**3. Discussion**

The present study describes a novel mechanism of KA-induced seizure which involves p38-dependent phosphorylation of Kv4.2 at T607. Both T602 and T607 are phosphorylated by KA but the induction timing is di fferent (Figure 2A–C). KA induces T607 phosphorylation relatively promptly (about 15 min) while T602 phosphorylation is relatively delayed (about 3 h). Both T602 and T607 phosphorylation were sustained for at least 5 days, the longest data point measured (Figure 2A–C). Kindling is a commonly used model for the development of seizures and epilepsy in which the duration and behavioral involvement of induced seizures increases after seizures are induced repeatedly [45]. Repeated seizure could boost Kv4.2 phosphorylation levels at T602 and T607, leading to downregulation and enhanced excitability. Therefore, the long-term phosphorylation response to seizure could be a mechanism of kindling.

We have found that Pin1 binds to T607 of Kv4.2 and isomerizes the T607-P bond to modulate the function of Kv4.2 [42]. The dual phosphorylation of T602 and T607 increases the Pin1 binding ability and therefore improves Kv4.2 modulation. KA triggers Kv4.2 phosphorylation at both sites after 3 h and lasts for 5 days (Figure 2A–C), suggesting that the Pin1 e ffect could be long lasting as well. The excitability of CA1 pyramidal neuron dendrites was increased in TLE because of the decreased availability of A-type potassium ion channels [28]. Pin1's long-lasting e ffect could eventually lead to reduced availability of Kv4.2, which fits with the notion that seizure decreases IA. Furthermore, the continued phosphorylation at these sites and persistent Pin1 activity in response to prolonged seizure may exacerbate their severity, manifesting as a positive feedback loop, promoting further downregulation of I A. Thus, the activation of p38 and subsequent isomerization of Kv4.2 may serve as a mechanism underlying the intractability of seizure progression often associated with severe epilepsy in the temporal lobe [1–3].

Kv4.2 T602 and T607 phosphorylation occur via both p38 MAPK (Figure 2A–C) and ERK MAPK [38,46]. However, the p38 inhibitor SB 203,580 blocked Kv4.2 phosphorylation induced by KA (Figure 3A,B), suggesting the primary role of p38 in seizure. Since ERK MAPK can be activated by KA [47], ERK may also have a contributing e ffect on Kv4.2. In addition, KA activates p38 in Kv4.2TA mice at similar levels as in WT mice, suggesting the reduced seizure phenotype in Kv4.2TA mice (Figure 1) is not because of the di fferential induction of p38 but the deficiency of Kv4.2 phosphorylation at T607.

At the cellular level, Kv4.2 phosphorylation by p38 at T607 alters the excitability of hippocampal pyramidal neurons of area CA1. The results presented here indicate that the role of p38 in altering membrane excitability is primarily through a reduction in suprathreshold excitability. While Kv4.2 can impact subthreshold properties in pyramidal neurons, including latency to AP onset [48] and rheobase [42], its role in regulating the frequency of repetitive AP firing is well documented [48,49]. Its robust control of membrane potential fluctuations in response to depolarizing currents from resting potential permits a significant functional interaction with voltage-gated Na<sup>+</sup> channels [48]. Indeed, we identified a significant increase in AP threshold in Kv4.2TA pyramidal neurons. Moreover, a slight increase in after-hyperpolarization amplitude in these cells is likely contributory in delaying the recruitment of Na<sup>+</sup> channel activation to stepped current injection, which was evidenced in this study and our previous analysis [42]. It is likely that the enhancement of I A amplitude that results with p38-Pin1 blockade reduces the precision of repetitive spiking through the modulation of voltage-gated channels, driving spiking in hippocampal pyramidal neurons.

It is clear that the T607A mutation induces a reduction of pyramidal cell excitability. The trend in reduced excitability of Kv4.2TA neurons was persistent in this analysis, strengthening our previous findings [42]. While slight variations in neuronal responsiveness to current injections of various magnitudes in Kv4.2TA mice relative to WT were observed in this study relative to [42], the overall trend toward a reduction in repetitive firing remained. Slight alterations in the slice recovery and recording conditions are likely to underlie these di fferences. Furthermore, by complementing the T607A mutation with pharmacological blockade of p38, we uncovered multiple means by which to lower hippocampal pyramidal cell excitability through Kv4.2. The application of SB 203580 altered the firing mode of WT neurons in a manner that mirrored that of Kv4.2TA neurons in its absence. This suggests that p38 regulation of membrane excitability is primarily mediated through Kv4.2. We did, however, note that, although p38 blockade did not significantly alter AP firing properties in Kv4.2TA slices, a reduction in excitability was generally augmented in its presence. This may imply that the inhibition of phosphorylation of both T602 and T607 sites may, together, produce an additive effect in Kv4.2 regulation of membrane properties. As noted, the dual phosphorylation of these motifs provides an environment particularly conducive to binding multiple domains of Pin1 upon p38 phosphorylation, which could facilitate dissociation of Kv4.2 from DPP6 and downregulation [42]. While this likely contributes to the observed phenotype, we cannot, however, rule out the involvement of additional ion channels that may also be impacted by broad pharmacological p38 blockade.

Furthermore, the breadth of this work and that of our previous study focuses on somatic Kv4.2 channel activity. While somatic Kv4 channels are capable of impacting neuronal firing modes, their privileged distribution in dendrites suggests their control of membrane potentials may be more impactful in these domains [23,50–52]. It is possible that p38 phosphorylation of Kv4.2 may contribute to the alteration of dendritic excitability by modulating the coupling of synaptic inputs and AP output. Indeed, in the context of the regulation of DPP6-Kv4.2 dynamics, p38 activity in the apical dendrites may be of particular significance. DPP6 knockout mice exhibit a non-uniformity in their manifestation of alteration in membrane excitability, with dendritic excitability being predominantly impacted [53]. Whether p38 phosphorylation of dendritic Kv4.2 channels may impact dendritic function, and further contribute to increased network hyperexcitability and seizure severity in the hippocampus, is a topic of future investigation.

#### **4. Materials and Methods**
