Astrocytes : a reservoir for new neurons

Abstract

The goal of regenerative medicine is to be able to help the body to replace worn-out cells and tissues. To achieve this goal, one approach being pursued today is to recruit and boost the body’s own mechanisms for cell replacement. In the brain, this approach is perhaps more challenging than in many other organs because the brain’s own ability to replace neurons is almost non-existent. Therefore, an alternative strategy is to genetically reprogram support cells in the living brain directly into neurons, or into progenitor cells which in turn would produce neurons. For such a strategy to work, a first step is to identify a cell type in the brain that has the capacity to act as a source for neurons. This thesis describes work that shows for the first time that astrocytes, one of the most abundant cell types in the brain, can generate neurons in vivo under certain circumstances and in certain brain regions. Thus, astrocytes represent a potential reservoir for new neurons, whose neurogenic capacity might be recruited to improve brain repair.

In Paper I, we describe the finding that some astrocytes generate neurons in response to stroke in the striatum of mice. We found that the neurogenic capacity of these astrocytes was under control of the Notch signaling pathway. By manipulating this pathway experimentally, we could activate the neurogenic program of these astrocytes even in the absence of stroke. Yet, not all astrocytes generated neurons. In fact, outside the striatum, most of them did not do so, neither after stroke nor after Notch manipulation.

In Paper II, we therefore used single-cell RNA sequencing to better understand how astrocytes respond to Notch manipulation and why this does not induce all astrocytes to undergo neurogenesis. We found that even astrocytes that do not generate neurons in fact initiate early steps of neurogenesis. However, they halt their lineage progression immediately before undergoing transit-amplifying divisions. In the striatum, exposure to a mitogen pushed such halted astrocytes into transit-amplifying divisions, and our results suggest that similar strategies could work also outside the striatum, given the right stimulus.

In Paper III, we asked whether stroke-induced striatal neurogenesis occurs also in humans. For this, we used radiocarbon dating to assess the age of striatal neurons isolated from postmortem samples of stroke patients. We found that the stroke-injured striatum contains a higher proportion of young neurons than the non-injured striatum of the same subjects. This could be explained either by neurogenesis or selective death of old neurons. Each of these possibilities would represent an interesting and previously undescribed biological scenario.

The work in these three papers suggests that endogenous brain cells exist whose neurogenic properties could be recruited to improve brain repair. However, one additional challenge with neuronal replacement strategies is that the great diversity of neurons in the brain is still incompletely characterized. Before any cell replacement interventions can be undertaken, the cell composition in the healthy brain must be known. In Paper IV, we describe a new method for performing RNA sequencing on intact tissue sections with retained spatial information. This technique, dubbed spatial transcriptomics, is useful for answering fundamental biological questions about cell distribution and gene expression. In addition, it could provide valuable information for disease diagnostics. In summary, the work presented in this thesis provides information that may prove crucial for bringing neuronal replacement strategies closer to the clinic.