Differential impacts of Cntnap2 heterozygosity and Cntnap2 null homozygosity on axon and myelinated fiber development in mouse

Abstract

Over the last decade, a large variety of alterations of the Contactin Associated Protein 2 (CNTNAP2) gene, encoding Caspr2, have been identified in several neuronal disorders, including neurodevelopmental disorders and peripheral neuropathies. Some of these alterations are homozygous but most are heterozygous, and one of the current challenges is to estimate to what extent they could affect the functions of Caspr2 and contribute to the development of these pathologies. Notably, it is not known whether the disruption of a single CNTNAP2 allele could be sufficient to perturb the functions of Caspr2. To get insights into this issue, we questioned whether Cntnap2 heterozygosity and Cntnap2 null homozygosity in mice could both impact, either similarly or differentially, some specific functions of Caspr2 during development and in adulthood. We focused on yet poorly explored functions of Caspr2 in axon development and myelination, and performed a morphological study from embryonic day E17.5 to adulthood of two major brain interhemispheric myelinated tracts, the anterior commissure (AC) and the corpus callosum (CC), comparing wild-type (WT), Cntnap2 ^-/- and Cntnap2 ^+/- mice. We also looked for myelinated fiber abnormalities in the sciatic nerves of mutant mice. Our work revealed that Caspr2 controls the morphology of the CC and AC throughout development, axon diameter at early developmental stages, cortical neuron intrinsic excitability at the onset of myelination, and axon diameter and myelin thickness at later developmental stages. Changes in axon diameter, myelin thickness and node of Ranvier morphology were also detected in the sciatic nerves of the mutant mice. Importantly, most of the parameters analyzed were affected in Cntnap2 ^+/- mice, either specifically, more severely, or oppositely as compared to Cntnap2 ^-/- mice. In addition, Cntnap2 ^+/- mice, but not Cntnap2 ^-/- mice, showed motor/coordination deficits in the grid-walking test. Thus, our observations show that both Cntnap2 heterozygosity and Cntnap2 null homozygosity impact axon and central and peripheral myelinated fiber development, but in a differential manner. This is a first step indicating that CNTNAP2 alterations could lead to a multiplicity of phenotypes in humans, and raising the need to evaluate the impact of Cntnap2 heterozygosity on the other neurodevelopmental functions of Caspr2.

Keywords: Caspr2; anterior commissure; axon diameter; corpus callosum; myelin thickness; neuronal activity; node of Ranvier; sciatic nerve.

FIGURE 1

Corpus callosum (CC) morphology in adult mice. Representative images of brain coronal sections (Bregma + 0.50) of wild-type (WT), HET, and KO P90 mice showing CC immunostained with anti-NF-M antibodies (A). CC thickness measured at different levels on coronal sections (B); inset, schematic representation of the levels (yellow lines) where CC thickness measurements were performed. CC thickness at the midline of the brains (C). Representative images of mid-sagittal brain sections of WT, HET, and KO P90 mice, showing the CC and the AC immunostained with anti-NF-M antibodies (D). CC area (E), minimum caliper (F), and maximum caliper (G) measured on mid-sagittal brain sections. (B,C,E–G) Six animals/genotype, (B) average of measurements on both hemispheres on three consecutive sections/animal, (C,E–G) average of measurements on three consecutive sections/animal. Statistical tests: (B) Two-way RM ANOVA test; (C,E–G) Unpaired t-test to compare HET or KO mice to WT mice, and one-way ANOVA test to compare the three genotypes; *P < 0.05, ^**P < 0.01, ^***P < 0.001.

FIGURE 2

Corpus callosum (CC) morphology during development. CC area, minimum caliper, and maximum caliper measured on mid-sagittal brain sections of wild-type (WT), HET, and KO P30 mice (A) and P7 pups (B). Representative images of brain coronal sections of WT, HET, and KO P2 pups (C) and E17.5 embryos (F) showing CC immunostained with anti-L1-CAM antibodies; the levels where CC thickness measurements were performed are schematized on the images of WT mice (yellow lines). CC thickness measured on coronal sections of P2 pups (D) and E17.5 embryos (G) at the levels indicated in panels (C,F). CC thickness at the midline of the brains of P2 pups (E) and E17.5 embryos (H). (A,B,D,E,G,H) Six animals/genotype, (A,B) average of measurements on three consecutive sections/animal, (D,G) average of measurements on both hemispheres on one section/animal, (E,H) measurements on one section/animal. Statistical tests: Unpaired t-test (A,B,E) or Mann–Whitney test (H) to compare HET or KO mice to WT mice, and one-way ANOVA test (A,B,E) or Kruskal–Wallis test (H) to compare the three genotypes; (D,G) Two-way RM ANOVA test followed by a Tukey’s multiple comparisons test (G); *P < 0.05, ^**P < 0.01.

FIGURE 3

Anterior commissure (AC) morphology during development and at adulthood. Representative phase-contrast images of horizontal brain sections of wild-type (WT), HET, and KO P90 mice, showing the two branches of the AC (A); yellow lines, levels where the thickness of the anterior branch (ACa) was measured. ACa thickness measured on horizontal brain sections of P90 mice (B). Representative images of mid-sagittal brain sections of WT, HET, and KO mice at P30, P7, and E17.5, showing AC immunostained with anti-NF-M (P30) or L1CAM (P7, E17.5) antibodies (C). AC area measured on mid-sagittal brain sections of P90 (D) and P30 (E) mice, P7 pups (F), and E17.5 embryos (G). (B,D–F) Six animals/genotype, (G) four animals/genotype, (B) average of measurements on both hemispheres on one section/animal, (D–G) average of measurements on three consecutive sections/animal. Statistical tests: Unpaired t-test (B,D,F,G) or Mann–Whitney test (E) to compare HET or KO mice to WT mice, and one-way ANOVA test (B,D,F,G) or Kruskal–Wallis test (E) to compare the three genotypes; *P < 0.05, ^**P < 0.01, ^****P < 0.0001.

FIGURE 4

Ultrastructural abnormalities of corpus callosum (CC) myelinated fibers in P30 mice. Electron micrographs of transversal sections of the CC at brain midline, from wild-type (WT), HET, and KO P30 mice (A). Axonal diameters of CC myelinated fibers (B). Cumulative frequency distribution of axonal diameters (in %) (C). Scatter plot graph displaying G-ratios of individual myelinated axons as a function of the respective axon diameters, and linear regression of the G-ratio measurements for each genotype (D). Percentage of myelinated axons containing mitochondria (E). (B–D) 450 myelinated fibers/genotype, 3 mice/genotype, 150 myelinated fibers/mouse; (E) 3 mice/genotype, % of mitochondria counted on ∼2000 axons/mouse. Statistical tests: (B) Mann–Whitney test to compare HET or KO mice to WT mice, and Kruskal–Wallis test to compare the three genotypes; (C) Kolmogorov–Smirnov test for pairwise comparisons; (D) Linear regression (WT R² = 0.2102, HET R² = 0.2871, KO R² = 0.1959) for pairwise comparisons, HET vs. WT elevation P < 0.0001, HET vs. KO elevation P = 0.0005; (E) Unpaired t-test to compare HET or KO mice to WT mice, and one-way ANOVA test to compare the three genotypes; *P < 0.05, ^**P < 0.01, ^****P < 0.0001; ns, non-significant.

FIGURE 5

Ultrastructural abnormalities of anterior branch (ACa) and posterior branch (ACp) myelinated fibers in P30 mice. Electron micrographs of transversal sections of the ACa (A) and ACp (F) at brain midline, from wild-type (WT), HET, and KO P30 mice; insets, semi-thin transversal sections of the AC stained with toluidine blue, showing the ACa and the ACp. Axonal diameters of ACa (B) and ACp (G) myelinated fibers. Cumulative frequency distribution of axonal diameters (in %) of ACa (C) and ACp (H) myelinated fibers. Scatter plot graphs displaying G-ratios of individual myelinated axons as a function of the respective axon diameters, and the linear regression of the G-ratio measurements for each genotype, for the ACa (D) and the ACp (I). Percentage of myelinated axons containing mitochondria in the ACa (E) and the ACp (J). (B–D,G–I) 441 myelinated fibers/genotype, 3 mice/genotype, 147 myelinated fibers/mouse; (E) 3 mice/genotype, % of mitochondria counted on ∼2000 axons/mouse; (J) 3 mice/genotype, % of mitochondria counted on ∼1000 axons/mouse. Statistical tests: (B,G) Mann–Whitney test to compare HET or KO mice to WT mice, and Kruskal–Wallis test to compare the three genotypes; (C,H) Kolmogorov–Smirnov test for pairwise comparisons; (D) Linear regression (WT R² = 0.08121, HET R² = 0.3458, KO R² = 0.1727) for pairwise comparisons, HET vs. WT slope P = 0.0024, HET vs. KO slope P = 0.0003; (I) Linear regression (WT R² = 0.1589, HET R² = 0.1062, KO R² = 0.3420) for pairwise comparisons, HET vs. WT elevation P = 0.0047, KO vs. WT slope P < 0.0001, HET vs. KO slope P < 0.0001; (E,J) Unpaired t-test (E) or Mann–Whitney test (J) to compare HET or KO mice to WT mice, and one-way ANOVA test (E) or Kruskal–Wallis test (J) to compare the three genotypes; ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001; ns, non-significant.

FIGURE 6

Anterior commissure (AC) and corpus callosum (CC) axon diameter abnormalities in P7 pups. Electron micrographs of transversal sections of the anterior branch (ACa), posterior branch (ACp), and CC at brain midline, from wild-type (WT), HET, and KO P7 pups; insets, semi-thin transversal sections of the AC stained with toluidine blue, showing the ACa and the ACp; the green asterisk indicates a growth cone (A). Axonal diameters of the ACa (B), ACp (C), and CC (D). Cumulative frequency distribution of axonal diameters (in %) of the ACa (E), ACp (F), and CC (G). (B,C,E,F) 900 axons/genotype, 3 pups/genotype, 300 axons/pup; (D,G) 1080 axons/genotype, 3 pups/genotype, 360 axons/pup. Statistical tests: (B–D) Mann–Whitney test to compare HET or KO mice to WT mice, and Kruskal–Wallis test to compare the three genotypes; (E–G) Kolmogorov–Smirov test for pairwise comparisons; *P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001; ns, non-significant.

FIGURE 7

Intrinsic excitability in cortical pyramidal cells from P10–P12 pups. Sample spike trains evoked by a 100-pA (left) or a 300-pA somatic current injection (right) in a WT (top), HET (middle) and KO (bottom) neuron (A). Mean f-i curve for WT (n = 27 cells from three mice), HET (n = 24 cells from three mice), and KO (n = 27 cells from three mice) (B). Resting membrane potential of the recorded neurons (C). Statistical tests: (B) Two-way RM ANOVA followed by Sidak’s post-hoc test; (C) One-way ANOVA to compare the three genotypes; ^**P < 0.01, ^***P < 0.001.

FIGURE 8

Caspr2, TAG-1, and Cntnap2 mRNA levels in mouse brains during development. Representative immunoblots showing Caspr2, TAG-1, and GAPDH levels in brain lysates of wild-type (WT), HET, and KO mice at different developmental stages and at adulthood (A). Levels of Caspr2 normalized to GAPDH levels and relative to mean WT (in %) (B). Levels of Cntnap2 mRNAs normalized to the Psap gene and relative to mean WT (ratio) in brain of WT, HET, and KO mice at E17.5, P7, P10, and P30 (C). Levels of TAG-1 normalized to GAPDH levels and relative to mean WT (in %) (D). (B–D) Six animals/genotype/age. Statistical tests: (B) Mann–Whitney test to compare Caspr2 levels in HET mice at stages E17.5, P2, P7, P10, P15, or P30, to Caspr2 level in HET mice at P90. (C) Unpaired t-test to compare Cntnap2 mRNA levels in HET mice at stages E17.5, P7, or P10, to Cntnap2 mRNA level in HET mice at P30. (D) One-way ANOVA test or Kruskal–Wallis test to compare the three genotypes at different ages; *P < 0.05, ^**P < 0.01, ^****P < 0.0001; ns, non-significant.

FIGURE 9

Myelinated fiber abnormalities in sciatic nerves of adult mice. Electron micrographs of transversal sections of the sciatic nerve of wild-type (WT), HET, and KO adult mice, showing a single myelinated fiber for each genotype (A). Axonal diameters of myelinated fibers (B). Scatter plot graph displaying G-ratios of individual myelinated axons as a function of the respective axon diameters, and the linear regression of the G-ratio measurements for each genotype (C). Representative confocal images of immunostainings of sciatic nerves fibers from WT, HET, and KO adult mice, for nodal (NrCAM), paranodal (Caspr, F3), and juxtaparanodal (Caspr2, TAG-1, Kv1.2) proteins (D). Length of the nodes measured on NrCAM immunostainings (E). Scatter plot graph displaying length of individual node as a function of the respective nodal diameters, and the linear regression of the node length measurements for each genotype (F). Number of slips per 100 steps made by WT, HET, and KO adult males during the grid-walking test (G). (B,C) 141 myelinated fibers/genotype, 4 mice/genotype, 47 myelinated fibers/mouse, (E,F) 243 nodes/genotype, 3 mice/genotype, 81 nodes/mouse, (G) 8–9 mice/genotype. Statistical tests: (B,E) Kruskal–Wallis test to compare the three genotypes; (C) Linear regression (WT R² = 0.2093, HET R² = 0.1143, KO R² = 0.2022) for pairwise comparisons, HET vs. WT elevation P < 0.0001, HET vs. KO slope P = 0.019; (F) Linear regression (WT R² = 0.2295, HET R² = 0.1368, KO R² = 0.1935) for pairwise comparisons, HET vs. WT elevation P = 0.0052, KO vs. WT slope P = 0.0315, HET vs. KO elevation P < 0.0001; (G) one-way ANOVA test to compare the three genotypes; *P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001. Bar scales, 10 μm (E).

FIGURE 10

Summary of changes observed in HET and KO mice as compared to wild-type (WT) mice. Changes observed in the brain during development and in adulthood (A). Upper part, timeline of axonal development and myelination in mouse. Changes observed in the sciatic nerves and in the grid test performance (B). Arrows indicate increase (↑) or decrease (↓) in the parameters; ns, non-significant changes. The colors of the boxes underline the nature of the differences between HET and KO (color code at the bottom of the figure). The P-values are indicated when HET mice present an intermediate phenotype or tend to be more affected than KO mice (*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001). Corpus callosum (CC) E17.5, P-values for coronal sections; posterior branch (ACp) myelin P30, P-values for G-ratios.