Many species of animals exhibit caregiving or aggression toward conspecific offspring. The neural mechanisms underlying the infanticide and pup care remain poorly understood. Here, using monogamous mandarin voles (Microtus mandarinus), we found that more oxytocin (OT) neurons in the paraventricular nucleus (PVN) were activated during pup caring than infanticide. Optogenetic activation of OT neurons in the PVN facilitated pup caring in male and female mandarin voles. In infanticide voles, optogenetic activation of PVN OT cells or PVN-medial prefrontal cortex (mPFC) OT projection fibers prolonged latency to approach and attack pups, whereas inhibition of these OT neurons or projections facilitated approach and infanticide. Optogenetic activation of PVN OT neuron projections to the mPFC in males shortened the latency to approach and retrieve pups and facilitated the initiation of pup care, but produced no effects on pup-care females. In addition, OT release in the mPFC increased upon approaching and retrieving pups, and decreased upon attacking pups. Intraperitoneal injection of OT promoted pup care and inhibited infanticide behavior. It is suggested that the OT system, especially PVN OT neurons projecting to mPFC, modulates pup-directed behaviors and OT can be used to treat abnormal behavioral responses associated with some psychological diseases such as depression and psychosis.

Neuropeptides and neurotrophins are stored in and released from dense core vesicles (DCVs). While DCVs and synaptic vesicles (SVs) share fundamental SNARE/SM proteins for exocytosis, a detailed understanding of DCV exocytosis remains elusive. We recently identified the RAB3-RIM1 pathway to be essential for DCV, but not SV exocytosis, highlighting a significant distinction between the SV and DCV secretory pathways. Whether RIM1 is the only RAB3 effector that is essential for DCV exocytosis is currently unknown. In this study, we show that rabphilin-3A (RPH3A), a known downstream effector of RAB3A, is a negative regulator of DCV exocytosis. Using live-cell imaging at single-vesicle resolution with RPH3A deficient hippocampal mouse neurons, we show that DCV exocytosis increased threefold in the absence of RPH3A. RAB3A-binding deficient RPH3A lost its punctate distribution, but still restored DCV exocytosis to WT levels when re-expressed. SNAP25-binding deficient RPH3A did not rescue DCV exocytosis. In addition, we show that RPH3A did not travel with DCVs, but remained stationary at presynapses. RPH3A null neurons also had longer neurites, which was partly restored when ablating all regulated secretion with tetanus neurotoxin. Taken together, these results show that RPH3A negatively regulates DCV exocytosis, potentially also affecting neuron size. Furthermore, RAB3A interaction is required for the synaptic enrichment of RPH3A, but not for limiting DCV exocytosis. Instead, the interaction of RPH3A with SNAP25 is relevant for inhibiting DCV exocytosis.