A new knockin mouse carrying the E364X patient mutation for CDKL5 deficiency disorder: neurological, behavioral and molecular profiling

Abstract

CDKL5 deficiency disorder (CDD) is a rare neurodevelopmental syndrome caused by mutations in the X-linked CDKL5 gene. Hundreds of pathogenic variants have been described, associated with a significant phenotypic heterogeneity observed among patients. To date, different knockout mouse models have been generated. Here we present a new knockin CDKL5 mouse model carrying a humanized, well-characterized nonsense variant (c.1090G > T; p.E364X) described in the C-terminal domain of the CDKL5 protein in a female patient with a milder phenotype. Both male and female Cdkl5 ^E364X mice were analyzed. The novel Cdkl5 ^E364X mouse showed altered neurological and motor neuron maturation, hyperactivity, defective coordination and impaired memory and cognition. Gene expression analysis highlighted an unexpected reduction of Cdkl5 expression in Cdkl5 ^E364X mice brain tissues, with a significant increase in overall neuron-specific gene expression and an area-dependent alteration of astrocyte- and oligodendrocyte-specific transcripts. Moreover, our results showed that the loss of CDKL5 protein had the most significant impact on the cerebellum and hippocampus, compared to other analyzed tissues. A targeted analysis to study synaptic plasticity in cerebellum and hippocampus showed reduced Gabra1 and Gabra5 expression levels in females, whereas Gabra1 expression was increased in males, suggesting an opposite, sex-dependent regulation of the GABA receptor expression already described in humans. In conclusion, the novel Cdkl5^E364X mouse model is characterized by robust neurological and neurobehavioral alterations, associated with a molecular profile related to synaptic function indicative of a cerebellar GABAergic hypofunction, pointing to Gabra1 and Gabra5 as novel druggable target candidates for CDD.

Keywords: CDD; CDKL5; CRISPR/Cas9 knockin; GABA; Gabra1; Gabra5.

Fig. 1

Colony survival and neurological maturation (A) Experimental design of neurological and behavioral monitoring of mice colonies. (B, D) Survival curve in male (B) and female (D) colonies. Curve comparison (Mantel-Cox and Gehan-Breslow-Wilcoxon tests) showed no statistical significance. (C, E) Neurological maturation. Neuroscores in WT males (C, n = 41) and females (E, n = 11) reach their maximum values faster than Cdkl5^E364X littermates (males HEM (+/Y), n = 38; females HET (X/+), n = 25; females HOM (+/+), n = 36); two-way ANOVA, factors: genotype and time (for statistical analyses data, see Result section); asterisks represent the multiple comparison analysis for the factor “genotype” with post hoc Sidak's test (males; day 21, WT (X/Y) vs HEM (+/Y), p = 0.0383) or Dunnett's test (females; day 7, WT (X/X) vs HOM (+/+), p = 0.0202; day 9, WT (X/X) vs HET (X/+), p = 0.0004, WT (X/X) vs HOM (+/+), p < 0.0001; day 11, WT (X/X) vs HET (X/+), p = 0.0124, WT (X/X) vs HOM (+/+), p = 0.0009; day 15, WT (X/X) vs HET (X/+), p < 0.0001, WT (X/X) vs HOM (+/+), p < 0.0001; day 17, WT (X/X) vs HET (X/+), p = 0.0037, WT (X/X) vs HOM (+/+), p = 0.0023; day 19, WT (X/X) vs HET (X/+), p = 0.0073; day 21, WT (X/X) vs HOM (+/+), p = 0.0380). (F, H) Hindlimb clasping is shown in both males (F) and females (H). The clasping score was significantly higher in male HEM (+/Y) mice (n = 38) than male WT (X/Y) mice (n = 41) (mixed-effect with Gaussier-Greenhouse correction, post hoc Sidak's test which showed no statistical significance among groups; for statistical analysis values, see Results section). Both Cdkl5^E364X female genotypes (HET (X/+), n = 25; HOM (+/+), n = 35) show a similar altered trend compared to their WT counterparts (n = 15), reaching statistical significance at PND 11 (WT (X/X) vs HOM (+/+), p = 0.0214), 13 (WT (X/X) vs HOM (+/+), p = 0.0134) and 17 (WT (X/X) vs HOM (+/+), p = 0.0092) (asterisks represents multiple comparisons significance with post hoc Dunnett's test; for statistical analysis values, see Results section). (G, I) Cliff aversion reflex in males (G) and females (I), evaluated on day of appearance. In both sexes, this reflex appears later in Cdkl5^E364X than in WT. Males HEM (+/Y) (n = 38) vs WT (X/Y) (n = 41), females HOM (+/+) (n = 36) and HET (X/+) (n = 25) vs WT (X/X) (n = 11). For statistical analyses values, see Result section. Asterisks represent the result of t-test in males and of Kruskal-Wallis multiple comparisons with post hoc Dunn's test in females (WT (X/X) vs HOM (+/+), p = 0.0008). (L) Grip strength test in females, evaluated in seconds, from PND 7 to PND 21. Females HOM (+/+) (n = 36) and HET (X/+) (n = 25) vs WT (X/X) (n = 11). For statistical analyses values, see Result section. Asterisks represent the result of mixed-effect model with post hoc Dunnett's test (PND 7, WT (X/X) vs HET (X/+), p = 0.0023, WT (X/X) vs HOM (+/+), p = 0.015; PND 9, WT (X/X) vs HET (X/+), p = 0.0014, WT (X/X) vs HOM (+/+), p = 0.0067; PND 13, WT (X/X) vs HET (X/+), p = 0.0139, WT (X/X) vs HOM (+/+), p = 0.0010). (M) Grasping reflex in females evaluated as day of appearance. Females HOM (+/+) (n = 36) and HET (X/+) (n = 25) vs WT (X/X) (n = 11). For statistical analyses values, see Result section. Asterisks represent the result of one-way ANOVA with post hoc Dunnett's test (hind paws, WT (X/X) vs HOM (+/+), p = 0.0058). ∗p < 0.05; ∗∗p < 0.005; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. All data are expressed as mean ± SEM.

Fig. 2

Locomotion, sensorimotor coordination, learning and memory (A) Spontaneous locomotion in an open-field arena in all genotypes in male and female mice, expressed as distance traveled. Statistical analysis showed no significance between male genotypes (HEM (+/Y), n = 38; WT (X/Y), n = 45), while female HOM (+/+) (n = 34) showed a significantly higher activity than sex-matched WT (n = 20) (p = 0.0051); no difference was observed between HET (X/+) (n = 24) vs WT and vs HOM (for statistical analysis values, see Results section; asterisks represent the multiple comparison with Tukey's test). (B) Sensorimotor coordination in the rotarod test in all genotypes in male and female mice, expressed as latency to fall. Males, asterisks represent the results of Mann-Whitney test on WT (X/Y) (n = 41) vs. HEM (+/Y) (n = 40); females, asterisks represent the results of multiple comparisons one-way ANOVA with Tukey's test on WT (X/X) (n = 19) vs. HET (X/+) (n = 22) (p = 0.0371) and HOM (+/+) (n = 41) (p = 0.0013) (for statistical analysis values, see Results section). (C) Spatial memory evaluated by Y-maze test in male and female mice, expressed as % of correct alternations. Males, asterisks represent the results of Student's t-test on WT (X/Y) (n = 36) vs. HEM (+/Y) (n = 30); females, asterisks represent the results of multiple comparisons one-way ANOVA with post hoc Tukey's test on WT (X/X) (n = 23) vs. HET (X/+) (n = 21) (p = 0.0186) and HOM (+/+) (n = 28) (p = 0.0006). For statistical analysis values, see Results section. (D) Experimental protocol of the Morris water maze test: violet, pre-training test; blue, acquisition phase; red, probe trial. Schematic representation of the MWM pool, with example of a good performance trace and a bad performance one; the platform zone (blue) and the quadrant (or target) zone (green) are indicated. (E, F, G, H) Spatial and long-term memory evaluated by Morris water maze (MWM) in male and female Cdkl5^E364X mice. (E, F) Graphs show the acquisition phase of the test in males (E) and females (F), which measures the latency to reach the platform; (G, H) graphs show the selected parameters of the probe trial for males (G) and females (H), which are the number of entries in platform zone, the time spent in platform zone and the latency to the first entry in platform zone.Acquisition phase of males, the asterisk represents the result of multiple comparison two-way ANOVA with post hoc Sidak's test in WT (X/Y) (n = 8) vs. HEM (+/Y) (n = 6) (multiple comparison for the factor “genotype”, day 5, WT (X/Y) vs HEM (+/Y), p = 0.0258); acquisition phase of females, no significant comparisons (two-way ANOVA with post hoc Sidak's test) in WT (X/X) (n = 10) vs. HOM (+/+) (n = 9). Probe trial, number of entries in platform zone: for both males and females, the asterisks represent the results of Mann-Whitney test in WT (X/Y) (n = 8) vs. HEM (+/Y) (n = 6) and WT (X/X) (n = 10) vs. HOM (+/+) (n = 9), respectively. Probe trial, time in platform zone: the asterisks represent the results of Student's t-test in males WT (X/Y) (n = 8) vs. HEM (+/Y) (n = 6) and of Mann-Whitney test in female WT (X/X) (n = 10) vs. HOM (+/+) (n = 8). Probe trial, latency to first entry in platform zone: for both males and females, the asterisks represent the results of Student's t-test in males WT (X/Y) (n = 8) vs. HEM (+/Y) (n = 7) and females WT (X/X) (n = 10) vs. HOM (+/+) (n = 9), respectively.For statistical analysis values, see Results section.∗p < 0.05; ∗∗p < 0.005; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.All data are expressed as mean ± SEM.

Fig. 3

Expression of the murine Cdkl5 gene in peripheral tissues and in the central nervous system The graphs show Cdkl5 expression in (A) the testis of male WT (X/Y) (n = 5) and HEM (+/Y) (n = 5) mice; (B) the liver of male WT (X/Y) (n = 5) and HEM (+/Y) (n = 4), and female WT (X/X) (n = 5) and HOM (+/+) (n = 5) animals; (C) different areas of the CNS in male WT (X/Y) and HEM (+/Y), and female WT (X/X) and HOM (+/+) mice (five animals per group were included in the analysis). Student's t-test was performed between Cdkl5^E364X and WT animals of the same sex and asterisks represent the related statistical significance (∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). For statistical analysis values, see Results section. Data are represented as mean ± SEM. Relative gene expression was calculated using 2^−ΔΔCT method using the WT of each sex as control for each tissue of interest.Abbreviations: CTX, cerebral cortex; CB, cerebellum; HIP, hippocampus; SC, spinal cord; CNS, central nervous system.

Fig. 4

The p.E364X mutation abolishes the protein expression of CDKL5 in female HOM (+/+) brain (A–D) Representative low magnification images showing CDKL5 (red) immunofluorescence and DAPI staining (cyan) throughout the layers of the S1 cortex (A, B) and in the CA1 region of the hippocampus (C, D) of both WT (X/X) and HOM (+/+) female mice. (A1-B1). An unspecific secondary fluorescence antibody staining of myelinated fibers in the corpus callosum (CC) is visible (A, B). Higher magnifications of the boxed areas in A and B are shown (below) and arrows point to immunostained apical dendrites (proximal portion). (E) Western blotting showing the expression of both CDKL5 and actin in total lysates of cortex, hippocampus and cerebellum from both WT (X/X) and HOM (+/+) female mice. Asterisks indicate an unspecific band detected by the anti-CDKL5 antibody. Abbreviations: CTX, cerebral cortex; HIP, hippocampus; CB, cerebellum.

Fig. 5

Histological analysis and cell composition of brain areas in CDKL5^E364X-mutated and C57BL6 WT mice (A–F) Low-power micrographs of toluidine blue-stained sections, showing the cerebral cortex (A, B), hippocampus (C, D) and cerebellum (E, F) in female WT (X/X) (A, C, E) and HOM (+/+) (B, D, F) mice. (G, L) High-power micrographs of toluidine blue-stained sections, showing the CA1 hippocampal field (G, H), dentate gyrus (I, J), and cerebellar cortex (K, L). Arrows in (K) show Purkinje cells. (M − P) Graphs show the gene expression of lineage-specific genes representing neurons (Eno2, M), astrocytes (Gfap, N), oligodendrocyte precursor cells (Pdgfra, O), and myelinating oligodendrocytes (Mbp, P) in different areas of the CNS (separated by a black continuous vertical line) of male WT (X/Y) and HEM (+/Y) and female WT (X/X) and HOM (+/+) animals. Student's t-test was performed between Cdkl5^E364X and WT animals of the same sex, and asterisks represent the related statistical significance among the sexes (male, female) separated by a grey dotted vertical line (∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). For statistical analysis values, see Results section. Data are represented as mean ± SEM. Relative gene expression was calculated using 2-ΔΔCT method normalizing on the WT of each sex for each tissue of interest. Number of animals included in the analysis:Eno2 in CTX, males WT (X/Y) n = 5, HET (+/Y) n = 5; females, WT (X/X) n = 5, HOM (X/X) n = 3.Eno2 in CB, males WT (X/Y) n = 5, HET (+/Y) n = 3; females, WT (X/X) n = 5, HOM (X/X) n = 5.Eno2 in HIP, males WT (X/Y) n = 4, HET (+/Y) n = 5; females, WT (X/X) n = 5, HOM (X/X) n = 5.Eno2 in SC, males WT (X/Y) n = 5, HET (+/Y) n = 5; females, WT (X/X) n = 5, HOM (X/X) n = 5.GFAP, five animals were included in the analysis for each group in all tissues and genes tested.Pdgfra, CTX, males WT (X/Y) n = 5, HET (+/Y) n = 5; females, WT (X/X) n = 5, HOM (X/X) n = 4.Pdgfra in CB, males WT (X/Y) n = 5, HET (+/Y) n = 5; females, WT (X/X) n = 5, HOM (X/X) n = 5.Pdgfra in HIP, males WT (X/Y) n = 5, HET (+/Y) n = 4; females, WT (X/X) n = 5, HOM (X/X) n = 4.Pdgfra in SC, males WT (X/Y) n = 5, HET (+/Y) n = 5; females, WT (X/X) n = 5, HOM (X/X) n = 5.Mbp, five animals were included in the analysis for each group in all tissues and genes tested.Abbreviations: ACAv, anterior cingulate area, ventral part; ACAd, anterior cingulate area, dorsal part; Arb, arbor vitae; cc, corpus callosum; CA1, hippocampus, CA1 field; CA2, hippocampus, CA2 field; CB, cerebellum; Ccg, genu of corpus callosum; CP, caudoputamen; CTX, cerebral cortex; DG, hippocampus, dentate gyrus; DGlb, DG, lateral blade; DGmb, DG, medial blade; Eno2, enolase 2; Gfap, glial fibrillary acidic protein; HIP, hippocampus; ILA, infralimbic area; Mbp, myelin basic protein; mo, DG, molecular layer; Pdgfra, platelet derived growth factor receptor alpha; po, DG, polymorph layer; RSPv, retrosplenial area, ventral part; SC, spinal cord; sg, DG, granule cell layer; slm, CA1, stratum lacunosum-moleculare; so, CA1, stratum oriens; sp, CA1, pyramidal layer; sr, CA1, stratum radiatum.

Fig. 6

PCR array analysis and validation of genes involved in synaptic plasticity processes (A–B) Cluster analysis of the PCR array results on 84 genes involved in synaptic plasticity in the hippocampus (A) and cerebellum (B). The magnitude of expression of each gene across the four analyzed groups is indicated via a reference color scale (minimum, green; maximum, red). Clusterization analysis was performed on the basis of expression pattern similarity. (C–E) Scatter plots of all PCR array results in the hippocampus, analyzed as the relative expression of male WT (X/Y) compared to female WT (X/X) animals (C), male HEM (+/Y) compared to WT (X/Y) animals (D), and female HOM (+/+) compared to WT (X/X) animals (E). Only genes with a fold change >2 are indicated. (F–H) Scatter plots of the entire PCR array in the cerebellum, analyzed as the relative expression of male WT (X/Y) compared to female WT (X/X) animals (F), male HEM (+/Y) compared to WT (X/Y) animals (G), and female HOM (+/+) compared to WT (X/X) animals (H). Only genes with a fold change >2 are indicated. (I) Values of fold change for the Gabra5 gene expression in the hippocampus and cerebellum between the analyzed groups. (J–K) Relative expression of Gabra1 (J) and Gabra5 (K) in male HEM (+/Y) compared to WT (X/Y) and female HOM (+/+) compared to WT (X/X) animals. Student's t-test was performed between Cdkl5^E364X and WT animals of the same sex and asterisks represent the related statistical significance among the sexes (male, female) separated by a grey dotted vertical line (∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). For statistical analysis values, see Results section. Five animals per group were pooled for the PCR array analysis (A–I), while five animals per group as single samples were included in the qPCR analysis (J–K). qPCR analysis data are shown as a mean value ± SEM, and relative gene expression was calculated using 2-ΔΔCT method normalizing on the WT of each sex.Abbreviations: Adcy8, adenylate cyclase 8; Egr2, early growth response 2; Fos, FBJ osteosarcoma oncogene; Gabra1, gamma-aminobutyric acid (GABA) A receptor, subunit alpha 1; Gabra5, gamma-aminobutyric acid (GABA) A receptor, subunit alpha 5; Junb, Jun-B oncogene; Nr4a1, nuclear receptor subfamily 4, group A, member 1.