An in vitro model of the impact of chemotherapy on neural stem cells and the protection provided by cells in the neurogenic niche

Chemotherapy has been highly successful in treating many forms of cancer; however there are increasing reports that this treatment causes cognitive declines in cancer survivors. These effects have been called “chemobrain” and while not affecting all patients, can persist for many years after the completion of treatment. The symptoms of chemobrain include a decline in concentration, memory and attention which are associated with a lower quality of life and in ability to return to work. Very little is known about the mechanisms behind these changes, or even the brain regions that are affected. Animal studies have found that systemic chemotherapy causes a decrease in the proliferation of neural stem cells (NSCs) in the subgranular zone (SGZ) neurogenic niche of the hippocampus and a decline in spatial memory. As hippocampal neurogenesis is required for a number of memory functions including the consolidation of long term memory, a decline in neurogenesis is likely to be one of the causes of the cognitive decline experienced by patients after chemotherapy. Previous animal work has shown that chronic treatment with the antidepressant fluoxetine prevents the decrease in neurogenesis and the associated cognitive decline. In the absence of fluoxetine chemotherapy spares dividing cells which are in contact with the surface of blood vessels in the SGZ of the hippocampus. This project used in vitro techniques to firstly look at whether fluoxetine has a direct effect on the sensitivity of neural cells to chemotherapy or whether treatment of astrocytes cells with fluoxetine produces an indirect effect on sensitivity. Secondly the effect of contact between NSCs and either astrocytes or endothelial cells was investigated to see if these cell types could provide protection to NSCs from chemotherapy. In the third part of the project, fluorescence activated cell sorting (FACS) and measurements of oxygen consumption after different drug treatments were used to investigate DNA damage, apoptosis and changes in the cell cycle and the metabolic response of different cell types to chemotherapy respectively. To do this we evaluated the relative sensitivity of primary NSCs, neural N2a cells, endothelial cells HBMEC, primary astrocytes, C6 astrocytes and 3T3 cells to the chemotherapy drug 5-fluorouracil (5-FU). NSCs were found to be more sensitive to 5-FU than other cell types. A concentration of 5 μM 5-FU was found to reduce NSC viability by 50% but largely to spare astrocytes and endothelial cells. FACS analysis showed that 5-FU was causing DNA damage and inducing apoptosis with some cells arresting in the G1 stage of the cell cycle. Measurements of oxygen consumption were not conclusive in explaining the differences in sensitivity between different cell types. Treatment with fluoxetine, either directly to neural cell cultures or via conditioned media from fluoxetine treated astrocytes or endothelial cells, was not found to be protective against 5-FU treatment. In contrast co-culture of NSCs with astrocytes or endothelial cells, protected NSCs from chemotherapy treatment. This effect was specific to cells in the neurogenic niche as co-culture with 3T3 fibroblasts did not provide protection. Direct cell contact between neural cells and astrocytes or endothelial cells was required for protection as neither conditioned media nor co-culture separated by a porous membrane was found to be effective. To investigate the mechanism behind astrocyte or endothelial contact dependent protection of neural cells, the gap junction protein Cx34 was stained for and found to be present on neural, astrocytic and endothelial cells. Fluorescent dye loading of marked cells showed that gap junctions were functional and dye could pass between them. Blocking gap junctions with the gap junction inhibitor carbenoxolone (CBX) abolished the protection provided by contact with astrocytes or endothelial cells. I hypothesise that the protection provided in vivo by fluoxetine may be mediated by its anti-inflammatory effect on the brain and is not a direct effect on cells in the stem cell niche. Contact with astrocytes or endothelial cells may provide Ca2+ buffering via gap junctions and in this way protect the more sensitive neural cells from the effects of chemotherapy