A Foxp2 Mutation Implicated in Human Speech Deficits Alters Sequencing of Ultrasonic Vocalizations in Adult Male Mice

Abstract

Development of proficient spoken language skills is disrupted by mutations of the FOXP2 transcription factor. A heterozygous missense mutation in the KE family causes speech apraxia, involving difficulty producing words with complex learned sequences of syllables. Manipulations in songbirds have helped to elucidate the role of this gene in vocal learning, but findings in non-human mammals have been limited or inconclusive. Here, we performed a systematic study of ultrasonic vocalizations (USVs) of adult male mice carrying the KE family mutation. Using novel statistical tools, we found that Foxp2 heterozygous mice did not have detectable changes in USV syllable acoustic structure, but produced shorter sequences and did not shift to more complex syntax in social contexts where wildtype animals did. Heterozygous mice also displayed a shift in the position of their rudimentary laryngeal motor cortex (LMC) layer-5 neurons. Our findings indicate that although mouse USVs are mostly innate, the underlying contributions of FoxP2 to sequencing of vocalizations are conserved with humans.

Keywords: FoxP2; KE family; song; speech apraxia; syntax; ultrasonic vocalizations.

Figure 1

Mouse song system anatomy and syllable types. (A) Proposed anatomy of the rudimentary mouse forebrain vocal communication circuit based on Arriaga et al. (2012). Not shown are other connected brainstem regions, the amygdala, and insula. (B) Comparison with human, based on Arriaga et al. (2012) and Pfenning et al. (2014). (C) Comparison with songbird. (D) Sonograms of examples syllables of the four syllable categories quantified from a C57 male mouse USV song, labeled according to pitch jumps. Anatomical abbreviations: ADSt, anterior dorsal striatum; Amb, nucleus ambiguous; ASt, anterior striatum; aT, anterior thalamus; Av, nucleus avalanche; HVC, a letter-based name; LArea X, lateral Area X; LMO, lateral mesopallium oval nucleus; LMAN, lateral magnocellular nucleus of the nidopallium; LMC, laryngeal motor cortex; LSC, laryngeal somatosensory cortex; M1, primary motor cortex; M2, secondary motor cortex; NIf, interfacial nucleus of the nidopallium; PAG, periaqueductal gray; RA, robust nucleus of the arcopallium; T, thalamus; VL, ventral lateral nucleus of the thalamus; XIIts, 12th vocal motor nucleus, tracheosyringeal part.

Figure 2

Syllables production rate and repertoire composition across contexts. (A) Syllable production rate for wildtype (n = 8) and FoxP2-R552H heterozygous (n = 10) mice in each context. Data are presented as mean ± SEM. ^*p < 0.05, ^**p < 0.005 for post-hoc Student's paired t-test after Benjamini-Hochberg correction. (B) Repertoire compositions of the four major syllable categories in each context. ^*p < 0.05 repeated measure ANOVA across contexts for a given syllable type and genotype.

Figure 3

Temporal organization of sequences in different contexts. (A) Distribution of the inter-syllables intervals, for the four context (colors), defining three types of silent intervals (horizontal red dashed lines) in sequences of syllables for wildtype (n = 8) and FoxP2-R552H heterozygous (n = 10) mice. (B) Sonogram showing example syllable sequence intervals of a C57 wildtype male.

Figure 4

Sequence measures for each context. (A) Ratio of complex song syllable sequences over simple songs in each context. Graphed are the number of sequences with two or more complex “m” syllables divided by the number of sequences with one or no “m” syllables in each context. Sequences with two syllables or less were not included. (B) Lengths of syllable sequences. Graphed are the average number of syllables per sequence, regardless of the total length of the syllables or sequence in seconds. Data are presented as mean ± SEM. ^*p < 0.05 using Wilcoxon-Mann-Whitney tests for independent samples (n = 8 WT; 10 heterozygous). The values for the AF group approached significance. (C,D) Example sonograms of longer complex and shorter simple syllable sequence differences between wildtype and Foxp2-R552H heterozygous mice, respectively, in the LF context.

Figure 5

Correlations between syllable sequence length and syllable rate across context. Shown are correlations in wildtype (n = 8) and Foxp2-R552H heterozygous (n = 10) mice in the four context: (A) Fresh female urine (UF); (B) Live female (LF); (C) Anesthetized female (AF); and (D) Anesthetized male (AM). The x-axes are not drawn to the same scale, since the males produced greater differences in ranges of syllable production rates than sequence lengths (y-axes) across contexts. The correlations in the AM context (D) are still significant when removing from the analyses animals that did not sing (0 syllables; +/+R = 0.976, p = 0.005; +/R552H R = 0.988, p = 0.0001). Statistics are Spearman's correlation.

Figure 6

Syntax analyses. (A–C) Diagrams representing conditional probabilities (for those produced at p = 0.05 or greater) of syllable transitions within song sequences in each context and genotype. Arrow thickness is proportional to probability value of going from one syllable type to another (averaged from n = 8 WT; 10 heterozygous males). Red colored arrows are transitions produced by wildtype in the LF context that add to increased diversity. (D–F) Heat map distributions of the statistical probabilities of differences between wildtype and Foxp2-R552H heterozygous mice for each transition type across contexts. For each of the 24 transition types we tested whether the corresponding group-specific distributions are equal between genotype (WMW). Combined p-values returned by these “local” WMW tests provide test statistics and p-values for testing the differences in the transition probabilities to (columns) and from (rows) different syllables. The individual cells within correspond to transitions from (start) a given syllable type to (end) a given syllable type. Figure S3 shows the to (columns) and from (rows) p-values for multiple tests using Benjamini-Hochberg correction.

Figure 7

Syntax comparisons across contexts. (A–C) Heat maps distributions of the statistical probabilities of differences between (A) UF and LF, (B) UF and AF, (C) LF and AF for wildtypes (WT; left columns) and Foxp2-R552H/+ heterozygotes (right columns). For each of the 24 transition types, we tested whether the corresponding group-specific distributions are equal between contexts (See Data Sheet 1 in Supplementary Material). Combined p-values returned by these “local” tests provide statistics and p-values for testing the differences in the transition probabilities to (columns) and from (rows) different syllables. The individual cells within correspond to transitions from (start) of a given syllable type to (end) of a given syllable type.

Figure 8

Retrograde tracing of the laryngeal motor cortex neurons. (A) Example of GFP-labeled (green) layer 5 neurons in mouse LMC-M1 from a pseudorabies virus (PRV) unilateral injection in the cricothyroid and cricoarytenoid lateralis larynx muscles (diagram to right) of a C57 male mouse. Roman numbers correspond to different layers of the cortex as determined in DAPI counterstaining. Section is coronal, contralateral hemisphere to muscle injection. Scale bar, 500 μm. Left image schematic from (Arriaga et al., 2015). (B) Total number of PRV-GFP positive cells labeled from all rostral to caudal coronal sections processed in wildtype and Foxp2-R552H heterozygous mice. No significant difference was found (p = 0.42; Wilcoxon-Mann-Whitney tests for independent samples). (C) Example double labeling of GFP-backfilled (green) LMC layer 5 neurons and Foxp2 protein expression (red Cy3). Layer 6, as known (Hisaoka et al., 2010), has the highest numbers of neurons with Foxp2 expression, followed by layer 5 in this particular region of the cortex. Arrow, example doubled labeled cell with intermediate levels of Foxp2 expression; arrowhead, example non-backfilled layer 5 cell with high Foxp2 expression. (D) Distribution, section-by-section, of the PRV positive cells in both genotypes. Data are presented as mean ± SEM normalized per number of section counted for wildtype and Foxp2-R552H heterozygous mice. Kolmogorov-Smirnov test was used to assess the difference between the two distributions (n = 6 males per genotype). Anatomical coronal diagrams below the graph show representative locations with coordinates relative to Bregma indicated; images used from The Mouse Brain in Stereotaxic Coordinates, Paxinos G. and Franklin K. B. J. with permission.