Centre for Discovery Brain Sciences, University of Edinburgh, Edinburgh, UK.
J Physiol. 2019 Jul;597(14):3657-3671. doi: 10.1113/JP278045. Epub 2019 Jun 9.
A quantitative model of oxytocin neurones that combines a spiking model, a model of stimulus-secretion coupling and a model of plasma clearance of oxytocin was tested. To test the model, a variety of sources of published data were used that relate either the electrical activity of oxytocin cells or the secretion of oxytocin to experimentally induced changes in plasma osmotic pressure. To use these data to test the model, the experimental challenges involved were computationally simulated. The model predictions closely matched the reported outcomes of the different experiments.
Magnocellular vasopressin and oxytocin neurones in the rat hypothalamus project to the posterior pituitary, where they secrete their products into the bloodstream. In rodents, both vasopressin and oxytocin magnocellular neurones are osmoresponsive, and their increased spiking activity is mainly a consequence of an increased synaptic input from osmoresponsive neurons in regions adjacent to the anterior wall of the third ventricle. Osmotically stimulated vasopressin secretion promotes antidiuresis while oxytocin secretion promotes natriuresis. In this work we tested a previously published computational model of the spiking and secretion activity of oxytocin cells against published evidence of changes in spiking activity and plasma oxytocin concentration in response to different osmotic challenges. We show that integrating this oxytocin model with a simple model of the osmoresponsive inputs to oxytocin cells achieves a strikingly close match to diverse sources of data. Comparing model predictions with published data using bicuculline to block inhibitory GABA inputs supports the conclusion that inhibitory inputs and excitatory inputs are co-activated by osmotic stimuli. Finally, we studied how the gain of osmotically stimulated oxytocin release changes in the presence of a hypovolaemic stimulus, showing that this is best explained by an inhibition of an osmotically regulated inhibitory drive to the magnocellular neurones.
我们测试了一种将尖峰模型、刺激-分泌偶联模型和催产素血浆清除模型相结合的催产素神经元定量模型。为了测试该模型，我们使用了各种来源的已发表数据，这些数据要么与催产素细胞的电活动有关，要么与催产素的分泌有关，与实验诱导的血浆渗透压变化有关。为了使用这些数据来测试该模型，我们对涉及的实验挑战进行了计算机模拟。该模型的预测结果与不同实验的报告结果非常吻合。
大鼠下丘脑的大细胞血管升压素和催产素神经元投射到垂体后叶，在那里它们将其产物分泌到血液中。在啮齿动物中，血管升压素和催产素大细胞神经元均对渗透压有反应，其尖峰活动的增加主要是由于来自第三脑室前壁相邻区域渗透压反应神经元的突触输入增加所致。渗透压刺激的血管升压素分泌促进抗利尿作用，而催产素分泌促进钠排泄。在这项工作中，我们针对不同渗透压挑战下尖峰活动和血浆催产素浓度变化的已有发表证据，测试了先前发表的催产素细胞尖峰和分泌活动计算模型。我们表明，将该催产素模型与催产素细胞的简单渗透压反应输入模型相结合，可以非常准确地匹配各种来源的数据。使用毒蕈碱型乙酰胆碱受体拮抗剂（如印防己毒素）阻断抑制性 GABA 输入，将模型预测与已发表数据进行比较，支持了这样的结论，即渗透压刺激会同时激活抑制性输入和兴奋性输入。最后，我们研究了在低血容量刺激存在的情况下，渗透压刺激的催产素释放增益如何变化，结果表明，这可以通过对大细胞神经元的渗透压调节性抑制驱动的抑制来最好地解释。
