Abstract
Successful lactation depends on the hormone, oxytocin, which activates oxytocin receptors in the mammary glands to induce myoepithelial smooth muscle contraction for milk ejection during suckling. Oxytocin secretion into the circulation is initiated by action potential (spike) firing in magnocellular oxytocin neurons that trigger exocytosis from axon terminals in the posterior pituitary gland. Milk ejection is episodic because suckling induces high frequency bursts of spikes that are coordinated across the oxytocin neuron population, causing a large bolus of oxytocin secretion. The anterior pituitary hormone, prolactin, inhibits the basal firing rate of oxytocin neuron activity in virgin rats. However, prolactin stimulates milk synthesis during lactation when oxytocin secretion is also required for milk ejection and prolactin increases the basal firing rate of oxytocin neurons in lactation in anaesthetized rats. Therefore, I hypothesised that prolactin would stimulate oxytocin neuron bursts in freely-behaving lactating mice.
In-vivo electrophysiology has been used to investigate the mechanisms that generate oxytocin neuron bursts in anaesthetised rats, but these investigations only allowed recording of pairs of oxytocin neurons and quickly stalled because of the technical challenges of such recordings. Fibre photometry has recently been developed to monitor the activity of neuronal populations through viral transduction of genetically-encoded Ca2+ indicators in freely-behaving mice, which is ideal for investigation of the mechanisms that generate oxytocin neuron bursts. Therefore, I aimed to characterise a mouse model for fibre photometry recording of oxytocin neuron activity in freely-behaving mice, and to use this model to determine whether prolactin stimulates oxytocin neuron bursts during suckling.
Oxytocin neurons also secrete oxytocin centrally to modulate behaviours, including stress and food intake. However, the temporal response of oxytocin neurons to stress and food intake is not well characterised. Furthermore, oxytocin neurons receive multiple inputs that relay thirst but the response of oxytocin neurons to fluid intake is also not well characterised. Therefore, I also aimed to record oxytocin neuron activity in response to stress, and to food and fluid intake in freely-behaving virgin mice.
First, an oxytocin-GCaMP6s mouse model was developed for visualisation of oxytocin neuron activity. To validate the model, I showed that oxytocin-Cre and oxytocin-Cre/td-Tomato mice expressed high levels of oxytocin colocalisation with Cre, and that stereotaxic injection of Cre-dependent virus transduced GCaMP6s specifically in oxytocin neurons. Next, I showed that GCaMP6s fluorescence increased concurrently with an increase in the firing rate of individual oxytocin neurons in brain slices. Finally, I confirmed that the GCaMP6s fluorescence from the oxytocin neuron population was evident in freely-behaving virgin mice, which exhibited clear increases in fluorescence in response to a mild stressor.
Fibre photometry recording of lactating mice showed large peaks of fluorescence that were followed closely by vigorous pup movement, indicating milk ejection. I found that reducing the number of pups suckling, to reduce suckling intensity, reduced the frequency of peaks but not the shape of individual peaks. Hence, suckling intensity appears to drive the generation of oxytocin neuron bursts. Fluorescence was then recorded on multiple days across lactation and showed increased peak frequency as lactation progressed, possibly due to increased suckling intensity as the pups grew.
Oxytocin neuron bursts depend on positive autocrine feedback by oxytocin secretion from the soma and dendrites of oxytocin neurons. Therefore, I tested whether oxytocin receptor antagonism blocks oxytocin neuron bursts to confirm that the model could be used to replicate known mechanisms that underpin bursts. Subcutaneous administration of L-368,899 (10 mg kg-1), a blood brain barrier permeable oxytocin receptor antagonist, had no effect on the frequency or shape of peaks in lactating mice oxytocin neuron bursts in freely-behaving mice. The antagonist was given peripherally because the fibre optic headpiece prevents central administration, which blocked milk ejection by antagonism of oxytocin receptors on mammary glands. Hence, the increased activation by pup suckling might have overcome any inhibition by central oxytocin receptor antagonism.
Subcutaneous prolactin (3 mg kg-1) was administered during suckling to investigate the role of prolactin in the generation of oxytocin neuron bursts. Prolactin had no significant effect on the frequency or shape of peaks in lactating mice. Hence, it appears that prolactin might have no effect on the generation of oxytocin neuron bursts. However, these experiments were impacted by COVID, with two cohorts of fibre-implanted, oxytocin-Cre mice being culled prior to lactation due to lockdowns. Because of the low numbers of observations, the statistical analyses were underpowered and must be interpreted with caution.
Mild stress, but not food intake and water intake, caused a small, rapid increase in fluorescence immediately after the onset of stress, with no other changes in signal associated with the stress. Hence, oxytocin neurons appear to be sensitive to the onset of stress and might contribute to the physiological responses to stress.
Overall, the research presented in this thesis characterises a new mouse model that can be used to monitor oxytocin neuron activity in freely-behaving mice by fibre photometry in oxytocin-Cre mice following viral transduction of GCaMP6s. However, it remains to be determined whether prolactin drives oxytocin neuron bursts during lactation because COVID lockdowns prevented full completion of this experiment. Nevertheless, I showed that oxytocin neuron activity increases at the onset of stress, and so might contribute to the behavioural and physiological responses to stress.