**2. Results**

#### *2.1. p38 MAPK Contributes to Kainic Acid-Induced Seizure in WT but Not Kv4.2TA Mice*

We have generated a mutant mouse, Kv4.2TA, with abolished dynamic Thr 607 phosphorylation of Kv4.2 and isomerization of Kv4.2 by Pin1 [42]. Kv4.2TA mice displayed increased IA, decreased neuronal excitability, and improved cognitive flexibility [42]. Here, we examine if acute behavioral seizure is altered following the systemic injection of kainic acid (KA) in the Kv4.2TA mice. KA (25 mg/kg) was injected intraperitoneally into Kv4.2TA mice (*n* = 13) and littermate controls (*n* = 15) and behavioral seizure responses were scored using the modified Racine scale [43] for 60 min post injection. We observed a significant difference in behavior seizure scores, with Kv4.2TA mice showing significantly reduced seizure intensity over the full 1 h period following KA injection (Figure 1A,B). Our previous work showed that p38 can phosphorylate Kv4.2 at T607 in response to seizure induced by pentylenetetrazol (PTZ) or KA [42]. In addition, acute behavioral seizure in p38 knockout mice is significantly decreased compared to WT mice [44]. Therefore, we hypothesized that the effect of p38 on seizure intensity is dependent on the T607 site of Kv4.2. We injected p38 inhibitor SB 203580 (20 mg/kg, i.p.) 15 min ahead of KA injection. The result showed that control mice with SB 203580 injection (*n* = 14) exhibited significantly reduced behavioral seizure intensity compared to those with the control injection (Figure 1A,B). Interestingly, Kv4.2TA mice with SB 203580 injection (*n* = 13) did not display a significant reduction in behavioral seizure intensity compared to those injected with vehicle (Figure 1A,B). This decrease in sensitivity to KA-induced seizure was also reflected in the latency to stage 3 seizure (Figure 1C). Taken together, these data support the notion that p38 phosphorylation of Kv4.2 at T607 contributes to KA-induced seizure.

#### *2.2. Seizure Induced by Kainic Acid Triggers Kv4.2 T607 Phosphorylation in a Time-Dependent Manner*

It has been reported that Kv4.2 availability is altered in response to seizure, suggesting these events may trigger a molecular cascade leading to the functional downregulation of IA [28,29]. Our collective analyses indicate that this cascade is initiated by p38 kinase. We reason that the prolonged activation of p38 and subsequent phosphorylation of Kv4.2 in response to continued seizures (kindling) may be an important factor underlying the intractability of seizures in the temporal lobe. Therefore, in order to study the timing of Kv4.2 phosphorylation in response to seizure, we examined Kv4.2 phosphorylation at various times following KA injection. Since p38 can phosphorylate both T602 and T607 of Kv4.2 [42], we examined both phosphorylation sites. KA induced Kv4.2 phosphorylation at T607 at 15 min after KA injection but not at 5 min (Figure 2A,B). T607 phosphorylation peaked 3 h after KA injection and the induction effect lasted, even at 5 days post injection (Figure 2A,B). Kv4.2 phosphorylation at T602 was also induced 3 h post KA injection but not at earlier time points (Figure 2A,C). These data show that seizure induced by KA triggers a long-lasting effect on Kv4.2 that may contribute to on-going seizure progression.

#### *2.3. Kainic Acid-Induced Kv4.2 Phosphorylation at T607 Is Dependent on p38 MAPK*

Next, we wanted to know if p38 is required for KA-induced Kv4.2 phosphorylation at T607. The p38 inhibitor SB 203580 (20 mg/kg, i.p.) was injected 15 min ahead of KA injection (25 mg/kg, i.p.). We found that the p38 inhibitor SB 203580 blocked the induction of Kv4.2 phosphorylation 15 min after KA injection in the mouse hippocampus (Figure 3A,B), suggesting p38 contributed to KA-induced Kv4.2 phosphorylation at T607.

**Figure 1.** p38 mitogen-activated protein kinase (MAPK) contributed to kainic acid-induced seizure in WT mice but not Kv4.2TA mice. (**A**) Time course of mean behavioral seizure score following kainic acid injection. The mean behavioral seizure score was significantly reduced in Kv4.2TA mice compared to WT mice. Furthermore, p38 inhibitor SB 203580 significantly reduced behavioral seizure score following kainic acid injection in WT mice but not in Kv4.2TA mice, *n* = 13–15 for each group, two-way ANOVA, \* *p* < 0.05. (**B**) Total behavioral seizure score for each group, *n* = 13–15 for each group, *t*-test, \* *p* < 0.05. (**C**) Latency to stage 3 seizure for each group. *n* = 13–15 for each group, *t*-test, \*\* *p* < 0.01.

#### *2.4. p38 MAPK Colocalizes with Kv4.2*

Since p38 phosphorylates Kv4.2, we wanted to see if it colocalized with Kv4.2 in a heterologous system and in the mouse brain. First, HEK 293T cells were double stained after 2 days of transfection with p38 and Kv4.2. The result showed that p38 partially colocalized with Kv4.2 (Figure 4A). High magnification images and line scan confirmed this result (Figure 4B,C). In addition, we performed the double staining of phosphorylated p38 (pp38) and Kv4.2 on mouse brain sections. Phospho-p38 is mainly localized in the cell body of hippocampal pyramidal neurons but also localized in dendrites, while Kv4.2 is mainly localized in dendrites (Figure 4C). High magnification images showed pp38 partially colocalized with Kv4.2, as indicated by arrow heads (Figure 4D).

**Figure 2.** Seizure induced by kainic acid triggers Kv4.2 T607 phosphorylation in a time-dependent manner in mouse hippocampus. (**A**) Time course of Kv4.2 phosphorylation at Thr602 and Thr607 by kainic acid administration (25 mg/kg, i.p.) in mouse hippocampus. (**B**) Statistical analysis of kainic acid-induced phosphorylation of Kv4.2 at Thr607 in mouse hippocampus, *n* = 3–8 in each group, *t*-test, \*\* *p* < 0.01. (**C**) Statistical analysis of kainic acid-induced phosphorylation of Kv4.2 at Thr602 in mouse hippocampus, *n* = 3–8 in each group, *t*-test, \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001.

**Figure 3.** p38 MAPK contributes to kainic acid-induced Kv4.2 phosphorylation at T607. (**A**) SB 203580, a potent p38 inhibitor (20 mg/kg, i.p., 15 min), blocked kainic acid-induced phosphorylation of Kv4.2 T607 in mouse hippocampus. (**B**) Statistical analysis of the effect of SB 203580 on kainic acid-induced phosphorylation of Kv4.2 at Thr607 in mouse hippocampus, *n* = 4–6 in each group, *t*-test, \*\* *p* < 0.01.

**Figure 4.** p38 MAPK colocalizes with Kv4.2. (**A**) HEK293T cells were transfected with GFP-Kv4.2 and Flag-p38. Cells were fixed and stained with GFP and Flag to show co-localization. Scale bar: 20 μm. (**B**) High magnification images and line scan analysis of colocalization. Scale bar: 5 μm. (**C**) Mouse brains were co-stained with Kv4.2 and pp38 antibody. Phosphorylated p38 is localized in the cell body and dendrites as well. Scale bar: 20 μm. (**D**) High magnification images showing Kv4.2 and pp38 colocalized in dendrites, as indicated with arrow heads. Scale bar: 5 μm.

#### *2.5. Kainic Acid Activates p38 MAPK in both WT and Kv4.2TA Mice*

We next assessed whether KA-induced seizure activated p38 in both WT mice and Kv4.2TA mice. Mouse brain sections were stained with pp38. The pp38 level is significantly increased after KA injection (1 h) in the cell body of hippocampal pyramidal neurons (Figure 5A). Furthermore, we examined pp38 by western blot. The pp38 level but not p38 level is significantly increased after KA injection (30 min) in WT mouse hippocampus (Figure 5B). A similar result was found in Kv4.2TA mice (Figure 5B). These data indicate that the initiation of p38 kinase activity by KA is similar in WT and Kv4.2TA mice, and the observed effects on seizure intensity can be ascribed to the inability of p38 to phosphorylate Thr607.

**Figure 5.** Kainic acid activates p38 MAPK in both WT and Kv4.2TA mice. (**A**) Immunostaining analysis showed p38 phosphorylation increased with kainic acid administration (25 mg/kg, i.p., 30 min) in mouse hippocampus, *n* = 26 cells in each group, *t*-test, \*\*\* *p* < 0.001. (**B**) Western blot analysis showed p38 phosphorylation increased with kainic acid administration (25 mg/kg, i.p., 30 min) in hippocampus in both WT and Kv4.2TA mice, *n* = 4–6 cells in each group, *t*-test, \*\*\* *p* < 0.001.

#### *2.6. p38 MAPK Modulates Neuronal Excitability through Kv4.2*

In our previous study, we reported that hippocampal pyramidal cells from acute hippocampal slices of Kv4.2TA mice exhibited a nearly two-fold reduction in AP firing frequency in response to somatic current injection relative to WT mice. We determined that this was due to an enhancement of IA as a result of the T607A mutation blocking the dissociation of Kv4.2 from its auxiliary subunit DPP6 and subsequent functional downregulation [42]. Furthermore, we found the slicing process largely activated p38 kinase and induced Kv4.2 phosphorylation, revealing that this procedure acts similarly to kainic acid-induced seizure in altering the phospho-state of Kv4.2 in the hippocampus [42]. Therefore, we sought to investigate how a pharmacological blockade of p38 (SB 203580) altered the excitability of the principal hippocampal neurons in area CA1 of acute hippocampal slices. In light of the time-dependency of phosphorylation of Kv4.2 at T602 and T607 in response to seizure induction (peak ~3 h), we incubated slices in recovery solution with pharmacological treatment or vehicle (0.1% DMSO) for 2 h and continued their exposure during electrophysiological recordings (3–4 h total exposure). In response to stepped somatic current injections, we identified that treatment with 5 μM SB 203580 reduced AP firing frequency of pyramidal neurons at each magnitude above rheobase in WT slices (Figure 6A,B). At maximum current injection (+300 pA), the difference in firing frequency reached a statistically significant level. Specifically, a +300 pA injection induced a firing rate of 23.7 Hz in the presence of DMSO, representing a~33% increase relative to the rate recorded in the presence of 5 μM SB 203580 (16.9 Hz), shown in Figure 6B. This alteration in excitability was limited to suprathreshold properties, as subthreshold excitability was largely unaffected by pharmacological p38 blockade (Table 1). Parameters measured from ramp current injections (400 pA/s), including AP threshold

(Table 1), rheobase (Figure 6D,E), and latency to AP onset (Figure 6F), were similar in conditions with or without SB 203580, although a modest, but non-statistically significant, increase in AP threshold and rheobase was observed in slices from WT mice with pharmacological p38 blockade (Table 1 and Figure 6E).

**Figure 6.** p38 impacts hippocampal pyramidal neuron excitability through Kv4.2. (**A**) Current step of +300 pA induces repetitive firing in pyramidal neurons recorded from WT and Kv4.2TA mice with or without SB 203580 treatment. Scale 40 mV/250 ms. Square current inset 300 pA. (**B**) Sequential somatic current injections increasing in magnitude reveal p38 kinase inhibition reduces AP firing frequency in WT hippocampal neurons at +300 pA relative to vehicle (*n* = 15 in vehicle, *n* = 19 in treatment; two-way ANOVA, \* *p* < 0.05). Kv4.2TA neurons display reduced firing frequency at +300 pA relative to WT in vehicle, which is augmented in the presence of SB 203580 such that current magnitudes of +200 and +250 pA also exhibit significant differences (*n* = 18 in vehicle, *n* = 14 in SB 203580; two-way ANOVA, \* *p* < 0.05; \*\*\* *p* < 0.001). (**C**) Inter-spike intervals measured between the first two spikes in a train evoked by 150 pA injection display no significant difference among groups. Kruskal–Wallis test, *p* > 0.05. (**D**) Ramp current injections evoke repetitive firing in all pyramidal neurons recorded in each condition. Arrow indicates point at which action potential (AP) threshold, rheobase, and latency to fire were measured. Ramp current inset 400 pA/s. (**E**) Minimum current to elicit AP firing at threshold (rheobase) is not significantly different among the populations. One-way ANOVA, *p* > 0.05. (**F**) Latency to fire in response to ramp injection is not significantly different among populations. Kruskal–Wallis test, *p* > 0.05.


**Table 1.** Passive membrane properties and single action potential (AP) parameters (mean ± SEM).


a

We next tested the e ffect of SB 203580 on slices obtained from Kv4.2TA mice. Because p38-mediated phosphorylation of Kv4.2 is significantly reduced in these mice, we anticipated the mutation would occlude the impact of SB 203580 on neuronal firing if p38 mediates excitability primarily through its regulation of I A. Generally consistent with previous observations, pyramidal neurons from Kv4.2TA slices exhibited reduced AP firing frequency relative to WT in multiple experimental conditions (Figure 6A,B). AP firing frequency was reduced at each current magnitude, with a statistically significant reduction exhibited at peak injection in the presence of DMSO (17.5 Hz vs. 23.7 Hz, Kv4.2TA vs. WT, respectively; *p* < 0.05, Figure 6B). The significant decrease in AP firing frequency corresponded with a significant increase in AP threshold in Kv4.2TA neurons relative to WT in this condition (Table 1). Additionally, AP firing mirrored that of WT neurons in the presence of 5 μM SB 203580 (Figure 6B). Importantly, SB 203580 treatment in Kv4.2TA slices did not significantly reduce suprathreshold excitability, contrary to its impact in WT pyramidal neurons (Figure 6B). This suggests that p38 modulation of AP firing can primarily be traced to its regulation of Kv4.2. Furthermore, as noted previously, Kv4.2TA neurons in this condition displayed significant pauses in repetitive firing, which correlated with slight, non-statistically significant, increases in fast after-hyperpolarization amplitudes and inter-spike intervals (Table 1 and Figure 6C, respectively). Therefore, taken together, pharmacological blockade of p38 kinase reduced the suprathreshold excitability of hippocampal pyramidal neurons. The T607A mutation occludes the action of SB 203580, suggesting its impact on AP firing frequency is mediated predominantly through its modulation of Kv4.2-mediated I A.
