Discrete subicular circuits control generalization of hippocampal seizures | Panda Anku


All procedures were approved by the guidelines of the Animal Advisory Committee of Zhejiang University and in complete accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. C57BL/6J (Wild-type) and CaMKIIα-Cre (Stock number 005539) mice were used and genotyped in line with the protocols provided by the Jackson Laboratory. Mature male mice (8–16 weeks) were used in viral tracing, multi-unit recording, and behavior experiments. Mice at 4- to 6-weeks-old were used for in vitro electrophysiology. All mice were group-housed (four to six) in plastic cages prior to surgery with a 12 h light/dark cycle. The ambient temperature was kept about 23–26 °C and humidity was about 50–60%. After surgery, they were housed two to three per cage for better recovery. Behavior tests were conducted during the light cycle.

Viral constructs

For calcium photometry, AAV2/9-CaMKIIα-GCaMP6(s) (viral titers: 5.6*1012 particles/mL) and AAV2/9-EF1α-DIO-Axon-GCaMP6(s) (viral titers: 5.6*1012 particles/mL) were injected into the subiculum of wild-type and CaMKIIα-Cre mice, respectively. For selective optogenetic activation or hyperpolarization of pyramidal neurons and their projection terminals, AAV2/8-CaMKIIα-hChR2-eYFP (viral titers: 1.7*1013 particles/mL) or AAV2/9-CaMKIIα-ArchT-eGFP (viral titers: 1.7*1013 particles/mL) were injected into the subiculum or ANT of wild-type mice. For the control group, AAV2/9-CaMKIIα-eGFP (viral titers: 1.7*1013 particles/mL) was injected. For chemogenetic inhibition of pyramidal neuron projecting terminals, AAV2/9-CaMKIIα-hM4Di-mCherry (viral titers: 1.7*1013 particles/mL) was injected into the subiculum of the wild-type mice. For the control group, AAV2/9-CaMKIIα-mCherry (viral titers: 1.7*1013 particles/mL) was injected. To selectively knock down HCN1 expression in the subiculum, we used the following sequence: CCTCCAATCAACTATCCTCAA69. Cav2-cre (viral titers: 3.0*1012 particles/mL) was injected into the ANT or EC of wild-type mice, then AAV2/9-EF1α-DIO-miR30shRNA (Hcn1)-mCherry (viral titers: 7.7*1012 particles/mL) was injected into the subiculum. Viruses were purchased from OBio Technology (China), except for AAV2/9-EF1α-DIO-Axon-GCaMP6(s) and Cav2-cre from Taitool Bioscience (China). For retrograde tracing, cholera toxin subunit B conjugated to Alexa-555 or Alexa-647 (CTB-555 and CTB-647, Thermo Fisher, USA) diluted in PBS solution at a concentration of 1% wt/vol was used. 4–7 days after injections, mice were perfused for histology.

Stereotactic injections and surgeries

For virus injections, mice were anesthetized with sodium pentobarbital (60 mg/kg, i.p.) and mounted in a stereotaxic apparatus (RWD Life Science, China). Injections were targeted into the SUB (antero-posterior (AP) −3.4 mm; lateral (L) −2.0 mm; ventral (V) −1.7/−1.8 mm), ANT (AP −0.6 mm; L −0.7 mm; V −3.5 mm), EC (AP −4.8 mm; L −4.0 mm; V −3.0 mm), MMB (AP −2.8 mm; L −0.2 mm; V −4.9 mm), and NAc (AP + 1.4 mm; L −0.8 mm; V −5.0 mm) using a glass micropipette attached to a 1 μL syringe at 60 nL/min. For behavioral experiments, injection volumes were 200 nL for the SUB, 150 nL for the ANT and EC. For AAV anterograde tracing, injection volumes were 100 nL for the SUB. For CTB retrograde tracing, injection volumes were 120 nL for the ANT and EC, 80 nL for the MMB and NAc. For Cav2-cre injection, volumes were 150 nL for the ANT and EC. After each injection, the needle was left in place for 10 min before withdrawal.

For KA injection, mice were anesthetized with ~2% isoflurane. KA (0.25 μg in 0.5 μL saline, ab120100, Abcam) was injected into the dorsal CA1 (AP −2.1 mm; L −1.2 mm; V −1.6 mm) with a glass micropipette attached to a 1 μL syringe at 100 nL/min. After each injection, the needle was left in place for 10 min before withdrawal.

For implanting electrodes or optical cannulas, mice were anesthetized with ~2% isoflurane three weeks after viral delivery, and then twisted-bifilar stainless electrodes (795500, 0.127 mm diameter, A.M Systems, USA) were implanted into the right hippocampal CA3 (AP −2.9 mm; L −3.1 mm; V −3.1 mm) for both kindling stimulation and EEG monitoring. Optical cannulas (0.2 mm diameter, Inper, China) were separately lowered into the following areas: SUB (AP −3.4 mm; L −2.0 mm; V −1.7/−1.8 mm), ANT (AP −0.6 mm; L −0.7 mm; V −3.5 mm), EC (AP −4.8 mm; L −4.0 mm; V −3.0 mm), MMB (AP −2.8 mm; L −0.2 mm; V −4.9 mm), and NAc (AP + 1.4 mm; L −0.8 mm; V −5.0 mm), allowing for light stimulation. Cannulas (0.41 mm diameter, RWD Life Science, China) were implanted into the SUB, ANT, and EC for drug delivery. Then three screws were placed over the skull to fix the dental cement, two of which were placed over the motor cortex and cerebellum to serve as the ground and reference electrodes, respectively. Viral expressions and locations of the electrodes were verified after all behavior tests. We only included mice with correct electrode, cannula emplacement, and viral expression (approximately 30% mice were excluded under these criteria).

Calcium fiber photometry

Fiber photometry was performed 1 week after electrode and fiber implantation surgeries in mice expressing Gcamp6(s). The fiber photometry system contains a 488 nm diode laser (OBIS 488LS, Coherent, USA), a dichroic mirror (MD498, Thorlabs, USA), and coupled into an 0.23 mm, 0.37 NA optical fiber with a 10× objective lens (Olympus) and a fiber launch (Thorlabs). The power of the laser intensity between the fiber tip and the mice brain regions ranged from 0.01 to 0.03 mW to avoid bleaching. The collected GCaMP fluorescence was converted to voltage signals by a digital amplifier (C7319, Hamamatsu, Japan). The converted signals were recorded at 100 Hz for 200 s (100 s baseline and 100 s after kindling stimulation). Data were further analyzed by MATLAB (version R2017b, MathWorks, USA). The values of fluorescence change were shown as ΔF/F with the following equation: (ΔF/F) = (F − F0)/F0, among which F represented the current value of signal, F0 represented the average value of baseline signals between 90 and 100 s. The data are presented as time-related peri-event plot and heatmap.

Hippocampal kindling model

For rapid hippocampal kindling model, after 1 week of recovery, EEG of the right hippocampal CA3 was recorded by a Neuroscan system (Compumedics, Australia). Kindling stimulations were all conducted during wakefulness. The after-discharge threshold (ADT) of each mouse was determined (monophasic square-wave pulses, 20 Hz, 1 ms/pulse, 40 pulses) by a constant current stimulator (SEN-7203, SS-202J, Nihon Kohden, Japan) as in our previous study16. The stimulation current started at 40 μA and then increased 20 μA each time, and the minimal current that produced at least 5 s ADD was defined as ADT. Only mice with ADT less or equal than 200 μA were used later. All mice received ten suprathreshold stimulations (monophasic square-wave pulses, 400 μA, 20 Hz, 1 ms/pulse, 40 pulses, 30 min-interval per stimulation) daily from the next day on. Seizure severity was scored following the criteria of the Racine scale70: 1. facial movement; 2. head nodding; 3. unilateral forelimb clonus; 4. bilateral forelimb clonus and rearing; and 5. rearing and falling. Stages 1–3 are considered as FSs and stages 4–5 are sGSs. The length of ADD was defined as the duration between the moment of kindling stimulation and the end of paroxysmal discharge event in EEG. These continuous discharges showed an average amplitude >3 times versus baseline and isolated post-paroxysmal spikes were not calculated in the ADD according to previous study70. Behavioral assessment and EEG analysis were performed by well-trained experimenters blinded to the group allocation. When mice had three successive stage 5 seizures, they were regarded as fully kindled. During this period, mice exhibited stable sGS after each kindling stimulation. To determine the effect of specific interventions on sGS, we used suprathreshold current (200 μA, 20 Hz, 1 ms/pulse, 40 pulses, 30 min-interval per stimulation) to induce sGS. The EEG power recorded was calculated and analyzed offline by a software package (Scan 4.5) in the Neuronscan System. For electrical lesion studies, we used 1 mA, 10 s, direct current stimulation in the subiculum.

Photo stimulation

Blue (473 nm) or yellow (589 nm) light was delivered by a 200 μm diameter optical fiber (Inper) connected to the laser by a Master-8 (AMPI, Israel) commutator. The optical fiber was cut flat, and power of the laser was adjusted to 5 mW. Immediately before the mouse was placed in the chamber, the cannula cap was removed, and an optical fiber was directly inserted. In the hippocampal kindling model, optical stimulation was delivered immediately after the end of kindling stimulations. For optogenetic activation experiments, 473 nm blue light (20 Hz, 10 ms/pulse and 600 pulses) was used. For optogenetic inhibition experiments, 589 nm yellow light (continuous 30 s) was used.

Intra-hippocampal KA model

In this model, SE is typically induced via intra-hippocampal injection of KA (as mentioned in ‘Stereotactic injections and surgeries’ section), which further caused spontaneous recurrent seizures in the following several months. Two months after KA and AAV-CaMKIIα-hM4Di-mCherry or AAV-CaMKIIα-mCherry injection, EEG of the hippocampal CA3 was continuously recorded in freely moving surviving mice through a Powerlab system (AD Instruments, Australia) at a sample rate of 1 kHz, which was synchronized with video monitoring 8 h/day for 3 days as baseline (Pre). During this period, each mouse was intra-ANT or EC injected with saline for 500 nL daily before recording start. A FS was defined as a sharp paroxysmal event that continued more than 10 s and had an average amplitude > 3 times versus baseline and frequency > 2 Hz. A sGS was defined as an event that continued more than 30 s and had a period of post-inhibition, accompanied by a marked tonic-clonic behavioral seizure71. Only mice with detectable seizures were given CNO injections (The percentage of mice with detectable seizures in the study was about 40–50%). These mice were locally injected with CNO (1.0 mM, 500 nL in the ANT or EC) daily for 3 days to test the effect of chemogenetic inhibition of ANT- or EC-projecting subicular terminals on spontaneous seizures. Mice with mCherry were also given CNO via cannula to test its effect on seizures. The concentration and volume of intra-brain CNO injections were referenced in previous studies72,73. Mice were then injected with saline in the next 3 days for post treatment.

In vivo multi-unit recording and analysis

Multi-unit recording16 was performed as following: the body temperature of mice was kept at 37 °C by a heating pad. 12 microelectrodes (761500, 0.025 mm, A.M Systems) were twisted into a bundle to form recording electrodes, which had an impedance of 1–2MΩ and were combined with an optical fiber terminal to keep stiff. Neuronal activities were sampled by the Cerebus acquisition system. Recorded units were identified by low firing rate (<10 Hz), wide spike waveform (>0.3 ms) and flat auto-correlograms. They were post positioned for their sub-regional locations in the SUBa (the deep part of the subiculum) or SUBb (the superficial part of the subiculum). In particular, when electrodes were positioned at AP −3.3 to −3.5 mm, units at depths of −1.30 to −1.65 mm were located in SUBa and units at depths of −1.65 to −1.90 mm were located in SUBb; when electrodes were positioned at AP −3.5 to −3.7 mm, units at depths of −1.70 to −1.80 mm were located in SUBa and units at depths of −1.80 to −2.00 mm were located in SUBb; the rest that were not in these locations or whose locations could not be identified were excluded. A firing pattern that had more than three spikes with less than 170 ms interval was defined as one burst. Each burst had more than 200 ms external interval. We considered units with >5 bursts per min as bursting units74.

Subiculum slice preparation

4 weeks old wildtype mice were injected with CTB-555 3–5 days before slice experiments. Under sodium pentobarbital (100 mg/kg, i.p.) anesthesia, they were decapitated, and the brain was quickly removed and placed in an ice-cold solution containing (in mM): 110 Choline chloride, 2.5 KCl, 1 NaH2PO4, 0.5 CaCl2, 7 MgCl2, 20 Glucose, 1.3 Ascorbate acid, 0.6 Na-pyruvate, 25 NaHCO3 (pH 7.4, oxygenated with 95% O2 and 5% CO2). The 300 μm coronal or sagittal slices containing the hippocampus were cut using a vibratome (VT1000, Leica, Germany). Slices were then placed into a chamber filled with artificial cerebrospinal fluid (ACSF) containing (in mM): 120 NaCl, 11 Dextrose, 2.5 KCl, 1.28 MgSO4, 2.5 CaCl2, 1 NaH2PO4, and 14.3 NaHCO3 (pH 7.4, oxygenated with 95% O2 and 5% CO2) and incubated at 34.7 °C for 0.5 h and then maintained at 25 °C for further experiments.

In vitro electrophysiology and analysis

For slice recordings, slices were kept at 25 °C in a recording chamber perfused with 3 mL/min ACSF, which contains the same substances as the storage ACSF. Patch pipettes were pulled from glass capillaries and at resistances of 6–9 MΩ, which contained (in mM): 35 K-gluconate, 110 KCl, 10 HEPES, 2 MgCl2, 2 Na2ATP, and 10 EGTA. Patch-clamp recordings were performed by an EPC10 patch-clamp amplifier (HEKA Instruments, Germany), with a low-pass filter at 3 kHz, and a sample rate of 10 kHz. The series resistance and capacitance were compensated after a stable Gigaseal. Recordings were typically performed 3 min after break-in. Data were analyzed by Clampfit software (Molecular Devices, USA).

For current–voltage (I/V) curves, hyperpolarizing to depolarizing current gradients (−250 to 250 pA, 50 pA each step, 500 ms) were injected. To plot the number of bursting spikes as a function of membrane potential, patch recordings were additionally performed varying the membrane potential from −55 mV to −70 mV at gradient of 5 mV. The voltage sag ratio was calculated with the following equation at −200 pA: sag ratio = (VbaseVmin)/(VbaseVsteady), among which Vbase represented the resting membrane potential, Vmin is the hyperpolarizing current that induced minimum voltage, and Vsteady is the averaged voltage caused by the inject current. The numbers and frequency of burst AP were calculated at the 100 pA depolarizing inject current. Input resistance (Rin) was calculated from the gradient in the linear phase of current–voltage plots following responses to hyperpolarizing step current injection. The membrane time constant was fit by an exponential function of the membrane potential change in response to rectangular hyperpolarizing current injection that induced small (3–5 mV) voltage deflections.

For AP properties, depolarizing current pulses (increased 5 pA each step) were applied to measure the threshold, amplitude, and the half-wave width of the APs. We analyzed the first spike (first single spike for bursting neurons) induced by the minimum depolarizing current (Rheobase). AP amplitude was defined as the voltage from the AP threshold to the AP peak, and the AP half-width was calculated as the duration at half-maximal amplitude.

For HCN channel properties, tail currents were measured by stepping the membrane potential from an initial holding potential of −60 mV to test potentials of −130 mV to −60 mV in 10 mV increments at an interval of 200 ms. In slice pharmacology, recordings were performed 20 min after incubation of 20 μM ZD7288 (ab120102, Abcam) or 5 μM TTA-P2 (T-155, Alomone) until an apparent steady-state effect was achieved.

Single pulse measurement

For single pulse measurement, an opto-electrode was implanted into the ANT of CaMKIIα-Arch SUB or CaMKIIα-eGFP SUB mice for EEG recording and light delivery, a cannula-electrode was implanted into the subiculum for single pulse stimulation and drug injection, and an electrode was implanted into the CA3 for kindling stimulation. After recovery for 1 week, experiments included three parts75. First, the test stimulation current was determined before CA3 kindling. The stimulation intensity in the subiculum that induced half of the maximum voltage in the ANT was considered as the test current. Second, all mice received ten suprathreshold stimulations (monophasic square-wave pulses, 400 μA, 20 Hz, 1 ms/pulse, 40 pulses) daily from the next day on until they were fully kindled (three successive stage 5 seizures). During this period, the first group of CaMKIIα-eGFP SUB mice and CaMKIIα-Arch SUB mice received 30 s yellow light stimulation in the ANT after each kindling, and the second group of CaMKIIα-eGFP SUB mice were intra-subicular injected with ZD7288 daily before the kindling stimulation. Finally, the single pulse test was performed (monophasic square-wave pulses 100 μs/pulse, 10 s interval, 10 pulses). The results of the single pulse test were compared before and after sGS acquisition in control, optogenetic inhibition in the ANT of CaMKIIα-Arch SUB mice, and ZD7288 (50 μM, 500 nL) injection in the subiculum. Peak potential amplitude was defined as the voltage from the field EPSP (fEPSP) threshold to the fEPSP peak. fEPSP slope was defined as the ratio from peak potential amplitude versus the time from the fEPSP threshold to the fEPSP peak.

Histology and quantification

Mice were deeply anesthetized with pentobarbital and transcranially perfused with 0.9% saline and followed by 4% paraformaldehyde in 0.1 M phosphate buffer. Notably, in c-Fos staining experiments, animals were sacrificed 1.5 h after seizures. Brains were removed and stored in the same fixative overnight at 4 °C and then dehydrated in 30% sucrose. Coronal or sagittal sections were cut on a freezing microtome (Themo Scientific, USA) at 40 μm. The primary antibodies used were as follows: rabbit monoclonal anti-CaMKII (1:800; ab134041, Abcam), guinea pig polyclonal anti-c-Fos (1:2000; 226004, Synaptic Systems), rabbit monoclonal anti-NeuN (1:800; MABN140, Millpore) and mouse monoclonal anti-HCN1 (1:500; MAB6651, Abnova). The primary antibodies were incubated overnight at 4 °C and rinsed with PBS. Then Alexa-488, Alexa-594, or Alexa-647 conjugated secondary antibodies were used at 1:800 for 2 h at room temperature. The secondary antibodies used were as follows: Alexa Fluor® 488 or 594 AffiniPure Donkey Anti-Rabbit IgG (H + L) (Jackson ImmunoResearch, 711-585-152 or 711-545-152), goat Anti-Guinea pig IgG H&L (Alexa Fluor® 647) (ab150187, Abcam), goat Anti-Mouse IgG H&L (Alexa Fluor® 647) (ab150115, Abcam). Sections were then washed and mounted on slides with media containing DAPI (Vectashield Mounting Media, Vector Labs). Images were captured by a confocal microscopy (SP8, Leica).

Image analyses and quantification were performed using ImageJ (version 1.52a) software. For c-Fos+ and NeuN+ cell quantification, we counted the number of cells that exhibited c-Fos+ or NeuN+ immunoactivity within the corresponding nucleus of three representative coronal slices (anterior, intermediate, and posterior), and then calculated the mean value for each mouse. The number of mice used in each experiment was indicated in figure legends.

Statistics and reproducibility

Data are presented as means ± SEM. All experiments were repeated at least two times independently with similar results. Data analysis was performed blind to the experimenters using SPSS (version 17.0, IBM) and Prism 8 (Graphpad Software). Data with Gaussian distributions were analyzed by one-way or two-way ANOVA followed by post hoc Dunnett’s, Tukey’s or Scheffe’s test for multiple comparisons, unpaired t-test, or paired t-test for statistical significance when appropriate. Non-normally distributed data were analyzed by Friedman with post hoc Dunn’s test for multiple comparisons, Kruskal–Wallis test or Wilcoxon test. Incidence data was analyzed by Fisher’s exact test. Significance is reported in the figure legends. For all analyses, the tests were two-tailed and the P-value <0.05 was considered statistically significant. Detailed statistic parameters are provided in the Supplementary Table S2 with the paper.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Leave a Comment