Radial Glia

Radial Glia is the principal class of bipolar neural stem and progenitor cells that populates the embryonic vertebrate central nervous system during neurulation and corticogenesis. Differentiating directly from pseudostratified neuroepithelial cells after closure of the embryonic neural tube, radial glial cells span the full radial thickness of the developing brain wall, anchoring an apical endfoot at the ventricular lumen and extending an elongated basal fiber to the pial basement membrane. Viewed for more than a century after Camillo Golgi and Santiago Ramón y Cajal as purely structural support fibers guiding immature postmitotic cells, live-imaging and retroviral lineage-tracing experiments led by Stephen C. Noctor and Arnold R. Kriegstein in 2001 proved that radial glia act directly as the primary mitotically active neural stem cells generating the majority of excitatory cortical projection neurons and macroglia.

In gyrencephalic mammals—and most prominently in anthropoid primates and humans—diversification of the radial glial lineage produced a basally positioned progenitor population termed outer radial glia (oRG, also designated basal radial glia, bRG). Residing within an expanded germinal compartment known as the outer subventricular zone (oSVZ), outer radial glia lack apical ventricular junctions yet retain a pial-directed basal process, amplifying neuronal output and fanning out radial migration trajectories that drive human gyrification.

Neuroepithelial Transition and Transcriptional Identity

At the onset of cortical neurogenesis (around embryonic day 10.5 in mice and gestational weeks 5 to 7 in humans), pseudostratified neuroepithelial cells lining the neural tube downregulate tight-junction occludin and acquire astroglial markers—including BLBP (FABP7), GLAST (EAAT1 / SLC1A3), vimentin, and in primates GFAP—transitioning into apical radial glia (aRG, or ventricular radial glia). Apical radial glia express a conserved core transcriptional network anchored by PAX6, SOX2, and canonical Notch effector repressors HES1 and HES5.

At the luminal border of the ventricular zone (VZ), adjacent apical endfeet are coupled by N-cadherin (CDH2) and β-catenin adherens junctions associated with the PAR3–PAR6–aPKC polarity complex and a primary cilium bathed in embryonic cerebrospinal fluid secreted by the choroid plexus. At the opposite pole of the cerebral wall, the slender basal radial process (0.5 to 1.5 μm wide, spanning millimeters in fetal primate cortex) terminates in α₆β₁ integrin and dystroglycan complexes bound to laminin and nidogen within the pial extracellular matrix.

Cell-Cycle-Coupled Interkinetic Nuclear Migration

A defining biomechanical hallmark of ventricular radial glia is interkinetic nuclear migration (IKNM), the oscillatory apicobasal translocation of the soma within its stationary cytoplasm in synchrony with phases of the mitotic cell cycle:

• G1 Ascent and Basal S Phase: Following mitosis at the ventricular lumen, the diploid nucleus ascends toward the basal boundary of the ventricular zone via actomyosin contractility and kinesin-3 (KIF1A) microtubule transport, replicating chromosomal DNA during S phase in the basal VZ so nuclei stagger across depths to maximize progenitor packing density.
• G2 Descent and Apical Mitosis: During G2, recruitment of cytoplasmic dynein–dynactin motors (regulated by LIS1 and NDE1) to nuclear-envelope LINC bridges pulls the nucleus downward along microtubules anchored at the apical centrosome. Because the centrosome remains tethered at the apical belt, mitosis (M phase) occurs at the ventricular surface: symmetric vertical cleavage yields two radial glia, whereas asymmetric division yields one self-renewing radial glial cell plus either a postmitotic neuron or a TBR2⁺ (EOMES⁺) intermediate progenitor cell (IPC) that delaminates into the subventricular zone (SVZ).

Radial Unit Scaffolding and Inside-Out Cortical Lamination

In Pasko Rakic's radial unit hypothesis (1972, 1988), the mosaic of radial glial clones in the ventricular zone maps into radial ontogenetic cortical columns. Newborn excitatory pyramidal neurons attach to the maternal basal fiber via astrotactin-1 and connexin-43/connexin-26 adhesive contacts and migrate radially into the cortical plate. Successive cohorts bypass earlier-born deep-layer neurons (Layer VI TBR1⁺ and Layer V BCL11B⁺ projection neurons) before detaching beneath Cajal-Retzius Reelin-secreting cells in the marginal zone to form superficial intracortical Layers IV, III, and II (SATB2⁺/CUX1⁺). Loss-of-function mutations in PAFAH1B1 (LIS1) or DCX arrest dynein-coupled nucleokinesis along radial fibers, causing human lissencephaly.

The Neurogenic-to-Gliogenic Switch

At the close of neurogenesis (around embryonic day 17.5 in mice and gestational weeks 20 to 26 in humans), promoter demethylation at GFAP and S100B paired with neuron-derived cardiotrophin-1 JAK–STAT3 and Notch/BMP signaling triggers the gliogenic switch. Radial glia retract their pial fibers and differentiate into parenchymal astrocyte populations, oligodendrocyte precursor cells, and multiciliated ventricular ependymal cells, while a quiescent subset persists into adulthood as subventricular Type B1 neural stem cells, retinal Müller glia, and cerebellar Bergmann glia.

Outer Subventricular Zone Expansion and Mitotic Somal Translocation

Whereas lissencephalic mouse neocortex relies mainly on ventricular radial glia and TBR2⁺ IPCs, anthropoid primates evolved a thick outer subventricular zone (oSVZ) that dominates neurogenesis from human gestational week 12 to week 24. Time-lapse fetal human slice imaging by David V. Hansen, Arnold R. Kriegstein, and colleagues (2010) showed that the human oSVZ is populated by outer radial glia (oRG) lacking ventricular contact yet retaining a long pial-directed basal process.

Single-cell transcriptomics by Alex A. Pollen and colleagues (2015) defined the human oRG molecular signature: HOPX, FAM107A, PTPRZ1, TNC (tenascin-C), and LIFR. Through autocrine extracellular-matrix production plus LIFR–STAT3 and mTORC1 signaling, oRGs self-renew outside the ventricular niche and undergo mitotic somal translocation (MST): within the hour preceding cytokinesis, RhoA–ROCK and non-muscle myosin II activation propel the soma 50 to 150 μm basally along its radial fiber. After gestational week 16.5, ventricular radial glia transition into truncated radial glia (CRYAB⁺ tRGs) terminating on oSVZ vessels, so upper-layer neurons ascend along a diverging fan of oRG fibers (discontinuous scaffold) whose tangential cortical-plate expansion drives gyrification.

Modeling Human oRG Biology in Cerebral Organoids

Because mice possess sparse bRG-like cells (<5% of progenitors) and lack an organized oSVZ or hominid gene duplications (ARHGAP11B, NOTCH2NL), rodent cortex cannot model human supragranular folding disorders. Three-dimensional human cerebral organoid cultures generated from induced pluripotent stem cell lines self-organize polarized PAX6⁺ ventricular zones surrounded by HOPX⁺/PTPRZ1⁺ oSVZ-like niches exhibiting authentic mitotic somal translocation, enabling mechanistic dissection of Zika virus radial-glial tropism and TSC1/TSC2 mTORopathies.

Why are radial glia classified as neural stem cells despite expressing astroglial proteins?

Radial glia express vimentin, GLAST (SLC1A3), BLBP, and primate GFAP because embryonic neural progenitors and adult neurogenic niche astrocytes belong to a continuous neuroepithelial-astroglial lineage. Lineage tracing proved that radial glia undergo self-renewing asymmetric divisions generating excitatory cortical neurons first and macroglia later, establishing them as bona fide multipotent neural stem cells rather than differentiated support glia.

How does mitotic somal translocation differ from interkinetic nuclear migration?

Interkinetic nuclear migration (IKNM) occurs in bipolar ventricular radial glia across interphase as the nucleus rises during G1, replicates DNA basally in S phase, and descends via dynein motors in G2 to divide at the ventricular lumen. Mitotic somal translocation (MST) occurs in unipolar outer radial glia right before cytokinesis: lacking an apical anchor, the soma surges 50 to 150 μm toward the cortical plate via RhoA–ROCK–myosin II contractility, stepping the germinal zone outward.

Why do human cerebral organoids model outer radial glia better than laboratory mice?

Mouse cortex is smooth (lissencephalic) and contains very few HOPX⁺/FAM107A⁺ outer radial glia. Because prolonged cell-cycle kinetics, elevated mTORC1 activity, autocrine tenascin-C/PTPRZ1 signaling, and hominid-specific genes (NOTCH2NL, ARHGAP11B) are encoded within the human genome, human stem-cell-derived cerebral organoids intrinsically generate abundant outer radial glia that perform mitotic somal translocation in vitro.

1. Noctor, S. C., Flint, A. C., Weissman, T. A., Dammerman, R. S., & Kriegstein, A. R. (2001). Neurons derived from radial glial cells establish radial units in neocortex. Nature, 409(6821), 714–720. https://doi.org/10.1038/35055553
2. Hansen, D. V., Lui, J. H., Parker, P. R. L., & Kriegstein, A. R. (2010). Neurogenic radial glia in the outer subventricular zone of human neocortex. Nature, 464(7288), 554–561. https://doi.org/10.1038/nature08845
3. Pollen, A. A., Nowakowski, T. J., Chen, J., Retallack, H., Sandoval-Espinosa, C., Nicholas, C. R., ... & Kriegstein, A. R. (2015). Molecular identity of human outer radial glia during cortical development. Cell, 163(1), 55–67. https://doi.org/10.1016/j.cell.2015.09.004
4. Rakic, P. (1972). Mode of cell migration to the superficial layers of fetal monkey neocortex. Journal of Comparative Neurology, 145(1), 61–83. https://doi.org/10.1002/cne.901450105
5. Kriegstein, A., & Alvarez-Buylla, A. (2009). The glial nature of embryonic and adult neural stem cells. Annual Review of Neuroscience, 32, 149–184. https://doi.org/10.1146/annurev.neuro.051508.135600