Dr. Pașca on Huberman Lab: How Assembloids Cure Autism

Autism currently affects close to 3% of the population and is diagnosed entirely on behavior—there is no biomarker. The spectrum ranges from mildly affected, fully functional individuals to those with profound autism who are nonverbal, have co‑occurring intellectual disability and epilepsy, and require lifelong care. Twin studies and gene discovery over the last 10–15 years show a very high heritability and hundreds of implicated genes (synaptic proteins, ion channels, chromatin regulators, etc.). This complexity means there will not be one “autism cure,” but rather many disease‑specific mechanisms and interventions.

Changes in diagnostic criteria, greater awareness, and diagnostic migration (e.g., cases previously labeled as intellectual disability now classified as autism) have clearly raised recorded prevalence. Yet the high heritability and consistent prevalence across countries (e.g., Scandinavia, Korea, U.S.) and across isolated populations argue against simple explanations like U.S.‑specific chemicals or vaccines. Environmental factors (e.g., historical thalidomide exposure, prematurity, infections) modulate risk, but no single modern environmental cause has strong causal evidence.

By reprogramming a patient’s skin cell into an induced pluripotent stem cell (via Yamanaka factors) and then into 2D neurons or 3D organoids, Pașca’s lab can watch human cortical development unfold over months to years in vitro. Organoids respect human developmental timing (e.g., NMDA receptor subunit switch at ~9 months in vitro), indicating intrinsic “timers.” Assembloids, created by fusing region‑specific organoids, self‑organize realistic interactions: inhibitory interneurons migrate into cortex; cortico‑spinal‑muscle circuits generate movement; four‑part somatosensory assembloids process pain signals. These preparations reveal circuit‑level phenotypes that cannot be seen in mouse knock‑outs alone.

Timothy syndrome is caused by a single base mutation in a calcium channel gene expressed in heart and brain, leading to cardiac issues, epilepsy, and profound autism. iPSC‑derived neurons from patients showed prolonged calcium influx; cortical organoids and transplantation into rats revealed dramatic, disease‑specific defects (e.g., markedly smaller pyramidal neurons) that don’t appear in simple 2D cultures. By deeply characterizing the channel’s misprocessing, Pașca’s group designed a short nucleic‑acid therapeutic that rewires splicing/processing of the channel mRNA and rescues every measured cellular phenotype in vitro and in vivo models. A first clinical trial is now in preparation, developed entirely from human stem‑cell models without a traditional animal disease model.

Parents are flying children to clinics in South America and parts of Europe for intravenous, intrathecal, or even intracranial “stem cell” infusions of unclear provenance—sometimes their own cells, sometimes umbilical cells, sometimes unknown donor cells. For autism, there is no plausible mechanism by which generic stem cells in the bloodstream would correct distributed, circuit‑level developmental abnormalities in the brain, especially since most administered cells are lineage‑restricted and not competent to become neurons. Risks include infection, immune reactions, and tumorigenesis, with essentially no controlled trials showing disease‑modifying benefit.