Patient-derived SLC6A1 variant S295L results in an epileptic phenotype similar to haploinsufficient mice

Abstract

The solute carrier family 6 member 1 (SLC6A1) gene encodes GAT-1, a γ-aminobutyric acid transporter expressed on astrocytes and inhibitory neurons. Mutations in SLC6A1 are associated with epilepsy and developmental disorders, including motor and social impairments, but variant-specific animal models are needed to elucidate mechanisms. Here, we report electrocorticographic (ECoG) recordings and clinical data from a patient with a variant in SLC6A1 that encodes GAT-1 with a serine-to-leucine substitution at amino acid 295 (S295L), who was diagnosed with childhood absence epilepsy. Next, we show that mice bearing the S295L mutation (GAT-1^S295L/+ ) have spike-and-wave discharges with motor arrest consistent with absence-type seizures, similar to GAT-1^+/- mice. GAT-1^S295L/+ and GAT-1^+/- mice follow the same pattern of pharmacosensitivity, being bidirectionally modulated by ethosuximide (200 mg/kg ip) and the GAT-1 antagonist NO-711 (10 mg/kg ip). By contrast, GAT-1^-/- mice were insensitive to both ethosuximide and NO-711 at the doses tested. In conclusion, ECoG findings in GAT-1^S295L/+ mice phenocopy GAT-1 haploinsufficiency and provide a useful preclinical model for drug screening and gene therapy investigations.

Keywords: SLC6A1; absence epilepsy; chronic wireless telemetry; pharmacosensitivity; spike-and-wave discharges.

Figure 1.

Patient-derived Slc6a1 variant S295L is associated with absence epilepsy in human and mouse. A-B. Scalp EEG traces from a 4-year-old boy with S295L variant SLC6A1-related disorder, illustrating: A. occipital intermittent rhythmic delta activity (dimensions proportional to sensitivity 15 μV/mm, paper speed 30 mm/s) and B. frontal intermittent rhythmic delta activity (sensitivity 30 μV/mm, paper speed 30 mm/s). For each, a limited montage is shown; see Supplemental Figure 3 for a standard 10–20 montage. C. ECoG and EMG recordings from representative wild-type and S295L variant mice during active (top traces) and quiet (bottom traces) behavioral states. D. Spectral density plot of ECoG, evaluated during the active segment shown in (A) in GAT-1^S295L/+ (orange) and GAT-1^+/+ (white) mice. E. Quantification of spikes and SWDs occurring during the light period, averaged over two consecutive 12-hour light periods (n = 5, 7, Mann-Whitney test). F. Quantification of spikes and SWDs occurring during the dark period (n = 5, 7, Mann-Whitney test). G. Circadian patterns of locomotor activity in GAT-1^+/+ (n = 6, p = 0.0002, paired t-test) and GAT-1^S295L/+ mice (n = 7, p = 0.0022, paired t-test). H. Nest building, scored at 48 hours on a 6-point (0–5) scale (n = 8, 7, p = 0.3457, unpaired t-test). This panel captures data from some mice without ECoG or locomotor activity (mice underwent device implantation surgery and recovery but no ECoG was recorded due to transmitter failure). I. Pharmacosensitivity of epileptic phenotype in GAT-1^S295L/+ mice, demonstrated by ECoG and EMG traces from a representative GAT-1^S295L/+ mouse. See also Supplemental Figure 2. J-K. Population data summarizing ECoG changes in GAT-1^+/+ (white bars) and GAT-1^S295L/+ mice (orange bars) in response to vehicle, NO-711, and ethosuximide, assessed at 7a.m. (I) and at 7p.m. (J). RM-ANOVA with Dunnett’s multiple comparisons tests, n = 5, 8. L-M. Population data summarizing locomotor activity changes in GAT-1^+/+ (white bars) and GAT-1^S295L/+ mice (orange bars) in response to vehicle, NO-711, and ethosuximide, assessed at 7a.m. (L) and at 7p.m. (M) Data represent the first hour post-injection. The sequence of drug administration (NO-711 vs ethosuximide) was alternated in interleaved animals to control for order effects, if any. * p < 0.05, ** p < 0.01. Each dot represents a single mouse (solid circles: males, open circles: females).

Figure 2.

Gradient of epileptic abnormalities with progressive GAT-1 deficiency. A. Simultaneously recorded ECoG (top) and EMG (bottom) activity from representative GAT-1^+/+ (left), GAT-1^+/− (middle), and GAT-1^−/− (right) mice. B. Spectral density plot of ECoG during the interval shown in (A) in GAT-1^+/+ (white), GAT-1^+/− (light green), and GAT-1^−/− (dark green) mice. C. Spike and SWD quantification averaged over two consecutive 12-hour light periods (spikes ANOVA p < 0.0001, Tukey’s multiple comparisons, n = 9,9,9; SWDs Brown-Forsythe p < 0.0005, Dunnett’s multiple comparisons, n = 9,9,9). D. Quantification of spikes and SWDs, averaged over two consecutive dark periods (spikes, ANOVA p = 0.04, Tukey’s multiple comparisons, n = 9,9,9; SWDs, Brown-Forsythe p < 0.0001, Dunnett’s multiple comparisons, n = 9,9,9). E. Circadian patterns of locomotor activity in the GAT-1 colony (GAT-1^+/+ p < 0.0001, GAT-1^+/− p < 0.0001, GAT-1^−/− p = 0.0003, paired t-tests, n = 9,9,9). F. Quantification of nesting behavior (ANOVA p = 0.0015; post-hoc t-tests with Dunnett’s correction, n = 8,8,8). G-J. Pharmacosensitivity of GAT-1 knock-out mice. G. Population data plotting spike burden under conditions of vehicle, NO-711, and ethosuximide administration at 7a.m. (See also Supplemental Figure 2. GAT-1^+/+ RM-ANOVA p = 0.005, Dunnett’s multiple comparison test; n = 8; GAT-1^+/− RM-ANOVA p = 0.002, Dunnett’s multiple comparisons; n = 8; GAT-1^−/− RM-ANOVA p = 0.37, n = 8). H. Population data plotting spike burden under conditions of vehicle, NO-711, and ethosuximide administration at 7p.m. (GAT-1^+/+ RM-ANOVA p = 0.002, Dunnett’s multiple comparisons, n = 8; GAT-1^+/− RM-ANOVA p = 0.0001, Dunnett’s multiple comparisons, n = 7; GAT-1^−/− RM-ANOVA p = 0.026, n = 7). Data shown represent the first hour post-injection. The sequence of drug administration (NO-711 vs. ethosuximide) was alternated in interleaved animals to mitigate any order effects, similar to Fig. 1. *p < 0.05, **p < 0.01. Each dot represents a single mouse (solid circles: males, open circles: females).