Dissecting Gene Regulatory Networks Governing Human Cortical Cell Fate

Jingwen W. Ding, Chang N. Kim, Megan S. Ostrowski, Yashodara Abeykoon, Bryan J. Pavlovic, Jenelle L. Wallace, Tomasz J. Nowakowski, Alex A. Pollen

Posted on: 14 November 2025

Preprint posted on 24 September 2025

A deep dive into how transcription factors shape radial glia fate and clonal dynamics

Selected by Manuel Lessi

Categories: evolutionary biology, genetics, neuroscience

Introduction

The development of the human cortex is a complex process that requires dynamic activation of gene regulatory networks within defined time windows. Radial glia are the developmental progenitors that give rise to the main cortical cell types, from excitatory and inhibitory neurons to glia such as astrocytes and oligodendrocytes. Thus, these neural progenitors are a key hub in which gene regulatory networks mastered by Transcription Factors (TFs) shape cell identity and differentiation trajectories. Pooled CRISPR screening is a powerful tool to interrogate such dynamics by perturbing the expression of desired genes and profiling the consequent developmental shifts at single cell resolution by transcriptome sequencing (a similar approach was described in another preLight post [link]). This technology is mainly used in human pluripotent stem cells because of their flexibility and their ability to generate most neural cell types with standardized differentiation protocols. However, stem cells fall short of fully recapitulating in vivo processes compared with primary cultures directly derived from human samples.

Experimental design

With a clever experimental design, the authors of this preprint studied the impact of inhibiting the expression of TFs that are key for neurodevelopmental processes by combining ex vivo primary cultures of radial glia isolated from prenatal human cortical samples with pooled genetic perturbation.

The protocol was straightforward: dissociate primary samples, plate and expand the cells for a few days, then deliver a fluorescent GFP construct that contains both dead Cas9 and sgRNAs to perform CRISPRi. After 7 days of differentiation, GFP positive cells were enriched by FACS and subjected to sc-RNAseq. This approach allowed the preprint authors to match transcriptomic profiles of differentiating cells with sgRNA identity, thereby understanding how perturbing specific TFs altered cellular dynamics and differentiation.

In a second round of experiments, a virus containing a lineage tracing barcode was co-transduced into the culture, allowing the authors to directly link TFs perturbations to the acquired fate of individual radial glia clones.

Figure 1: Preprint figure 1A describing the experimental design carried out for the experiments described in the preprint. Briefly, primary cortex samples were dissociated, and the radial glia population was enriched through 5 days of culturing supplemented with FGF2 and EGF. Lentiviral pool containing GFP – CRISPRi guide library and dead Cas9 were then added to the dish. After 7 days, cells were sorted to retain a GFP+ population and left to differentiate for another week in presence of neurotrophic factors. Finally, cells were sorted to enrich for GFP+ again and subjected to scRNA-seq. DAPI marks the nuclei, EOMES marks intermediate progenitors and NEUROD2 marks excitatory neurons. Preprint figure 1A was made available under a CC-BY-NC-ND 4.0 International license.

Main results

The preprint authors, Ding and colleagues, started this journey by confirming that radial glia were actually able to give rise to the expected cell types. Moreover, the established protocol was found to be well suited to study TFs inhibition, as the CRISPR construct did not induce detectable stress or toxicity. Adding the lineage profiling layer allowed the authors to proof that single cortical radial glia clones can generate both excitatory and inhibitory neurons, overall providing a high-resolution view of how TFs influence the balance between different lineages that arise from a single progenitor.
A new uncharacterized function for ZNF219 was uncovered, as this TF was found to act as a repressor of neural differentiation.
ARX and NR2E1 regulate radial glia fate transitions and clonal dynamics, with ARX speeding up and NR2E1 slowing down developmental tempo, resulting in enrichment of different neuronal subtypes upon perturbation. Additional validation screening on rhesus macaque identified conserved and reproducible impact of ARX and NR2E1 on cortical interneuron neurogenesis.
The postconceptional week of the primary sample from which radial glia were isolated affected perturbation outcomes, highlighting that TFs act in defined temporal windows and their perturbation results in stage-specific phenotypes.

Conclusion: Why I highlight this preprint

I liked this preprint because it manages to combine the strengths of primary cultures and pooled CRISPRi perturbation, resulting in a useful platform to investigate gene perturbation effects on both lineage progression and identity acquisition. Key results were validated in organotypic 3D slice cultures and in another primate species, further enhancing the robustness of the results and demonstrating conserved gene effects across species.

Questions and future directions

Have you considered performing the analysis at later stages to understand effects on more mature neurons and glia? Are there any technical difficulties in maintaining primary cultures that would hinder you from going in this direction?
You highlighted convergence in genes that are differentially expressed upon perturbations of TFs that are linked to neurodevelopmental disorders. Have you considered examining the expression of such TFs or their binding sites accessibility levels in patient datasets, such as PsychENCODE?
What follow-up experiments would you recommend to continue this story, for example functional assays, or in vivo validation?
How much variability have you observed in the perturbation response across donors? Is this system overall more robust compared to iPSC also from this standpoint?

Tags: crispr, crispri, neurodevelopment, radial glia

doi: https://doi.org/10.1242/prelights.42125

Read preprint

(No Ratings Yet)

Author's response

Jingwen W. Ding, Alex A. Pollen shared

We designed the study to primarily investigate the effects TF perturbations on progenitor differentiation and fate transition, but we agree that assessing long-term outcomes by extending the in vitro culture could provide insights into the neuronal maturation gradient. In preliminary studies, we have cultured these dissociated cells for an additional 5-10 weeks and observed morphological and physiological maturation of pyramidal neurons and astrocytes.
We initially generated the candidate list of TFs from public datasets of single cell multiome data in neurotypical individuals. Many prioritized TFs in this study have been linked to neurodevelopmental or neuropsychiatric disorders. Therefore, we expect misexpression of many of these TFs may drive disease, and examining changes in the accessibility of their binding sites and the expression of their downstream target genes in patients would be a worthwhile next step. In the case of ARX, several of the aberrant genes most strongly induced in inhibitory neurons correspond to those described in mouse models of ARX-linked disorders.
As a specific example, in Figure 5, we observed that repression of ARX in inhibitory neurons leads to an “ectopic” cluster of inhibitory neurons that we could not find in atlases of normal brain development. We performed a pilot rescue experiment to study TF epistasis, finding that repression of aberrantly expressed LMO1 can partially rescue this ectopic state. We would be interested in extending this experiment to a broader genetic modifier screen to examine how different ARX downstream targets influence the ectopic cell state induced by ARX repression. Ultimately, we would be interested in generalizing from such examples to inform perturbations that could help to course correct developmental deficits in post-mitotic neurons.
Both iPSC models and primary cultures have strengths and limitations. In the primary cell model, we observed consistent responses to TF perturbations across cell lines and species. Extended Data Fig. 1 of the study describes increased fidelity to in vivo cortical development, scalability advantages, improvements in cell culture health, and increased reproducibility and temporal precision of cortical patterning in primary cultures that supported sensitive detection of cell fate consequences and comparison at different RG maturation stages.

Have your say Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sign up to customise the site to your preferences and to receive alerts

Also in the evolutionary biology category:

A high-coverage genome from a 200,000-year-old Denisovan

Stéphane Peyrégne, Diyendo Massilani, Yaniv Swiel, et al.

AND

A global map for introgressed structural variation and selection in humans

PingHsun Hsieh, Natthapon Soisangwan, David S. Gordon, et al.

Selected by 02 December 2025

Siddharth Singh

Discussion

Beyond venomous fangs: Uloboridae spiders have lost their venom but not their toxicity

Xiaojing Peng, Ludwig Dersch, Josephine Dresler, et al.

Selected by 13 November 2025

Daniel Fernando Reyes Enríquez, Marcus Oliveira

Getting a head start: Craniofacial heterochrony in marsupials involves dynamic changes to molecular and cellular mechanisms underlying neural crest development

Axel H Newton, Ella R Farley, Andrew T Major, et al.

Selected by 12 November 2025

Heather Pollington

Discussion

Also in the genetics category:

A high-coverage genome from a 200,000-year-old Denisovan

Stéphane Peyrégne, Diyendo Massilani, Yaniv Swiel, et al.

AND

A global map for introgressed structural variation and selection in humans

PingHsun Hsieh, Natthapon Soisangwan, David S. Gordon, et al.

Selected by 02 December 2025

Siddharth Singh

Discussion

VANGL2 shapes the mouse heart tube from adjacent epithelia and without planar polarity

Paul Palmquist-Gomes, Gaëlle Letort, Ayushi U. Hegde, et al.

Selected by 19 November 2025

Anubhav Prakash