Network and Population Function in Neuroendocrine Systems

The central control of hormone signaling is through neuroendocrine neurons of the hypothalamus which generate hormonal signals that control the pituitary gland. The anterior pituitary contains endocrine cells that manufacture hormones which regulate yet other hormone-producing tissues, while the posterior pituitary contains the nerve terminals of magnocellular oxytocin and vasopressin neurons that release their hormones directly into the blood. It is possible to record the electrical activity of these neurons in conjunction with measuring their secretory output and the physiological effect of that output, and this has made them important “model systems” in neuroscience. Their intrinsic properties have been studied extensively, and computational models have been developed that give a detailed understanding of their behavior. The vasopressin neurons have highly non-linear, bistable electrical properties. They generate phasic patterns of spiking activity that, in conjunction with complex non-linearities in stimulus-secretion coupling, optimize the efficiency of hormone release. Individually, the neurons respond non-linearly to increases in plasma osmotic pressure, yet the population as a whole generates an appropriate, linear secretory response. By contrast, oxytocin neurons communicate with each other to generate intermittent sharp pulses of secretion in response to suckling. Here we review how computational models have provided insights into how populations of neurons have signal processing properties that the individual neurons lack, and how intercommunication within these populations can generate complex modes of behavior.