Research Article

A latent lineage potential in resident neural stem cells enables spinal cord repair

See allHide authors and affiliations

Science  02 Oct 2020:
Vol. 370, Issue 6512, eabb8795
DOI: 10.1126/science.abb8795

You are currently viewing the abstract.

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution

Spinal cord stem cells

Injuries to the mammalian spinal cord do not heal easily. Llorens- Bobadilla et al. studied mouse ependymal cells, which function as stem cells for the spinal cord (see the Perspective by Becker and Becker). Chromatin accessibility and transcriptomic assays revealed that these cells carry a latent ability to differentiate into oligodendrocytes, which is much needed for remyelination of axons around an injury. The ependymal cells were triggered to differentiate into oligodendrocytes by expression of the oligodendrocyte lineage transcription factor OLIG2. Expression of OLIG2 in ependymal cells specifically and inducibly enabled the production of oligodendrocytes. The ependymal-derived oligodendrocytes aided axon remyelination and improved axon conduction after spinal cord injury.

Science, this issue p. eabb8795; see also p. 36

Structured Abstract


The capacity of a tissue to regenerate itself rests on the potential of its resident cells to replace cells lost to injury. Some tissues, such as skin or intestine, do this remarkably well through the activation of tissue-specific stem cells. Injuries to the central nervous system (CNS), in contrast, often lead to permanent functional impairment; some cells lost to injury are never replaced. Neural stem cells have been identified in the adult brain and spinal cord and are activated by injury. However, injury-activated neural stem cells predominantly produce scar-forming astrocytes, and the contribution of neural stem cells to cell replacement is insufficient for regeneration. To design regenerative strategies aimed at recruiting resident neural stem cells for repair, it is essential to know whether greater regenerative potential exists and how to elicit such potential.


The spinal cord is a great system to study neural stem cell recruitment for repair. The neural stem cell potential of the spinal cord resides in a well-characterized population of ependymal cells. Ependymal cells, normally quiescent, are activated by injury to generate almost exclusively scar-forming astrocytes. Ependymal-derived astrocytes help to preserve tissue integrity, but other cell types, such as myelin-forming oligodendrocytes, are insufficiently replaced. In parallel, neural stem cell transplantation has proven to be beneficial to recovery after spinal cord injury—a benefit that is associated with the increased supply of oligodendrocytes able to remyelinate demyelinated axons. Ependymal cells share a developmental origin with spinal oligodendrocytes, which led us to explore whether a latent potential for expanded oligodendrocyte generation might exist.


We integrated single-cell RNA sequencing (scRNA-seq) and single-cell assay for transposase-accessible chromatin using sequencing (scATAC-seq) to study lineage potential in adult ependymal cells of the mouse spinal cord. We found that the genetic program for oligodendrocyte generation is accessible in ependymal cells. However, this program is latent, as oligodendrocyte genes are not expressed. In particular, we found that a large fraction of binding sites for OLIG2, the transcription factor that initiates developmental oligodendrogenesis, had basal accessibility, despite OLIG2 and its key target genes not being expressed in adult ependymal cells. To study whether this latent accessibility was associated with a greater capacity to produce oligodendrocytes, we genetically engineered a mouse model to express OLIG2 in adult ependymal cells. We found that OLIG2 expression was compatible with ependymal identity during homeostasis. However, after injury, OLIG2 expression led to the increased accessibility of the latent program and subsequent expression of genes specifying oligodendrocyte identity. Unfolding of the latent program was followed by efficient oligodendrocyte production from ependymal cells, but not from astrocytes, after injury. Using scRNA-seq of ependymal-derived cells, we found that new oligodendrocytes followed the developmental program of oligodendrocyte maturation, including a self-amplifying oligodendrocyte progenitor cell–like state. These cells later matured to acquire the identity of resident mature myelinating oligodendrocytes. Further, ependymal oligodendrocyte generation occurred in parallel and not at the expense of astrocyte scarring. Newly recruited ependymal-derived oligodendrocytes migrated to sites of demyelination, where they remyelinated axons over the long term. Finally, using optogenetics, we found that ependymal-derived oligodendrocytes contributed to normalizing axon conduction after injury.


Adult neural stem cells have a greater potential for regeneration than is normally manifested. Targeted activation of such potential leads to the recruitment of neural stem cells for the generation of remyelinating oligodendrocytes in numbers comparable to those obtained via cell transplantation. Resident stem cells can thus serve as a reservoir for cellular replacement and may offer an alternative to cell transplantation after CNS injury.

Latent potential in neural stem cells.

Through the integration of different layers of genomic information in single cells, we found that the genetic program for oligodendrocyte generation is latently accessible in ependymal neural stem cells of the adult spinal cord. After injury, activating the latent potential by forced OLIG2 expression unfolds efficient oligodendrocyte generation, leading to enhanced repair.


Injuries to the central nervous system (CNS) are inefficiently repaired. Resident neural stem cells manifest a limited contribution to cell replacement. We have uncovered a latent potential in neural stem cells to replace large numbers of lost oligodendrocytes in the injured mouse spinal cord. Integrating multimodal single-cell analysis, we found that neural stem cells are in a permissive chromatin state that enables the unfolding of a normally latent gene expression program for oligodendrogenesis after injury. Ectopic expression of the transcription factor OLIG2 unveiled abundant stem cell–derived oligodendrogenesis, which followed the natural progression of oligodendrocyte differentiation, contributed to axon remyelination, and stimulated functional recovery of axon conduction. Recruitment of resident stem cells may thus serve as an alternative to cell transplantation after CNS injury.

View Full Text

Stay Connected to Science