A growing body of evidence implicates aberrant mTORC1 signaling in the neuropathology of many neurological disorders, in particular genetic disorders associated with cognitive and social deficits and epilepsy 
. Furthermore, cognitive impairments are a common comorbidity in acquired epilepsy that follows SE 
. Previous studies have shown aberrant mTORC1 activation following SE in rodent models 
. In our studies we build upon this previous work, extending these findings to show for the first time a link between mTOR dysregulation and hippocampal-dependent learning and memory deficits associated with SE. Rats that experienced SE and subsequently were treated with rapamycin 2 weeks later, performed significantly better than vehicle-treated SE rats in two hippocampal-dependent memory tasks, MWM and NOR (–). Additionally, we describe mTOR-dependent activation of microglial cells in association with structural and molecular dendritic alterations following SE that are reversed with rapamycin (, , , ). Here we show that at the molecular level, the SE-induced increase in mTORC1 activation through P-S6 occurs in reactive microglial cells and neurons (), and that inhibition of mTORC1 attenuated the SE-induced microgliosis (), improved dendritic arborization (–), and restored protein levels of hippocampal ion channels (). In parallel, we found that SE-induced microgliosis correlated with dendritic damage in the area CA1 of hippocampus (). Taken together, our findings implicate aberrant mTORC1 activation as a candidate mechanism involved in hippocampal-dependent cognitive deficits and dendritic pathology associated with SE.
The use of genetic models and pharmacological tools to manipulate mTORC1 activation implicate this pathway as an essential player in the modulation of memory 
. Transgenic rodents with loss of at least one allele of the upstream mTORC1 negative regulators the phosphatase and tensin homolog (Pten) or the tuberous sclerosis complex 1 and 2 (TSC1/2) proteins are characterized by mTOR pathway hyperactivity, and social and memory deficits, which are reversed with rapamycin 
. These studies suggest that up-regulation of mTOR signaling alters memory processing, strongly supporting a critical role for the mTOR pathway in memory modulation. For this reason we evaluated whether the SE-induced mTOR hyperactivation 
contributes to the memory deficits associated with SE and found that rapamycin rescued SE-induced hippocampal-dependent spatial learning and memory deficits (–). Our findings implicate mTORC1 dysregulation in the memory dysfunction that develops early after SE.
The effects of rapamycin in behavioral deficits in several developmental epilepsy models have been evaluated 
. A recent study shows that rapamycin improves learning and memory in a rat model of infantile spams 
, further supporting a role for mTORC1 dysregulation in the cognitive deficits associated with epilepsy. In a model of early-life seizures induced by acute hypoxia in neonatal rodents rapamycin rescued aberrant social behavior 
. However, in our study SE also was associated with altered social interaction, but rapamycin did not rescue the abnormal social behavior (Fig. S3
), suggesting different mechanisms for the altered social behavior in these models. Furthermore, a recent study showed that aggressive behavior in pilocarpine-treated epileptic rats is reversed by rapamycin 
We found that rapamycin had no significant effect in the performance of sham-treated rats in the behavioral tests (–), consistent with other studies. For instance, rapamycin rescued deficits in hippocampal-dependent spatial learning tasks in a TSC2 (+/−) mouse model but had no effect on the performance of the wild type (WT) mice on those tests 
. Several other studies have reported that rapamycin impairs learning and memory in normal rats and WT mice, but much higher doses (e.g. 150 mg/kg) or direct CNS infusion (e.g. 5–1000 ng) of rapamycin were required for this effect 
. Taken together these studies indicate that a higher level of mTORC1 inhibition is required to disrupt learning and memory in normal rodents, thus explaining the lack of effect of rapamycin on the behavioral tests in the sham rats described here.
Previous findings showing that rapamycin suppresses spontaneous seizures in epileptic rodents 
would have led us to predict a reduction in epileptiform activity in the SE group treated with rapamycin; however, in our studies acute rapamycin treatment given weeks 2–3 following SE had no effect on epileptiform activity during the period of our studies (). Furthermore, we cannot exclude the possibility that seizures might have occurred at times when the rats were not being monitored. In line with our findings, recent studies showed that short-term rapamycin treatments have limited anticonvulsive effects in rodent models of acute seizures 
. In addition, Buckmaster and Lew (2011) showed that long-term rapamycin treatment beginning one day after SE induction was not sufficient to reduce seizure frequency in the pilocarpine model of epilepsy 
. The discrepancies between these studies may be attributed to the timing of rapamycin administration (i.e. pre- vs. post-SE induction), length of rapamycin treatment (acute vs. long-term), different rodent species, and the epilepsy models used. Interestingly, it has been shown that rapamycin did not alter interictal epileptiform activity in a rat model of infantile spasms 
. Similarly, some anticonvulsants are suboptimal in reducing interictal spikes 
Our finding that SE-induced altered mTOR signaling occurs in both neurons and microglia suggests that mTOR effectors are activated in diverse cells types following SE and may thereby change neuronal and glial properties. In parallel with a reduction in P-S6 staining (–), there was a decrease in microgliosis in the CA1 region of the SE+Rap group when compared to the SE+Veh group (). In a separate cohort of rats, we found that these alterations were already evident 2 weeks after SE, suggesting that rapamycin reversed the changes (Fig. S4
). These data also suggest that SE-induced dendritic damage is mTOR-dependent and that microgliosis may play a role in this process. Indeed, it has been shown that lipopolysaccharide- and cytokine-induced pro-inflammatory activation of microglia is mTOR-dependent 
and microgliosis is suppressed with rapamycin 
. Under physiological conditions microglia play critical roles in synaptic pruning during development 
and in mature synaptic connections 
. Furthermore, prolonged microglia-synapse contacts induced by ischemia are followed by a large loss of synapses 
, and lipopolysaccharide-induced microgliosis results in loss of spine density and dendritic branching 
. Based on these studies we speculate that the presence of reactive microglia within CA1 following SE may contribute to the significant loss of dendrites and spines in this region (). Hippocampal microgliosis has been widely reported in humans and animal models of epilepsy 
and existing evidence suggests that microgliosis in the hippocampus contributes to cognitive deficits 
Altered cognitive behavior that occurs in association with epilepsy is often associated with structural 
and molecular dendritic abnormalities 
. Thus, it is possible that the rapamycin-mediated reduction in the SE-induced microgliosis promotes dendritic repair and memory improvement in SE rats. In fact, rapamycin has been shown to reduce microgliosis and in parallel improve functional recovery in several neurobehavioral tasks following traumatic brain injury in rodents 
. Astrogliosis also may contribute to CA1 dendritic damage following SE; however the SE-induced astrogliosis was evident throughout the hippocampus () and microgliosis was concentrated in areas of high Map2 loss. While the finding that rapamycin reduced SE-induced gliosis in the hippocampus was unexpected, it was not surprising, as rapamycin is a potent immunosuppressant 
. This finding is important because it opens up the possibility that the beneficial effect of rapamycin on SE-induced phenotypes is at least in part due to an effect on glial responses in the hippocampus. Neuronal death associated with seizures also may play a role in the altered cognitive behavior 
. Previous studies have shown that rapamycin treatment did not reverse the SE-induced hippocampal cell loss when assessed early or late following SE 
. Thus, the rapamycin-dependent reduction in SE-induced microgliosis and the improvement of dendritic arborization after SE may at least in part compensate for the loss of neurons to improve spatial memory. Future studies are needed to further evaluate these possible mechanisms following SE.
In neurons, mTORC1 regulates dendritic architecture 
and modulates dendritic synthesis of Map2 
. Association of Map2 with microtubules, kinases, and other proteins is critical for dendritic stability, protein trafficking, neurite initiation, and local signal transduction in neurons 
. For this reason we speculate that by modulating Map2 levels or other proteins such as ion channels, rapamycin may promote dendritic structural and functional integrity following SE. This is particularly important because studies in vitro
have shown mTORC1-dependent modulation of dendritic surface expression and protein synthesis of Kv1.1 and Kv4.2 channels, respectively 
. Our finding that rapamycin restored basal levels of Kv4.2 and HCN1 channels and partially rescued SK2 levels in hippocampus of rats subjected to SE () suggests that mTOR signaling may contribute to the translational regulation of these channels or associated regulatory proteins in neurons. Under physiological conditions the currents generated by Kv4.2, HCN1, and SK2 contribute to shaping the response to synaptic inputs, modulating neuronal excitability, and learning and memory 
. Dysregulation of these channels following SE is thought to alter post-synaptic responses through the modulation of the respective underlying currents 
. This in turn may alter limbic circuitry and thereby contribute to SE-mediated hippocampal-dependent learning and memory deficits. Indeed, genetic alterations in the levels of these channels cause aberrant synaptic plasticity and learning and memory 
. Based on these studies, we propose that the rapamycin-mediated dendritic repair and reversal of ion channel abnormalities contributes to improvement in the SE-induced memory deficits. In addition, to the dendritic channels, we found a rescue in Kv1.4 protein levels. Because Kv1.4 channel localization also has been described presynaptically within the hippocampus 
, our finding opens up the possibility of an effect of SE-induced mTORC1 dysregulation presynaptically. Additional studies are required to dissect out the specific effects of the SE-induced mTOR dysregulation in neuronal vs. glial cells, and the specific roles of these cell types may play in the associated memory and dendritic dysregulation.
In conclusion, our findings suggest that aberrant mTORC1 signaling contributes to SE-induced hippocampal-dependent spatial learning and memory deficits. While additional studies are required to determine direct molecular targets of this pathway within neurons and microglia, our findings indicate that following SE mTORC1 dysregulation contributes to memory deficits, microgliosis, and dendritic abnormalities that are associated with SE. Given that suppression of mTORC1 signaling improves cognition in humans 
, our findings suggest that the mTORC1 pathway may be a potential therapeutic target for cognitive dysfunction associated with epilepsy.