Astroglial Cx30 sustains neuronal population bursts independently of gap‐junction mediated biochemical coupling

Abstract Astroglial networks mediated by gap junction channels contribute to neurotransmission and promote neuronal coordination. Connexin 30, one of the two main astroglial gap junction forming protein, alters at the behavioral level the reactivity of mice to novel environment and at the synaptic level excitatory transmission. However, the role and function of Cx30 at the neuronal network level remain unclear. We thus investigated whether Cx30 regulates neuronal population bursts and associated convulsive behavior. We found in vivo that Cx30 is upregulated by kainate‐induced seizures and that it regulates in turn the severity of associated behavioral seizures. Using electrophysiology ex vivo, we report that Cx30 regulates aberrant network activity via control of astroglial glutamate clearance independently of gap‐junction mediated biochemical coupling. Altogether, our results indicate that astroglial Cx30 is an important player in orchestrating neuronal network activity.

While astroglial Cx43 is a well-known regulator of neuronal physiology through channel and non-channel functions (Chever, Lee, & Rouach, 2014;Chever, Pannasch, Ezan, & Rouach, 2014;Clasadonte, Scemes, Wang, Boison, & Haydon, 2017;Meunier et al., 2017;Retamal & Sáez, 2014;Roux et al., 2015;Samoilova, Wentlandt, Adamchik, Velumian, & Carlen, 2008;Stehberg et al., 2012;Theis et al., 2003;Wang, Xu, Wang, Takano, & Nedergaard, 2012), the contribution of Cx30 to physiological and pathological neuronal activity as well as associated behavior remains poorly investigated. At the behavioral level, Cx30 alters the reactivity of mice to novel environments and object recognition memory (Dere et al., 2003). At the synaptic level, Cx30 tunes hippocampal excitatory synaptic transmission by determining the efficacy of astroglial glutamate clearance through an unprecedented regulation of astroglial morphology independent of gap-junction-mediated biochemical coupling. In fact, by regulating astroglia ramification and the extent of astroglial processes contacting synaptic clefts, Cx30 directly sets synaptic glutamate levels through clearance . Cx30 is thus a molecular determinant of astroglial synapse coverage controlling synaptic efficacy. In all, astroglial Cx30 functions are complex and can either promote or dampen synaptic transmission. However, what is the specific role of Cx30 on neuronal network activity and which function of Cx30 is involved are unknown. We, thus, investigated whether Cx30 favors or limits aberrant neuronal network activity and associated convulsive behavior. We here found in vivo that Cx30 expression is increased by kainate-induced seizures and regulates in turn the associated convulsive behavior in mice.
Accordingly, we show ex vivo that Cx30 modulates aberrant neuronal population bursts induced by increased excitability via control of astroglial glutamate uptake independently of gap-junction mediated biochemical coupling.

| MATERIALS AND METHODS
Experiments were carried out according to the guidelines of the European Community Council Directives of January 1st 2013 (2010/63/ EU) and of the local animal welfare committee (certificate A751901, Ministère de l'Agriculture et de la Pêche), and all efforts were made to minimize the number of animals used and their suffering. Cx30 −/− (−/−) mice were generated as previously described  and provided by K. Willecke, University of Bonn, Germany. C57Bl6 (+/+) mice were supplied by Charles River, France. Heterozygous mice carrying the knockout mutation were interbred to obtain homozygous strain.

| Immunoblot of Cx30 expression
Immunoblotting and quantification were performed as previously described . Hippocampi were collected in a small volume of cold SDS 2% containing a cocktail of protease inhibitors and phosphatase inhibitors (β-glycerophosphate (10 mM) and orthovanadate (1 mM)) to which Laemmli 5× buffer was added. Samples were sonicated, boiled 5 min and loaded on 4-12% polyacrylamide gels. Equal amounts of proteins were separated by electrophoresis and transferred onto nitrocellulose membranes. Membranes were saturated with 5% fat-free dried milk in triphosphate buffer solution and incubated overnight at 4 C with primary antibodies (GAPDH rabbit monoclonal antibody [Sigma], Cx30 rabbit polyclonal antibody). They were then washed and exposed to peroxidase-conjugated secondary antibodies (donkey anti-rabbit IgG HRP-conjugated secondary antibodies, Amersham Biosciences). GAPDH was used as loading control. Specific signals were revealed with the chemiluminescence detection kit (ECL, GE Healthcare). Semi-quantitative densitometric analysis was performed after scanning the bands with the imageJ software.

| Ex vivo electrophysiology
Acute transverse hippocampal slices (400 μm) were prepared as previously described (Chever et al., 2016) from 20 to 25 days-old wild type (+/+), Cx30 knockout (−/−) , and Cx30 T5M mice (Schütz et al., 2010). Slices were maintained at room temperature in a storage chamber that was perfused with an artificial cerebrospinal fluid (ACSF) containing (in mM): 119 NaCl, 2.5 KCl, 2.5 CaCl 2 , 1.3 MgSO 4 , 1 NaH 2 PO 4 , 26.2 NaHCO 3 , and 11 glucose, saturated with 95% O 2 and 5% CO 2 , for at least 1 hr prior to recording. Slices were transferred to a submerged recording chamber mounted on an Olympus BX51WI microscope equipped for infrared-differential interference (IR-DIC) microscopy and were perfused with ACSF at a rate of 1.5 ml/min at room temperature. Extracellular field and whole-cell patch-clamp recordings were performed. Stratum radiatum astrocytes were identified by their small cell bodies, low input resistance (~20 MΩ), high resting potentials Recordings were acquired with Axopatch-1D amplifiers (Molecular Devices, San Jose, CA), digitized at 10 kHz, filtered at 2 kHz, stored and analyzed on computer using Pclamp9 and Clampfit9 software (Molecular Devices).

| Immunohistochemistry
Saline and KA-injected mice were perfused with phosphate buffered saline (PBS) 4 hr after injection and their brain rapidly removed and frozen. Cryostat brain slices were then cut and fixed for 10 min at room temperature with 4% paraformaldehyde (PFA), washed three times with PBS and pre-incubated 1 hr with PBS-1% gelatin in the presence of 1% Triton-X100. Brain slices were then immunostained overnight at 4 C for GFAP (1:500, mouse anti-GFAP antibody, Sigma-Aldrich) and Cx30 (1:500, rabbit anti-Cx30 antibody, ThermoFisher) and washed in PBS three times. Appropriate secondary antibodies (goat anti-mouse IgG conjugated to Alexa 488 and goat anti-rabbit IgG conjugated to Alexa 561, 1:200, ThermoFisher) were finally applied for 1-2 hr at room temperature, followed by DAPI staining (1:2,000, ThermoFisher).
After several washes, brain slices were mounted in fluoromount (Southern Biotechnology) and examined with a spinning-disk confocal microscope (Eclipse Ti, Nikon) equipped with CMOS camera (Photometrics). Stacks of consecutive images were taken with a 60× objective at 500 nm intervals and acquired sequentially with 3 lasers (405, 488, and 561 nm). Z-projections were then reconstructed using ImageJ software and average fluorescence intensity for Cx30 was measured in hippocampal CA1 stratum radiatum astrocytes.

| Statistics
All data are expressed as mean ± SEM. Statistical significance for between-group comparisons was determined by unpaired or paired ttests. Fisher exact test was used to compare distributions.

| Drugs
Picrotoxin was obtained from Sigma and all other drugs were from Tocris.
These data indicate that Cx30 deficiency decreases behavioral seizures severity.

| Cx30 controls neuronal population bursts
To investigate whether Cx30 alters neuronal network activity, we recorded neuronal population bursts in the hippocampal CA1 area. were comparable in T5M and +/+ mice (Figure 3a,c).

| DISCUSSION
We here show in vivo that Cx30 expression is increased after kainic acid injection and that it regulates the severity of kainate-induced behavioral seizures. Accordingly, we also found that Cx30-deficient mice display ex vivo less frequent neuronal population bursts due to an enhanced glial glutamate clearance, which tunes down neuronal network activity. Interestingly, we observed that the effect on neuronal burst firing is independent from gap junction-mediated biochemical coupling.  Conversely, decreased Cx30 levels have been found during excitotoxic brain injury in reactive astrocytes located in the area of neuronal death (Koulakoff, Ezan, & Giaume, 2008) and during astrocyte transformation into highly motile glioma cells (Princen et al., 2001). Furthermore, patients with major depression disorders or suicide completers show decreased brain levels of Cx30 ( (Condorelli et al., 2002;Takahashi, Vargas, & Wilcox, 2010

| Cx30-mediated regulation of bursts and behavioral seizures
We have shown that Cx30, which is upregulated after kainate injection, worsen kainate-induced behavioral seizures and regulates neuronal population bursts. These findings, together with physiological and pathological modulations of Cx30 expression (Condorelli et al., 2002;Koulakoff et al., 2008;Liu et al., 2013;Rampon et al., 2000;Roux et al., 2011), suggest reciprocal regulations between neuronal activity and astroglial Cx30. Indeed, enhanced neuronal activity may boost Interestingly, we have found that Cx30-deficient mice display enhanced glial glutamate clearance, which contributes to decrease the frequency of hippocampal neuronal population bursts. Therefore, our results identify the Cx30-mediated control of astroglial glutamate clearance as a critical factor controlling neuronal bursting activity. The enhanced astroglial glutamate transport may have the potential to tune the threshold for neuronal burst initiation and the recovery of neurons to the resting state by removing rapidly glutamate from synaptic and extrasynaptic sites. These effects most likely result from a reduction in glutamate receptor activation and the accompanied depolarization of neurons.
We previously showed that the increased glutamate clearance in Cx30 deficient mice results from changes in astrocytic morphology and coverage of synapses independently of gap junction-mediated biochemical coupling. This leads structurally to a closer proximity of GLT1 to synaptic active zones and functionally to a depression of basal excitatory synaptic transmission . In the present study, revealing the role of Cx30-mediated enhanced glutamate clearance in bursting activity independently of biochemical coupling, a similar mechanism is likely to be at play. Yet, we cannot exclude that Cx30 hemichannels also contribute to the control of bursting activity. However, although Cx30 can form hemichannels (Nielsen, Alstrom, Nicholson, Nielsen, & MacAulay, 2017), to date, their presence and functional activation have not been reported in astrocytes. Furthermore, there is presently no pharmacological tool to selectively and acutely inhibit their activity. These limitations hinder the study of their activation and selective implication in the regulation of neuronal activity. The development of specific Cx30 hemichannel blockers should allow future investigation of their contribution to bursting activity.
Finally, our in vivo results indicate that astroglial glutamate clearance play a crucial role in reducing the severity of seizures. Consistent with our data, a reduction in astroglial glutamate transporter expression associated with increased extracellular glutamate levels has been found in the sclerotic hippocampus of patients with temporal lobe epilepsy (Cavus et al., 2005(Cavus et al., , 2008Jabs, Seifert, & Steinhäuser, 2008;Mathern et al., 1999;Proper et al., 2002) as well as in a tuberous sclerosis epilepsy model (Wong et al., 2003). Furthermore, knockout mice for the glial glutamate transporter GLT-1 develop spontaneous seizures and hippocampal pathology resembling those observed in temporal lobe epilepsy patients with hippocampal sclerosis (Petr et al., 2015;Sugimoto et al., 2018;Tanaka et al., 1997). In addition, pharmacological block of GLT-1 decrease the threshold to trigger epileptiform activity and increase the occurrence of spontaneous epileptiform discharges in the rat cortex (Campbell & Hablitz, 2004. Conversely, increase in GLT-1 expression with the beta-lactam antibiotic ceftriaxone (Rothstein et al., 2005) has already been shown to have anti-convulsant effects (Jelenkovic et al., 2008). In all, these findings suggest that efficient astrocytic glutamate uptake by GLT-1 may be essential to counteract epileptogenesis. Thus, disrupting astroglial Cx30 to enhance GLT-1 activity could represent a novel therapeutic strategy against seizures.