Curing Autism, Epilepsy & Schizophrenia with Stem Cells | Dr. Sergiu Pașca

KEYPOINTS

• 01

  Autism is a behaviorally defined spectrum, not one disease

  Pașca notes that autism has no biomarker and is diagnosed by the presence or absence of behaviors by a certain age. Prevalence is now close to 3%, whereas it was considered rare when he was in medical school. He compares today's 'autism' to 19th-century 'fever,' which could be an infection, cancer or an autoimmune disease needing very different treatments. His lab focuses on profound autism, the severe forms that often come with intellectual disability or epilepsy and may require lifelong care.

• 02

  Genetics are strong; the 'refrigerator mother' idea was wrong

  Psychoanalysts in the 1950s–60s blamed emotionally cold mothers. Twin studies in the 1970s showed that if one identical twin has autism, the other very likely does too, and heritability is among the highest of psychiatric conditions. Only in the last 10–15 years have specific genes been tied to profound forms. Prevalence in Scandinavia is around 1 in 30 to 1 in 40, similar to the US, which Pașca says weakens US-specific explanations like glyphosate.

• 03

  The male bias is about 4:1 and may reflect developmental resilience

  The ratio varies with severity but is generally around 1 female to 4 males. Pașca points out that being female is one of the best predictors of good outcomes in premature babies, and that male and female nervous systems mature at different paces and may handle injury differently. With autism being many conditions, he cautions against pinning the bias on any one factor.

• 04

  Vaccines and the 'fever effect'

  Huberman states for the record that he sees no solid evidence that vaccines cause autism, and Pașca agrees there is none epidemiologically. Parent reports that non-verbal children briefly speak in sentences during high fevers are 'mostly anecdotal.' Proposed explanations include noradrenergic activation, cytokines reaching the brain and temperature-sensitive ion channels.

• 05

  Gene therapy options and their limits for the brain

  Approaches include adding a missing gene via engineered, non-disease-causing adeno-associated viruses, delivering a missing enzyme directly, or editing DNA with CRISPR (already used for sickle cell disease in blood). The brain is hard to reach: delivery is by surgery or intrathecal injection, viral payloads and efficiency are limited, large genes such as calcium channels don't fit in a virus, and immune responses often allow only one shot. Pașca's lab instead targets the RNA step downstream.

• 06

  Yamanaka factors turn skin cells into stem cells that can become brain tissue

  About 20 years ago, Shinya Yamanaka showed that four factors can revert a specialized cell, such as a skin fibroblast, into an induced pluripotent stem (iPS) cell. This sidestepped embryonic-cell ethics, and Pașca likens it to alchemy. Without signals pushing them elsewhere, these cells become neural cells 'almost by default' within days. That lets researchers grow neurons from any patient, including those with rare mutations.

• 07

  Organoids keep an intrinsic developmental clock

  Floating 3D balls of cells, inspired by Yoshiki Sasai's work, survived far longer than expected, and the lab kept cultures for 500–800+ days. Without hormone changes or any change in the culture medium, NMDA-receptor subunits (2B down, 2A up) switched to a postnatal signature around nine months, matching birth. Pașca wants to learn whether this timer could be sped up to model aged neurons, for example for Parkinson's, or reversed to rejuvenate them, but says they only have clues so far.

• 08

  Assembloids rebuild circuits by combining parts in the right order

  Combining a cortical organoid, a spinal cord organoid and a ball of human muscle yields a circuit in which stimulating the cortex makes the muscle contract within weeks. A four-part sensory assembloid (pain-sensing neurons, spinal cord, thalamus, cortex) only self-assembles in that order. After hundreds of days, the whole pathway flickers in sync. It's being used to study sodium-channel mutations that cause either excessive pain or a dangerous inability to feel pain.

• 09

  Transplanted human organoids integrate into rat brains

  Cortical organoids placed into the somatosensory cortex of newborn rats became vascularized, were colonized by rat microglia, and grew to occupy about a third of one hemisphere. They responded when the rat's whiskers on the opposite side were moved. Human neurons grew six to eight times larger than in a dish, making disease features visible: Timothy syndrome neurons are about half normal size. Integration works because the young rat brain is still 'eager to connect.'