Vasopressin neurons generate distinctive phasic patterned spike activity in response to elevated extracellular osmotic pressure. These spikes are generated in the cell body and are conducted down the axon to the axonal terminals where they trigger Ca 2+ entry and subsequent exocytosis of hormone-containing vesicles and secretion of vasopressin. This mechanism is highly non-linear, subject to both frequency facilitation and fatigue, such that the rate of secretion depends on both the rate and patterning of the spike activity. Here we used computational modelling to investigate this relationship and how it shapes the overall response of the neuronal population. We generated a concise single compartment model of the secretion mechanism, fitted to experimentally observed profiles of facilitation and fatigue, and based on representations of the hypothesised underlying mechanisms. These mechanisms include spike broadening, Ca 2+ channel inactivation, a Ca 2+ sensitive K + current, and releasable and reserve pools of vesicles. We coupled the secretion model to an existing integrate-and-fire based spiking model in order to study the secretion response to increasing synaptic input, and compared phasic and non-phasic spiking models to assess the functional value of the phasic spiking pattern. The secretory response of individual phasic cells is very non-linear, but the response of a heterogeneous population of phasic cells shows a much more linear response to increasing input, matching the linear response we observe experimentally, though in this respect, phasic cells have no apparent advantage over non-phasic cells. Another challenge for the cells is maintaining this linear response during chronic stimulation, and we show that the activity-dependent fatigue mechanism has a potentially useful function in helping to maintain secretion despite depletion of stores. Without this mechanism, secretion in response to a steady stimulus declines as the stored content declines.
Vasopressin is a hormone that is secreted from specialised brain cells into the bloodstream; it acts at the kidneys to control water excretion, and thereby help to maintain a stable ‘osmotic pressure’. Specialised cells in the brain sense osmotic pressure, and generate electrical signals which the thousands of vasopressin neurons process and respond to by producing and secreting vasopressin. In response to these signals, vasopressin neurons generate complex “phasic” patterns of electrical activity, and this activity leads to vasopressin secretion in a complex way that depends on both the rate and pattern of this activity. We have now built a computational model that describes both how the vasopressin neurons generate electrical activity and also how that activity leads to secretion. The model, which gives a very close fit to experimental data, allows us to explore the adaptive advantages of particular features of the vasopressin neurons. This analysis reveals the importance of heterogeneity in the properties of vasopressin neurons, and shows how the vasopressin system is optimally designed to maintain a consistent hormonal output in conditions where its stores of releasable hormone are severely depleted.