Viral vector–mediated expression of NaV1.1, after seizure onset, reduces epilepsy in mice with Dravet syndrome

Dravet syndrome (DS), an intractable childhood epileptic encephalopathy with a high fatality rate, is typically caused by loss-of-function mutations in one allele of SCN1A, which encodes NaV1.1, a 250-kDa voltage-gated sodium channel. In contrast to other epilepsies, pharmaceutical treatment for DS is limited. Here, we demonstrate that viral vector–mediated delivery of a codon-modified SCN1A open reading frame into the brain improves DS comorbidities in juvenile and adolescent DS mice (Scn1aA1783V/WT). Notably, bilateral vector injections into the hippocampus and/or the thalamus of DS mice increased survival, reduced the occurrence of epileptic spikes, provided protection from thermally induced seizures, corrected background electrocorticographic activity and behavioral deficits, and restored hippocampal inhibition. Together, our results provide a proof of concept for the potential of SCN1A delivery as a therapeutic approach for infants and adolescents with DS-associated comorbidities.

tract. Analgesia was achieved with subcutaneous injections of meloxicam (5 mg/kg) at the end of the surgical procedure and repeated 24 h post-surgery.
For thalamic-hippocampal co-injections, the needle was first lowered to: AP -1.8 mm; ML ± 1.8 mm; DV -3.5 mm, and a volume of 0.5 µl was injected. After a ~5 min, the injection needle was raised to DV 3 mm, and another volume of 0.5 µl was injected. The AP coordinates were slightly modified for P21 mice in which the distance between bregma and lambda was less than 3 mm, and reduced to AP -1.7 mm. The DV coordination was measured from the tip of the beveled injection needle. CAV-2 vectors were injected at a rate of 100 nl/min (Quintessential Stereotaxic Injector, Stoelting, Wood Dale, IL, USA). After injection, the syringe was kept in place for at least 5 min to prevent backflow before it was slowly retracted. The skin was then closed with sutures, and the mice were isolated for a period of 7 days according to TAU BSL-2 safety instructions.

Immunohistochemistry
Ten-days post-surgery, the animals were injected with an overdose of ketamine and xylazine and perfused transcardially with saline buffer (0.9%) followed by a fixative solution containing 4% Scientific Cat# A48270, RRID:AB_2896336) ; 4′,6-diamidino-2-phenylindole dihydrochloride (Sigma D8417). The sections processed for colorimetric immunohistochemistry were incubated with avidin-biotin-complex (Vector Laboratories PK-6100, RRID:AB_2336819) for 1 h at room temperature. Following washing with TBS, the peroxidase reaction was visualized using 0.05% 3,3′-diaminobenzidine (Sigma, D5637) and 0.03% hydrogen peroxide. Finally, sections were rinsed in TBS and mounted on SuperFrost Ultra Plus® slides (Epredia J1800AMNZ), dried at room temperature and counterstained using cresyl violet or hematoxylin, dehydrated and coverslipped with Eukitt (Sigma, 03989). The sections processed for immunofluorescence were rinsed in TBS and H2Od, let dry overnight, covered using DAKO mounting media (S3023) and kept at 4°C. To test the specificity of the secondary antibodies, we omitted the primary antibodies in some sections while maintaining the rest of the procedures. All the control sections exhibited a lack of specific staining.
The colorimetric signals were visualized using a NIKON ECLIPSE NI-E and a color camera DS-Ri2 (4908*3264 px de 7.3 µm), a Leica Thunder microscope and a Leica K3C USB3 color camera (3072 X 2048 interlines, 2.4 µm pixel size) or Nanozoomer (Hamamatsu). Immunofluorescence signals were visualized using a Zeiss LSM980 Airyscan laser-scanning microscope. Images were adjusted for brightness and contrast using ImageJ. Picture setup was achieved with Adobe Illustrator CS6. Full resolution was maintained until the micrographs were cropped and assembled; at which time they were adjusted to a resolution of 300 dpi. The brain regions were identified using a mouse brain atlas (6).

In situ hybridization using RNAScope technology and immunofluorescence
RNA in situ hybridization using RNAScope technology was performed as previously described (7). We used parallel series of the ones used for IHC and immunofluorescence experiments. On day one, slides were fixed in 4% PFA for 1 h at 4°C. Slides were then rinsed in PBS, followed by dehydration in increasing concentrations of ethanol (50%, 70%, 100% and 100%). Next, the sections were placed in the RNAScope oven at 60°C for 1 h, followed by the H2O2 blocking treatment and the antigen retrieval treatment included in the kit and washed in ethanol 100%.
Finally, the sections were left to dry overnight at RT. On the next day, a hydrophobic barrier was drawn around the brain sections. Sections were processed for the protease treatment III (provided in the kit) for 30 min at 40°C in a humid chamber using an RNAScope oven, followed by incubation with the probes for 2 h at 40°C and incubation with preamplifier and amplifier probes (AMP1, 40°C for 30 min; AMP2, 40°C for 30 min; AMP3, 40°C for 15 min). Slides were incubated in fluorescently labeled probes by selecting a specific combination of colors associated with each channel Opal 570 for Gad1 and Opal 690 for Gad2 mRNAs, following the instruction provided by the kit.
Once the in situ hybridization was finished, the sections were processed for immunofluorescence.
Positive and negative controls for the RNAScope were performed in parallel.
Immunofluorescence signals were visualized using a Confocal Zeiss LSM980 Airyscan 8Y or Dragonfly spinning disk coupled to a Nikon inverted microscope. Images were adjusted for brightness and contrast by using Imaris (RRID:SCR_007370) and Fiji. Picture setup was achieved with Adobe Illustrator CS6. Full resolution was maintained until the micrographs were cropped and assembled; at which time they were adjusted to a resolution of 300 dpi.

Voltage clamp recording in DK cells
DK cells (9) were infected with 200 physical particles/cell. The recordings were made 8-12 h postincubation, as described previously (8). Briefly, recordings were made using a Sutter IPA amplifier (Sutter Instrument, Novato, CA, USA). The pipette solutions contained: 140 mM CsF, 10 mM NaCl, 1 mM EGTA, 10 mM HEPES, 10 mM glucose, adjusted to pH 7.3 with CsOH. The external solution contained: 140 mM NaCl, 20 mM glucose, 10 mM HEPES, 1 mM MgCl2, 3 mM KCl, 1 mM CaCl2, adjusted to pH 7.35 with NaOH. The voltage dependence of activation was measured from a holding potential of -120 mV. Cells were depolarized for 20 ms to potentials ranging from -80 to +30 mV in 10 mV increments. The voltage dependence of inactivation was measured from a holding potential of -120 mV. Cells were depolarized for 500 ms to potentials ranging from -140 to +0 mV in 10 mV increments, followed by test pulses to -10 mV.

ECoG and depth electrode and recordings
Seven to ten days following vector injections, cortical or depth electrodes were implanted as previously described (2). Briefly, a midline incision was made above the skull, and fine silver wire electrodes (130 µm diameter bare; 180 µm diameter coated) were implanted. We used the previously formed injection holes for ECoG or depth electrodes. For hippocampal depth recordings, the wire electrodes (coated Platinum/Iridium wires,75 µm diameter bare) were lowered using the same stereotactic coordinates used for injection. A reference electrode was placed on the cerebellum; a ground electrode was placed behind the neck. The electrodes were connected to a Mill-Max connector and secured with dental cement before the skin was closed with sutures.
Following the surgery, the mice were given two to five days to recover before recording.
Video-depth/EcoG recordings lasting 2-4 h were obtained during the light period from freely moving mice, connected to a T8 Headstage (Triangle BioSystems, Durham, NC, USA), using a PowerLab 8/35 acquisition hardware and the LabChart 8 software (ADInstrumnts, Sydney, Australia). The electrical signals were recorded and digitized at a sampling rate of 1 kHz with a notch filter at 50 Hz. The analysis was performed using LabChart 8 (ADInstrumnts, Sydney, Australia). The ECoG signal was processed offline with a 0.5-100 Hz bandpass filter. Power spectral density was calculated using fast Fourier transform, with the Hann (cosine-bell) data window set to 50% overlap. We calculated the average of four to eight 30 s segments of the wakefulness, immobility and epileptic-free activity, but following a movement for each mouse, as determined by the video recording.
For the long-term video-EEG recordings depicted in Fig. 9B, the electrodes were implanted immediately following vector injection, and the mice were recorded for 5 days using a Neuralynx Digital Lynx data acquisition system (Neuralynx, Inc., Bozeman, MT, USA)

Thermally-induced seizures
Thermal induction was done about 1 month post-injection as described previously (1). Briefly, the mice were given 10 min to habituate to the thermal probe and the recording chamber. The baseline body temperature was measured, followed by an increase of 0.5°C every 2 min until 40.5°C or until a seizure was generated.

Behavioral analyses
Behavioral experiments were performed as described previously (2). Spontaneous alternation in the Y maze was assessed 5-14 days post-injection into the hippocampus and the open field test was conducted 9-17 days after treatment (hippocampal injection). Briefly, for the Y maze, the mice were placed in a symmetrical maze composed of three opaque white Plexiglas arms (each 35 cm L x 7.6 cm W x 20 cm H) and allowed free exploration for 10 min. For the open field test, mice were placed in the center of a square (50 x 50 cm) Plexiglas apparatus and their activity was recorded for 10 min. The novel object recognition test was performed 12-16 days after co-injection into the thalamus and hippocampus. On the first day, the mice were placed in a square (50×50 cm) Plexiglas apparatus for 15 min to allow for habituation. Then, two identical objects were introduced (small Lego blocks) and the mice were allowed to explore these objects for 10 min. On the second day, one of the objects was swapped with a novel object (a different Lego block) and the mouse was allowed to explore both objects for 5 min. One hour late, the novel object was moved to a new location in the arena, and the mouse was allowed to explore both objects for 5 min. Exploration of the object was counted when the mouse's nose was within 3 cm of the object.

Acute brain slice recordings
Hippocampal injections of CAV-GFP or CAV-SCN1A were performed at P21. Acute brain slices were made 72 -96 h later, as described before (10)       (B) Injection of CAV-SCN1A into the thalamus and hippocampus provides greater protection from thermally-induced seizures (sz). The solid lines are the same data presented in Figs. 6C, 8B, 9F, respectively, the shaded areas depict 95% confidence intervals. Note the lack of overlap of 95% confidence intervals following combined injection into the thalamus and the hippocampus, demonstrating greater protection from thermally-induced seizures.