Electrophysiological differentiation of oxytocin-and vasopressin-secreting neurones

Antidromically identified neurosecretory cells of the paraventricular (p.v.) and supraoptic (s.o.) nuclei of the hypothalamus were recorded in lactating rats under urethane anaesthesia during reflex milk ejection (m.e.) and haemorrhage. Eighty one p.v. and s.o. neurones were studied. Their background firing rates ranged from < 0.1 to 6.3 spikes/s and three distinct patterns of activity were encountered: slow irregular (73%), fast continuous (10%) and phasic (17%). Forty units (49%) displayed a brief (2-4 s) high-frequency discharge (30-60 spikes/s) correlated with suckling-induced m.e., and these were classified as m.e. (oxytocin-secreting) neurones. The remainder of the cells showed no activation at this time and were classified as non-m.e. neurones. Ten m.e. neurones tested through haemorrhage (5 ml of blood) showed a gradual acceleration of firing rates, reaching a maximum of 3.7 $\pm $ 0.7 spikes/s (mean $\pm $ s.e.) about 20 min after blood withdrawal. The firing pattern of the m.e. neurones therefore changed from a slow irregular to a fast continuous type. By contrast, 11 non-m.e. neurones tested with the same procedure showed a rapid activation reaching a maximum of 6.4 $\pm $ 0.6 spikes/s by the fourth minute. Non-m.e. neurones which were initially of the slow irregular type, first became fast continuous and later evolved into a highly characteristic phasic pattern of activity which was never induced in the m.e. neurones. After the blood was replaced, all the cells returned to their original firing pattern. In a parallel series of experiments, plasma samples taken 5 and 25 min after haemorrhage showed a ten-fold elevation in antidiuretic activity. A slight but non-significant increase in m.e. activity was also observed. Thus p.v. and s.o. neurosecretory cells may be electrophysically differentiated into two functionally distinct populations: (1) oxytocin releasing neurones which show a high-frequency discharge before m.e. induced by suckling, and (2) vasopressin-releasing neurones which adopt a phasic pattern of firing during vasopressin release induced by haemorrhage. We suggest that the rate of vasopressin secretion into the circulation largely depends on the proportion of vasopressin neurones firing phasically, their firing rates within the phases and the duration and degree of synchronization of the phases.