A transcriptional blueprint of neurogenesis from human stem cells

Stem cell-derived neurons serve as attractive human neuronal model systems in health and disease. In order to understand in vitro neurogenesis, transcriptional changes over the time course of differentiation were analyzed and a network of key transcription factors that promoted the loss of pluripotency and rapid neurogenesis via progenitor states was revealed.

HFSP Long-Term Fellow Volker Busskamp and colleagues
authored on Mon, 17 November 2014

Until now it has been challenging to employ systems biology methodologies to study stem cells as they differentiate because they often require long and delicate protocols to attain differentiated cells. Furthermore, differentiated cells usually are obtained with low yields and found in heterogeneous colonies with many undifferentiated cells. Direct induction of two transcription factors of the Neurogenin family in human induced pluripotent stem cells overcame these challenges. This improved protocol resulted in a rapid and highly efficient neuronal differentiation. The induced neurons (iNGN cells) had a bipolar morphology in four days at greater than 90% purity. The speed, high yield and homogeneity of differentiation provided the opportunity to subject the cells to transcriptomic analysis to identify many key regulatory factors that underlie this differentiation process.

Figure: Rapid neuronal differentiation from stem cells is achieved by Neurogenin overexpression. The iNGN cells express neuronal markers such as MAP2 (cyan) and TUBB3 (magenta) after four days of induction. Nuclei are stained by DAPI (yellow).

This work highlights that by directly controlling transcription factors, neuronal differentiation from stem cells can be extremely efficient without any additional bioactive factors in the culturing media. Neuronal maturation and electrical activity needed additional extrinsic factors despite expression of the synaptic machinery within four days in stem cell media. The transcriptomic changes over the time course of differentiation were measured by RNA sequencing and microRNA profiling and revealed that the iNGN cells differentiated continuously via unstable progenitor states.

Transcriptomic comparison to human samples from the BrainSpan atlas assigned iNGN cells higher correlations to prenatal human tissues whereas the spatial mapping suggested that iNGN cells do not resemble cortical neurons.

MicroRNA levels also dynamically changed from stem cell profiles towards neuronal ones but microRNA regulation took place mostly downstream of the neuronal differentiation initiation phase.

The systems biological approach revealed a robust gene regulatory network driving stem cells to neurons. This differentiation program was validated by individual and combinatorial perturbations of key transcription factors.  Perturbations to this network resulted in variations in neuronal yield, morphology, and gene expression. Notably, most transcription factors that are frequently used to induce neurogenesis in stem cells were also activated by the Neurogenins. This could explain why most protocols to date result in similar neuronal cell types. The obtained coding and non-coding transcriptomic blueprint of a neuronal differentiation program primes further targeted manipulations of iNGN cells and/or the usage of different transcription factors to increase the variety of human neurons.

This work was carried out in collaboration with George Church’s laboratory at the Harvard Medical School (Boston, USA) and Ron Weiss’ laboratory at the MIT (Cambridge, USA).


Rapid neurogenesis through transcriptional activation in human stem cells. Volker Busskamp*, Nathan E Lewis*, Patrick Guye*, Alex HM Ng, Seth L Shipman, Susan M Byrne, Neville E Sanjana, Jernej Murn, Yinqing Li, Shangzhong Li, Michael Stadler, Ron Weiss & George M Church. Mol Syst Biol. (2014) 10: 760. DOI 10.15252/msb.20145508  (*, equal contribution)

Link to article