In the search for specific inhibitors of eEF2K, also known as CaMKIII (Ryazanov et al., 1997; Middelbeek et al., 2010) the small molecule inhibitor A-484954 was identified from an Abbott compound library using high throughput screening (Chen et al., 2011). This compound possesses a half-maximal inhibitory concentration (IC50) against eEF2K of 0.28 μM.

To validate the inhibitory effect of A-484954 on eEF2K, we performed a biochemical protein analysis of the eEF2K substrate by eEF2 phosphorylation. We did not determine the total amount of eEF2 because the time between drug application and protein phosphorylation analysis was short and significant de novo protein synthesis or degradation of eEF2 was unlikely to have taken place. To this end, 5 μM A-484954 was applied to the eEF2K substrate for 8, 16, or 32 min, followed by snap freezing and storage at -80°C. On the day of analysis, the CA1 region was isolated and the resulting eEF2 phosphorylation level was examined (Yuanxiang et al., 2014). The western blots indicated that A-484954 significantly prevented the phosphorylation of eEF2 (Figure 1A). After verification of the efficient inhibition of eEF2K by A-484954, we looked into the effects of eEF2K inhibition on hippocampal synaptic transmission. We observed that the inhibition of eEF2K by A-484954 (5 μM) 20 min after stable baseline recordings resulted in a fast potentiation of fEPSPs (131 ± 3.8% at 40 min; n = 9) (Figure 1B), which differed significantly from drug-free experiments from 20 min onward. The fEPSP values were 105 ± 2.0% at 40 min (n = 9; Figure 1C).

To determine whether the potentiation mediated by eEF2K inhibition required evoked synaptic activity, we took a fEPSP recording without stimulation of the fEPSPs during the initial 20 min of drug application (Figure 1D). Under these conditions, we observed a significantly different (p < 0.01) fEPSP potentiation (125 ± 3.6% at 45 min; n = 4) in comparison with drug-free experiments (control: 102 ± 2.3% at 45 min; n = 4). Because the effect of A-484954 on synapses was independent of evoked synaptic transmission, all synaptic inputs within the stratum radiatum were probably changed.

To identify the origin of the effect of the drug on synaptic transmission, we analyzed the paired-pulse (PP) ratio of fEPSPs and the frequency and amplitude of miniature excitatory postsynaptic currents (mEPSCs). Using field potential recordings of synaptic transmission, we detected that the PP ratio of fEPSPs decreased significantly after drug application for all inter-stimulus intervals tested in comparison with PP ratios before drug application (p < 0.05; Figure 2). We concluded that these data indicated a modulation of the presynaptic release probability, which we aimed to investigate in detail using whole-cell voltage clamp analysis of evoked EPSCs and mEPSCs (Figures 3 and 4).

To verify whether a presynaptic origin of the potentiation mediated by eEF2K inhibition underlies the potentiation of synapses, we performed whole-cell voltage clamp recordings of evoked EPSCs using a paired pulse stimulation protocol (interstimulus interval: 50 ms) in hippocampal CA1 neurons. The recordings revealed that the application of A-484954 significantly reduced the paired pulse ratio of evoked EPSCs (Figure 3). We also performed whole-cell voltage clamp recordings of mEPSCs in CA1 pyramidal cells. Tetrodotoxin (1 μM) and bicuculline (10 μM) were added to prevent action potential-mediated glutamate release and to block inhibitory postsynaptic currents, respectively. The recordings revealed that the application of A-484954 evoked an increase in mEPSC events over time (Figures 4A,B) with no alteration of their amplitudes (Figure 4C) in comparison with drug-free conditions. Thus, inter-event intervals became smaller (Figure 4D), and the maximal inter-event intervals for before and during drug application shifted from 16.5 to 6.5 s (Figure 4E) respectively. The significantly higher probability of mEPSC appearance after drug application was also represented by an increase in mEPSC frequency in comparison with the value before drug application (Figure 4F).

eEF2K and one of its substrate – the eEF2 – has been linked to the regulation of protein synthesis, however, also other substrates of eEF2K has been identified that might promote protein-synthesis independent regulation of cellular processes. Thus, we were interested to determine whether the effect of A-494854 on synaptic transmission was dependent on protein synthesis. To this end, we pre-incubated acute hippocampal slices with the protein synthesis inhibitor anisomycin (20 μM) for 1 h before the baseline recording. After an additional 20-min baseline recording in the presence of the inhibitor, A-484954 (5 μM) was applied to the artificial cerebrospinal fluid causing a fast potentiation of fEPSP. No significant difference (p = 0.47) was found between the A-484954-treated group (119 ± 1.5% at 40 min; n = 4; Figure 5A) and pretreated group (131 ± 8.8% at 40 min; n = 4; Figure 5A), indicating that anisomycin did not prevent eEF2K-inhibition mediated fEPSP potentiation.

To verify the previous data, we pretreated hippocampal slices with a different protein synthesis inhibitor, cycloheximide (100 μM), for 1 h. After 20 min of a stable baseline recording, A-484954 (5 μM) was applied and induced a fast potentiation of fEPSP. No significant difference (p = 0.18) was found between the A-484954-treated group (119 ± 1.5% at 40 min; n = 4; Figure 5B) and pretreated group (117 ± 2.1% at 40 min; n = 4; Figure 5B), indicating that cycloheximide did not prevent eEF2K-inhibition mediated fEPSP potentiation. These experiments provided evidence that eEF2K-inhibition mediated potentiation does not rely on protein synthesis.

Presynaptically localized CaMKII, PKC, and PKA have been reported to be involved in the regulation of vesicle release probabilities (Abeliovich et al., 1993; Mayford et al., 1996; Patterson et al., 2001). A simplified schemata of potential interactions is shown in Figure 6A (modified from Leenders and Sheng, 2005). To investigate the potential involvement of these key pathways we perfused acute hippocampal slices with specific inhibitors for 1 h before A-484954 application. We did not observe any change in the potentiation of fEPSP mediated by eEF2K inhibition by specific inhibitors of CaMKII, PKC, or PKA (Figure 6). Specifically, the CaMKII inhibitor KN-93 (20 μM) did not attenuate A-484954 induced fEPSP potentiation (Figure 6B); the PKC inhibitor Gö 6983 (Simons et al., 2012; Vasefi et al., 2013) did not alter the dynamic of the A-484954 mediated effect on fEPSPs (Figure 6C); and the inhibition of PKA by KT5720 did not result in any modulation of the drug-induced synaptic potentiation (Figure 6D).

Intracellular calcium release from endoplasmic reticulum (ER) has been reported to play an important role in the regulation of neural properties, such as excitability of hippocampal pyramidal cells (Borde et al., 2000; Shah and Haylett, 2000; Shirasaki et al., 2001), frequency and amplitude of mEPSCs (Sharma and Vijayaraghavan, 2003), and presynaptic plasticity (Emptage et al., 2001; Liang et al., 2002; Martin and Buno, 2003). We therefore investigated whether the potentiation of fEPSP mediated by eEF2K inhibition requires calcium release from intracellular calcium stores. Firstly, we tested thapsigargin, a non-competitive inhibitor of sarco-endoplasmic reticulum Ca²⁺-ATPases (SERCA) that causes depletion of intracellular calcium stores. We observed that 10 μM thapsigargin provoked an enhancement of basal synaptic transmission that stabilized within 1 h of incubation (data not shown). After a 20-min stable baseline recording under thapsigargin, application of A-484954 (5 μM) enhanced the fEPSP slopes (121 ± 4.8% at 40 min; n = 4; Figure 7A). The increase in fEPSP slopes did not differ from the potentiation induced by A-484954 alone (126 ± 3.9% at 40 min; n = 7; Figure 7A), but was significantly larger than that observed in drug free experiments (Figure 7D).

In a second step, we evaluated the involvement of the inositol trisphosphate receptor (InsP3R) in the drug-induced potentiation. The intracellular calcium release activated by InsP3R plays important roles in proliferation, development, learning, and memory (Bosanac et al., 2002). We tested the InsP3R antagonist aminoethoxydiphenyl borate (2-APB) and observed that pretreatment and co-application with A-484954 did not affect the potentiation of fEPSPs. The co-application of A-484954 with an InsP3R blocker induced a fast fEPSP potentiation of 132 ± 7.1% at 40 min (n = 6; Figure 7B). This potentiation was similar to that observed with the A-484954 alone (126 ± 3.9% at 40 min; n = 7; Figure 7B) and was significantly larger than baseline values in drug-free experiments (104 ± 1.7% at 40 min; n = 7; Figure 7D). Therefore, the data suggest that calcium release from InsP3R dependent calcium stores is not involved in the eEF2K-inhibition mediated effect on synaptic potentiation.

Ryanodine receptors (RyRs) at the ER are sensitive to caffeine and ryanodine and are linked to thapsigargin non-sensitive calcium stores. These channels respond to minor variations in intracellular calcium concentrations that causes further release of calcium (Zucchi and Ronca-Testoni, 1997). Blockage of RyRs by dantrolene (20 μM) prevented A-484954 (5 μM) induced fEPSP potentiation to some degree. Co-application of dantrolene and A-484954 resulted in a potentiation of fEPSP of 117 ± 2.2% at 40 min (n = 7; Figure 7C) in comparison with A-484954 alone (126 ± 3.9% at 40 min, n = 7; Figure 7C). The remaining fEPSP potentiation was still significantly larger than the fEPSP values of drug free experiments (Figure 7D). Thus, only the calcium release mediated by the activation of RyRs played a partial role in eEF2K-inhibition mediated potentiation.

We looked into various other signaling pathways also known to be involved in the regulation of vesicle release and conducted experiments using specific inhibitors against PI3K-Akt and MEK/ERK (English and Cobb, 2002). LY294002 is an inhibitor of PI3K that has been shown to be effective at a concentration of 10 μM in acute hippocampal slices (Chang et al., 2014). In addition, we tested U0126 as an inhibitor of MEK 1/2 at a concentration of 10 μM that also has been shown to be effective in hippocampal slices (Kidambi et al., 2010). Pretreatment of hippocampal slices with a cocktail of LY294002 and U0126 for 1 h did not alter fEPSP potentiation induced by A-484954. The resulting fEPSP potentiation was 123 ± 2.0% (n = 4) at 40 min (Figure 8A) that did not differ (p = 0.24) from the fEPSP values of the A-484954 group (131 ± 4.2%, n = 9), but was still significantly larger than that in the control group (drug free; 102 ± 2.4%, n = 9).

p38 MAPKs are stress-activated protein kinases that are activated through extracellular stress, cytokines, growth factor, and osmotic shock (Cuenda and Rousseau, 2007). In addition, the p38 MAPK pathway plays an important role in synaptic plasticity and neurodegenerative diseases (Correa and Eales, 2012). The p38 MAPK family comprises four isoforms: p38α, p38β, p38γ and p38δ that act on a variety of substrates, e.g., microtubule-associated protein tau and transcription factors. SB203580 is a highly specific cell-permeable inhibitor (English and Cobb, 2002) of p38α and p38β that can inhibit these isoforms at 10 μM in hippocampal slices (Capron et al., 2006; Hasegawa et al., 2014). We tested the involvement of p38 MAPK in the eEF2K-inhibition mediated fEPSP potentiation by adding 10 μM SB203580 to the Akt/MEK 1/2 inhibitor cocktail (Figure 8B). We noticed that this cocktail efficiently eliminated the previously robust A-484954 induced fEPSP potentiation. The fEPSPs (105 ± 4.2% at 40 min; n = 4; Figure 8B) differed significantly (p = 0.0024) for all time points tested in comparison with A-494854 alone experiments (131 ± 3.8% at 40 min; n = 9; Figure 8B). These data suggested that p38 MAPK is required for the eEF2K-inhibition mediated fEPSP potentiation.

To confirm the role of p38 MAPK, we incubated the acute hippocampal slices for 1 h in 10 μM SB203580 containing ACSF (Figure 8C). Consistent with our previous result, 10 μM SB203580 alone was sufficient to attenuate the A-484954 induced fEPSP potentiation, with a significant difference (p = 0.0003) between the pretreated group (107 ± 0.4% at 40 min; n = 6; Figure 8C, gray circle) and the non-pretreated group (131 ± 3.8% at 40 min; n = 9; Figure 8C, black circle). Application of SB203580 alone did not show a significant (p = 0.07) effect on the fEPSP baseline (113 ± 5.6% at 45 min; n = 8; Figure 8D, black circles). To further determine whether the inhibition of eEF2K results in an increase of p38 MAPK phosphorylation, we collected slices after 8-, 16-, or 32-min applications A-484954 (5 μM) for whole-protein biochemical analysis. The western blot data indicated that A-484954 (5 μM) induced a rapid phosphorylation of p38 MAPK (Figure 8E), suggesting that p38 MAPK phosphorylation is altered in response to eEF2K inhibition and that the p38 MAPK takes part in the A-484954 induced enhancement of synaptic transmission.

p38 MAPK has been reported to regulate hyperpolarization-activated cyclic nucleotide-gated (HCN) channels in hippocampal pyramidal neurons (Poolos et al., 2006; Jung et al., 2010). We tested whether HCN channels were involved in eEF2K-inhibition mediated fEPSP potentiation. In these experiments, the blocker (ZD7288 or Zatebradine) was applied from the start of baseline recording to the end of the experiment and showing no effect on fEPSP potentiation (data not shown).

We then tested the broad-spectrum potassium channel blocker cesium chloride (CsCl). After a stable baseline recording with CsCl (2 mM), we washed in A-484954 (5 μM) and observed a fast potentiation of fEPSP (117 ± 4.0% at 40 min; n = 4; Figure 9A) that was significantly smaller (p = 0.031) than the values observed in the A-484954 group (135 ± 4.2% at 40 min; n = 4; Figure 9A). Likewise, we tested the effect of 10 mM TEA (Figure 9B) and 400 μM BaCl₂ (Figure 9C) on eEF2K-inhibition mediated potentiation. These experiments indicated that TEA only partially (p = 0.046) affected the drug induced fEPSP potentiation reaching 119 ± 3.5% (TEA, 40 min, n = 4). However, similar experiments with BaCl₂ attenuated the drug induced fEPSP potentiation fully as indicated by the fEPSP slope value of 107 ± 2.3% at the 40 min (n = 6, p = 0.0003). Statistical analysis revealed a significant difference for all blockers tested in comparison with experiments with the drug alone. However, the resulting potentiation for TEA and CsCl, but not for BaCl₂, remained significantly different in comparison with the drug-free experiments (Figures 9D,E). These results suggest that the eEF2K inhibition mediated fEPSP potentiation requires especially the contribution of barium sensitive potassium channels that are not affected by TEA or CsCl (Wulff and Zhorov, 2008).

Insight into barium sensitive potassium channel family (Wulff and Zhorov, 2008) gave us a candidate – the outward-rectifier background potassium channel TREK-1 (also named KCNK2), which is involved in the control of resting membrane potential (Ma et al., 2011). This channel has been previously linked to play important roles in neuroprotection against epileptic seizures and ischemia. To study a potential involvement of TREK-1 in the effects of A-484954 on hippocampal synaptic transmission, we investigated whether fluoxetine, which has been proven to block TREK-1 (Kennard et al., 2005; Eckert et al., 2011; Moha ou Maati et al., 2011; Sandoz et al., 2011; Xi et al., 2011) in a voltage-independent manner (Honore et al., 2002; Kennard et al., 2005) and also interferes with other signaling pathways (Edgar et al., 1999; Fumagalli et al., 2005; Stepulak et al., 2008), attenuates the potentiation of synapses. This compound has been widely prescribed as a fast-acting treatment of depression disorders, however, with psychotropic side effects (Alboni et al., 2015; Bastos et al., 2015; Liang et al., 2015). We observed that pre-incubation and co-application of fluoxetine with A-484954 prevented the potentiation of fEPSP-slopes (Figures 10A,B).

Increased synchrony of neuronal network has been implicated as one of the mechanisms to counteract depression and it has been shown that a ketamine metabolite with anti-depressant effects increased Gamma oscillation in vivo (Zanos et al., 2016). Here we tested whether the application of this eEF2K inhibitor would enhance neuronal network synchronization. We transduced neurons of primary hippocampal cell cultures with the genetically encoded calcium sensor GCamp6f and analyzed the changes in spontaneous fluorescence intensity within individual neurons of the network (Figure 11). As it is indicated in the fluorescence images at different time points in Figure 11A, only a few neurons had a synchronized activity pattern under drug-free conditions (Figure 11A, left). The application of 5 μM A-484954 enhanced the number of synchronized neurons per frame (Figure 11A, right). The peaks of the fluorescence intensity traces of these three cells (circled in Figure 11A, left) appear simultaneously after drug application (Figure 11B), indicating the highest degree of synchronized activity pattern.

To quantify the degree of synchronization of the neuronal pairs within one culture dish, a Pearson’s correlation matrix was calculated and the individual coefficients were presented as a color-coded heat map matrix (Figure 11C). The coefficients are color-coded from dark red over white to dark blue, which correspond to -1, 0, and 1, respectively. This matrix illustrates that the degree of synchronization increased during drug application.

We conducted analyses of two more cell cultures and analyzed more than 130 neurons. The calculation of the Pearson’s correlation coefficients and construction of the Gaussian fit for the frequency distribution histograms (Figure 11D) revealed that A-484954 efficiently increased the coefficients and therefore the synchronicity of neuronal activity patterns in the cultures. In addition, a comparison of the averaged coefficients, for the three cultures, confirmed the significant difference between the results before and during drug application (Figure 11E). Thus, we conclude that A-484954 not only alters the efficiency of synaptic transmission, but also increases the level of synchronicity of neuronal activity pattern.

Ketamine and MK-801 have been found to alter the postsynaptic AMPA receptor compensation under the involvement of mTOR, eEF2K and BDNF (Autry et al., 2011; Duman et al., 2012; Dwyer and Duman, 2013; Monteggia et al., 2013; Nosyreva et al., 2013). We like to note that the differences between our and ketamine related study is that the compound A-484954 is inhibiting eEF2K without direct activation of upstream signaling pathways through the interference with glutamate receptors. This resulted in a very fast synaptic response to the compound, while Ketamine and MK-801 related experiments require at least 30 min to alter synaptic transmission.

One other difference of the effect of direct eEF2K inhibition on synaptic responses in comparison with data from ketamine and MK-801 related experiments (Nosyreva et al., 2013) is its protein synthesis independency. We address this discrepancy to the complexity of interconnected signaling pathways when compounds have a large number of interaction partners. Ketamine itself is well known to affect a wide range of cellular processes by modulation, e.g., of cholinergic (Yamakura et al., 2000; Lydic and Baghdoyan, 2002), or aminergic system (dopamine and noradrenaline) (Kubota et al., 1999; Kamiyama et al., 2011; Wang et al., 2012; Sleigh et al., 2014) and for its derivative the targets are even unknown (Zanos et al., 2016). Thus, a distinct modulation of different signaling pathways might promote protein synthesis dependent modulation of synaptic transmission. Only proteomic or phospho-proteomic studies might be able to visualize the complexity of altered pathways, e.g., for MK-801 (Guest et al., 2015). In addition, many studies of the eEF2K function relied on the inhibitors NH125 and Rottlerin that have been shown recently to act also in an unpredicted fashion indicating the necessity of further studies to clarify the interaction partners (Davies et al., 2000; Bender et al., 2016). One of the identified eEF2K substrates (Newman et al., 2013; Hu et al., 2014) has been shown to potentially modulate synaptic transmission in a protein synthesis independent manner (Jiao and Li, 2011). Another substrate is involved in regulation of the activity of protein phosphatases (Comerford et al., 2006), which could in turn modulate the phosphorylation levels of other key enzymes such as p38 MAPK (Kitatani et al., 2006). Many indicated interactions need still to be confirmed to take place in the hippocampus, but the few presented examples and our own data point toward protein synthesis independent modulation of cellular processes by eEF2K. However, the specific network of signaling pathway with its final targets requires further studies.

To discover a new branch of the eEF2K signaling pathway that might have been altered by the compound, we tested a variety of key enzymes that have been shown to be involved in activity-dependent synaptic plasticity and more importantly in the regulation of neurotransmitter release (Capogna et al., 1995; Hilfiker et al., 1999). For example, the activation of CaMKII and PKC have been shown to increase vesicle release by the phosphorylation of synapsin-I (Ohyama et al., 2002; Ninan and Arancio, 2004) or Munc-18 (Fujita et al., 1996; Dulubova et al., 1999) respectively. In addition, PKA can also contribute to the vesicle release by phosphorylating RIM1α (Lonart et al., 2003), SNAP-25 (Hepp et al., 2002), Snapin (Thakur et al., 2004) and Syntaphilin (Boczan et al., 2004). However, in our experiments, no effects of specific inhibitors of these kinases on the degree of compound-induced fEPSP potentiation were observed.

A brief rise in the intracellular calcium concentration is known to be a key step for the evoked release of glutamate from presynapses. The origin of the increase in intracellular calcium is manifold and can rely on intracellular calcium release from ER. This mechanism plays an important role in regulating neural function, including the excitability of hippocampal pyramidal cells (Borde et al., 2000; Shah and Haylett, 2000; Shirasaki et al., 2001), and might determine the frequency and amplitude of mEPSCs (Sharma and Vijayaraghavan, 2003) and the degree of presynaptic plasticity (Emptage et al., 2001; Liang et al., 2002; Martin and Buno, 2003). The release of calcium from the ER can be induced by the activation of calcium transporter SERCA (Zucchi and Ronca-Testoni, 1997) or the activation of InsP3 receptors by elevated levels of InsP3. However, depletion of the calcium contents of the ER by thapsigargin or blockage of InsP3 receptors did not prevent the compound-induced potentiation. Only the inhibition of RyRs by dantrolene showed a tendency to attenuate the potentiation of synapses. Interestingly, Reese and Kavalali (2015) reported a critical role of RyRs in amplifying NMDA receptor-driven Ca²⁺ signals at rest for the maintenance of synaptic potentiation.

As discussed above, signaling pathways relevant to classical plasticity are not involved in the compound-mediated effects on synaptic transmission. In addition, we showed that the activity of the PI3K-Akt and MEK/ERK pathways was not required for this synaptic potentiation. Only an inhibitor against the p38 MAPK was shown to prevent the potentiation of synaptic transmission mediated by eEF2K inhibition efficiently. The p38 MAPK is activated by multiple mechanisms, including extracellular stresses, cytokines, growth factors, and osmotic shocks (Cuenda and Rousseau, 2007). The regulation of p38 MAPK also provokes much interest due to its role in synaptic plasticity and neurodegenerative disorders (Correa and Eales, 2012). Indeed, we were able to prove that the inhibition of p38α and p38β by SB203580 (Capron et al., 2006; Kidambi et al., 2010; Jin et al., 2013; Hasegawa et al., 2014) prevents the synaptic potentiation caused by the inhibition of eEF2K. In addition, we showed that the inhibition of eEF2K did not only attenuate the phosphorylation of eEF2 but also significantly increased the phosphorylation level of p38 MAPK within a short time, which was consistent with the results of another study on eEF2K (Usui et al., 2015). However, how the inhibition of eEF2K can cause an enhancement of p38 phosphorylation remains unknown. To date, we have not found any study that addresses a potential interaction of the compound A-494854 with upstream components of the p38 MAPK pathway. A reversed interaction of p38 MAPK on eEF2K has been shown by the p38-dependent phosphorylation of MK2/3 and subsequent phosphorylation of Ser-377 of the eEF2K that slightly inhibited its activity (Knebel et al., 2002).

In the search for potential targets that mediated the synaptic potentiation, we also focused on potassium channels. For example, p38 MAPK regulates HCN gate channels (h-channels; I_h) in hippocampal pyramidal neurons (Poolos et al., 2006; Jung et al., 2010). However, in our experiments, the inhibition of the h-channel did not prevent the potentiation of synaptic transmission mediated by eEF2K inhibition, either the cesium- and TEA-sensitive potassium channels. Only prevention of the activation of barium-sensitive potassium channels attenuated the effect of eEF2K inhibition on synaptic transmission. A channel that belongs to the two-pore potassium channel family with specific Ba²⁺ selectivity is the TREK-1 channel (Wulff and Zhorov, 2008; Ma et al., 2011). Indeed, blockage of the TREK-1 channel by fluoxetine prevented the eEF2K-inhibition mediated potentiation. TREK-1 channels are expressed both pre- and post-synaptically (Honore, 2007; Westphalen et al., 2007; Sandoz et al., 2009; Veale et al., 2014) and contribute to the maintenance of the neuronal resting potential and membrane excitability by opposing depolarization. In response to its phosphorylation, the outflow of potassium ions is reduced and leads to a slight depolarization of the membrane potential. This feature may account for the effect on the synaptic response by eEF2K inhibition. Since this receptor can be phosphorylated through phosphorylation by PKC and PKA (Murbartian et al., 2005), it may also represent a direct target of p38 MAPK which need further validation. The interpretation of the effects of the compound fluoxetine on eEF2K mediated synaptic responses is complicated by its interference with pathways such as PKC, PKA, or ERK (Edgar et al., 1999; Fumagalli et al., 2005; Stepulak et al., 2008). Nevertheless, we think that we can rule out some of its side effects due to our experiments that excluded the involvement of PKC, PKA, and ERK in the eEF2K-inhibition effect. Further research is required using genetically modified animals or more specific antagonist against TREK-1. With regard to the potential involvement of TREK-1 in neurological diseases, it has been already shown that TREK-1 can be beneficial against mood disorders. Genetic deletion of the background potassium channel TREK-1 creates immunity to depression-inducing paradigms (Heurteaux et al., 2006; Mazella et al., 2010; Borsotto et al., 2015; Vivier et al., 2016). Therefore, these results suggest a possibility to regulate TREK-1 by specific modulation of eEF2K activity.

We showed that A-484954 induced a synchronization of the neural network in primary hippocampal cell cultures. A previous study showed that the regulation of dendrotoxin-sensitive potassium channels contributes to the pattern of neuronal activity (Cudmore et al., 2010). Similarly, our data show the potential implication of barium-sensitive potassium channel TREK-1 in the regulation of neural network activity.

The synchrony of neuronal network activity is an important parameter for physiological function of cognitive abilities (Llinas and Ribary, 2001). Many studies have reported that alteration in the mechanisms of synchrony of neuronal networks could alter the cognitive abilities of schizophrenic and autistic and Alzheimer’s disease patients and can induce motor dysfunctions (Uhlhaas and Singer, 2006). In addition, high levels of p-eEF2 were also reported in a mouse model of Alzheimer’s disease (Li et al., 2005) and in brain material of patients affected by Alzheimer’s disease, suggesting a particular role of the eEF2K/eEF2 pathway in mental disorders. Moreover, the pharmacological and genetic inhibition of eEF2K was recently reported to be able to reverse the epileptic phenotype in a mouse model of human epilepsy (Heise et al., 2016). These data show a novel role of eEF2K in the regulation of brain network activity, providing a novel target for brain disorder treatment.

Sprague Dawley rats (aged 6–10 weeks) were provided from the animal center of the Chinese Academy of Sciences (CAS, Shanghai, China). The animals were maintained in accordance with the established standards of animal care and procedures of the Institutes of Brain Science and State Key Laboratory of Medical Neurobiology of Fudan University, Shanghai, China. Efforts were made to minimize the number of animals studied.

All data are expressed as mean ± standard error of mean (SEM). The values of different experimental conditions were compared using two-sided paired or non-paired t-test. A p < 0.05 was considered to indicate a statistically significant difference between two groups. Drug experiments were interspersed with drug-free controls.