H3 lysine 4 (H3K4), instead of two. Histones are large molecules that act as spools when DNA coils up into its inactive state. They also allow the DNA to unspool so that it can be transcribed and produce the proteins and regulatory elements that keep living things running. A few people diagnosed with autism spectrum disorder have KMT2C haploinsufficiency, meaning that they have only one active gene, like the mice, but most do not.

When the mice were old enough to run a maze, the researchers ran exercises to assess their sociability, rigidity, cognitive function, and other traits that they defined as associated with autism (the mice were also hypersensitive to sound, but it’s not clear if the scientists measured this). For example, they put the mice in enclosures that gave them the choice of either spending time with another mouse or hanging out alone. The Kmt2c^+/fs mice chose to spend time alone significantly more often than wild-type (Kmt2c^+/+) mice did.

Then the researchers examined gene expression a little further. They found the experimental mice also showed upregulation of the expression of other genes already known to be associated with ASD, something the scientists had not included in their design.

“This was somewhat unexpected,” said study co-author, Dr. Atsushi Takata. “KMT2C mediates H3K4 methylation, which is thought to activate gene expression, and thereby KMT2C haploinsufficiency was expected to cause reduced expression of target genes.”

In other words, less expression of KMT2C appears to mean more expression of genes affirmatively associated with ASD.

The researchers then gave the mice vafidemstat, a brain-penetrating agent that acts on lysine-specific histone demethylase 1A (LSD1A) and has been found to render histone problems less severe. The treated mice showed less social impairment, and the differences in their gene expression relative to wild-type mice became less pronounced.

As the research paper says as it draws to a close, “This indicates that histone-modifying drugs would be effective not only in patients with KMT2C haploinsufficiency, which represents a tiny proportion of individuals diagnosed with ASD, but also in a broader patient population with specific types of transcriptomic and epigenomic dysregulation.” Or, translated into press release, “The results open doors to future research to strengthen the foundation for the pharmacologic treatment of ASD and other neurodevelopmental disorders.” Give an autistic organism a drug that affects one gene, and it will in fact affect many genes, making the organism act more neurotypical.

And we’d better start the hard conversations about that now.

Each study is one tiny step toward understanding, and the researchers are professionally clear about the work that remains to be done before their findings can be used in drug discovery, but there are a few things to keep in mind as we head down this road:

First, ASD itself is poorly understood, partially due to unclear diagnostic criteria that were only somewhat remedied when the DSM-IV became the DSM-V. In a hundred years, we may find that today’s ASD is really five or six different neurotypes with overlapping presentations.

Second, scholars of prehistory are beginning to speculate that ASD, ADHD, dyslexia, and other conditions considered disabling in today’s 9-5+ job market may have been beneficial to early human communities, even if they pose a problem at the individual level today. Some people with ADHD feel more alert and productive at night. There was one guy in the village who didn’t mind night watch duty. A big part of autism is about seeing the world differently. Only one adult needs hear the tiger break a twig before it pounces on a posse of tribal toddlers. Medicating ASD out of existence could rob humankind of something that we don’t yet fully understand.

Third, our school systems already have a habit of reaching for drugs out of convenience. The Body Keeps the Score cites the use of antipsychotic meds “to render ‘difficult’ children more tractable.”* Autistic students are often inconvenient. There are autistic adults who describe the applied behavior analysis (ABA) interventions they received as youngsters (sometimes the only intervention covered by insurance) as functionally torture to make them pretend to be normal. The capacity for abuse of any autism-ameliorating drug treatment already has a fully developed infrastructure.

If these findings described an avenue for treatment of epilepsy, we’d shout hooray and move on. Offer epileptics a way to no longer have epilepsy, and almost all of them will cheerfully agree. Offer the parents of epileptic children a cure for epilepsy, and they will gladly consent on their child’s behalf. The patient will likely grow through all stages of youth and adulthood with few conflicted feelings about this. The thing that is making an epileptic person unhealthy is often epilepsy.

This is because epilepsy is a disease, not an identity. Ask an epileptic if their epilepsy is an essential part of who they are as a human being, and most will give some version of “no.” Ask an autistic? If you’ve met one person with autism, you’ve met one person with autism, but you’re more likely to get a “yes.”

Autism is part of how autists see, hear, think about, and interact with the world. It can be a disease, but not always. It can be a disability, but not always. It is an identity. The thing making an autistic person unhappy can be autism, but more often it’s the way other people react to autism.

Even if we assume that some autistic individuals would benefit from drug intervention that brings them closer to what we think of as normal health, there is likely to be a large swath of autistic individuals who would be better off without it but are not in a position to refuse. Autistic children are often more work than their parents and teachers expected and do not always involve the same emotional and social rewards as their neurotypical counterparts. We should expect a large cohort of parents ready to jump at the word “cure” and a large number of strapped school systems ready to demand they do so. It isn’t as if no one’s doing it now.

We may be a long way from “Here is a pill you can take that will make you not autistic any more,” but it’s not the first time many autistic individuals will have heard it. Autistic adults can tell stories of being medicated or threatened with medication for failure to hide their neurodiversity thoroughly enough.

Some of the nineteenth-century scientists discovering genes for the first time were asked whether their work could be used to improve humanity. The response, imperfectly paraphrased from my years-ago read-through of The Gene by Siddhartha Mukherjee, was that we would need “radical social equality” before we should even attempt radical genetic manipulation.* We had to know not only what each gene does but how it got there and what the true implications are. Similar caution is called for here.

Nakamura, T., Yoshihara, T., Tanegashima, C. et al. Transcriptomic dysregulation and autistic-like behaviors in Kmt2c haploinsufficient mice rescued by an LSD1 inhibitor. Mol Psychiatry (2024). https://doi.org/10.1038/s41380-024-02479-8

*Quotations are from memory and may be imperfect.