“The point is, if you down-regulate one autism gene, you down-regulate a whole bunch of them,” said Andersen. The finding, detailed in this week’s issue of Nature, reveals that mutations in hundreds of different autism-linked genes may funnel into a shared molecular network—and identifies the gene ZFX as a potential explanation for why autism is diagnosed far less often in females.
The study, led by researchers at Boston Children’s Hospital and published in Nature, used CRISPR-based tools to knock down individual autism-risk genes in human neural cells and cerebral organoids. They found that silencing one gene often silenced many others in the same network, a phenomenon the authors call “transcriptomic convergence.” More than sixty genes on the X chromosome escape the usual silencing of one X chromosome in females, and among them is the transcriptional activator ZFX. Because females carry two active copies of ZFX, they express it at higher levels, which the researchers suggest may buffer them against mutations in autism-related genes.
What the ZFX Discovery Means for Patients
- Autism-linked genes may act as a connected network rather than isolated switches, meaning one therapeutic target could influence many genes at once.
- Females appear to have higher baseline expression of ZFX, a gene that helps regulate autism-related genes, which may partly explain the male bias in autism diagnoses.
- Early-stage research suggests boosting ZFX in males could offer partial protection, though no clinical therapy exists yet.
How Hundreds of Mutations Produce One Condition
The study described the convergence as a way to understand how diverse genetic mutations lead to similar clinical presentations.
“The most important take-home messages are the idea of this transcriptomic convergence in autism and our identification of the ZFX gene as a potential driver of autism’s male bias,” said Walsh.
The researchers employed CasRx (NLS-RfxCas13d-NLS), an RNA-targeting CRISPR system, to reduce expression of specific genes in neural progenitor cells derived from human induced pluripotent stem cells. They designed guide RNAs using deep-learning tools, filtered out sequences with off-target risks, and delivered them via lentiviral vectors. They also grew cerebral organoids—three-dimensional brain-like structures embedded in Matrigel—to observe these effects in a more complex tissue context. When they knocked down one autism-associated gene, RNA sequencing showed that numerous other genes in the same network were also down-regulated, supporting the convergence model.
Why Females Appear Protected
In typical female cells, one X chromosome is largely inactivated to balance gene dosage with males, who have one X and one Y. However, more than sixty X-chromosome genes escape this inactivation, including ZFX. Females therefore produce ZFX from both X chromosomes. The study found that ZFX regulates the expression of many autism-related genes, so higher ZFX levels in females could compensate for harmful mutations elsewhere in the network. The full organoid protocol is available via protocols.io.
Excited to share our new paper out today in @Nature ! With @BeckyEAndersen and @ChrisAWalsh1 , we find that disrupting diverse autism genes converges on strong, shared changes in gene expression. A quick 🧵… (1/n)
| Finding | Detail from the Study | Clinical Relevance |
|---|---|---|
| Transcriptomic convergence | Knocking down one autism-risk gene down-regulates many others in the same network | Explains why hundreds of different mutations can cause similar autism phenotypes |
| ZFX escapes X-inactivation | Females express ZFX from both X chromosomes; males from one | Higher ZFX in females may buffer against autism-linked mutations |
| Therapeutic implication | Upregulating ZFX in males could mimic the female protective effect | Suggests a single target could address multiple genetic causes |
Regulatory Pathways in the US, Europe, and UK
For health systems worldwide, the convergence model suggests that future therapies might not need to be tailored to each of the hundreds of individual autism-risk genes. Instead, treatments that boost master regulators like ZFX could theoretically address multiple genetic subtypes simultaneously. Currently, however, this remains preclinical research, and no ZFX-targeting drug has entered human trials. The complete study is available in Nature.
Funding Behind the Boston Children’s Study
The study was supported by the Autism Speaks Postdoctoral Fellowship 13008, NIGMS T32 GM007748, NIH T32 GM007753 and T32 GM144273, the Simons Foundation Autism Research Initiative, the Howard Hughes Medical Institute, Autism BrainNet (953759), the Yang Tan Collective at Harvard University, and the Hock E. Tan and K. Lisa Yang Center for Autism Research at Harvard University. D.C.P. and C.A.W. are Investigators of the Howard Hughes Medical Institute.
When to Seek Evaluation for Autism Concerns
This research does not change current diagnostic or treatment guidelines. Parents or caregivers who notice developmental delays, limited eye contact, or repetitive behaviors in a child should consult a pediatrician or developmental specialist. There is no approved therapy that boosts ZFX, and no genetic test currently uses this finding to predict autism risk. Anyone considering genetic testing should speak with a genetic counselor to understand the limitations and implications of results.
Looking Toward Convergent Therapies
Walsh suggested the convergence model opens the door to treatments that upregulate multiple autism-associated genes at once, rather than developing a separate drug for each mutation.
“Because one gene can regulate many, targeting fewer genes could be effective. For example, upregulating ZFX in males might bring them partial protection against autism closer to what females experience,” Walsh said.
The researchers caution that this remains a theoretical framework requiring extensive validation in animal models and eventually human trials before it could translate to clinical care. Detailed methods for the neural progenitor and organoid work are published on protocols.io.
References
- Andersen, R. E., et al. (2026). Transcriptomic convergence and the female protective effect in autism. Nature. DOI: 10.1038/s41586-026-11236-6
- Boston Children’s Hospital. Research identifies specific gene linked to female protection from autism.
- Protocols.io. Cerebral organoid generation and neural progenitor protocols. DOIs: 10.17504/protocols.io.x54v95p5ql3e/v1; 10.17504/protocols.io.5qpvoj5pzg4o/v1
