Neural network simulation with hyperploid neurons

When subjected to stress, terminally-differentiated neurons are susceptible to reactivate the cell cycle and become hyperploid. This process is well documented in Alzheimer’s disease (AD), where it may participate in the etiology of the disease. However, despite its potential importance, the effects of neuronal hyperploidy (NH) on brain function and its relationship with AD remains obscure. An important step forward in our understanding of the pathological effect of NH has been the development of transgenic mice with neuronal expression of oncogenes as model systems of AD. The analysis of these mice has demonstrated that forced cell cycle reentry in neurons results in most hallmarks of AD, including neurofibrillary tangles, Abeta peptide deposits, gliosis, cognitive loss, and neuronal death. Nevertheless, in contrast to the pathological situation, where a relatively small proportion of neurons become hyperploid, neuronal cell cycle reentry in these mice is generalized. We have recently developed an in vitro system in which cell cycle is induced in a reduced proportion of differentiated neurons, mimicking the in vivo situation. This manipulation reveals that NH correlates with synaptic dysfunction, and that membrane depolarization facilitates the survival of hyperploid neurons. This suggests that the integration of synaptically-silent, hyperploid neurons in electrically-active neural networks allows their survival while perturbing the normal functioning of the network itself, a hypothesis that we have tested in silico. To this aim, an `Integrate-and-fire´ simulation of neural networks containing hyperploid neurons was implemented using the Python-based Brian 2 simulator.