Oxygen Shapes the Developing Cortex, but HIF-2α Turns Out Not to Be the Messenger

Oxygen is usually cast in the brain as a fuel, the invisible gas that keeps neurons firing and mitochondria humming. But over the past two decades, developmental neuroscientists have come to appreciate a second, subtler role: oxygen acts as a signaling molecule, one whose concentration fluctuates naturally as an embryo grows and whose levels can steer the behavior of the stem cells that build the cerebral cortex. The developing mammalian brain is anything but uniformly oxygenated. Early in gestation, the embryonic brain exists in a state of profound hypoxia, and as the placental circulation matures and vascular networks penetrate the neural tissue, oxygen tensions rise and shift in dynamic patterns. These changing oxygen landscapes are now understood to be part of the developmental program itself, not merely a background condition that the growing brain must tolerate.

The best-characterized translator of oxygen signals into genetic instructions is hypoxia-inducible factor 1α, or HIF-1α, a transcription factor that accumulates when oxygen is scarce and activates a broad program of genes involved in angiogenesis, glycolytic metabolism, and cell survival. HIF-1α has been implicated in nearly every stage of cortical development, from the maintenance of neural stem cells in the ventricular zone to the timing of neuronal differentiation and the migration of newborn neurons into their final laminar positions. Yet HIF-1α has a close sibling, HIF-2α, encoded by the EPAS1 gene, which binds a nearly identical DNA motif and partners with the same HIF-1β subunit. Because the two isoforms overlap so heavily in their downstream targets, a persistent question has hung over the field: if oxygen shapes cortical development through HIF signaling, which of the two isoforms is actually doing the work?

A new study published in PLOS One by Jennifer Lanto, Laura Rohmann, Tim Theege, Valeriia Abramenko, Lisa Große, Andreas Madlener, Julia Anna Elisabeth Neitzel, Alexander Storch, and Franz Markert set out to answer that question directly, and the answer it delivers is a decisive negative for HIF-2α. Using conditional knockout mice in which the HIF-2α gene was deleted specifically in neural tissue, the team exposed pregnant dams and their developing fetuses to three distinct oxygen environments and then examined the resulting cortical architecture in detail. Their findings, while confirming that oxygen powerfully modulates cortical development, rule out HIF-2α as the mediator of those effects, redirecting attention toward other oxygen-dependent mechanisms that remain to be identified.

The experimental design was deliberately ambitious in its coverage of the physiological oxygen range. Starting at embryonic day 14.5, a stage at which the cerebral cortex is in the midst of its most intense neurogenic activity, the researchers maintained the animals for 48 hours under one of three conditions: maternal hypoxia at 10 percent oxygen, physioxia at ambient 21 percent oxygen, or hyperoxia at 75 percent oxygen. This window matters. Between E14.5 and E16.5, the cortex is generating deep-layer projection neurons, including the layer IV and layer V populations that will form the output highways of the neocortex, while the upper-layer neurons that define the more recently evolved superficial strata are just beginning to appear. Perturbing oxygen across this interval is therefore a sensitive probe of the mechanisms that govern cortical lamination and neurogenesis.

When the researchers analyzed the E16.5 fetuses by immunohistochemistry, the oxygen manipulations produced clear and interpretable effects. Markers of newborn neurons, detected through BrdU incorporation, and of layer-specific neuronal populations, identified through the transcription factors Tbr1 for layer IV, Ctip2 for layer V, and Satb2 for upper-layer neurons, all shifted in response to altered oxygen. The most striking observation concerned hyperoxia, which particularly increased the number of Ctip2-positive, Tbr1-negative layer V neurons in the E16.5 fetuses. In other words, breathing air enriched with oxygen pushed the developing cortex toward an expanded production of a specific deep-layer neuronal subtype, a measurable remodeling of the cortical output architecture driven purely by the ambient gas mixture.

Against this backdrop of oxygen sensitivity, the HIF-2α knockout results came as a genuine surprise. Neither the number of newborn neurons nor the number of layer-specific neurons was affected by the loss of HIF-2α, regardless of whether the animals carried the deletion under the Emx1 promoter, which targets the cortical primordium, or the Nestin promoter, which targets broader neural stem and progenitor populations. The oxygen-related effects on neural stem cells and neuroprogenitor cells observed in the E16.5 fetuses likewise persisted without any detectable contribution from HIF-2α. Whatever oxygen was doing to the developing cortex, it was not doing it through this transcription factor, despite the molecule’s established role as a bona fide oxygen sensor in other tissues and developmental contexts.

To probe the question from a second angle, the team turned to an in vitro system in which neural stem cells could be cultured under controlled oxygen tensions. Here again the oxygen effect was unmistakable: the expansion and stem cell maintenance of neural stem cells were accelerated under physioxia at 3 percent oxygen compared with hyperoxia at 21 percent oxygen. This result is consistent with a growing body of literature suggesting that the low oxygen tensions characteristic of the embryonic niche are not a stress to be overcome but an actively favorable environment for stem cell self-renewal. Yet, mirroring the in vivo findings, HIF-2α was not involved under either condition. The stem cell response to oxygen tension operated independently of the isoform, reinforcing the conclusion drawn from the intact animals.

The implications of this double negative are worth unpacking. On one hand, the study strengthens the case that oxygen is a genuine developmental signal in the cortex, capable of altering the balance of neuronal subtypes produced during neurogenesis and modulating the behavior of the stem cells that supply them. On the other hand, it removes HIF-2α from the list of candidate mediators, at least for the developmental window and cell populations examined. The field is left with a sharpened question: if not HIF-2α, then what? The obvious remaining candidate is HIF-1α, whose involvement in cortical development is already well documented, but the authors’ findings also leave open the possibility of HIF-independent oxygen sensing pathways, including those involving prolyl hydroxylase domain enzymes, oxygen-sensitive ion channels, reactive oxygen species signaling, or metabolic reprogramming that does not route through canonical HIF transcription factors at all.

There are also important caveats and boundaries to the conclusion. The study focused on a specific developmental window, E14.5 to E16.5 in the mouse, and on the cortical neuroepithelium targeted by the Emx1 and Nestin promoters. It remains possible that HIF-2α plays a role at earlier embryonic stages, in other brain regions, or in cell types outside the targeted lineages. Conditional knockout strategies, for all their power, also leave open the possibility of compensatory upregulation of HIF-1α in the absence of HIF-2α, a scenario that could mask a genuine requirement for the deleted isoform. These are the standard qualifications of developmental genetics, and they do not diminish the central result: within the tested parameters, the oxygen-dependent modulation of cortical neurogenesis proceeds normally without HIF-2α.

For researchers working on cortical development, the study is a reminder that the oxygen signaling landscape of the embryonic brain is more complex than the HIF-centric model suggests, and that negative results can be as informative as positive ones. For the broader audience fascinated by how the brain takes shape, the work offers a vivid illustration of a principle that developmental biology keeps reasserting: the environment in which an organ grows is part of its blueprint. The air a pregnant mouse breathes, at concentrations spanning from 10 percent to 75 percent oxygen, can measurably alter the composition of her offspring’s cortex within 48 hours. Decoding exactly how that environmental information reaches the genome, now with HIF-2α crossed off the list of suspects, is the task the field must take up next.

Subject of Research: The role of hypoxia-inducible factor 2α in oxygen-dependent regulation of cortical brain development in mice

Article Title: Hypoxia-inducible factor 2α does not regulate cortical brain development

Article References: Lanto, J., Rohmann, L., Theege, T., Abramenko, V., Große, L., Madlener, A., Neitzel, J. A. E., Storch, A., & Markert, F. (2026). Hypoxia-inducible factor 2α does not regulate cortical brain development. PLOS One, 21(10), e0359507. https://doi.org/10.1371/journal.pone.0359507

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0359507

Keywords: HIF-2α, cortical development, oxygen signaling, neural stem cells, neurogenesis, hypoxia, hyperoxia, physioxia, conditional knockout mice, cerebral cortex, developmental neuroscience, PLOS One