We have recently developed a novel reporter mouse line designed to help identify neural stem and progenitor cells in the adult nervous system and accurately quantify changes in selected classes of precursor cells induced by pro- or anti-neurogenic stimuli. In these mice, regulatory elements of the nestin gene, a marker of neural stem and progenitor cells, drive the expression of cyan fluorescent protein (CFP) fused to a nuclear localization signal from SV40 (nestin-CFPnuc mice; (Encinas, et al., 2006
). Nestin-driven transgene expression visualizes several subclasses of neural progenitors in the adult SVZ and DG, whereas the nuclear representation of the transgene-expressing cells greatly facilitates enumeration of these subclasses. We use this reporter line to identify the precursor cell populations targeted by antidepressant fluoxetine (Encinas, et al., 2006
), by deep brain stimulation (Encinas et al., submitted) and by electroconvulsive shock (Park et al., unpublished), thus demonstrating its utility for the studies of neurogenesis.
To simulate the space radiation environment, we exposed nestin-CFPnuc mice to 100 cGy doses of 1 GeV/n 56Fe ions. After irradiation animals were injected with thymidine analogue 8-Bromo-deoxyuridine (BrdU) to label dividing cells and selected cell populations in the brain were analyzed 24 hours later. We focused on neurogenesis in the hippocampus, since this region of the brain is involved in learning, memory, and emotions.
Exposure to radiation resulted in a dramatic decrease (to 55%) in the number of BrdU-positive cells in the DG (), in line with other reports demonstrating the radiation-induced decrease in the number of dividing cells in the DG (Mizumatsu, et al., 2003
, Monje, et al., 2002
, Rola, et al., 2004
). In all cases, BrdU-positive cells were positioned in the subgranular zone (SGZ) and only occasionally in the hilus (), indicating that radiation did not induce aberrant ectopic neurogenesis in the hippocampus.
Figure 1 A–D.
56Fe radiation decreases the number of BrdU-labeled cells (A), nestin-CFPnuc cells (B), QNP cells (C), and ANP cells (D) in the DG of nestin-CFPnuc mice. Mice were irradiated, injected with BrdU, and sacrificed 24 hours later. White bars (more ...)
Radiation also decreased the number of nestin-CFPnuc-expressing neural precursor cells in the DG (). In the nestin-CFPnuc reporter animals, expression of CFP marks two classes of precursors in the DG, the quiescent neuronal progenitors (QNPs) and the amplifying neuronal progenitors (ANPs); these classes can be distinguished by their morphology and by their expression of glial fibrillary acidic protein (GFAP) and vimentin (Encinas, et al., 2006
). The number of QNPs was decreased by irradiation by 45%; the reduction in the number of ANP cells, which are derived from the QNP cells, was much less (25%; ). There were no additional morphological or anatomical changes in the population of progenitor cells that were apparent as a result of irradiation (). Thus, 56
Fe radiation drastically decreased the number of dividing cells and neural progenitor cells in the DG, having the greatest effect on the quiescent population of neural progenitors.
The loss of neural progenitors was also reflected in the highly increased number of dying and dead cells in the brain. These changes were evident when revealed both by amino cupric silver (a postmortem cell marker, showing disintegrative debris) and caspase-3 staining (a pre-mortem marker of apoptosis). The increase in cell degeneration was observed selectively in the neurogenic areas of the brain, the DG (), the RMS (), and the SVZ (data not shown); there were only a few caspase-3- or amino cupric silver-positive cells seen outside of these neurogenic areas of the brain, e.g., in the cortex (not shown). Within the DG, the dying cells were detected in the SGZ, where QNP and ANP cells are located (). At this point, we cannot determine the identity of dying cells; thus the evidence for the death of QNP cells is indirect and is based on a dramatic decrease in their number. This leaves open several alternative interpretations such as decrease in the levels of expression of markers used to identify QNPs (note, however, that we probed these cells using several markers, GFP, GFAP, and vimentin); changes in the QNP morphology (note, however, that morphology of the remaining QNPs cells does not change); or their rapid transformation into other cells types e.g., ANPs (note, however, that the number of ANPs is also decreased).
A–C. Amino-cupric-silver stain for cell degeneration (A, B) and anti-caspase-3 staining for apoptotic cells (C) in the DG of control (A) and irradiated (B, C) mice, showing cell degeneration and apoptosis after irradiation.
The increase in the number of dying cells was observed 6 hours after irradiation, but not after 24 hours or 3 weeks, indicating a rapid clearance of radiation-damaged cells from the neurogenic areas, consistent with the observed apoptosis. Supporting these findings, we observed activated Iba1-positive microglial cells and cells resembling infiltrating macrophages selectively located in the SGZ (); they were also seen in the SVZ and the RMS () but not in the cortex of the irradiated animals. As with the number of dying cells, these changes, normally associated with inflammation and tissue degeneration, were observed 6 h, but not 24 h or 3 weeks, after irradiation. Thus, the observed reduction in neural progenitor cells in the neurogenic areas is reciprocated by an increase in the number of dying and dead cells in these regions.
Together, our results demonstrate that QNPs, a population of quiescent stem-like cell in the hippocampus, are selectively killed by radiation. The concomitant loss of ANPs was expected since they represent a rapidly dividing cell population that is thus susceptible to various types of radiation. However, the finding that QNP cells, despite their low division rate, are particularly vulnerable to radiation, was unexpected as the current view holds that proliferating cells are more sensitive to radiation than quiescent ones. This suggests that additional mechanisms, not directly related to cell replication, may be responsible for this selective loss.
Our finding that a quiescent population of stem cells is killed by radiation is surprising in the light of the emerging concept of cancer stem cells (Dalerba, et al., 2007
, Ignatova, et al., 2002
, Jordan, et al., 2006
, Vescovi, et al., 2006
). This concept holds that tumors may re-arise after chemo- or radiotherapy from a small population of transformed cells with stem cell properties; it is assumed that these cells escape the cytotoxic regimens directed against rapidly dividing cells by virtue of their quiescence and then re-initiate the tumor. Our results on the exceptional sensitivity of QNPs to irradiation suggest that quiescence alone may not provide protection from radiation-induced cell death. Thus, there may be some other additional mechanisms operating in cancer stem cells (e.g., elevated repair capabilities or ability to exclude chemotherapeutic drugs) that are crucial for their apparent resistance to cytotoxic therapy.
Importantly, if the loss of QNPs, the most primitive and normally non-self-renewable progenitor class in the hippocampus, is not compensated (e.g., by increasing the rate of asymmetric divisions of the remaining QNPs or restoring their number through symmetric divisions), the number of new neurons and later, of all granule neurons in the DG may decline Such a delayed effect on neurogenesis has been observed in animals subjected to irradiation with 12
C and 56
Fe ions (Rola, et al., 2005
) (at this point we are limited in our access to the radiation beam and will be able to address this important issue in our system in the future). Given the increasing evidence pointing to the role of adult neurogenesis in memory and mood control (e.g., the association between reduced neurogenesis in the DG and impairments in hippocampus-dependent cognitive tasks or the deficient response to antidepressant in mice with suppressed hippocampal neurogenesis (Santarelli, et al., 2003
, Saxe, et al., 2006
), the risk to stem-like QNP cells represents an important factor to consider when planning manned space missions or considering radiation exposure for therapeutic purposes. Further investigations should address the issue of whether a reduced flux expressed over a longer period that manned missions may expect during space exploration would have the same effects as the acute exposure levels used in these experiments. Our model offers a ground-based radiation-exposure test system that will help to assess radiation risk and to develop countermeasures, such as shielding and radioprotective drugs.