PLPP/CIN-mediated NEDD4-2 S448 dephosphorylation regulates neuronal excitability via GluA1 ubiquitination

Neuronal precursor cell expressed developmentally downregulated 4-2 (NEDD4-2) is an E3 ubiquitin ligase to regulate ion transport by controlling cellular trafficking/endocytosis and lysosomal degradation of ion channels and transporters. Thus, NEDD4-2 is relevant to neuronal excitability and epileptic encephalopathies in human patients. However, the regulatory molecules for NEDD4-2 dephosphorylation have been still elusive. Here, we demonstrate that pyridoxal-5′-phosphate phosphatase/chronophin (PLPP/CIN) specifically dephosphorylated NEDD4-2 serine (S) 448 site. PLPP/CIN deletion inhibited NEDD4-2 ubiquitination, and diminished the responsiveness of α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid receptor (AMPAR) by facilitating NEDD4-2-mediated ubiquitination of GluA1 subunit under physiological condition. PLPP/CIN overexpression reversed these effects. These PLPP/CIN-mediated processes were required for the increased seizure severity and its progression in response to kainic acid (KA). Therefore, we suggest the novel function of PLPP/CIN as a NEDD4-2 phosphatase, which may be a potential therapeutic target for NEDD4-2-associated diseases as well as various neurological and psychiatric disorders, including epilepsy.


Introduction
Epilepsy is one of the common neurological disorders, suffering about 1% of all people, which is characterized by the periodic occurrence of seizures exhibiting abnormal synchronized neuronal discharges. Unprovoked recurrent seizures contribute to a cyclical or progressive process of worsening epilepsy and neurological deficits including learning disabilities and memory problems. The pathophysiology underlying seizure susceptibility is relevant to channelopathy, aberrant synaptic organization, impaired glial function, inflammation, and neuronal loss [1][2][3][4][5][6][7][8][9] . However, the cellular and molecular mechanisms in epilepsy remain unclear.
Ubiquitination is a rapid, local, and reversible posttranslational modification that involves the covalent conjugation of ubiquitin to target proteins, which assigns misfolded cytosolic proteins for degradation by the 26S proteasome and regulates physiological processes. Ubiquitination is involved in modulation of synaptic function via nonproteolytic processes such as endocytosis, protein localization/targeting, complex assembly and regulation of the duration, and intensity of signaling by effector molecules. Generally, ubiquitin is bound to lysine residues on target proteins by a cascade of reactions carried out sequentially by the ubiquitin-activating enzyme (E1), ubiquitin-conjugating enzyme (E2), and ubiquitin ligase enzyme (E3). In the final step, an E3 ubiquitin ligase transfers the ubiquitin to recognition motifs in the target proteins and thus confers target protein specificity 10,11 .
Here, we demonstrate that PLPP/CIN bound to NEDD4-2 and dephosphorylated its S448 site, which accelerated NEDD4-2 ubiquitination, and subsequently enhanced the responsiveness of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor by inhibiting NEDD4-2-mediated ubiquitination of GluA1 subunit, but not voltage-gated K + channels KCNQ2/3/5. We further describe that NEDD4-2 knockdown increased seizure severity and facilitated its progression in response to KA, like PLPP/CIN overexpression. These findings indicate the novel function of PLPP/CIN as a NEDD4-2 phosphatase, which may regulate neuronal excitability by an inhibitory mechanism for E3 ubiquitin ligase activity of NEDD4-2. Therefore, we suggest that PLPP/CIN may be a potential therapeutic target for epileptogenesis and NEDD4-2-associated diseases.

PLPP/CIN deletion attenuates seizure-induced NEDD4-2 ubiquitination
Recently, we have reported that PLPP/CIN −/− mice show the shorter latency of seizure onset, but interrupt seizure progression in response to KA 30 . In contrast, spontaneous NEDD mutation in NEDD4-2 andi mice increases seizure susceptibility in response to KA 14 . Since 2h after KA injection is the suitable time point to compare the time of seizure onset, total EEG power and the changes in anatomical or biochemical profiles in PLPP/CIN Tg and PLPP/CIN −/− mice 30 , we investigated whether PLPP/CINmediated NEDD4-2 ubiquitination reciprocally regulates seizure activity 2 h after KA injection. Consistent with our previous report 30 , PLPP/CIN deletion reduced the latency of seizure onset, seizure intensity (severity), and seizure duration in response to KA (p < 0.05 vs. WT mice, respectively; Fig. 2a-c). After 2 h of KA injection, SGK1 a-c Effect of PLPP/CIN deletion on expression and phosphorylation levels of SGK1 and NEDD4-2 in vivo. Total NEDD4-2 protein and pNEDD4-2 S448 phosphorylation levels are higher in PLPP/CIN −/− mice, as compared to WT mice. a Representative western blots of SGK1 and NEDD4-2. b, c Quantification of SGK1 and NEDD4-2 expression, their phosphorylation levels, and the ratio of pNEDD4-2 to total NEDD4-2. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; *p < 0.05 vs. WT animal; n = 7, respectively). d Representative co-immunoprecipitation data revealing PLPP/CIN-NEDD4-2 bindings. e-g In vitro assay using recombinant proteins. e, f Representative co-immunoprecipitation and western blot data demonstrating PLPP/CIN-NEDD4-2 bindings, and PLPP/CINmediated NEDD4-2 S448 dephosphorylation. g Quantification of NEDD4-2 phosphorylation levels based on western blot data. phosphorylations were diminished in WT mice without changing its protein level and PLPP/CIN activity (p < 0.05 vs. control animals; Fig. 2d-f). NEDD4-2 protein level and its phosphorylations were decreased to~0.6-fold of control level (p < 0.05 vs. control animals; Fig. 2d, g). KA also decreased SGK1 phosphorylations in PLPP/CIN −/− mice without altering its protein level (p < 0.05 vs. control animals; Fig. 2d, f). However, PLPP/CIN deletion attenuated the reductions in NEDD4-2 protein and S448 (not S342) phosphorylation levels (p < 0.05 vs. WT mice; Fig. 2d, g). Because of the reduced NEDD4-2 protein level, the phosphorylation ratios of S342 and S448 in WT mice were similar to those in WT control animals following KA injection (Fig. 2d, h). In PLPP/CIN −/− mice, S342 phosphorylation ratio was lower than that in WT mice, while S448 phosphorylation ratio was higher than WT mice (p < 0.05 vs. WT mice, respectively; Fig. 2d, h). Consistent with seizure severity, PLPP/CIN deletion attenuated seizureinduced neuronal damage in the CA3 region (p < 0.05 vs. WT mice; Fig. 2i, j). These findings indicate that PLPP/ CIN may reduce NEDD4-2 protein level and its S448 phosphorylation induced by seizure activity independent of SGK1 activity, and that upregulated NEDD4-2 protein level in PLPP/CIN −/− mice may abrogate seizure progression in response to KA.

PLPP/CIN overexpression increases NEDD4-2 ubiquitination and facilitates seizure progression induced by KA
To further elucidate the role of PLPP/CIN in NEDD4-2mediated GluA1 ubiquitination, we employed PLPP/CIN transgenic (PLPP/CIN Tg ) mice. Consistent with our previous study 30 , PLPP/CIN Tg mice showed the increases in the seizure latency and seizure intensity (severity) in response to KA (p < 0.05 vs. WT mice, respectively; Fig.  6a-c). Under physiological condition, the PLPP/CIN activity in PLPP/CIN Tg mice was 7.5-fold higher than that in WT mice under physiological condition (p < 0.05 vs. WT mice, Fig. 6d). PLPP/CIN Tg mice also demonstrated the reductions in NEDD4-2 protein and S448 (not S342) phosphorylation levels without altering SGK1 activity (p < 0.05 vs. WT mice, respectively; Fig. 6e-g). Since PLPP/ CIN Tg mice showed the downregulation of NEDD4-2 protein level (p < 0.05 vs. WT mice, Fig. 6f-g), S342 phosphorylation ratio in PLPP/CIN Tg mice was higher than that in WT mice, but the S448 phosphorylation ratio in PLPP/CIN Tg mice was similar to that in WT mice (p < 0.05 vs. WT mice, respectively; Fig. 6g, h). After 2 h of KA injection, PLPP/CIN activity was slightly decreased, but not significantly, in PLPP/CIN Tg mice (Fig. 6d). SGK1 phosphorylations were decreased to 0.5-fold-0.6-fold of control level in both WT and PLPP/CIN Tg mice without changing its expression level (p < 0.05 vs. control animals, respectively; Fig. 6e, g). There was no difference in the altered SGK1 phosphorylations between both groups (Fig.  6e, g). KA decreased NEDD4-2 expression and its phosphorylations to about half-fold of control level in both WT and PLPP/CIN Tg mice (p < 0.05 vs. control animals; Fig. 6f, g). KA did not affect the S342 and S448 phosphorylation (see figure on previous page) Fig. 2 Effects of PLPP/CIN deletion on seizure activity and NEDD4-2 phosphorylations in response to KA. a-c Effect of PLPP/CIN deletion on seizure activity in response to KA. PLPP/CIN −/− mice demonstrate the reductions in the latency of seizure onset, seizure intensity, and its duration. a Representative EEG traces and frequency-power spectral temporal maps in response to KA. b Quantification of latency of seizure onset. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; *p < 0.05 vs. WT animals; n = 7, respectively). c Quantification of total EEG power (seizure intensity) in response to KA (mean ± S.E.M.; n = 7, respectively). d-h Effects of KA on PLPP/CIN activity, SGK1, and NEDD4-2 expressions and their phosphorylation levels. KA does not affect PLPP/CIN activity. PLPP/CIN deletion attenuates the reductions in total NEDD4-2 protein and pNEDD4-2 S448 phosphorylation levels, but not SGK1 and its phosphorylations, following KA injection. d Representative western blots of SGK1 and NEDD4-2. e Quantification of PLPP/CIN activity (*p < 0.05 vs. control WT animals; n = 7, respectively). f Quantification of SGK1 expression and its phosphorylation levels. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; *p < 0.05 vs. control animals; n = 7, respectively). g, h Quantification of NEDD4-2 expression, its phosphorylation levels and the ratios of pNEDD4-2 to total NEDD4-2. PLPP/CIN Tg mice did not show the changes in total protein levels and membrane expressions of KCNQ2/3/5 channels under physiological condition, as compared to WT mice (Fig. 7a-c). However, PLPP/CIN Tg mice showed the enhanced membrane GluA1 expression without altering total GluA1 protein level (p < 0.05 vs. WT mice, Fig. 7a-c). The binding of NEDD4-2 to PLPP/CIN in PLPP/CIN Tg mice was~1.4-fold higher than that in WT mice (p < 0.05; Fig. 7d, e), while the GluA1-NEDD4-2 bindings was~0.5-fold of WT mouse level (p < 0.05; Fig.  7f, g). The ubiquitination levels of NEDD4-2 and GluA1 in PLPP/CIN Tg mice were 1.3-fold and 0.4-fold of WT mouse levels, respectively (p < 0.05; Fig. 7h-j). KA did not affect total and membrane KCNQ2/3/5 and GluA1 levels in both groups (Fig. 7a-c). However, KA increased PLPP/ CIN-NEDD4-2 binding in both groups (p < 0.05 vs. control animals; Fig. 7d, e). Furthermore, KA decreased the GluA1-NEDD4-2 bindings in WT mice (p < 0.05 vs. control animals; Fig. 7f, g), but not in PLPP/CIN Tg mice (Fig. 7f, g). KA increased NEDD4-2 ubiquitination in both groups (p < 0.05 vs. control animals; Fig. 7h, i). NEDD4-2 ubiquitination in PLPP/CIN Tg mice was higher than that in WT mice (p < 0.05 vs. WT mice; Fig. 7h, i). In contrast, KA decreased GluA1 ubiquitination in WT mice, but not in PLPP/CIN Tg mice (p < 0.05 vs. control animals; Fig. 7h, j). Focal AMPA injection revealed that the efficacy of AMPA-mediated neuronal discharge in PLPP/CIN Tg mice was higher than that in WT mice (p < 0.05 vs. WT mice, Fig. 8a, b). These findings indicate that impaired NEDD4-2-mediated GluA1 ubiquitination may be responsible for seizure progression in PLPP/CIN Tg mice. Taken together, our findings suggest that PLPP/CIN-mediated NEDD4-2 S448 dephosphorylation may facilitate NEDD4-2 ubiquitination, which may enhance surface AMPAR expression and seizure progression (Fig. 8c).
Aforementioned, S448 site plays a role in NEDD4-2 protein stability independent of 14-3-3 bindings, since inhibition of NEDD4-2 S448 phosphorylation reduces NEDD4-2 protein level 12,25 . In the present study, PLPP/ CIN deletion increased NEDD4-2 protein level with enhancing NEDD4-2 S448 phosphorylation and reducing NEDD4-2 ubiquitination under physiological condition and after KA injection, while PLPP/CIN overexpression showed the reverse effects. With respect to ubiquitination of NEDD4-2 32,33 , these findings suggest that PLPP/CINmediated NEDD4-2 S448 dephosphorylation may be one of inhibitory mechanisms for regulation of NEDD4-2 activity via its ubiquitination. In the present study, furthermore, PLPP/CIN deletion reduced the phosphorylation ratio (the ratio of phosphoprotein to total protein) of S342, but elevated that of S448, concomitant with the increase in total NEDD4-2 protein level, which were reversed by PLPP/CIN overexpression. Since NEDD4-2 phosphorylation inhibits ubiquitinations of various ion channel and maintains its protein level 12,21,23,25,[32][33][34]39,40 , it is plausible that PLPP/CIN would inhibit NEDD4-2 E3 ubiquitin ligase activity by reducing target channel-NEDD4-2 interaction and its protein level by increasing S342 and diminishing S448 phosphorylation ratio, respectively. However, the present data could not provide the evidence for the role of S342 phosphorylation ratio in NEDD4-2-substrate interactions. Furthermore, the present study could not verify whether NEDD4-2 ubiquitin modification results from self-ubiquitination or the activations of other unknown E3 ligases, although NEDD4-2 is auto-ubiquitinated like other E3 ligases 32,33 . Thus, it would be needed to elucidate the role of the relative NEDD4-2 phosphorylation ratio in its activity, and the specific E3 ligase for NEDD4-2 ubiquitination in future studies.
NEDD4-2-mediated ubiquitination is one of the underlying machineries for synaptic homeostatic scaling, which is a feedback mechanism for maintenance of a physiological rage of neuronal excitability 24,42,43 . Thus, perturbation of the homeostatic scaling following stimulation may trigger a prolonged neuronal hyperexcitability inducing seizure activity 24 . In the present study, PLPP/ CIN deletion interrupted seizure progression induced by KA, which was reversed by NEDD4-2 knockdown. Consistent with the present data, Nedd4-2 andi mice show the elevated KA-induced seizure susceptibility 13,14 . The present study also reveals that the enhancements of PLPP/ CIN-NEDD4-2 bindings and NEDD4-2 ubiquitination in PLPP/CIN Tg mice increased seizure intensity and its progression in response to KA. Thus, our findings indicate that PLPP/CIN-mediated NEDD4-2 ubiquitination may enhance seizure activity and lead to seizure progression.
Recently, it has been reported that the higher seizure susceptibility due to insufficient function of NEDD4-2 in Nedd4-2 andi mice is recovered by the genetically reducing GluA1 level 14 . Furthermore, Fmr1 −/− mice (the fragile X syndrome mouse model) show neuronal hyperexcitability, because loss-of-function dephosphorylation of NEDD4-2 impairs NEDD4-2-GluA1 interaction and GluA1 ubiquitination after chronic activity stimulation, but not under physiological condition 24 . Therefore, it is likely that NEDD4-2 may play an important role in downregulation of synaptic homeostatic scaling via GluA1 ubiquitination 13,14 . In the present study, PLPP/CIN regulated membrane GluA1 expression and its ubiquitination under physiological condition. Furthermore, PLPP/CIN increased the seizure intensity due to the reduced GluA1 ubiquitination following KA injection. Since PLPP/CIN enhances the synaptic strength during high-frequency stimulation 28 , our findings suggest a novel mechanism by which PLPP/CIN inhibits synaptic downscaling by dephosphorylating NEDD4-2 S448 site and disrupting the subsequent GluA1 ubiquitination during neuronal hyperexcitability. Conversely, it is plausible that GluA1-NEDD4-2 bindings would be affected by GluA1 phosphorylation level, since the Na v 1.6 S553 phosphorylation by p38 mitogen-activated protein kinase is required for its binding to NEDD4-2 44 . Under physiological conditions, however, PLPP/CIN deletion/overexpression cannot affect GluA1 S831 and S845 phosphorylation levels 28 . Therefore, our findings also indicate that NEDD4-2mediated GluA1 ubiquitination may be independent of GluA1 phosphorylation.
In the case of Fmr1 −/− mice 24 , the activity-dependent dephosphorylation of NEDD4-2 is caused by p53medaited AKT destabilization, which is an upstream pathway of SGK1 12,45 . In the present study, KA reduced SGK1 activity and S342 and S448 phosphorylations of (see figure on previous page) Fig. 4 Effects of PLPP/CIN deletion on surface KCNQ and GluA1 expression, GluA1-NEDD4-2 binding, GluA1 ubiquitination, and neuronal activity in response to AMPA. a-c Effects of PLPP/CIN deletion on surface KCNQ and GluA1 expression following KA injection. As compared to WT mice, PLPP/CIN −/− mice show the diminished surface GluA1, but not KCNQ2/3/5, expression under physiological condition and after KA injection. a Representative western blots of surface and total KCNQ and GluA1 expressions. b, c Quantification of surface and total KCNQ and GluA1 expressions based on western blot data. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; * ,# p < 0.05 vs. saline-treated and WT animals, respectively; n = 7). d-g Effects of PLPP/CIN deletion on GluA1-NEDD4-2 binding and GluA1 ubiquitination. As compared to WT mice, PLPP/CIN −/− mice show the increased GluA1-NEDD4-2 bindings and GluA1 ubiquitination under physiological condition and after KA injection. d, e Representative western blots of GluA1-NEDD4-2 binding and GluA1 ubiquitination. f, g Quantification of GluA1-NEDD4-2 binding and GluA1 ubiquitination based on western blot data. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; * ,# p < 0.05 vs. saline-treated and WT animals, respectively; n = 7). h, i Effect of PLPP/CIN deletion on neuronal activity in response to AMPA. PLPP/CIN −/− mice demonstrate the reduced neuronal activity in response to AMPA. h Representative EEG traces and frequency-power spectral temporal maps in response to AMPA. i Quantification of total EEG power in response to AMPA (mean ± S.E.M.; n = 7, respectively) NEDD4-2. Thus, it is likely that activity-dependent SGK1 inhibition would also be involved in NEDD4-2 dephosphorylation. However, PLPP/CIN expression level influenced only NEDD4-2 S448 phosphorylation, but not NEDD4-2 S342 and SGK1 phosphorylation, under physiological and postseizure conditions. Therefore, our findings suggest that PLPP/CIN may fine-tune NEDD4-2-GluA1 interaction and GluA1 ubiquitination by S448 dephosphorylation under basal and pathological conditions independent of SGK1 activity.
Impaired KCNQ regulation causes spontaneous seizures due to reduction in M-current that exerts a stabilizing effect on the resting membrane potential and attenuates neuronal excitability 46,47 . NEDD4-2 phosphorylation at S448 site is essential for the SGK1dependent regulation of KCNQ2/3 and KCNQ3/5 channels in Xenopus oocytes 39 . Considering E3 ubiquitin ligase activity of NEDD4-2 for KCNQ 34 , it is likely that PLPP/ CIN-mediated NEDD4-2 ubiquitination would affect the surface KCNQ2/3/5 expressions. However, PLPP/CIN deletion/overexpression and NEDD4-2 knockdown did not show any significant effect on total or surface KCNQ2/3/5 expression, suggesting the dominant role of NEDD4-2 in GluA1 ubiquitination. These data demonstrate that PLPP/CIN may not influence the general ubiquitination process mediated by NEDD4-2. Since NEDD4-2 ubiquitinates multiple membrane receptors, including voltage-gated Na + channel 16 , neurotrophin receptors 17 , and glutamate transporters 15 , however, the possibility is not excluded that changes in ubiquitination of other substrates could contribute to seizure activity when PLPP/CIN are deleted or overexpressed. Thus, further studies are needed to elucidate whether and how PLPP/CIN mediates ubiquitination status of other NEDD4-2 substrates, which contributes to seizure generation and/or its progression.
On the other hand, Zhu et al. 14 have revealed that NEDD4-2 mutation increases seizure susceptibility in response to KA. Interestingly, the same group has also reported that murine double minute-2 triggers the degradation of p53 and subsequent induction of NEDD4-2 as an adaptive response in neuronal culture upon chronic elevation of neuronal activity by picrotoxin, a GABA A receptor antagonist 13,38 . In the present study, however, NEDD4-2 expression level was reduced in the hippocampus of WT mice 2 h after KA injection. Similar to the present data, pilocarpine (a muscarinic acetylcholine receptor agonist)-induced acute seizure activity rapidly degrades NEDD4-2 in the rat hippocampus 2 h after injection 48 . These discrepancies would result from the differences in methods (neuronal culture study vs. in vivo study), seizure severity in response to convulsants (picrotoxin vs. pilocarpine or KA), or dose of KA (60 vs. 25 mg/kg). However, the data from these previous reports and the present study identically suggest that the reduced NEDD4-2 expression may increase seizure severity and its duration. Indeed, chronic epilepsy rat shows downregulation of NEDD4-2 expression 48 , and NEDD4-2 knockdown increased seizure severity and its duration in PLPP/CIN −/− mice in the present study. Furthermore, the present data demonstrate that PLPP/CIN activity in PLPP/CIN Tg mice was higher than that in WT mice under physiological condition. PLPP/CIN overexpression also diminished NEDD4-2 protein level with enhancing NEDD4-2 S448 dephosphorylation and its ubiquitination under physiological condition. In addition, KA enhanced NEDD4-2 S448 dephosphorylation in WT and PLPP/ CIN Tg mice without altering PLPP/CIN activity, indicating that seizure activity may increase NEDD4-2 dephosphorylation via promoting physical interaction between PLPP/CIN and NEDD4-2. Therefore, our findings suggest that PLPP/CIN-mediated NEDD4-2 dephosphorylation may play an important role in the seizure intensity and its progression, regardless of the presence of compensatory upregulation of NEDD4-2 expression in response to seizures 13,38,48 .
In conclusion, we describe a new paradigm for the regulation of NEDD4-2 activity by PLPP/CIN-mediated S448 dephosphorylation (Fig. 8c). To our knowledge, these findings provide the novel PLPP/CIN-mediated mechanism underlying seizure progression by modulating NEDD4-2-mediated GluA1 ubiquitination and suggest the development of therapeutic strategies for various neurological and psychiatric disorders, including epilepsy.

Experimental animals and chemicals
PLPP/CIN −/− (129/SvEv-C57BL/6J background) and PLPP/CIN Tg (C57BL/6J background) mice were used in the present study. Each background WT mice were used as control animals, respectively. Animals were provided with a commercial diet and water ad libitum under controlled temperature, humidity, and lighting conditions (22 ± 2°C, 55 ± 5%, and a 12:12 light/dark cycle). All experimental protocols were approved by the Animal Care and Use Committee of Hallym University. All reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA), except as noted.

Seizure induction and EEG Recording
After baseline recording for at least 30 min, animals were given KA (25 mg/kg, i.p.). Control animals received an equal volume of normal saline instead of KA. EEG signals were recorded with a DAM 80 differential amplifier (0.1-1000 Hz bandpass; World Precision Instruments, USA) and the data were digitized (1000 Hz) and analyzed using LabChart Pro v7 software (AD Instruments, Australia). Latency of seizure onset was defined as the time point showing more than 3 s and consisting of a rhythmic discharge between 4 and 10 Hz with amplitude of at least two times higher than the baseline EEG 4,30 . Total EEG power was normalized by the baseline power obtained from each animal. Spectrograms were automatically calculated using a Hanning sliding window with 50% overlap by LabChart Pro v7. Diazepam (Valium; Roche, France; 10 mg/kg, i.p.) was administered 2 h after KA injection and repeated, as needed. After recording, animals were quickly decapitated, and their hippocampi were dissected out in the presence of cooled artificial cerebrospinal fluid (in mM: 124 NaCl, 5 KCl, 1.25 NaH 2 PO 4 , 26 NaHCO 3 , 10 (see figure on previous page) Fig. 6 Effects of PLPP/CIN overexpression on seizure activity and NEDD4-2 phosphorylations in response to KA. a-c Effect of PLPP/CIN overexpression on seizure activity in response to KA. PLPP/CIN Tg mice demonstrate the increases in the latency of seizure onset and seizure intensity. a Representative EEG traces and frequency-power spectral temporal maps in response to KA. b Quantification of total EEG power in response to KA (mean ± S.E.M.; n = 7, respectively). c Quantification of the latency of seizure onset. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; *p < 0.05 vs. WT animals; n = 7, respectively). d-h Effects of PLPP/CIN overexpression on PLPP/CIN activity, SGK1, and NEDD4-2 phosphorylations in response to KA. PLPP/CIN overexpression increases PLPP/CIN activity under physiological condition, and facilitates the reductions in total NEDD4-2 and pNEDD4-2 S448 phosphorylation levels, but not SGK1 and its phosphorylations, under physiological condition and after KA injection. d Quantification of PLPP/CIN activity (*p < 0.05 vs. control WT animals; n = 7, respectively). e Quantification of SGK1 and its phosphorylation levels (*p < 0.05 vs. control animals; n = 7, respectively). f Quantification of NEDD4-2 expression and its phosphorylation levels. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; * ,# p < 0.05 vs. WT and saline-treated animals; n = 7, respectively). g Representative western blots of SGK1 and NEDD4-2. h Quantification of the ratios of pNEDD4-2 to total NEDD4-2.

Analysis of neuronal activity in responses to AMPA
After baseline recording for at least 30 min, animals implanted with a guide-electrode-combo (C313G-MS303/ 2/SPC, Plastics One, USA) were directly infused AMPA (20 μM) over a 1-min period using a microinjection pump (1 μl/min, KD Scientific, USA) into the hippocampus with an internal infusion cannula (C315IA, Plastics One, USA) 29 . EEG signals were digitized and analyzed using LabChart Pro v7 (AD Instruments) by the same methods aforementioned.

Co-immunoprecipitation and membrane fraction
The hippocampal tissues were lysed in the radioimmunoprecipitation assay buffer (RIPA: 50 mM Tris-HCl pH 8.0; 1% Nonidet P-40; 0.5% deoxycholate; 0.1% SDS, Thermo Fisher Scientific, USA) containing protease inhibitor cocktail (Roche Applied Sciences, USA), phosphatase inhibitor cocktail (PhosSTOP ® , Roche Applied Science, USA) and 1 mM sodium orthovanadate. Protein concentrations were calibrated by the BCA protein assay (Pierce, USA) and equal amounts of total proteins were incubated with NEDD4-2, Ubiquitin, or GluA1 antibody (Supplementary Table 1) and protein G sepharose beads at 4°C overnight. Beads were collected by centrifugation, eluted in 2 × SDS sample buffer, and boiled (see figure on previous page) Fig. 7 Effects of PLPP/CIN overexpression on surface KCNQ and GluA1 expression, GluA1-NEDD4-2 bindings, and GluA1 ubiquitination. a-c Effects of PLPP/CIN overexpression on surface KCNQ and GluA1 expression following KA injection. As compared to WT mice, PLPP/CIN Tg mice demonstrate increases surface GluA1, but not KCNQ2/3/5, expression under physiological condition and after KA injection. a Representative western blots of surface and total KCNQ and GluA1 expressions. b, c Quantification of surface and total KCNQ and GluA1 expressions based on western blot data. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; * ,# p < 0.05 vs. saline-treated and WT animals, respectively; n = 7). d, e Effects of PLPP/CIN overexpression on PLPP/CIN-NEDD4-2 bindings. As compared to WT mice, PLPP/CIN Tg mice show the increased PLPP/CIN-NEDD4-2 binding under physiological condition and after KA injection. d Representative western blots of PLPP/CIN-NEDD4-2 binding. e Quantification of PLPP/CIN-NEDD4-2 binding based on western blot data. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; * ,# p < 0.05 vs. saline-treated and WT animals, respectively; n = 7). f, g Effects of PLPP/CIN overexpression on GluA1-NEDD4-2 binding. PLPP/CIN Tg mice demonstrate the diminished GluA1-NEDD4-2 binding under physiological condition and after KA injection. f Representative western blots of PLPP/CIN-NEDD4-2 bindings. g Quantification of PLPP/CIN-NEDD4-2 bindings based on western blot data. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; * ,# p < 0.05 vs. saline-treated and WT animals, respectively; n = 7). h-j Effects of PLPP/CIN overexpression on the ubiquitinations of NEDD4-2 and GluA1. PLPP/CIN Tg mice demonstrate the increased NEDD4-2 ubiquitination, but the reduced GluA1 ubiquitination under physiological condition and after KA injection. h Representative western blots of PLPP/ CIN-NEDD4-2 bindings. i, j Quantification of ubiquitinations of NEDD4-2 and GluA1 based on western blot data. Open circles indicate each individual value. Horizontal bars indicate mean value (mean ± S.E.M.; * ,# p < 0.05 vs. saline-treated and WT animals, respectively; n = 7) Fig. 8 Effect of neuronal activity in response to AMPA and the hypothesized role of PCPP/CIN-mediated GluA1 ubiquitination. a, b Effects of PLPP/CIN overexpression on neuronal activity in response to AMPA. PLPP/CIN Tg mice demonstrate the increased neuronal activity in response to AMPA. a Representative EEG traces and frequency-power spectral temporal maps in response to AMPA. b Quantification of total EEG power in response to AMPA (mean ± S.E.M.; p < 0.05 vs. WT animals; n = 7, respectively). c Scheme of inhibitory role of PLPP/CIN in NEDD4-2-mediated GluA1 ubiquitination. Protein kinases including SGK1 phosphorylate NEDD4-2 S342 and S448 sites, which facilitate its binding to and ubiquitination of GluA1. However, PLPP/CIN dephosphorylates NEDD4-2 S448 site, which leads to ubiquitination of NEDD4-2 and subsequently inhibits GluA1 ubiquitination, independent of SGK1 activity. These PLPP/CIN-mediated functional couplings of NEDD4-2 and GluA1 increase seizure intensity and its progression at 95°C for 5 min. To analyze membrane expressions of KCNQ and GluA1, we used a subcellular Protein Fractionation Kit for Tissues (Thermo Scientific, USA), according to the manufacturer's instructions.

PLPP/CIN activity assay
The hippocampal tissues were homogenized in TE buffer. Total protein content was measured by the BCA protein assay kit. To measure PLPP/CIN activity in lysates, we used PLPP/CIN activity assay kits (#OKEH01229 and #OKEH03398; Aviva systems biology, USA), according to the manufacturer's instructions.

Western blot
Western blotting was performed according to standard procedures. The list of primary antibody used in the present study is Supplementary Table 1. The rabbit antiβ-actin (input) or N-cadherin (membrane fraction) was used as internal reference. The signals were scanned and quantified on ImageQuant LAS 4000 system (GE health, USA). The values of each sample were normalized with the corresponding amount of β-actin or N-cadherin.

Fluoro-Jade B (FJB) Staining and cell counting
One day after KA injection, animals were perfused transcardially with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) under urethane anesthesia (1.5 g/kg, i.p.). Brains were postfixed in the same fixative overnight and then cryoprotected and sectioned at 30 μm with a cryostat. Thereafter, tissues were used for a conventional FJB staining according to previous studies 49,50 . All images were obtained using an AxioImage M2 microscope and AxioVision Rel. 4.8 software. Areas of interest (1 × 10 5 μm 2 ) were selected in the captured images of the CA3 region of the hippocampus proper (ten sections per each animal), and the number of FJB-positive neurons was counted 49,50 .

Statistical analysis
After evaluating the values on normality using Shapiro-Wilk W-test, Student's t test or ANOVA were used to analyze statistical significance. Bonferroni's test was applied for post hoc comparisons. A p-value below 0.05 was considered statistically significant. Quantitative data were expressed as mean ± standard error of the mean.