X-linked intellectual disability gene CASK regulates postnatal brain growth in a non-cell autonomous manner

Abstract

The phenotypic spectrum among girls with heterozygous mutations in the X-linked intellectual disability (XLID) gene CASK (calcium/calmodulin-dependent serine protein kinase) includes postnatal microcephaly, ponto-cerebellar hypoplasia, seizures, optic nerve hypoplasia, growth retardation and hypotonia. Although CASK knockout mice were previously reported to exhibit perinatal lethality and a 3-fold increased apoptotic rate in the brain, CASK deletion was not found to affect neuronal physiology and their electrical properties. The pathogenesis of CASK associated disorders and the potential function of CASK therefore remains unknown. Here, using Cre-LoxP mediated gene excision experiments; we demonstrate that deleting CASK specifically from mouse cerebellar neurons does not alter the cerebellar architecture or function. We demonstrate that the neuron-specific deletion of CASK in mice does not cause perinatal lethality but induces severe recurrent epileptic seizures and growth retardation before the onset of adulthood. Furthermore, we demonstrate that although neuron-specific haploinsufficiency of CASK is inconsequential, the CASK mutation associated human phenotypes are replicated with high fidelity in CASK heterozygous knockout female mice (CASK ((+/-))). These data suggest that CASK-related phenotypes are not purely neuronal in origin. Surprisingly, the observed microcephaly in CASK ((+/-)) animals is not associated with a specific loss of CASK null brain cells indicating that CASK regulates postnatal brain growth in a non-cell autonomous manner. Using biochemical assay, we also demonstrate that CASK can interact with metabolic proteins. CASK knockdown in human cell lines cause reduced cellular respiration and CASK ((+/-)) mice display abnormalities in muscle and brain oxidative metabolism, suggesting a novel function of CASK in metabolism. Our data implies that some phenotypic components of CASK heterozygous deletion mutation associated disorders represent systemic manifestation of metabolic stress and therefore amenable to therapeutic intervention.

Keywords: CASK; Cerebellar hypoplasia; MAGUK; Metabolism; Non-cell autonomous; X-linked intellectual disability.

Fig. 1

CASK deletion from Purkinje cells or granule cells does not alter cerebellar development. a Representative brain sections from two months old control mice carrying a floxed CASK gene (X^flCASK/Y) and mice with CASK knocked out from Purkinje cells (X^flCASK/Y; PCP2-cre); sections are stained with green (nucleic acid stain) and red (calbindin). Arrows indicate Purkinje cells. b Quantitation of molecular layer thickness and (c) Quantitation of Purkinje cell diameter; data are plotted as mean ± SEM, n = 3 mice. d, e, f Locomotion of mice analyzed in home cage using an overnight force plate experiment indicating no obvious change in locomotion. Average speed, average highest speed and average % activity are quantified and data are plotted as mean ± SEM, n = 3 mice. g Quantitation of retention time of mice on fixed speed rotarod, data are plotted as mean ± SEM, n = 4 mice. Floxed indicates CASK ^floxed mice and PCKO indicates Purkinje cell-specific CASK knockout mice. h Representative brain sections from two month old mice with indicated genotypes showing that CASK deletion from cerebellar granule cells does not lead to neuronal death or alteration in layering. LSL-tdTomato is a Cre indicator line in which a loxP-flanked stop cassette prevents expression of a red fluorescent protein; Math5-Cre mouse line has Cre recombinase expressed under the Math5 promoter. Arrows indicate inner granular layer where Cre-recombinase is expressed. Please note that the top two panels are cerebellar sections from mice with unperturbed CASK gene. i Representative brain images of two month old mice with indicated genotype. j Parasaggital sections of cerebellum showing that CASK deletion from granule cells does not lead to cerebellar hypoplasia. k Quantitation of cerebellar area in parasagittal sections; data are plotted as mean ± SEM, n = 3. Floxed indicates CASK ^floxed mice and GCKO indicates granule cell-specific CASK knockout mice

Fig. 2

Survival and normal brain formation in CASK neuronal knockout mice. a Shows the expected (grey bar) and observed (black bar) percentage of mice with each genotype obtained from crossing the X/Y;synapsin-Cre males with X^flCASK/X females; n = 62. Arrow indicates the neuronal CASK knockout group. b Weight of CASK ^floxed (floxed) and neuronal CASK knockout male mice (NKO) at 20 days of age. (* indicates p < 0.05; n = 5) (c) Representative brain halves and brain weight at 21 days of age from CASK (X^flCASK/Y) (floxed) and neuronal CASK knockout (NKO) mice. (* indicates p < 0.05; n = 4). d Ratio of brain to body weight of CASK (X^flCASK/Y) (floxed) and neuronal CASK knockout mice (NKO). (* indicates p < 0.05; n = 4). e Image showing hippocampus from 21 day old CASK ^floxed and CASK neuronal knockout mice. Green staining indicates nucleic acid and red staining indicates synaptophysin. Note that there is no significant change in the lamination and CA3 region synaptophysin staining in CASK neuronal knockout mice. f Representative Western blot showing endogenous levels of CASK, PSD95, synaptophysin and tubulin from 21 day old mice. g Western blot protein quantitation where data is normalized to tubulin and expressed relative to wild-type levels. Bar graphs are plotted as mean ± SEM. (* indicates p < 0.05; n = 3). Wild-type represents littermate mice with unperturbed CASK gene, floxed indicates CASK ^floxed mice and NKO indicates neuron-specific CASK knockout mice

Fig. 3

CASK ^(+/-) heterozygous mutant mice display postnatal microcephaly, cerebellar hypoplasia and optic nerve hypoplasia. a Representative brain images of the sex-matched CASK ^(+/+) and CASK ^(+/-) mutant littermates at postnatal day 1 and day 75 (P1 and P75), respectively. b Quantitation of brain weights from CASK ^(+/+) and CASK ^(+/-) mice at P1 and P75 (* indicates p < 0.05; n = 4). c Hematoxylin and eosin (H&E) stained sections of hippocampus derived from CASK ^(+/+) and CASK ^(+/-) mice at P75; note the hippocampi sizes are comparable. d H&E stained sections of the cerebellum at P75 showing pronounced cerebellar hypoplasia in CASK ^(+/-) mice relative to the CASK ^(+/+) control. e High magnification of the indicated square regions from panel D showing relatively fewer cells in the cerebellar folia of CASK ^(+/-) mice compared to the CASK ^(+/+) control. f Left panel shows representative images of the optic globe and optic nerve derived from three month old CASK ^(+/+) and CASK ^(+/-) mice . Right panel shows the optic nerve cross section stained with anti-tubulin antibody from CASK ^(+/+) and CASK ^(+/-) mice; note the decrease in optic nerve diameter . Scale bar = 100 μm. g Quantification of the cerebellar and hippocampal areas and the optic nerve and optic globe diameters obtained from CASK ^(+/+) and CASK ^(+/-) mice. Measurements were made using Image J software and normalized to the sex-matched wild-type littermate controls. Bar graphs are plotted as mean ± SEM; (* indicates p  < 0.05; n = 4)

Fig. 4

CASK ^(+/-) mice display musculoskeletal and locomotion defects. a Dissected torso from three month old CASK ^(+/+) and CASK ^(+/-) mice; CASK ^(+/-) mice display scoliosis. b CASK ^(+/-) mice display hind-limb clasping phenotype. c Activity of CASK ^(+/+) and CASK ^(+/-) mice measured in metabolic cages. Data plotted as mean ± SEM; (* indicates p < 0.05; n = 5) (d) CASK ^(+/-) mice were trained on a fixed speed rotorod and retention time noted. On each day, fresh sets of sex-and-age matched CASK ^(+/+) littermate mice were used as controls. Data are plotted as mean ± SEM; (* indicates p < 0.05; n = 5). e Representative 5-min trace of positions of CASK ^(+/+) and CASK ^(+/-) mice in home cage detected by a force plate. f Quantitation of locomotion observed on a force plate (area covered in 5 min/net distance traveled). Data are plotted as mean ± SEM; (* indicates p < 0.05; n = 5). All locomotion experiments were performed on mice between 25-30 days of age

Fig. 5

CASK is present in cytosol as a part of large protein complexes. a Western blot showing endogenous level of CASK, synaptophysin, ATP synthase β subunit and tubulin from CASK ^(+/+) and CASK ^(+/-) mutant mice (n = 3). b Western blot quantitation in CASK ^(+/-) mice relative to CASK ^(+/+) mice. Data is normalized to tubulin and expressed relative to wild-type levels. Bar graphs are plotted as mean ± SEM. (* indicates p < 0.05; n = 3). c Wild-type mice brains were initially fractionated to separate synaptosomal and cytosolic fractions. The synaptosomal fraction was solubilized initially in Triton X-100 and then in deoxycholic acid (DCA). Samples were blotted for the indicated antigen. d Wild-type mouse brain homogenate was solubilized in Triton X-114 at 4 °C, samples were warmed up to 37 °C and centrifuged to separate into an aqueous phase, a detergent phase and a pellet. Equal dilutions of each phase were blotted for the indicated proteins. e Aqueous phase (after Triton X-114 phase separation) from brain containing CASK was carefully layered on a 10-40 % continuous glycerol gradient. Following ultracentrifugation, samples were blotted for CASK and liprin-α3. Fig. also shows pure recombinant CASK. Arrows indicate the fractions where protein standards of 150 kDa and 460 kDa were observed. f GST-pulldown assay performed from rat brain homogenates using either full-length GST-CASK or GST (negative control) immobilized on glutathione resin. Synaptophysin (negative control); liprin α2 (positive control); ATPsynβ (ATP synthase β subunit); and IDH (isocitrate dehydrogenase)

Fig. 6

CASK ^(+/-) mutant mice exhibit growth retardation and skeletal muscle metabolic defects. Comparison of (a) body weight, (b) percentage lean mass, (c) percentage fat mass and (d) percentage fluid mass. All comparisons are presented as mean ± SEM, (* indicates p < 0.05; n = 5). e Energy expenditure/fat-free mass/hr in a 24-h period from CASK ^(+/-) mice compared to sex-matched CASK ^(+/+) littermate control. Data is presented as mean ± SEM (* indicates p < 0.05; n = 5). f Respiratory exchange ratio from CASK ^(+/-) mice compared to sex-matched CASK ^(+/+) littermate control. Data is presented as mean ± SEM. g Fatty acid oxidation rate and (h) glucose oxidation rate measured in skeletal muscle homogenates from CASK ^(+/+) and CASK ^(+/-) sex-matched littermate mice. Bar graphs are plotted as mean ± SEM. (* indicates p < 0.05; n = 5). All experiments were performed on one month old mice

Fig. 7

Reduction of CASK expression in cells and brain reduces oxidative metabolism. a Representative Western blot showing endogenous CASK expression in human embryonic kidney (HEK293-PLKO) cells as well as CASK knockdown in HEK293 cells using shRNA692 (HEK-692). b Quantitation of CASK knockdown (KD) in HEK293 cells using shRNA692 (* indicates p < 0.05; n = 3) (WT = wildtype). c Hemocytometer cell count in CASK knockdown (KD) HEK293 cells compared to wild-type control cells (WT). Bar graphs are plotted as mean ± SEM (n = 4; * indicates p < 0.05). d Lactate measurement from the culture media of wild-type control (WT) or CASK knockdown (KD) HEK293 cells. Bar graphs are plotted as mean ± SEM (* indicates p < 0.05; n = 3). e Total cellular respiration rate in CASK knockdown (KD) HEK293 cells and CASK knockdown HEK293 cells overexpressing recombinant rat CASK protein (KD + CASK) compared to the wild-type control cells (WT). Respiration is normalized to total protein levels (n = 4; * indicates p < 0.05, ** indicates p < 0.01). f Oxygen consumption rate measured in brain homogenates isolated from one month old CASK ^(+/+) and CASK ^(+/-) mice. Respiration rate is normalized to the total protein content (* indicates p < 0.05; n = 4). g Glucose oxidation rate measured in brain homogenates isolated from CASK ^(+/+) and CASK ^(+/-) sex-matched littermate mice (* indicates p < 0.05; n = 4)