A Shared Druggable Switch: How Nitric Oxide, TSC2, and mTOR Connect Multiple Forms of Autism

Here is something that has puzzled autism researchers for decades. We know that autism has an enormous genetic diversity. Hundreds of different genes have been linked to it, each working through different mechanisms, in different families. And yet, when you sit across from a child in clinic, many of the core features look remarkably similar. The social communication differences. The sensory sensitivities. The repetitive behaviours. If the genetics are so varied, why does so much of the biology end up in the same place?

A study published in Molecular Psychiatry in early 2026 has brought us significantly closer to answering that question. And unusually for a basic science paper, it points directly toward a potential treatment target.

A pathway you have probably heard of

If you have spent any time reading about the genetics of autism, you will have come across the mTOR pathway. It is one of the most important molecular switches in brain development, controlling cell growth, protein production, and the formation of synapses, the connections between neurons.

When mTOR works properly, it responds to signals and switches on and off as needed. When it gets stuck in the “on” position, brain development goes awry. This is well established in tuberous sclerosis complex (TSC), a genetic condition caused by mutations in the TSC1 or TSC2 genes that directly regulate mTOR. A significant proportion of children with TSC also have autism.

But here is the question that has nagged at researchers: is mTOR dysregulation just a TSC problem, or does it play a role in other forms of autism too?

A molecular chain reaction

Dr Haitham Amal and his team at the Hebrew University of Jerusalem set out to answer exactly that question. What they found was a specific chain of molecular events that leads to mTOR overactivation, and it works like this:

1. Nitric oxide, a signalling molecule produced naturally by neurons, chemically modifies the TSC2 protein at a specific site (cysteine 203). This process is called S-nitrosylation.
2. That chemical tag marks TSC2 for destruction by the cell’s waste disposal system.
3. Without TSC2 acting as a brake, mTOR runs unchecked, leading to the kind of synaptic disruption seen in autism.

Now, that on its own would be interesting but not extraordinary. What makes this study stand out is what happened next.

The same mechanism, again and again

Most findings in autism research apply to a single genetic model. You see something in one type of mouse, and it does not replicate in another. This study was different. The same nitric oxide/TSC2/mTOR mechanism was confirmed in:

• Shank3 mutant mice, one of the most widely studied genetic models of autism
• Cntnap2 mutant mice, a model associated with language and social difficulties
• Blood samples from children with autism, including both those with a known genetic cause and those with idiopathic autism (where no single gene has been identified)

That last point deserves underlining. The same molecular signature turned up in human blood samples from children whose autism has no identified genetic cause. This is not a niche finding relevant only to rare syndromes. It may be part of the broader biology of autism itself.

And when they blocked it?

This is the part that gets genuinely exciting. When the researchers used a drug to block neuronal nitric oxide synthase (nNOS), the enzyme that produces nitric oxide in neurons, the entire cascade was prevented. TSC2 was preserved. mTOR activity returned to normal. The downstream effects were reversed.

In other words, they found the switch, and they found a way to flip it back.

What “druggable” really means

You will see the word “druggable” in the press coverage of this study, and it is worth being precise about what that means. It means the chemistry is feasible. There is a specific enzyme (nNOS) acting on a specific protein (TSC2) at a specific site (cysteine 203). That is exactly the kind of target that pharmaceutical development can work with.

It does not mean a drug exists today. It does not mean clinical trials have started. It does not mean a treatment is around the corner. What it means is that we can see a specific point in the molecular chain where intervention might work, and that is a genuinely important step. Many aspects of autism biology are too diffuse to target without unacceptable side effects. A specific enzyme, a specific protein, a specific amino acid: that is actionable.

How this fits with the bigger picture

This study did not arrive in isolation. A landmark paper published in Nature around the same time, led by Daniel Geschwind at UCLA, showed that eight different autism-linked mutations, when modelled in brain organoids, start out producing different effects but gradually converge on the same developmental pathways (1). Geschwind found convergence at the developmental level. Amal found convergence at the molecular level. Together, they paint a compelling picture: autism’s enormous genetic diversity funnels into a surprisingly small number of shared biological bottlenecks.

And if the bottlenecks are shared, treatments targeting those bottlenecks could potentially help across multiple genetic subtypes. That is a fundamentally different proposition from needing a bespoke therapy for each of the hundreds of genes involved.

The important caveats

This is a preclinical study. The mouse models and plasma samples are strong evidence, but we are still a long way from a treatment for children. Nitric oxide does many things in the body beyond the brain. It helps regulate blood pressure, immune function, and gut motility. Any drug targeting nNOS would need to be exquisitely specific to avoid unwanted effects elsewhere.

The human plasma findings come from a relatively small sample. Larger studies will be needed to confirm whether this mechanism is a consistent feature across the autism spectrum or is more relevant in certain subgroups.

What this means for families

If your child has idiopathic autism, where no single genetic cause has been found, this research is particularly relevant. It suggests that shared molecular mechanisms may be at work even without a specific gene diagnosis, and those mechanisms may eventually be targetable.

If your child has a known genetic cause such as a Shank3 or Cntnap2 variant, this study provides direct evidence that the mTOR pathway is disrupted through an identifiable, specific mechanism.

This is not a treatment today. But it is the kind of finding that treatments are eventually built upon. A specific target, validated across multiple models, confirmed in human samples. That is how the journey from laboratory to clinic begins.

References

1. Geschwind, D.H. et al. (2026). Developmental convergence and divergence in human stem cell models of autism. Nature. Full text
2. Ojha, S.K., Kartawy, M., Hamoudi, W., Tripathi, M.K., Aran, A. & Amal, H. (2026). Nitric Oxide-Mediated S-Nitrosylation of TSC2 Drives mTOR dysregulation across Shank3 and Cntnap2 Models of Autism Spectrum Disorder. Molecular Psychiatry. Full text

---

Dr Odet Aszkenasy is a Consultant Community Paediatrician and the author of The Genetics of Autism: A Guide for Parents and Professionals.