Developmental social environment imprints female preference for male song in mice

Abstract

Background: Sexual imprinting is important for kin recognition and for promoting outbreeding, and has been a driving force for evolution; however, little is known about sexual imprinting by auditory cues in mammals. Male mice emit song-like ultrasonic vocalizations that possess strain-specific characteristics.

Objectives: In this study, we asked whether female mice imprint and prefer specific characteristics in male songs.

Methods and findings: We used the two-choice test to determine the song preference of female C57BL/6 and BALB/c mice. By assessing the time engaged in searching behavior towards songs played back to females, we found that female mice displayed an innate preference for the songs of males from different strains. Moreover, this song preference was regulated by female reproductive status and by male sexual cues such as the pheromone ESP1. Finally, we revealed that this preference was reversed by cross-fostering and disappeared under fatherless conditions, indicating that the behavior was learned by exposure to the father's song.

Conclusions: Our results suggest that female mice can discriminate among male song characteristics and prefer songs of mice from strains that are different from their parents, and that these preferences are based on their early social experiences. This is the first study in mammals to demonstrate that male songs contribute to kin recognition and mate choice by females, thus helping to avoid inbreeding and to facilitate offspring heterozygosity.

Figure 1. Strain-specific characteristics of male mouse songs.

(a) Representative spectrogram of the songs from B6, KJR, BALB, and ICR. Arrows and arrowheads indicate the “Jump” (B6) and “Harmonic” waveforms (BALB), respectively. The dotted lines indicate the average peak frequency of each song (B6, 79 kHz; KJR, 75 kHz; BALB, 55 kHz; ICR, 66 kHz). (b) Cluster analysis using the following parameters: mean peak frequency of syllables, interval between syllables, duration of syllables, and percentage composition, for each waveform category among B6, BALB, ICR, and KJR male individuals (B6, n = 6; BALB, n = 7; ICR, n = 5; KJR, n = 3). There were 2 major song groups: BALB and ICR, and B6 and KJR.

Figure 2. Female mice prefer songs of males from different strains.

(a) Schematic of the apparatus used for the playback experiment. (b) B6 (n = 6) and BALB (n = 10) females in diestrus exposed to male-soiled bedding before testing showed longer-duration search times for other strains' songs. (c) Duration of time searching during diestrus in B6 (n = 5) and BALB (n = 7) females in the absence of male odor before testing. (d) B6 (n = 11) and BALB (n = 13) females in diestrus exposed to male pheromone ESP1 before testing showed longer duration search times for other strains' songs. Values represent means+standard error. Asterisks indicate significant differences p<0.05.

Figure 3. Female song searching response to playback with KJR and ICR strain male song.

(a) When B6 and ICR male songs were presented, ESP1-treated B6 females showed longer search times for ICR songs than for B6 songs (n = 13), whereas there was no difference in search time when B6 and KJR male songs were presented (n = 12). (b) When BALB and KJR male songs were presented, ESP1-treated BALB females showed longer search times for KJR songs than for BALB songs (n = 12), whereas there was no difference when BALB and ICR male songs were presented (n = 11). Values represent means+standard error. Asterisks indicate significant differences p<0.05.

Figure 4. Cross-fostering reversed, and fatherless rearing diminished, female song preference.

(a) Experimental scheme of the cross-fostering paradigm. Part of the litter (B6-fosterer and BALB-fosterer) was reciprocally cross-fostered to parents of the other strain of mice. (b–e) B6 sister (n = 10) and BALB-fosterer (n = 15) showed longer search times for BALB song. In contrast, B6-fosterer (n = 16) and BALB sister (n = 14) treated with ESP1 prior to the experiment showed longer search times for B6 song. GLM with fosterer/sister and male song as factors revealed significant effects of the interaction. (f) Fatherless B6 and fatherless BALB ware raised by mothers only. (g–h) There was no significant difference in search time for B6 and BALB songs in fatherless B6 and BALB females. Values represent means+standard error. Asterisks indicate significant differences p<0.05.