Identifying a population of glial progenitors that have been mistaken for neurons in embryonic mouse cortical culture

Abstract

Experiments in primary culture have helped advance our understanding of the curious phenomenon of cell cycle-related neuronal death. In a differentiated postmitotic cell such as a neuron, aberrant cell cycle re-entry is strongly associated with apoptosis. Indeed, in many pathological conditions, neuronal populations at risk for death are marked by cells engaged in a cell cycle like process. The evidence for this conclusion is typically based on finding MAP2-positive cells that are also positive for cell cycle-related proteins (e.g., cyclin D) or have incorporated thymidine analogs such as BrdU or EdU into their nuclei. We now report that we and others may have partly been led astray in pursuing this line of work. Morphometric analysis of mouse embryonic cortical cultures reveals that the size of the 'cycling' MAP2-positive cells is significantly smaller than those of normal neurons, and their expression of MAP2 is significantly lower. This led us to ask whether, rather than representing fully developed neurons, they more closely resembled precursor-like cells. In support of this idea, we find that these small MAP2-positive cells are immunopositive for nestin, a neuronal precursor marker, Olig2, an oligodendrocyte lineage marker, and NG2, an oligodendrocyte precursor marker. Tracking their behavior in culture, we find that they predominantly give rise to GFAP+ astrocytes instead of neurons or oligodendrocytes. These findings argue for a critical reexamination of previous reports of stimuli that lead to neuronal cell cycle-related death in primary cultures.

Significance While many laboratories use cultures of rodent embryonic cortex to study the cell biology of brain function, few attend to the identity of the cells in their cultures. We find that a subpopulation of small MAP2-weakly-positive cells, presumed for decades to be neurons, are actually NG2⁺Olig2⁺Nestin⁺ precursor cells that give rise over time mostly to astrocytes. Subjecting the cultures to a challenge meant to mimic an in vivo cell cycle-related neuronal death, we find the percentage of 'cycling' MAP2⁺ cells does indeed increase, but only because large MAP2⁺ cells (true neurons) die with no change in their mitotic activity. The MAP2⁺ precursors remain constant in number and in cell cycle activity. This argues for re-interpretation of experimental neuronal culture data.