Cadm1-expressing synapses on Purkinje cell dendrites are involved in mouse ultrasonic vocalization activity

Abstract

Foxp2(R552H) knock-in (KI) mouse pups with a mutation related to human speech-language disorders exhibit poor development of cerebellar Purkinje cells and impaired ultrasonic vocalization (USV), a communication tool for mother-offspring interactions. Thus, human speech and mouse USV appear to have a Foxp2-mediated common molecular basis in the cerebellum. Mutations in the gene encoding the synaptic adhesion molecule CADM1 (RA175/Necl2/SynCAM1/Cadm1) have been identified in people with autism spectrum disorder (ASD) who have impaired speech and language. In the present study, we show that both Cadm1-deficient knockout (KO) pups and Foxp2(R552H) KI pups exhibit impaired USV and smaller cerebellums. Cadm1 was preferentially localized to the apical-distal portion of the dendritic arbor of Purkinje cells in the molecular layer of wild-type pups, and VGluT1 level decreased in the cerebellum of Cadm1 KO mice. In addition, we detected reduced immunoreactivity of Cadm1 and VGluT1 on the poorly developed dendritic arbor of Purkinje cells in the Foxp2(R552H) KI pups. However, Cadm1 mRNA expression was not altered in the Foxp2(R552H) KI pups. These results suggest that although the Foxp2 transcription factor does not target Cadm1, Cadm1 at the synapses of Purkinje cells and parallel fibers is necessary for USV function. The loss of Cadm1-expressing synapses on the dendrites of Purkinje cells may be associated with the USV impairment that Cadm1 KO and Foxp2(R552H) KI mice exhibit.

Figure 1. Abnormal cerebellum development of Cadm1 KO.

(A) Wild-type, heterozygote, and homozygous Cadm1 KO mice. (P50) (B) The difference in mean weight between homozygous Cadm1 KO mice and their wild-type littermates (five each) was significant at P10 and increased over the next 20 days (A, B); at P30, the mean weight of the homozygous Cadm1 KO mice was 20–25% less than that of the wild-type mice. In addition, the brains of homozygous Cadm1 KO mice were smaller (C, n = 22), and the cerebellums of homozygous Cadm1 KO mice had an approximately 20% reduction in size and weight (D, n = 10). Bars in the graph indicate mean±standard error (SEM). Student's t-test (*p<0.05). Bars in the pictures indicate 1 cm (A), 5 mm (C), and 0.75 mm (D), respectively.

Figure 2. Analysis of ultrasonic vocalizations (USVs) of Cadm1 KO mice (P8).

(A) Real-time spectrography of the USVs by pups after separation from the dam. (B) Major vocalization patterns of Cadm1 KO and wild-type pups. Wild-type vocalization was mainly whistle-type USVs, but Cadm1 KO mice exhibited only a small number of click-type vocalizations. (C) The number of whistle-type USVs per min by pups. Vocalizations were recorded for 3 min. Experiments were done three times for 5 pups in each group, and an example of typical results is shown. Values are mean±standard error (SEM). Student's t-test (**p<0.01).

Figure 3. Distribution of Cadm1 in the cerebellum (P11).

The Cadm1 intensity preferentially distributed in an apical–distal dendritic portion. Wild-type (A, B, upper panel) and Cadm1 KO mice (B, lower panel). Green, Cadm1. Red, Calbindin. Blue, Hoechst. Bars, 30 µm.

Figure 4. Developmental changes of Cadm1, VGluT1, and VGluT2 in wild-type pups.

Alteration of the distribution of Cadm1, VGluT1, and VGluT2 was examined in the molecular layer of the developing cerebellum (P6–14). VGluT2 first appeared in the molecular layer in the early postnatal cerebellum (P6–8), in which Cadm1 co-localized with VGluT2, and then the level of VGluT2 decreased. VGluT1 increased in the later postnatal cerebellum (P11–14), in which Cadm1 co-localized with VGluT1. Green, Cadm1. Red, VGluT1 or VGluT2. Blue, Hoechst. Bar, 30 µm. Values are mean±standard error (SEM). Student's t-test (*p<0.05, **p<0.01). Pups: n = 3. Images: n = 8.

Figure 5. The influence of Cadm1 deficiency on the expression of synaptic proteins and Foxp2 in the cerebellum.

(A) Immunoblot analysis of the influence of the deficiency in the cerebellums of Cadm1 KO and wild-type pups (P10). An example of the typical immunoblotting results is shown. Values are mean±standard error (SEM). Student's t-test (*p<0.05, **p<0.01). Pups: n = 5. All experiments were performed three times. (B) RT-PCR analysis of the influence of the Cadm1 deficiency on the expression of Foxp2 in the cerebellum of wild-type and Cadm1 KO pups (P10). Values are mean±standard error (SEM). Pups: n = 5. All experiments were performed three times. A comparison showed no significant difference (Student's t-test; p<0.05).

Figure 6. Altered distribution of Cadm1 in the molecular layer of wild-type and Foxp2(R552H) KI mice (P11).

Cadm1 preferentially distributed in the apical–distal dendritic portion in the molecular layer. The immunoreactivity of Cadm1 as well as that of Synaptophysin and VGluT1, pre-synaptic markers, was decreased in the molecular layer of the Foxp2(R552H) KI mice. Green, Calbindin. Red, Cadm1, VGluT1, Synaptophysin. Blue, Hoechst. Bar, 30 µm.