UCLA Maps Two Hidden Signals That Shape the Developing Human Cortex
Before birth, the human cortex is built by a narrow population of stem cells that must choose, again and again, what to make next. The new UCLA work places radial glia at the center of that decision-making process and shows that their fate is influenced by two distinct forces: how they handle glucose and whether they physically touch incoming thalamic projections. Those findings come from paired studies published in Cell and Science, and they recast early cortical development as a process governed by both metabolism and direct cell-to-cell contact.
Radial glia are not passive scaffolds. They generate much of the cerebral cortex, help drive the expansion that makes the human cortex unusually large, and later disappear before birth. Related cells can reappear in brain tumors, which makes their developmental logic clinically relevant well beyond embryology. Aparna Bhaduri, the UCLA investigator leading the work, framed the stakes plainly: understanding how these cells decide what to become offers a route into neurodevelopmental and neuropsychiatric disorders, as well as cancer biology.
Glucose Handling Rewrites the Fate of Cortical Stem Cells
The Cell study built a metabolic atlas of the early human cortex by combining donated human tissue with brain organoids grown from stem cells. Instead of treating metabolism as background housekeeping, the analysis showed that it actively shapes lineage decisions. Radial glia depended heavily on the pentose phosphate pathway, a glucose-processing route that supplies materials needed by rapidly dividing cells.
When investigators reduced glucose availability or disrupted that pathway, the stem cells shifted their output. They began generating more inhibitory neurons and other cell types that normally appear later in development. One sentence captures the biological implication starkly: changing fuel supply changed cell identity.
The mechanism makes developmental sense. Rapidly dividing progenitor cells need more than energy; they need nucleotide precursors, redox balance, and biosynthetic capacity to keep pace with cortical growth. The pentose phosphate pathway sits at exactly that intersection. In practical terms, the study suggests that maternal nutrition, metabolic disease, or other environmental influences could alter the developing cortex by nudging progenitor metabolism away from its normal trajectory.
The work also expands a basic research toolset. A detailed atlas of early cortical metabolism gives investigators a reference map for future studies of human brain development, especially in situations where animal models may not capture the same cellular sequence or metabolic dependence.
Thalamic Fibers Reach Cortical Progenitors Earlier Than Expected
The second study, published in Science, followed a very different signal. This time the focus was the thalamus, a deep brain structure whose neurons send long projections toward the cortex. For years, those projections were known to reach the cortical region early in development, before mature synaptic connections were fully established. The unanswered anatomical question was simple: why arrive so soon?
Using human stem cell-derived brain assembloids, the UCLA team found that the thalamic projections do more than pass nearby. They physically touch radial glia while the cortex is still taking shape. That contact altered the behavior of the progenitor cells and pushed them toward producing more excitatory neurons, the principal signal-carrying neurons of the cortex. The effect was strongest for upper-layer neurons, a population especially expanded in humans.
This is a direct developmental instruction, not a vague influence. Physical contact from thalamic projections appears to tell cortical progenitors what kind of neurons to make.
The finding also helps explain a longstanding species difference. The study suggests that this specific point of contact may not exist in rodents, which means some elements of human cortical specification may rely on interactions that are difficult to reproduce in standard animal models. For researchers trying to understand human brain development, that distinction matters immediately because it defines where animal data can guide interpretation and where they may miss the mark.
NRXN1 Links a Known Autism Gene to Early Cortical Signaling
The thalamic study also tied the developmental interaction to NRXN1, a gene already linked to synaptic organization and associated with autism spectrum disorder when mutated. Patient-derived assembloids carrying an NRXN1 mutation showed altered thalamic signaling behavior compared with unaffected cells. Those changes shifted the balance between stem cells and the neurons they produced.
The result does not establish a clinical pathway by itself, but it does create a testable model for how a mutation tied to autism could influence cortical formation much earlier than the stage at which many clinical symptoms become obvious. Developmental disorders often begin with subtle disruptions in cell fate, timing, or connectivity long before postnatal circuitry is mature. Here, the disruption appears tied to how incoming thalamic signals are received by cortical progenitors.
That model is particularly valuable because it uses human cells. Human assembloids can preserve features of species-specific development that are difficult to study in rodents, especially when the question involves upper-layer cortical neurons and early thalamocortical interactions.
Metabolism and Physical Contact Act as Parallel Developmental Inputs
The two studies answer different questions, yet they converge on the same core idea: radial glia integrate information rather than simply following a preset program. One axis is metabolic. The other is structural and signaling-based. Together, they indicate that the earliest stages of cortical development depend on a constant exchange between the internal state of the stem cell and the outside world around it.
For clinicians and translational researchers, the relevance lies in what this opens up, not in any immediate therapeutic claim. Metabolic conditions during pregnancy, inherited mutations that alter synaptic genes, and disturbances in the timing of long-range brain projections could all influence the balance of neurons produced during development. Those pathways may help explain why the cortex is vulnerable to disorders that arise from early wiring defects, and why the same stem-cell programs can later reappear in brain cancer.
The studies are also a reminder of how much human organoid technology has changed developmental neuroscience. A decade ago, these questions were far harder to ask directly in human tissue. Now they can be approached with metabolic atlases, fused organoids, and patient-derived models that make cellular decisions visible in the lab.
What remains unresolved is the extent to which these mechanisms operate in vivo with the same strength and timing seen in organoids, and how far they extend across diverse human developmental contexts. The work provides a detailed starting point, but the next step will be defining how these nutrient and contact-dependent signals interact across gestation, disease risk, and later neural circuit formation.