Re-expression of SynGAP protein in adulthood improves translatable measures of brain function and behavior

Abstract

It remains unclear to what extent neurodevelopmental disorder (NDD) risk genes retain functions into adulthood and how they may influence disease phenotypes. SYNGAP1 haploinsufficiency causes a severe NDD defined by autistic traits, cognitive impairment, and epilepsy. To determine if this gene retains therapeutically-relevant biological functions into adulthood, we performed a gene restoration technique in a mouse model for SYNGAP1 haploinsufficiency. Adult restoration of SynGAP protein improved behavioral and electrophysiological measures of memory and seizure. This included the elimination of interictal events that worsened during sleep. These events may be a biomarker for generalized cortical dysfunction in SYNGAP1 disorders because they also worsened during sleep in the human patient population. We conclude that SynGAP protein retains biological functions throughout adulthood and that non-developmental functions may contribute to disease phenotypes. Thus, treatments that target debilitating aspects of severe NDDs, such as medically-refractory seizures and cognitive impairment, may be effective in adult patients.

Keywords: autism; biomarker; circuits; epilepsy; human; mouse; mouse model; neuroscience; reversal.

Figure 1.. Seizure threshold is improved after adult restoration of SynGAP expression.

(A) Syngap^Cre-;+/ls and Syngap^Cre+;+/ls mice exhibit hyperexcitability in two of the three events without Cre activation (No TMX) Main effects-1^st clonus: Cre F(1,24)=2.13, p=0.157, Genotype F = 117.73, p=9.75E-11, Interaction F(1,24)=1.69, p=0.206); Cre- Cohen’s d = 3.855, Cre +Cohen’s d = 4.737. TC: Cre F(1,24)=722, p=0.404, Genotype F(1,24)=40.05, p=1.53E-6), Interaction F(1,24)=.257, p=0.617); Cre- Cohen’s d = 2.396, Cre+ Cohen’s d = 2.405. THE: Cre F(1,24)=9.99E-6, p=0.998, Genotype F(1,24)=.320, p=0.577), Interaction F(1,24)=.420, p=0.523. (B) Syngap^Cre+;+/ls mice exhibit thresholds comparable to those of Syngap^Cre-;+/ls mice after Cre activation (TMX-treated) in two of the three events Main effects-1^st clonus: Cre F(1,71)=2.59, p=0.112; Genotype F(1,71)=58.328, p=7.86E-11, Interaction F = 1 (1,71)=18.84 p=4.62E-5; Cre- Cohen’s d = 3.329, Cre+ Cohen’s d = 0.674; TC: Cre F(1,71)=4.53, p=0.037, Genotype F(1,71)=26.15, p=2.57E-6, Interaction F(1,71)=6.50, p=0.013; Cre- Cohen’s d = 2.040; Cre+ Cohen's d = 0.540; THE: Cre F(1,71)=.037, p=0.847, Genotype F(1,71)=1.15E-5, p=0.997, Interaction F(1,71)=.049, p=0.826. Data points (and numbers) in bars represent biological replicates (animals). Data from panel B are pooled from two separate experiments.

Figure 1—figure supplement 1.. TMX-induced restoration of SynGAP protein levels in adult Syngap1^Cre+;+/ls mice.

(A) Western blot demonstrating expression levels of total SynGAP in Cre(-) or Cre(+) heterozygous Lox-Stop mice and WT littermates. (B) Densitometric analysis of SynGAP. Band intensities were normalized to total protein levels and transformed to % of the Syngap1^Cre-;+/+ group mean. Two-factor ANOVA. Main effects: Cre p=0.198, Genotype p=0.007, Interaction p=1.554E-4.=0.=0. Pairwise comparisons from posthoc tests can be found in Supplementary file 1. Data points (and numbers) in bars represent biological replicates (animals).

Figure 2.. Rectification of state-dependent paroxysmal spiking events in Syngap1 mutants after adult-initiated gene therapy.

(A) Representative EEG/LFP traces from a WT [Cre(+); +/+] and Syngap1 heterozygous mutant mouse [Cre(+); +/ls]. After initial recordings (pre-TMX), all animals were injected with TMX. Post-TMX recordings were acquired 30 days after the last TMX injection. TMX rescued low levels of SynGAP protein in +/ls animals (see Figure 1—figure supplement 1). Highlighted areas correspond to periods of sleep (see Materials and methods). Phase I and Phase II recordings are from the same animals. (B) Frequency of spiking events observed in the hippocampal LFP channel during the wake phase (i.e. non-highlighted areas in panel A) from both pre- and post-TMX recording sessions in each animal. Two-way repeated measures ANOVA.:Main genotype effects: F(1,11)=10.1, p=0.00879, Main TMX effects: F(1,11)=12.088, p=0.00518. Interaction between genotype and TMX: F(1,11)=9.777, p=0.00963. Cre(+);+/+n = 6, Cre(+);+/ls n = 7. (C) Comparison of the spiking frequency from the hLFP channel in Cre(+);+/ls mice during wake and sleep before TMX injections, paired-t test t(5)=-5.6007, p=0.002507 (n = 5). Data points in plots represent biological replicates (animals).

Figure 2—figure supplement 1.. Generalization of high-amplitude spikes across the forebrain.

Representative traces from all channels during a Phase I recording from a Cre(+) Lox-Stop mouse.

Figure 3.. Representative EEG recordings taken from SYNGAP1 patients during wake and sleep.

Ten second epochs of electroencephalograms from patients with SYNGAP1 pathogenic variants. (A) Patient S3-060 while awake (B) Patient S3-060 while asleep (C) Patient S3-080 while awake (D) Patient S3-080 while asleep. Shaded areas indicate bursts of generalized epileptiform activity.

Figure 4.. Long-term memory can be improved in adult mice with Syngap1 pathogenicity.

(A) Syngap1^+/+ and Syngap1^+/- ± were trained in the remote contextual fear conditioning paradigm and tested one month later for activity suppression levels. Activity of the Syngap1^+/- was suppressed significantly less than that of the Syngap1^+/+ group indicating compromised remote memory for the mutant group. Unpaired t test (t(19)=-2.567, p=0.019). Cohen’s d = 1.150. (B) Syngap1^+/+ and Syngap1^+/ls mice were trained in the contextual fear conditioning paradigm and tested one month later for activity suppression levels. Activity of the Syngap1^+/ls group was suppressed significantly less than that of the Syngap1^+/+ group indicating compromised remote memory for the mutant group. Wilcoxon rank sum test W = 19, p=2.82E-5, Cohen’s d = 1.676. (C) Syngap1^+/+ and Syngap1^+/- were tested, firstly, 1d after training, followed by another testing one month later. Activity suppression levels were not significantly different between the groups for either testing (unpaired t test,1-day t(13)=-0.033, p=0.974; 26 days t(13)=-1.068, p=0.305). (D) Experimental schematic depicting the breeding strategy for generation of Cre-inducible Syngap1^Cre+;+/ls mice and Cre induction with TMX treatment for restoration of Syngap1 expression and subsequent remote fear conditioning testing. (E–F) Syngap1^Cre-;+/+, Syngap1^Cre-;+/ls, Syngap1^Cre+;+/+, and Syngap1^Cre+;+/ls mice were run in the remote contextual fear conditioning paradigm without (E) and with (F) TMX administration. Activity suppression values from mice without TMX administration (No TMX) were assessed (2-factor ANOVA: Main Effects-Cre F(1,90)=0.030, p=0.864, Genotype F(1,91)=46.78, p=9.28E-10, Interaction F(1,91)=6.81, p=0.011; Cre- Cohen’s d = 1.725, Cre+ Cohen’s d = 0.910. With TMX administration (2-factor ANOVA: Main Effects- Cre F=(1,73)=0.019, p=0.891, Genotype F(1,73)=27.49, p=1.48E-6, Interaction F(1,73)=14.75, p=2.59E-4; Cre- Cohen’s d = 2.167). Data points (and numbers) in bars represent biological replicates (animals). Data from panels E-F are pooled from at least two separate experiments.

Figure 5.. Increased amplitude of theta oscillations after SynGAP re-expression in adult Syngap1 mutant mice.

(A–B) CA1 LFP traces from a WT (A) and a Syngap1 mutant (B) mouse during Phase I and Phase II sessions. (C) Grand average of within-subjects changes in signal amplitude across the full spectrum of hippocampal rhythms. The amplitude change was normalized by the average amplitude during Phase I sessions. The shaded areas represent 95% bootstrapped confidence intervals. Significant increases in amplitude in Phase II were detected in the 6–12 Hz theta range (Permutation test: p=0.0128, 5000 shuffles). N’s are biological replicates (animals). Legends for Figure Supplements.

Figure 5—figure supplement 1.. Amplitude of theta oscillations in each mouse during Phase I and Phase II recording sessions.

(A–B) CA1 LFP recordings from WT (A) and Syngap1 mutant (B) mice during Phase I and Phase II sessions. (C–D) Average amplitude spectra for each mouse during Phase I (C) and Phase II (D) sessions. Individual mice are indicated with individual lines, and WT and Syngap1 mutant spectra are depicted in blue and red, respectively.

Figure 5—figure supplement 2.. Effect of genotype, but not phase, on horizontal activity during neurophysiological recordings.

Cre(+) WT and Cre(+) Lox-Stop mice were video tracked for distances traveled during the first ten minutes of recording during Phase I (TMX-) and Phase II (TMX+) sessions. RMANOVA- Group: F(1,12)=16.527, p=0.002; Phase: F(1,12)=0.164, p=0.692; Group x Phase: F(1,12)=3.521, p=0.085. Data points in bars represent biological replicates (animals). Legends for Supplementary Files.