doi: 10.1093/braincomms/fcaa028.
Epub 2020 Mar 10.
GABAA receptor β3 subunit mutation D120N causes Lennox-Gastaut syndrome in knock-in mice
Affiliations
- PMID: 32467926
- PMCID: PMC7238755
- DOI: 10.1093/braincomms/fcaa028
Item in Clipboard
GABAA receptor β3 subunit mutation D120N causes Lennox-Gastaut syndrome in knock-in mice
Brain Commun.
2020.
Abstract
The Lennox-Gastaut syndrome is a devastating early-onset epileptic encephalopathy, associated with severe behavioural abnormalities. Its pathophysiology, however, is largely unknown. A de novo mutation (c.G358A, p.D120N) in the human GABA type-A receptor β3 subunit gene (GABRB3) has been identified in a patient with Lennox-Gastaut syndrome. To determine whether the mutation causes Lennox-Gastaut syndrome in vivo in mice and to elucidate its mechanistic effects, we generated the heterozygous Gabrb3+/D120N knock-in mouse and found that it had frequent spontaneous atypical absence seizures, as well as less frequent tonic, myoclonic, atonic and generalized tonic-clonic seizures. Each of these seizure types had a unique and characteristic ictal EEG. In addition, knock-in mice displayed abnormal behaviours seen in patients with Lennox-Gastaut syndrome including impaired learning and memory, hyperactivity, impaired social interactions and increased anxiety. This Gabrb3 mutation did not alter GABA type-A receptor trafficking or expression in knock-in mice. However, cortical neurons in thalamocortical slices from knock-in mice had reduced miniature inhibitory post-synaptic current amplitude and prolonged spontaneous thalamocortical oscillations. Thus, the Gabrb3+/D120N knock-in mouse recapitulated human Lennox-Gastaut syndrome seizure types and behavioural abnormalities and was caused by impaired inhibitory GABAergic signalling in the thalamocortical loop. In addition, treatment with antiepileptic drugs and cannabinoids ameliorated atypical absence seizures in knock-in mice. This congenic knock-in mouse demonstrates that a single-point mutation in a single gene can cause development of multiple types of seizures and multiple behavioural abnormalities. The knock-in mouse will be useful for further investigation of the mechanisms of Lennox-Gastaut syndrome development and for the development of new antiepileptic drugs and treatments.
Keywords: GABRB3; Lennox–Gastaut syndrome; early-onset epileptic encephalopathy; genetic epilepsies; seizures.
© The Author(s) (2020). Published by Oxford University Press on behalf of the Guarantors of Brain.
Figures
Creation of the Gabrb3+/D120N KI mouse. (A) Scheme for targeted KI of the Gabrb3 locus. A vector was constructed to replace Gabrb3 Exon 4 genomic DNA with an Exon 4 containing the G358A mutation and the positive selection cassette (PGK-Neo). (B) Long-range PCR was used to identify the correctly targeted clones, G3 and F5. The 6- and 4.6-kb fragments were amplified from correctly targeted clones using primers A/B and C/D, respectively. The PCR fragments were digested with the HindIII restriction enzyme (to generate the predicted 4.2- and 1.8-kb fragments for the 5′ end and 3.6- and 1-kb fragments for the 3′ end) to further confirm correct homologous recombination. (C) The KI mice were genotyped. PCR amplified a 392-bp fragment from the wild type β3 subunit allele and a 480-bp fragment from the β3 subunit KI allele due to the insertion of FRT sequences and other modifications. (D) Sequencing chromatogram of RT-PCR derived from total brain RNA showed the presence of the G358A mutation and corresponding equal level of wild type and mutant nucleotide expression. (E1) Semi-quantitative RT-PCR showed that the KI allele did not affect Gabrb3 gene expression, using GAPDH as a loading control. (E2) Band intensities of the RT-PCR products were first normalized to GAPDH for quantification and then to wild type levels. (F1) Immunoblotting showed equal β3 subunit expression in KI and wild type littermate mouse brain whole-cell lysates. GAPDH served as a loading control. (F2) Protein levels were normalized first to internal GAPDH and then to wild type levels. (G) The body weight of KI and their littermate mice at weaning were plotted (n = 22 wild type, 18 KI). GAPDH: glyceraldehyde 3-phosphate dehydrogenase; RT-PCR: reverse transcriptase PCR.
Spontaneous seizures were observed in adult KI mice. Representative EEG traces for typical absence seizures seen in wild-type mice, and atypical absence, myoclonic, tonic and GTCS seen in KI mice. The arrowheads indicate the start and end of the seizure. Note that the atypical absence seizure extends beyond the selected time frame, so no endpoint is shown. (A) Typical absence seizures in wild-type mice were brief and behavioural arrest was time-locked to SWD onset and offset. (B) Power spectral density for typical absence seizure EEGs in wild-type mice with peak frequency at 7.6 Hz. (C) Atypical absence seizure EEG trace; atypical absence seizures were not always time-locked with the behavioural onset and/or had brief movements during the seizure (see Video 1 for full seizure). (D) Power spectral density of atypical absence EEGs often started with (i) higher-frequency components that (ii) decreased over time; however, the frequency over the entire seizure (iii) was low. (E) Myoclonic seizure EEG discharges were brief (∼300 µs), and both EEG and EMG traces had prominent spikes. (F) EEG during tonic seizures did not stand out from the baseline in amplitude but showed an increased frequency compared with the baseline. (G) GTCS EEGs were striking with the largest increase in EEG amplitude. The tonic phases during GTCS had slightly lower amplitude but higher waveform frequencies compared with the clonic phase, which was followed by electrodecrement (see Video 1 for examples of KI video-EEG recordings for each seizure type). (H) Bar graph showing average number seizures in wild type and KI mice in a 24-h period (note: y-axis is log scale). Atypical absence seizures and myoclonic seizures were the dominant seizures in KI mice. (I, J) Most atypical absence seizures in KI mice occurred during the dark (active) period with a peak incidence at the light-to-dark period transition (paired t-test, n = 7, light mean = 142, dark mean = 297.7, *P = 0.011). As the light cycle does not change with daylight savings, all times are set to the appropriate non-daylight savings time (6:00–18:00 light, 18:00–6:00 dark). Video-EEG recordings (for panels A–J) were done in n = 4 wild type (two males, two females) and 7 KI (three males, four females) mice at age 4.5–6.5 months. (K) Spontaneous spasms in wild type and KI mice were observed in pups between P12 and P18. Shown are the percentages of wild type and KI pups that exhibited spasms at each age. Observers were blind (five litters, with n = 16 wild type and n = 19 KI mice).
KI mice have spatial learning and memory deficits. The Barnes maze test demonstrated a delay in spatial learning and a deficit in spatial memory in the KI mice at both P49 and P200. Eight days of learning trials depict spatial learning abilities in P49 (A–C) and P200 (D–F) wild type and KI mice. Each day represents the average across all mice of each genotype with the data point of each mouse being the average of the four trials conducted that day. (A, D) The time it takes each animal (latency) to find the target hole for each day was plotted. (B, E) The number of non-target hole zones entered (errors) was plotted for each day. (C, F) Search strategy scores for the 8 days of learning trials were plotted for each day. Mice earned zero points for a direct path to the target hole, one point for a serial path and two points for a random path. Higher scores represented less efficient searching. (G–J) Five minute probe trial for spatial memory performed on Day 8 after the learning trials were plotted for (G, H) P49 and (I, J) P200 mice. The target hole was then covered and appeared identical to the other 11 holes. (G, I) Time spent near the target hole and (P49 wild type 26.25 ± 2.27 s, KI 18.44 ± 1.45 s; P200 wild type 26.50 ± 2.28 s, KI 22.27 ± 2.02 s). (H, J) Number of non-target hole zones entered (errors) (P49 wild type 64.14 ± 5.74, KI 87.10 ± 8.06; P200 wild type 66.36 ± 3.29, KI 102.1 ± 5.88). (A–F) Values expressed as mean ± SEM. (A–F) Two-way ANOVA for repeated measures: genotype effect #P < 0.05, ###P < 0.001, ####P < 0.0001. Bonferroni post-tests: *P < 0.05, **P < 0.01, ***P < 0.001. (G–J) Graphed values are expressed as median ± SEM. Unpaired two-tailed Student’s t-test, *P < 0.05, ***P < 0.001. n = 21–22 for all panels.
Gabrb3+/D120N KI mice exhibited hyperactivity, social deficits and anxiety. (A–D) KI mice had a mild hyperactive phenotype at P49, which became more pronounced at P200. Data shown were collected from 60 min in the locomotor activity chambers. (A) Vertical counts, or the number of rearings, detected at P49 were plotted (group means: wild type 293.6 ± 29.10, KI 417.0 ± 44.42). (B) Total distance travelled (centre and surround included) in the locomotor activity chambers at P49 (group means: wild type 49.03 ± 3.22 m, KI 58.71 ± 4.38 m). (C) Vertical counts at P200 (group means: wild type 373.3 ± 32.00, KI 527.5 ± 63.17). (D) Total distance travelled at P200 (group means: wild type 52.68 ± 3.97 m, KI 70.71 ± 7.09 m). Graphed values are expressed as median ± SEM. Unpaired two-tailed Student’s t-test, *P < 0.05. n = 20–25. (E–H) The 3CST showed social deficits in KI mice that begin at P49 and become more pronounced at P200. Not shown is Stage 1, a 10-min familiarization stage for the test mouse to acclimate to the set-up. (E, F) Stage 2 of the 3CST was a 10-min trial in which subject mice had two socialization options: (i) an empty inverted pencil cup in one side chamber or (ii) an inverted pencil cup containing a novel age-matched female mouse in the other side chamber. Two-way ANOVA was not significant for a genotype effect at P49 in E, and genotype effect was significant at P200 at P < 0.05 in F (group means: P49, novel mouse: wild type 1.89 ± 0.13 min, KI 1.86 ± 0.18 min; novel object: wild type 1.19 ± 0.08 min, KI 1.06 ± 0.14 min; total exploration: wild type 3.00 ± 0.24 min, KI 2.64 ± 0.18 min. P200, novel mouse: wild type 1.60 ± 0.17 min, KI 1.00 ± 0.14 min; novel object: wild type 0.79 ± 0.08 min, KI 0.76 ± 0.11 min; total exploration: wild type 2.39 ± 0.21 min, KI 1.73 ± 0.22 min). (G, H) Stage 3 of the 3CST was a 10-min trial in which test mice had two socialization options: (i) familiar socialization, in which the novel mouse from Stage 2 remained where she was or (ii) novel socialization, in which a new novel mouse was placed under the previously empty pencil cup. Two-way ANOVA showed a significant genotype effect P < 0.01 at both (G) P49 (group means: familiar mouse: wild type 1.11 ± 0.14 min, KI 0.72 ± 0.07 min; novel mouse: wild type 1.54 ± 0.12 min, KI 1.27 ± 0.14 min; total exploration: wild type 2.65 ± 0.21 min, KI 2.00 ± 0.17 min) and (H) p200 (group means: familiar mouse: wild type 0.89 ± 0.08 min, KI 0.82 ± 0.14 min; novel mouse: wild type 1.49 ± 0.19 min, KI 0.72 ± 0.09 min; total exploration: wild type 2.38 ± 0.21 min, KI 1.58 ± 0.17 min). Graphed values expressed as median ± SEM. Statistical tests shown on graphs are a priori Bonferroni post-tests after two-way ANOVAs. Four comparisons were made in each graph; however, only significant or otherwise relevant results were shown. Bonferroni multiple-comparison correction significance levels for E–H were: *P < 0.05, **P < 0.01, ***P < 0.001. n = 21–23. (I–L) KI mice displayed elevated anxiety in both the locomotor activity chambers and the elevated zero maze, with a mild anxiety phenotype at P49 and more pronounced anxiety at P200. (I, J) Each subject spent 60 min in the locomotor activity chamber. The chamber was divided into two 50% sections by area: the centre and the surround. Distance travelled within the surround (edge) area as a percentage of total distance travelled was shown for (I) P49 (group means: wild type 44.80 ± 1.92, KI 37.84 ± 1.67) and (J) P200 (group means: wild type 53.63 ± 2.49, KI 44.96 ± 2.08). (K, L) Each subject spent 5 min in the elevated zero maze. The distance travelled in the open arms at (K) P49 (group means: wild type 6.87 ± 0.43, KI 6.39 ± 0.33) and (L) P200 (group means: 5.82 ± 0.39, KI 4.386 ± 0.58). Data not shown: total distance travelled was not changed for the KI mice at either age in the zero maze. Graphed values are expressed as median ± SEM. Unpaired two-tailed Student’s t-test, for I–L, *P < 0.05, **P < 0.01. n = 22–28. 3CST: three-chamber socialization test.
Biogenesis and trafficking of receptors containing mutant β3(D120N) subunits were unaltered in Gabrb3+/D120N KI mice. The mutant β3(D120N) subunits did not affect expression, surface trafficking or distribution in synaptosomes of either their own or other GABAA receptor subunits. (A) Whole-cell lysates from different brain regions [cortex (Co), cerebellum (Ce), hippocampus (Hi) and thalamus (Th)] were collected from KI and wild type littermate mice and subjected to SDS–PAGE and immunoblotted with anti-α1, -β3 and -γ2 subunit antibodies. GAPDH served as a loading control. (B) Band intensities of α1, β3 and γ2 subunits were normalized to GAPDH signal and, then, KI was compared with wild type (n = 3 KI: wild type pairs). One-way ANOVA followed by the Dunnett’s multiple-comparison test was used to determine significance; however, no changes were identified. (C) Plasma membrane proteins from mouse brain were isolated, analysed by SDS–PAGE and immunoblotted with anti-α1, -β3 and -γ2 subunit antibodies. Na+/K+ ATPase served as a loading control, and GAPDH served as cytosolic protein contamination control. (D) Band intensities of α1, β3 and γ2 subunits were normalized to ATPase signal and, then, KI bands were compared with wild type bands (n = 3 KI: wild type pairs). (E) Proteins from synaptosomes were analysed by SDS–PAGE and immunoblotted with anti-α1, -β3 and -γ2 subunit and gephyrin antibodies. Na+/K+ ATPase served as a loading control. The distribution of GABAA receptor subunits was not changed in KI or wild type littermate mice. (F) Band intensities of α1, β3 and γ2 subunits and gephyrin were normalized to the ATPase signal and, then, KI bands were compared with wild type bands (n = 3 KI: wild type pairs). Student’s t-tests were used in D and F with a significance level of α = 0.05. GAPDH: glyceraldehyde 3-phosphate dehydrogenase.
mIPSCs recorded from KI mouse SS cortex layer V/VI neurons were altered. mIPSCs were recorded from SS cortex layer V/VI neurons (voltage-clamped at −60 mV with equal chloride concentration inside and outside cells) in thalamocortical slices from littermate (A) wild type and (B) KI mice. (C) Normalized mIPSC events for wild type and KI mice showed the frequency of mIPSCs in each amplitude bin. (D) Normalized cumulative distribution was plotted for wild type and KI mouse mIPSCs. Compared with mIPSCs recorded from wild type littermate mice, those from SS cortex layer V/VI neurons in KI mice had (E) significantly reduced amplitudes (wild type −49.67 ± 2.71 pA; KI −31.48 ± 1.76 pA) and (F) slowed mIPSC decay (wild type 25.05 ± 2.87 ms; KI 39.13 ± 2.72 ms). wild type, n = 6 mice; KI, n = 5 mice. Student’s t-test, **P < 0.01, ***P < 0.001. SS: somatosensory.
Spontaneous thalamocortical network oscillations from KI mice were longer and more frequent than those from wild type littermate mice. Representative extracellular multiple unit recordings from VBn in slices showing (A) spontaneous brief thalamocortical bursts in wild-type mice and (B) spontaneous prolonged thalamocortical oscillations in KI mice. One short burst from a wild type and one oscillation from a KI mouse were expanded in each panel to show the multiple spikes in the burst or oscillation. Scale bars are as indicated. (C) Average duration of spontaneous bursts for P42 wild-type mice (0.78 ± 0.71 s, n = 6 mice) and spontaneous oscillations from P42 KI mice (6.88 ± 1.39 s, n = 9 mice) were plotted (**P < 0.01, Student’s t-test). VBn: thalamic ventrobasal nucleus.
Ethosuximide, clobazam and a cannabinoid, but not topiramate, reduced atypical absence seizures in KI mice. (A, B) Three AEDs were administered individually to KI mice. Total observation period by video-EEG lasted 4 h. At Hour 0, an injection of each drug’s respective vehicle was given as a control observation. Vehicle and vehicle washout were found to be stable and were therefore averaged as a baseline. Drug + vehicle were administered at Hour 2. The 2 h after the drugs were administered were split into single hour observations: the first hour after administration and the second hour after administration. A one-way ANOVA was performed on each drug individually (shown above each drug set), and Tukey post-tests were conducted comparing all three pairs of time points (shown on bar graphs where significant). (A) Both ethosuximide and clobazam, but not topiramate, significantly reduced the cumulative time spent in atypical absence seizures. (B) Ethosuximide resulted in a significant reduction in the number of atypical absence seizures, while clobazam and topiramate did not. n = 8–9, 6–8 months old, males and females. Repeated-measures one-way ANOVA #P < 0.05; Tukey post-tests adjusted P-value *P < 0.05 compares drug to baseline, @P < 0.05 compares drug to drug washout. (C, D) Win 55,212-2 was administered daily for 1 week. KI mice had a significant reduction in both (C) cumulative time spent in atypical absence seizures and (D) atypical absence seizure count after 1 week of dosing. n = 6, paired Student’s t-test, 5–8 months old, males and females, *P < 0.05, **P < 0.01.
References
Grants and funding
LinkOut - more resources
Full Text Sources
Molecular Biology Databases
Research Materials