Social experience dependent plasticity of mouse song selectivity without that of song components

Abstract

Syllable sequences in male mouse ultrasonic-vocalizations (USVs), "songs", contain structure -quantified through predictability, like birdsong and aspects of speech. Apparent USV innateness and lack of learnability, discount mouse USVs for modelling speech-like social communication and its deficits. Informative contextual natural sequences (SN) were theoretically extracted and they were preferred by female mice. Primary auditory cortex (A1) supragranular neurons show differential selectivity to the same syllables in SN and random sequences (SR). Excitatory neurons (EXNs) in females showed increases in selectivity to whole SNs over SRs based on extent of social exposure with male, but syllable selectivity remained unchanged. Thus mouse A1 single neurons adaptively represent entire order of acoustic units without altering selectivity of individual units, fundamental to speech perception. Additionally, observed plasticity was replicated with silencing of somatostatin positive neurons, which had plastic effects opposite to EXNs, thus pointing out possible pathways involved in perception of sound sequences.