Receptor tyrosine kinase EphA7 is required for interneuron connectivity at specific subcellular compartments of granule cells

Neuronal transmission is regulated by the local circuitry which is composed of principal neurons targeted at different subcellular compartments by a variety of interneurons. However, mechanisms that contribute to the subcellular localisation and maintenance of GABAergic interneuron terminals are poorly understood. Stabilization of GABAergic synapses depends on clustering of the postsynaptic scaffolding protein gephyrin and its interaction with the guanine nucleotide exchange factor collybistin. Lentiviral knockdown experiments in adult rats indicated that the receptor tyrosine kinase EphA7 is required for the stabilisation of basket cell terminals on proximal dendritic and somatic compartments of granular cells of the dentate gyrus. EphA7 deficiency and concomitant destabilisation of GABAergic synapses correlated with impaired long-term potentiation and reduced hippocampal learning. Reduced GABAergic innervation may be explained by an impact of EphA7 on gephyrin clustering. Overexpression or ephrin stimulation of EphA7 induced gephyrin clustering dependent on the mechanistic target of rapamycin (mTOR) which is an interaction partner of gephyrin. Gephyrin interactions with mTOR become released after mTOR activation while enhanced interaction with the guanine nucleotide exchange factor collybistin was observed in parallel. In conclusion, EphA7 regulates gephyrin clustering and the maintenance of inhibitory synaptic connectivity via mTOR signalling.

Briefly, the maze was placed 50 cm above the floor and consisted of two open arms and two closed arms (with 30 cm high Plexiglas walls and no roof), arranged in a way that similar arms are opposite to each other. After 5 min habituation to the room, each animal was placed in the center of the maze facing an open arm and was allowed to explore the arena freely for 5 min.
Behaviour was recorded and analyzed via the EthoVision XT8 tracking system.
For the two way shuttle avoidance task, adult rats were conditioned in a two way shuttle avoidance box (Panlab, Harvard apparatus, Spain) placed in a dimly-lit, ventilated and soundattenuated cupboard. The rectangular chamber (60 x 26 x 28 cm) was divided by an opaque partition with a small passage (10 x 8 cm) into two equal sized cubicles. Both compartments' metal grid floors are weight sensitive; micro-switches transmit information on the rat's location to a computer-controlled and automated data collection program managing the conditioned stimulus (CS) and the unconditioned stimulus (US) presentations. The CS was a tone produced by loudspeakers located on the distal walls of the compartments and the US is an electric foot shock. Information was also recorded concerning the rats' responses, which can be avoidance (shuttle during the CS), escape (shuttle during the US) or no escape (no shuttle during the US).
Two training sessions were performed on two consecutive days. The first session began with 10 minutes of habituation to the apparatus and the second session began with one minute of habituation. Sessions consisted of 75 trials; each starting with the delivery of the CS for 10 sec (75 db, 3000 Hz) immediately followed by the US (0.8mA foot shock, 10 sec maximum) with an inter-trial interval of 30 ± 7.5 sec.

In vivo electrophysiology
About 10 days after behavioral testing, in vivo electrophysiology was performed on all rats.
Adult rats were anesthetized (40% urethane, 5% chloral hydrate in saline, max. 0.5 ml/100 g i.p.) and placed in a stereotaxic apparatus. After electrode insertion, recording was allowed to stabilize for 30 min. Input-output relations were then examined, using average of 5 successive responses at 0.1 Hz for increasing stimulus intensities. Baseline field potential responses were recorded with a stimulus intensity of 50 % of the intensity that evoked maximum spike amplitude (monopolar pulses, 100 µs duration).
Baseline measurements were then taken for 30 min (at 0.1 Hz) followed by three local circuit activity protocols.
Local circuit activity and frequency dependent modulation: frequency dependent inhibition was determined as described previously (Rosenblum et al., 1999;Sloviter, 1991). Ten baseline pulses were delivered to the perforant path at 0.1 Hz, followed by 10 pulses delivered at 1 Hz.
Inhibition index was measured by the average population spike amplitude of the 1 Hz stimulation divided by that of prior stimulation at 0.1 Hz.
Local circuit activity and paired-pulse inhibition: As described in earlier studies (Andersen et al., 1966;Richter-Levin and Segal, 1991;Sloviter, 1991), paired-pulse inhibition was measured by applying five pairs of two constant stimuli to the perforant path at inter-stimuli interval of 15 ms. The inhibition index was measured by the averaged population spike amplitude of the response to the second stimulus divided by that of the first stimulus.
Local circuit activity and commissural modulation: The DG commissural pathway was activated by stimulating the contralateral DG at intervals of 15 and 30 ms prior to perforant path stimulation as described before (Richter-Levin and Segal, 1991;Yarom et al., 2008). This was repeated five times. The inhibition/excitation index was measured by the averaged population spike amplitude of the five second pulses divided by the average of the baseline measurement.

LTP induction
Following the application of local circuit activity protocols, some of the rats were also tested for LTP induction. Baseline measurements were again taken for 30 min. Theta burst stimulation (TBS) of the perforant path was then used to induce LTP. The TBS protocol consisted of three sets of 10 trains each, each train consisting of 10 pulses at 100 Hz, at baseline stimulation intensity (inter-train interval: 200 ms; inter-set interval: 1 min). Following TBS, recording at 0.1 Hz continued for 90 min.

Assessment of injection site
After completion of electrophysiological recordings, rats were anesthetized with a pentobarbital overdose and transcardially perfused with 200 ml of 0.9% sodium chloride, followed by 250 ml of 4% paraformaldehyde (4ºC) in 0.01 M phosphate buffered saline (PBS). Brains were removed, post-fixed overnight at 4ºC in the same fixative, and immersed in a 30% sucrose/PBS solutions. Free floating 30 µm coronal sections were collected with a cryostat (Leica, Wetzlar, Germany) in PBS azide 0.05% and stored at 4ºC until use. Sections were mounted on glass slides, let dry and coverslipped with immu-mount media (Thermoscientific, Waltham, and United State). EGFP expression was checked via epifluorescence microscopy (excitation at 480 nm) to validate appropriate transduction in the dentate gyrus for both control-and EphA7 knockdown virus injected rats. Only rats with bilateral infection restricted to the dentate gyrus were included in the electrophysiological and behavioral analyses.