Balancing memory fidelity and representational stability in the female mouse accessory olfactory bulb

Sensory systems must balance the value of efficient coding schemes against the need to update specific memorized representations without perturbing other memories. Here we describe a unique solution to this challenge that is implemented by the vomeronasal system (VNS) to encode and remember multiple conspecific individuals as part of the Bruce Effect (BE). In the BE, exposure of a pregnant female mouse to the odors of an unfamiliar male leads to failure of the pregnancy ( pregnancy block) via the VNS. Following mating and sensory exposure, however, the female becomes protected from a pregnancy block by the stud individual. While this form of natural learning has been proposed to depend on changes in the representation of his odors in her accessory olfactory bulb (AOB), a key VNS structure, there are no direct comparisons of in vivo sensory responses before and after imprinting. It has further been suggested that these changes simply render the AOB insensitive to stud odors. However, the combinatorial odor code used by the AOB and the significant overlap in the odor composition of different males means that silencing responses to one individual is likely to degrade responses to others, posing potential problems for more general sensory encoding. To identify the neuronal correlates of learning in the context of the BE, we recorded extracellular responses of AOB neurons in vivo in mated and unmated female mice upon controlled presentation of urinary chemosignals, including urine from both the stud and males of a distinct strain. We find that while initial sensory responses in the AOB (within a timescale required to guide social interactions) remain stable, responses to extended stimulation (as required for eliciting the pregnancy block) display selective attenuation of stud-responsive neurons. Based on our results, we propose a model that reconciles the formation of strong, selective memories with the need to sustain robust representational bandwidth by noting a distinction between the representations of brief and extended stimuli. This temporal disassociation allows attenuation of slow-acting endocrine processes in a stimulus-specific manner, without compromising consistent ongoing representations of stimuli that guide behavior.