Dynamic modulation of pulsatile activities of oxytocin neurons in lactating wild-type mice

**Background:** Oxytocin (OT) is a nine-amino-acid peptide hormone essential for milk ejection during lactation. The milk ejection reflex is mediated by synchronous burst activities of OT neurons in the paraventricular (PVH) and supraoptic (SO) hypothalamus. Previous studies relied on Cre-dependent double knock-in mice for calcium imaging, which limited the use of other neuroscience toolkits. The dynamics of OT neuron activity across different lactation stages, including weaning, remained unknown.

**Methods:** The authors developed an adeno-associated virus (AAV) vector driving GCaMP6s under a 2.6 kb mouse oxytocin mini-promoter (AAV9 OTp-GCaMP6s). This was injected into the PVH of wild-type C57BL/6N and ICR female mice, followed by optic fiber implantation. Fiber photometry recordings were conducted for 6-hour sessions in light and dark phases across multiple lactation stages: early (PPD2-4), middle (PPD12-14), late (PPD18), and peri-weaning (PPD18-27). Video recordings were synchronized to analyze mother-pup interactions. In isolation experiments, pups were placed in a wire mesh enclosure within the home cage for 6 hours to prevent physical contact while allowing visual, auditory, and olfactory exposure. Fostering experiments involved introducing elder pups (PND17-19) to early-lactation mothers (PPD2-4) and younger pups (PND2-4) to late-lactation mothers (PPD17-19).

**Key Results:** The AAV9 OTp-GCaMP6s showed high specificity: 96.7% ± 0.37% of GCaMP6s+ neurons were OT+ in virgin mice and 94.6% ± 1.2% in mother mice. No significant differences were found in peak number or FWHM between C57BL/6N and ICR strains, though a slight but significant difference in IPI distribution was observed (p<0.01, Kolmogorov-Smirnov test). During physical isolation of pups in a wire mesh, no pulsatile OT neuron activity was observed, even when mothers actively investigated the enclosure. After reunion, the number of peaks per 2-hour window increased significantly by an average of 1.9-fold (p<0.05, Wilcoxon signed-rank sum test), with a significant shift toward shorter IPIs (p<0.01, Kolmogorov-Smirnov test). Throughout natural lactation, pulsatile activity was closely correlated with mothers being in the nest. The number of pulses per hour in the nest during the dark phase increased significantly in later lactation stages (p<0.05, two-way ANOVA with Tukey's HSD). IPI distributions showed significant changes across lactation stages (p<0.01, Kolmogorov-Smirnov test with Bonferroni correction), with IPIs within 450 s at PPD18 following an exponential distribution consistent with a Poisson process. Fostering elder pups to early-lactation mothers increased the proportion of short IPIs, mimicking late-lactation patterns. During peri-weaning (PPD22-26), the number of peaks decreased gradually, disappearing by PPD26, with the dark phase showing earlier reduction. Peak intensities and waveforms remained constant throughout weaning.

**Clinical Implications:** This study establishes that in mice, unlike in humans and livestock, the milk ejection reflex strictly requires direct nipple stimulation by pups, with no conditional learning to pup-associated sensory cues. The Cre-free AAV method enables oxytocin neuron imaging in any mouse strain without genetic crossing, facilitating future studies using various genetically modified mice. The finding that OT neuron activity follows a Poisson process within clusters suggests stochastic recruitment mechanisms. The gradual rather than sudden decline in OT neuron activity during weaning, with dark-phase precedence, may reflect parent-offspring conflict. This tool is compatible with Cre-based optogenetic and pharmacogenetic manipulations, enabling future dissection of afferent circuitry from nipple to OT neurons. The approach is potentially applicable to other mammalian species for comparative studies of lactation.