Conceived and designed the experiments: TK KN. Performed the experiments: KN RN. Analyzed the data: KN RN MN KM. Wrote the paper: TK KN RN KO.
The authors have declared that no competing interests exist.
Vocal learning is a central functional constituent of human speech, and recent studies showing that adult male mice emit ultrasonic sound sequences characterized as “songs” have suggested that the ultrasonic courtship sounds of mice provide a mammalian model of vocal learning.
We tested whether mouse songs are learned, by examining the relative role of rearing environment in a cross-fostering experiment.
We found that C57BL/6 and BALB/c males emit a clearly different pattern of songs with different frequency and syllable compositions; C57BL/6 males showed a higher peak frequency of syllables, shorter intervals between syllables, and more upward frequency modulations with jumps, whereas BALB/c males produced more “chevron” and “harmonics” syllables. To establish the degree of environmental influences in mouse song development, sons of these two strains were cross-fostered to another strain of parents. Songs were recorded when these cross-fostered pups were fully developed and their songs were compared with those of male mice reared by the genetic parents. The cross-fostered animals sang songs with acoustic characteristics - including syllable interval, peak frequency, and modulation patterns - similar to those of their genetic parents. In addition their song elements retained sequential characteristics similar to those of their genetic parents' songs.
These results do not support the hypothesis that mouse “song” is learned; we found no evidence for vocal learning of any sort under the conditions of this experiment. Our observation that the strain-specific character of the song profile persisted even after changing the developmental auditory environment suggests that the structure of these courtship sound sequences is under strong genetic control. Thus, the usefulness of mouse “song” as a model of mammalian vocal learning is limited, but mouse song has the potential to be an indispensable model to study genetic mechanisms for vocal patterning and behavioral sequences.
Many animals, including humans, use vocal signals to communicate with conspecifics.
Song is a long, complex vocalization of several acoustic elements arranged in
specific sequences
The mouse,
The B6D2F1strain of male mice showed individual differences in syllable usage and the
temporal structure of their songs as reported by Holy and Guo
To elucidate genetic and environmental effects on mouse songs, we conducted a cross-fostering study to understand the effects of the social experience during the juvenile developmental period on song development. First, we compared 2 strains of inbred C57BL/6 and BALB/c males and found that these 2 strains of male mice emitted a different pattern of songs with regard to frequency, inter-syllable intervals, and syllable composition. C57BL/6 males showed a higher peak frequency of syllables and more frequency-modulated syllables with 1 or multiple jumps and short- and upward syllables, whereas BALB/c males produced more chevron-, flat-, and harmonics-syllables. None of these strain-specific parameters were affected by cross-fostering. Therefore, developmental social environments appear to have no significant role in adult male songs of mice. In other words, mouse songs do not seem to involve imitative learning.
When a male subject encountered a female, he emitted complex ultrasounds.
Sound spectrograms demonstrated that B6 males showed a peak at 70–80
kHz, and BALB males at 50–60 kHz (
(a) Sound spectrograms of ultrasonic songs in B6 (upper) and BALB (lower) male mice. B6 males showed a higher peak frequency of syllables ranging from 70–110 kHz, shorter intervals between syllables, and more upward frequency modulations with jumps (arrows), whereas BALB males produced more “chevron” and “harmonics” syllables (arrow head). (b) The mean syllable peak frequency and inter-syllable interval significantly differed between B6 and BALB mice, but syllable duration was not. Data are expressed as mean ± SEM; *p<0.05 between strains. (c) Pie graphs showing percentages of the 10 categories of song syllables in B6 and BALB mice. Percentages were calculated in each strain as the number of syllables in each category for each subject/total number of syllables analyzed in each subject. The number of total syllables analyzed was: 6179 for B6 mice and 6244 for BALB mice. B6 mice produced more “short,” “one jump,” and “more jumps” syllables than BALB mice, whereas BALB mice produced more “flat”, “chevron”, “complex”, and “harmonics” syllables; *p<0.05 between strains. (d) In the sequential analysis, we divided all syllable types into 2 categories, namely, A (syllables with frequency jumps) and B (syllables without jumps). Z indicates silent gaps longer than 0.25 s. Circles represent the percentage of syllable types, and the thickness of the arrows represents the transition probabilities. The sequential analyses of syllables demonstrated strain-specific patterns; B6 mice showed more transition from A to A, A to B, A to Z, B to A, and Z to A than BALB mice and BALB mice showed more B to B self transition compared to that in B6 mice; *p<0.05 between strains.
According to previous studies
The sequential patterns of B6 and BALB mice songs are shown in
Sound spectrograms demonstrated that B6-sons and B6-foster males showed a
peak at 70–80 kHz, whereas BALB mice showed a peak at 50–60 kHz
(
Sonograms of ultrasonic songs recorded from B6-son, B6-foster, BALB-son, and BALB-foster male mice. Cross-fostered mice showed similar patterns to those of normally reared mice, and the effects of the rearing environment were not obvious. B6-son and B6-foster mice showed a higher peak frequency of syllables, shorter intervals between syllables, and more upward frequency modulations with jumps (arrows), whereas BALB-son and BALB-foster males produced more “chevron” and “harmonics” syllables (arrow head).
We compared songs between fostered groups, and found that the main strain
differences we quantified were not affected by fostering. BALB
cross-fostered males still showed a lower peak frequency
(F(1,20) = 106.5, p<0.0001) and longer
inter-syllable intervals (F(1,20) = 9.67, p<0.01)
than B6-fostered males (
Song parameters in B6-son, B6-foster, BALB-son, and BALB-foster male mice. The distribution histogram of the peak frequency (a) and intervals (b), but not the duration (c), of the syllables demonstrated significant strain differences, regardless of the fostering. Mean peak frequency (d) and interval (e) significantly differed between genetic B6 and BALB groups. Data are expressed as mean ± SEM; *p<0.05 vs. B6-son and B6-foster mice.
MANOVA revealed a significant effect of strain
(F(9,9) = 25.9, p<0.0001), but not of fostering
(F(9,9) = 0.91, p = 0.55) or an
interaction of strain and fostering (F(9,9) = 0.41,
p = 0.89). Regardless of fostering experience, B6 mice
produced more “short,” “one jump,” and “more
jumps” syllables than BALB mice (
The appearance ratio of each of the 10 syllable categories in B6-son, B6-foster, BALB-son, and BALB-foster mice. Genetic B6 groups produced more “short,” “one jump,” and “more jumps” syllables than BALB/c mice, whereas genetic BALB groups produced more “flat,” “chevron,” “complex,” and “harmonics” syllables. Data are expressed as mean ± SEM; *p<0.05 vs B6-son and B6-foster mice.
Pie graphs showing the percentages of the 10 categories of song syllables in B6-son (a), BALB-son (b), B6-foster (c), and BALB-foster (d) mice. Percentages were calculated in each strain as the number of syllables in each category for each subject/total number of syllables analyzed in each subject. The total syllables determined are as follows: 5487 syllables; B6-son; 6414 syllables, B6-foster; 4973 syllables, BALB-son; 6963 syllables, BALB-foster.
Regardless of fostering, B6 and BALB mice showed distinct transitional
patterns of the song syllables, and these characteristics were displayed by
cross-fostered males. MANOVA revealed a strain difference
(F(6,12) = 24.6, p<0.0001), but no fostering effect
(F(6,12) = 0.655, p = 0.687) and
no interaction between these (F(6,12) = 1.56,
p = 0.241). A Bonferonni post hoc test revealed that
sons of BALB mice showed a greater occurrence of B to B self-transitions, B
to Z and Z to B transitions as well as a lower occurrence of A to A
self-transitions, A to B, B to A, A to Z, and Z to A transitions compared to
sons of B6 and B6-foster male mice (p<0.05,
Sequential analyses of syllables demonstrated strain-specific patterns; BALB-son mice showed a greater occurrence of B to B self transitions, and B to Z and Z to B transitions, as well as a lower occurrence of A to A self-transitions and A to B, B to A, A to Z, and Z to A transitions compared to B6-son and B6-foster mice. BALB-foster mice demonstrated greater occurrence of type B to B self-transitions and a lower occurrence of A to A, A to B, B to A, A to Z, and Z to A transitions compared to B6-son and B6-foster mice. Circles represent the percentage of syllable types, and the thickness of the arrows represents the transition probabilities; *p<0.05 vs. B6-son and B6-foster mice.
In the present study, we revealed that B6 and BALB male mice showed distinct patterns
and sound profiles of songs when encountering a female. Our syllable categories are
similar to those reported in earlier studies
Studies of the natural history of mice have demonstrated that a pair of male and
female mice lives in a nest together with their juveniles
Several studies have demonstrated that female mice show attraction to male songs
Recent studies have demonstrated that ultrasonic vocalization of mouse pups is
affected by genes related to neuropsychiatric disorders such as Autism
Here we showed that imitative vocal learning is not involved in the strain
specificity of mouse songs. Vocal learning requires two independent processes.
First, the animal must have voluntary control over the vocal output. Second, the
animal should be able to match its vocal output with the externally acquired
auditory memory. For the first process, the existence of the direct motor pathway
connecting the oro-facial motor cortex and the medullar phonatory and respiratory
areas, including the nucleus ambiguus, has been suggested as an anatomical substrate
responsible for vocal plasticity
Arriaga et al. reported singing-related gene expression in mice cingulated, motor
cortex and basal ganglia
Our results show that the auditory environment does not affect song phenotypes in mice, and, thus, vocal learning does not appear to be involved in mouse songs. Nevertheless, mouse song is a very complex behavior, with at least 10 categories of vocal tokens and complex note-to-note transition rules. Even if this phenotype is largely controlled by genetic factors and only limited learning is involved, we can still pose interesting questions regarding the genetic encoding of acoustic categories and the neural mechanisms involved in sequence generation. Thus, the mouse song should remain an important model in which to study the biological basis of complex communicative behavior, including spoken human language.
BALB/cAJcl (BALB) and C57BL/6JJcl (B6) mice were originally obtained from Japan Clea Co. Ltd. (Japan Clea, Yokohama, Japan) and bred in our laboratory. Food and water were given ad libitum, and all the animals were kept at a constant temperature (23±1°C) and humidity (40%±10%) under a 12-h light:dark cycle (light on at 0600). All experiments were conducted in accordance with the guideline of the "Policies Governing The Use of Live Vertebrate Animals" by Azabu University, and were approved by The Ethical Committee for Vertebrate Experiments (ID# 070418).
A male and a female mouse of the same strain were pair-housed in a cage (17.5 cm
× 24.5 cm × 12.5 cm) for breeding. When the female was pregnant,
delivery was examined every 6–8 hours. When newly born pups were found at
the same time in both strains of parents, a part of the litter was reciprocally
cross-fostered to parents of the other strain of mice (B6-foster and
BALB-foster). The control mice were handled in the same manner as fostered pups
but returned to their own parents (B6-son and BALB-son). All litters were left
undisturbed until weaning (postnatal day (PD) 21). After PD21, they were housed
with males of the non-cross fostered controls of the different strain until
ultrasound recording at 10–20 weeks of age (
This figure illustrates the case of cross-fostering from BALB to B6. When newly born pups were found at the same time in both strains of parents, a part of the litter was reciprocally cross-fostered to parents of the other strain of mice. The control mice were handled in the same manner as fostered pups but returned to their own parents. All litters were left undisturbed until weaning (PD21). After weaning,they were housed with males of the non-cross fostered controls of the different strain until ultrasound recording at 10–20 weeks of age.
All experiments were carried out in a soundproof chamber (Muromachi Kikai, Tokyo,
Japan) under a red dim light, from 1300 to 1700 hours. Ultrasonic sounds were
detected using a condenser microphone (UltraSoundGate CM16/CMPA, Avisoft
Bioacoustics, Berlin, Germany) designed for recordings between 10 and 200 kHz.
The microphone was connected to an A/D converter (UltraSoundGate 116, Avisoft
Bioacoustics, Berlin, Germany) with a sampling rate of 300 kHz and acoustic
signals were transmitted to a sound analysis system (SASLab Pro, Avisoft
Bioacoustics, Berlin, Germany). During the recording, a subject male mouse was
individually housed in a test cage (12.5 cm × 20.0 cm × 11.0 cm) and
kept there for at least 2 h for habituation. The test cage was placed in the
soundproof chamber, and a female mouse, devocalized by unilateral sectioning of
the inferior laryngeal nerve
Spectrograms were generated with an FFT-length of 1024 points and a time-window overlap of 75% (100% frame, Hamming window). The spectrogram was produced at a frequency resolution of 488 Hz and a time resolution of 1 ms. A lower cut-off frequency of 20 kHz was used to reduce background noise outside the relevant frequency band. Parameters analyzed for each subject included the number of syllables, duration of syllables, and qualitative and quantitative analyses of sound frequencies measured in terms of frequency at the maximum of the spectrum.
Waveform patterns of calls collected from every group (B6, 6179 syllables; BALB,
6244 syllables; B6-son, 5487 syllables; B6-foster, 6414 syllables; BALB-son,
4973 syllables; BALB-foster, 6963 syllables) were analyzed in detail. Each
syllable was identified as 1 of 10 distinct categories, based on internal pitch
change, length, and shape, according to previously reported categories with
minor modifications (
The prevalence of a syllable type was defined as follows on the basis of a
previous study
(TIF)
(TIF)
(WAV)
(WAV)
(WAV)
(WAV)
(WAV)
(WAV)
We thank Drs. Erich Jarvis and Gustavo Arriaga of Duke University for generously showing their preliminary data and providing the discussion on this experiment. We are grateful to Professor Björn Brembs for suggestions and encouragement and to Dr. Olga Feher and Benjamin Treuhaft for English proof reading.