Promotie Bas Jan Zandt - 21 februari 2014

The interruption of blood flow to the brain as occurs in cardiac arrest and stroke results within minutes in irreversible damage. The development of neuroprotective treatments that prevent cell damage after stroke has so far largely been unsuccessful, while we still have an incomplete understanding of the dynamics of the processes involved. This thesis has focused on the dynamics of ion concentrations and neuronal activity during and after hypoxia.

In chapter 3, it was shown that the extracellular potassium concentration has a crucial role in anoxic depolarization. If potassium fluxes are insufficiently compensated by the ATP-dependent Na/K pumps, the extracellular concentration may reach a critical concentration, resulting in an additional, significant increase in potassium efflux. This efflux induces massive depolarization of the neurons, reflected as a transient wave of activity on the scalp EEG. This explains the misnamed ``Wave of Death'' that is observed in rats after decapitation.

The released potassium does not only excite the neurons that released it, but also diffuses to neighboring cells, causing a chain reaction or ``reaction-diffusion'' process. Chapter 4 has presented expressions for the initiation, propagation and wave shape of idealized spreading depolarization (SD). The predicted shape of the onset of the wave was validated with potassium measurements in vivo in rat from literature.

The model used to calculate the behavior of depolarizing single neurons in the previous two chapters, the Hodgkin-Huxley (HH) model with dynamic ion concentrations, is experimentally validated in Chapter 5. Several time courses of depolarizing pyramidal cells were obtained in in-vitro experiments after blocking the Na/K-pump with ouabain. Five different types of membrane voltage dynamics were observed. These correspond to different trajectories of the sodium and potassium Nernst potentials in a bifurcation diagram of the HH model.

In chapter 6, an initial effort was made in modeling the behavior of large populations of neurons (neural mass models), rather than single ones, with pathological transmembrane ion gradients. The firing rate curve was used as link between the single cell and neural mass model. This model excellently reproduces the dynamics observed in a simulated network of HH-neurons.