September 28, 2016;
Gain-of-function defects of astrocytic Kir4.1 channels in children with autism spectrum disorders and epilepsy.
Dysfunction of the inwardly-rectifying potassium channels Kir4.1 (KCNJ10
) represents a pathogenic mechanism contributing to Autism-Epilepsy comorbidity. To define the role of Kir4.1 variants in the disorder, we sequenced KCNJ10
in a sample of affected individuals, and performed genotype-phenotype correlations. The effects of mutations on channel activity, protein trafficking, and astrocyte
function were investigated in Xenopus laevis oocytes, and in human astrocytoma cell lines. An in vivo model of the disorder was also explored through generation of kcnj10a morphant zebrafish overexpressing the mutated human KCNJ10
. We detected germline heterozygous KCNJ10
variants in 19/175 affected children. Epileptic spasms with dysregulated sensory processing represented the main disease phenotype. When investigated on astrocyte
-like cells, the p.R18Q mutation exerted a gain-of-function effect by enhancing Kir4.1 membrane expression and current density. Similarly, the p.R348H variant led to gain of channel function through hindrance of pH-dependent current inhibition. The frequent polymorphism p.R271C seemed, instead, to have no obvious functional effects. Our results confirm that variants in KCNJ10
deserve attention in autism-epilepsy, and provide insight into the molecular mechanisms of autism and seizures. Similar to neurons, astrocyte
dysfunction may result in abnormal synaptic transmission and electrical discharge, and should be regarded as a possible pharmacological target in autism-epilepsy.
Disease Ontology terms:
autism spectrum disorder
[+] show captions
References [+] :
Figure 1. Whole-cell Kir4.1 currents recorded from cells expressing WT and R18Q channels.Sample current traces for (a) WT and (b) R18Q channels expressed in U251 astrocytoma cells, obtained in response to voltage steps from −140 to +20 mV in 10 mV steps, from a holding potential of −30 mV. (c) Mean current density (pA/pF) as a function of voltage obtained from cells expressing the indicated channels. Inset: scatter plot of the single-cell current densities at −120 mV. The open symbols represent the mean ± SEM. (d) The resting membrane potentials were evaluated in current clamp conditions from cells expressing WT (black bar) or R18Q (grey bar) channels (data are mean ± SEM; **p < 0.01). (e) Sample inward-current densities evoked by puffs of an external solution containing 20 mM KCl from astrocytoma cells expressing the indicated channels (holding potential of −80 mV). (f) WT (black circles) and R18Q (grey circles) current densities elicited from each cell using the experimental protocol described in panel e which were subsequently averaged and depicted (mean ± SEM; *p < 0.05).
Figure 2. WT and mutated Kir4.1 expression and distribution in U251 astrocytoma cells.(a) Co-immunofluorescence staining of Kir4.1 WT and R18Q expressing cells using Anti-Xpress epitope Ab to stain recombinant Kir4.1 (red) and phallacidin to stain actin filaments (green) shows that WT channels are mostly localized in the cytoplasm, and at plasma membrane in a low percentage of cells (top panels, arrows), while R18Q mutant channels are mainly distributed along cell membranes, filopodia-like structures and cell-cell contacts (bottom panels, arrowheads), and partially co-localizes with actin, in the majority of cells (70%). (b,c) Western Blot (WB) analysis of total membrane and cytosolic protein extracts of astrocytoma cells expressing WT and R18Q Kir4.1, probed with anti Kir4.1 Ab (b) and Anti-Xpress epitope tag Ab (c) revealed that R18Q is more abundantly expressed in the cytoplasm (CYT) and particularly in the total membrane protein fraction (MEM), than the WT protein. Anti-Xpress epitope tag Ab does not detect Kir4.1 protein in the cytoplasm. Arrowheads on the right of panel b highlight the monomeric and oligomeric forms of the Kir4.1 channel. Actin is used as loading control. Molecular weight markers are on the left (kDa). (d) Densitometric analysis of recombinant Kir4.1 bands derived from total membrane protein extracts from WT or R18Q Kir4.1 expressing cells, detected by anti-Xpress Ab and normalized to the corresponding actin value (mean ± SEM, expressed as arbitrary units; *P < 0.05 from three independent experiments). (e) WB of total cell proteins (SM) and enriched fraction of plasma membrane proteins after biotinylation experiments (Eluates) probed with anti–Kir4.1 Ab. Higher amount of R18Q Kir4.1 is expressed at the plasma membrane (arrow) when compared to Kir4.1 WT. Among Kir4.1 associated proteins α-syntrophin (α-synt) is found at plasma membrane of Kir4.1 R18Q mutant expressing cells (asterisk) but not in WT expressing cells. No differences are observed in β-dystroglycan (β-DG) and AQP4 association to astrocytoma plasma membrane. One representative experiment out of two performed with the same results has been shown.
Figure 3. Degradation kinetics, membrane compartmentalization and molecular interactions of WT and R18Q Kir4.1.(a) Left panel: WB analysis of protein extracts obtained from cells expressing WT and R18Q Kir4.1 channels treated with the protein synthesis inhibitor cycloheximide (CHX) for 3 and 6 h to inhibit protein synthesis and evaluate protein degradation kinetic. Anti-Xpress epitope Ab was used to detect WT and mutant Kir4.1. Actin is used as loading control. Molecular weight markers are on the left (kDa). Right panel: densitometric analysis of WT or R18Q Kir4.1 protein bands after CHX treatment, normalized to the corresponding untreated controls, indicates no significant differences in the degradation kinetics between WT and R18Q Kir4.1 proteins. Data are expressed as mean ± SEM from three independent experiments. (b) WB analysis of cholesterol-rich (Triton insoluble fractions: 4–7) and cholesterol-poor membrane fractions (Triton soluble fractions: 10–13) obtained from WT or R18Q Kir4.1 expressing cells shows that WT and mutant Kir4.1 channels are both distributed in cholesterol-poor membrane fractions. Caveolin-1 and flotillin identify the caveolar raft fractions in cells expressing Kir4.1 WT. Molecular weight markers are on the left (kDa). (c) WB analysis of Kir4.1 channel interactors selected by Histidine (His) affinity chromatography. Eluates derived from astrocytoma cells infected with the empty vector (U251) were used as controls for unspecific binding to NiNTA-resin. The Input lane indicates the total protein extracts before His pull-down assay and 200 mM and 50 mM the proteins eluted from NiNTA-resin using imidazole (50, 200 mM, respectively). Kir2.1, aquaporin-4 (AQP4), Kir5.1, TRPV4, but not connexin-43 were found among the proteins co-eluted with Kir4.1. No difference was observed between WT and R18Q Kir4.1 interactors. One representative experiment out of four is shown. Molecular weight markers are indicated on the left (kDa).
Figure 4. WT, R271C and R348H channels expression pattern and pH-gating in X.laevis oocytes.Mean current amplitudes during 10 days post mRNA injection: (a) WT vs R271C, and (b) WT vs R348H. Sample current traces recorded at pHi 7.4 and 6.1 from WT (c), R348H (d), and R271C (e) expressing oocytes. Time course of current inhibition (%) of R271C vs WT (f), and R348H vs WT (g) upon K+-acetate pHi acidification from 7.4 to 6.1. All traces were recorded at −100 mV from a holding potential −10 mV. (h) Bar graph of pHi 6.1-induced current inhibition (%) of the indicated channels. (i) pHi/inhibition relationships for WT (square) vs R348H (circles) assessed by using X. laevis oocytes and the K+-acetate pHi acidification method121314.
Figure 5. pH-gating of R348H channels expressed in astrocyte-like cells.Sample current traces for WT (a) and R348H (b) channel expressed in U251 cells, and recorded in the whole-cell configuration by applying voltage ramps from −140 to +30 mV (holding potential of −80 mV), immediately after the establishment of the whole-cell configuration (pHi7.2), after 10 min (pHi6.1), and following external BaCl2 application (pHi6.1 + 100 μM Ba2+). Both cells were recorded using an intracellular pipette solution having a pH of 6.1. (c,d) Representative time courses of WT (c) and R348H (d) currents recorded at −100 mV (normalized to the current recorded immediately after the establishment of the whole-cell configuration), using a pipette solution with pH of either 7.2 (black circles) or 6.1 (grey circles). (e) Plot of the mean fractional WT and R348H current inhibition by an intracellular pH of either 7.2 or 6.1. The fractional current reduction was estimated at −100 mV and calculated as 1− (I10 min/I0 min), in experiments similar to those shown in panels a–d. (f) Plot of the mean WT and R348H current density at −100 mV of applied potential.
Figure 6. In vivo modelling of Kir4.1 mutations in zebrafish.(a) Transient kcnj10a knockdown zebrafish show macroscopic abnormalities in organ development compared to wild-type (WT). In morphant embryos, the pronephric duct is visible because it is dilated (black arrow); the swim bladder is not visible (white arrow). (b–e) Number of spontaneous tail flicks (registration time 30sec) seen in WT and mutant 30 hpf embryos. Values are expressed as percent of flicks counted in uninjected embryos. Embryos injected with MO (b–e), and with equimolar amount of MO and either R18Q (b), V84M (c), and R348H (d) mRNA show an increased rate of spontaneous contractions, compared to uninjected embryos (b–e), and to either MO+WT (b–e) and MO+R271C (e) mRNA injected embryos. **p < 0.01; ***p < 0.001.
Genetically induced dysfunctions of Kir2.1 channels: implications for short QT3 syndrome and autism-epilepsy phenotype.