Abstract
Yuskaitis CJ, Jones BM, Wolfson RL, Super CE, Dhamne SC, Rotenberg A, Sabatini DM, Sahin M, Poduri A. Neurobiol Dis 2018;111:91–101.
DEPDC5 is a newly identified epilepsy-related gene implicated in focal epilepsy, brain malformations, and Sudden Unexplained Death in Epilepsy (SUDEP). In vitro, DEPDC5 negatively regulates amino acid sensing by the mTOR complex 1 (mTORC1) pathway, but the role of DEPDC5 in neurodevelopment and epilepsy has not been described. No animal model of DEPDC5-related epilepsy has recapitulated the neurological phenotypes seen in patients, and germline knockout rodent models are embryonic lethal. Here, we establish a neuron-specific Depdc5 conditional knockout mouse by cre-recombination under the Synapsin1 promotor. Depdc5flox/flox-Syn1Cre (Depdc5cc+) mice survive to adulthood with a progressive neurologic phenotype that includes motor abnormalities (i.e., hind limb clasping) and reduced survival compared to littermate control mice. Depdc5cc+ mice have larger brains with increased cortical neuron size and dysplastic neurons throughout the cortex, comparable to the abnormal neurons seen in human focal cortical dysplasia specimens. Depdc5 results in constitutive mTORC1 hyperactivation exclusively in neurons as measured by the increased phosphorylation of the downstream ribosomal protein S6. Despite a lack of increased mTORC1 signaling within astrocytes, Depdc5cc+ brains show reactive astrogliosis. We observed two Depdc5cc+ mice to have spontaneous seizures, including a terminal seizure. We demonstrate that as a group Depdc5cc+ mice have lowered seizure thresholds, as evidenced by decreased latency to seizures after chemoconvulsant injection and increased mortality from pentylenetetrazole-induced seizures. In summary, our neuron-specific Depdc5 knockout mouse model recapitulates clinical, pathological, and biochemical features of human DEPDC5-related epilepsy and brain malformations. We thereby present an important model in which to study targeted therapeutic strategies for DEPDC5-related conditions.
Commentary
Despite being a relative newcomer to the epilepsy mTORopathy neighborhood, DEPDC5 (DEP Domain-Containing Protein 5) is rapidly making its presence felt. DEPDC5 mutations are being increasingly identified in a wide variety of familial and sporadic epilepsy syndromes, including familial focal epilepsies such as familial focal epilepsy with variable foci (FFEVF), autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE), familial mesial temporal lobe epilepsy (FMTLE), and familial focal epilepsy with focal cortical dysplasia (FCD; 1–3). DEPDC5 mutations have also been found in sporadic FCD, hemimegalencephaly, and infantile spasms (4, 5). Importantly, DEPDC5-related epilepsy has been associated with sudden unexpected death in epilepsy (SUDEP), even in individuals with infrequent seizures (6, 7).
DEPDC5, along with its partners NPRL2 and NPRL3, make up the GATOR1 protein complex, which acts as a negative regulator of activation of the mTORC1 complex by Rag GTPases (8). Inactivating mutations in DEPDC5 therefore lead to mTORC1 hyperactivation, similar to that caused by mutations in TSC1/2 or in PTEN, long known to be associated with epilepsy (9). An interesting unique function of the GATOR1 complex is its role in the sensing of amino acid availability; under normal circumstances, the mTORC1 pathway is suppressed by amino acid deprivation, while in the setting of DEPDC5 mutation, mTORC1 is hyperactive despite amino acid starvation (8).
Epilepsy-associated mutations in DEPDC5 are loss-of-function in nature, generally involving truncating, frameshift, or splice-site mutations (9). Familial DEPDC5-related epilepsies are inherited in an autosomal dominant manner with reduced penetrance. Surgical tissue from an individual with FCD has revealed a germline DEPDC5 mutation plus a second-hit loss of function somatic mutation (1), but it is currently unknown if a second hit is necessary for all cases of epilepsy in the setting of germline DEPDC5 mutation.
Similar to what has been seen with other mTORopathy genes such as Tsc1 and Tsc2, germline homozygous Depdc5 knock-out mice or rats exhibit embryonic lethality (10, 11). Heterozygous Depdc5 knock-out animals are generally healthy and fertile, with either cytomegalic and dysmorphic neurons (11) or no brain pathologic abnormalities (10) and no increase in seizure susceptibility. Therefore, to better assess the neurologic phenotype of Depdc5 dysfunction in mice, Yuskaitis et al. employed a similar approach to what their group has used previously in the study of Tsc1 and Tsc2 (12, 13), conditional neuronal knockout of Depdc5 using synapsin I-cre/flox technology. Homozygous mixed-strain background male mice with loxP sites flanking exon 5 of the Depdc5 gene were bred with female synapsin I-cre mice to generate homozygous (Depdc5cc+) and heterozygous (Depdc5cw+) neuronal Depdc5 knock-out mice. Homozygous Depdc5cc+ mice exhibited a shortened life span (median 115 days) than their heterozygous and wild-type littermates, and reduced weight gain (in males only). Adult Depdc5cc+ mice exhibited limb-clasping behavior on tail suspension, indicating a degree of motor dysfunction.
The neuropathology of adult Depdc5cc+ mice was investigated in detail, revealing increases in brain weight, cortical thickness, neuronal soma size, and neuronal phosphorylated ribosomal protein S6 (p-S6) expression as compared with wild-type littermates. Dysmorphic cortical neurons were identified, as was a decrease in neuronal density in cortical layers IV-VI. Western blot analysis of cortical lysates from Depdc5cc+ mice confirmed increased levels of p-S6 and decreased levels of S473 phosphorylated AKT, indicating mTORC1 activation with decreased mTORC2 activity, similar to that seen in Tsc1 and Tsc2 conditional knockouts. Interestingly, reactive astrogliosis was identified in Depdc5cc+ cortex in the absence of mTORC1 activation in astrocytes as demonstrated by the absence of p-S6 staining. Unfortunately, given previous discrepancies reported in heterozygous Depdc5 knock-out mice versus rats (10, 11), neuropathologic characterization of heterozygous Depdc5cw+ mice was not reported.
Importantly, tonic or tonic–clonic seizures were witnessed during routine handling in 3 of 43 Depdc5cc+ mice. One of these mice died during a witnessed seizure, while the other two died within 2 weeks of their seizures. Attempts were made to characterize spontaneous seizures in adult Depdc5cc+ mice through video EEG recording of 7 mice for a duration of 60 hours to 10 days. No electroclinical seizures were captured in any of these mice, while interictal activity was identical between knock-out and wild-type mice. These findings indicate that spontaneous seizures either occur in a small proportion of Depdc5cc+ mice or at a low frequency. To further characterize seizure predisposition, Depdc5cc+ and wild-type mice were injected with pentylenetetrazol (PTZ) while undergoing EEG monitoring. Depdc5cc+ mice were found to have a significantly shorter latency for development of tonic–clonic seizures after PTZ injection and were much more likely to die during their seizures.
The neuronal Depdc5 knock-out mouse model developed by Yuskaitis et al. represents a valuable tool for the study of DEPDC5-related epilepsy and SUDEP. More extensive and prolonged video-EEG monitoring could be performed to better define the incidence of spontaneous seizures in these mice and their association with early death. Other potential mechanisms of sudden death could be investigated, as the synapsin I-cre would be expected to knock out Depdc5 in brainstem respiratory and cardiovascular control centers as well as in cortex. Behavioral phenotypes in these animals could be determined. Mechanistic reasons could be studied for the relatively milder phenotype of Depdc5 neuronal knock-out mice as compared with Tsc1 or Tsc2 conditional knock-out mice, which often exhibit more prominent seizures and earlier death. These mechanisms may be related to the specialized role of Depdc5 in amino acid-induced regulation of mTORC1 as opposed to the more pleiotropic role of Tsc1/2 in mTORC1 inhibition. Finally, and perhaps most clinically relevant, treatment of Depdc5 neuronal knock-out mice with mTOR inhibitors such as rapamycin or everolimus may provide insight into the potential utility of these agents in the management of DEPDC5-related epilepsy in humans.