Increased KV2.1 Channel Clustering Underlies the Reduction of Delayed Rectifier K+ Currents in Hippocampal Neurons of the Tg2576 Alzheimer’s Disease Mouse

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by the progressive deterioration of cognitive functions. Cortical and hippocampal hyperexcitability intervenes in the pathological derangement of brain activity leading to cognitive decline. As key regulators of neuronal excitability, the voltage-gated K+ channels (KV) might play a crucial role in the AD pathophysiology. Among them, the KV2.1 channel, the main α subunit mediating the delayed rectifier K+ currents (IDR) and controlling the intrinsic excitability of pyramidal neurons, has been poorly examined in AD. In the present study, we investigated the KV2.1 protein expression and activity in hippocampal neurons from the Tg2576 mouse, a widely used transgenic model of AD. To this aim we performed whole-cell patch-clamp recordings, Western blotting, and immunofluorescence analyses. Our Western blotting results reveal that KV2.1 was overexpressed in the hippocampus of 3-month-old Tg2576 mice and in primary hippocampal neurons from Tg2576 mouse embryos compared with the WT counterparts. Electrophysiological experiments unveiled that the whole IDR were reduced in the Tg2576 primary neurons compared with the WT neurons, and that this reduction was due to the loss of the KV2.1 current component. Moreover, we found that the reduction of the KV2.1-mediated currents was due to increased channel clustering, and that glutamate, a stimulus inducing KV2.1 declustering, was able to restore the IDR to levels comparable to those of the WT neurons. These findings add new information about the dysregulation of ionic homeostasis in the Tg2576 AD mouse model and identify KV2.1 as a possible player in the AD-related alterations of neuronal excitability.


Introduction
Alzheimer's disease (AD) is a devastating neurodegenerative disorder representing the most common cause of dementia in elderly population. The AD pathology consists of the progressive deterioration of cognitive functions, resulting from an irreversible neuronal loss and the pathological derangement of brain circuits and networks, among other mechanisms [1,2]. According to a growing amount of evidence, the dysregulation of brain activity observed in AD is correlated with the alteration of neuronal excitability. In particular, network hyperexcitability and hypersynchrony, and epileptiform activity have been observed in AD mouse models [3][4][5] and proposed as crucial factors causing the malfunction of hippocampal circuitry and subsequent memory impairment [6,7]. Moreover, an increased risk of developing seizures has been observed in the AD patients with mutations in presenilin 1, presenilin 2, and amyloid precursor protein (APP), or with APP duplications [8,9]. Of note, numerous studies have reported that the Amyloid-β 1-42 (Aβ 1-42 ) peptide, the main component of both the neuritic plaques and intracellular aggregates found in the brain of AD patients, can trigger neuronal hyperactivity [10][11][12]. Accordingly, hyperactive neurons and WT females were purchased from a commercial source (B6; SJLTg(APPSWE) 2576 Kha, model 1349, Taconic Biosciences, Hudson, NY, USA).

Genotyping: PCR Analysis
Genomic DNA from embryonic brain tissues was isolated by salt precipitation as previously described [47]. Briefly, embryonic brain tissues were harvested during cerebral dissection and then thawed and homogenized with the TRI-reagent (Sigma-Aldrich, Milan, Italy). After adding one volume of chloroform to each sample, the DNA was precipitated with 100% ethanol and centrifuged at 4 • C for 5 min at 16,000× g. The DNA pellet was dried at room temperature and then re-suspended in Tris-EDTA buffer. The following primers were used to amplify the DNA region with the human APP Swedish mutation on both types of genomic DNA: 5 -CTGACCACTCGACCAGGTTCTGGGT-3 and 5 -GTGGAT AACCCCTCCCCC AGCCTAGACCA-3 (Eurofins Genomics, Ebersberg, Germany). The DNA was amplified to detect the transgenic genotype as previously described [13].

Primary Hippocampal Neurons
Primary neuronal cultures were prepared from the hippocampi of embryonic day-15 WT and Tg2576 mice as described by Ciccone et al. [13].The cells were plated on 35 mm culture dishes coated with poly(D)-lysine hydrobromide with a molecular weight >300,000 (Sigma-Aldrich, Milan, Italy) or onto 25 mm glass coverslips (Glaswarenfabrik Karl Hecht KG, Sondheim, Germany), coated with 100 µg/mL poly(D)-lysine hydrobromide with a molecular weight of 30,000-70,000 (Sigma-Aldrich, Milan, Italy) at a density of one embryo hippocampus/1 mL. Cytosine β-D-arabinofuranoside (10 µM, Sigma-Aldrich, Milan, Italy) was added 3 days after plating to inhibit non-neuronal cell growth. The neurons were cultured at 37 • C in a humidified 5% CO 2 atmosphere. The experiments were performed between 10-12 days in vitro (DIV).

Electrophysiological Recordings
K + currents were recorded from WT and Tg2576 primary hippocampal neurons with the patch-clamp technique in whole-cell configuration using a commercially available amplifier Axopatch200B and Digidata1322A interface (Molecular Devices, San Jose, CA, USA), as previously described [48,49]. In most experiments, the whole-cell configuration of the patch-clamp technique was adopted using glass micropipettes of 2.5-5 MΩ resistance. Currents were filtered at 5 kHz and data were acquired and analyzed using the pClamp software (version 9.0, Molecular Devices, CA, USA). The pipette solution contained (in mM): 140 KCl, 2 MgCl 2 , 10 HEPES, 10 glucose, 10 EGTA, and 1 Mg-ATP adjusted at pH 7.4 with KOH. The extracellular solution contained (in mM): 150 NaCl, 5.4 KCl, 3 CaCl 2 , 1 MgCl 2 , 10 HEPES, and adjusted pH 7.4 with NaOH. A total of 50 nM tetrodotoxin (TTX) and 10 µM nimodipine were added to Ringer's solution to abolish TTX-sensitive Na + , and L-type Ca 2+ current. Monoclonal K V 2.1 antibody (10 µg mL −1 ; Neuromab, Davis, CA, USA) was added to the pipette solution to block K V 2.1 currents [22]. The total K + currents (I K ) are composed of the fast-inactivating (I A ) currents and the delayed currents (I DR ).
To discriminate K + current components, I K = I A + I DR , appropriate electrophysiological protocols were used [48,49]. In particular, I DR were isolated by stepping from −80 mV to +40 mV for 250 ms after conditioning pulses at −40 mV, lasting 1.5 s to fully inactivate I A . Current amplitudes were measured for I DR at +40 mV at the end of the depolarizing pulse. The steady-state properties of I DR activation were measured as previously described [49]. The current amplitudes recorded upon depolarizing steps from −100 mV to +40 mV (10 mV) were normalized to peak currents at +40 mV and then fitted to the following form of the Boltzmann equation: where V is the pulse voltage, V 1/2 is the half-inactivation voltage, and k is the slope factor. Possible changes in cell size were taken into account by measuring, in each cell, the membrane capacitance, which is directly related to membrane surface area, and by expressing the current amplitude data as current densities (pA/pF). Capacitive currents were elicited by 5 mV depolarizing pulses from −80 mV and acquired at a sampling rate of 50 kHz. The capacitance of the membrane was calculated according to the following equation: C m = τ c · I o /∆E m (1 − I ∞ /I o ), where C m is the membrane capacitance, τ c is the time constant of the membrane capacitance, I o is maximum capacitance current value, ∆E m is the amplitude of the voltage step, and I ∞ is the amplitude of the steady state current.

Immunocytochemistry
Hippocampal neurons were washed twice in cold 0.01 M of PBS (pH 7.4) and fixed at room temperature in 4% of paraformaldehyde for 20 min. Following three washes in PBS, cells were blocked in PBS containing 3% of BSA for 30 min, and then incubated overnight at 4 • C with the rabbit anti-K V 2.1 antibody (1:1000, Alomone Labs, Jerusalem, Israel). The control for the specificity of the antibody against K V 2.1 was performed with its replacement with normal serum as previously described [50]. Then, cells were washed with PBS and incubated with anti-rabbit Cy2-conjugated antibody (1:200; Jackson Immuno Research Laboratories, Inc., West Grove, PA, USA) for 1 h at room temperature under dark conditions. Cells were finally incubated for 5 min with Hoechst 33342. Cover glasses were mounted with a SlowFade Antifade Kit (Molecular Probes, Life Technologies, Milan, Italy) and acquired by a 63× oil immersion objective using a Zeiss inverted 700 confocal microscope.

Immunohistochemistry
Confocal immunofluorescence procedures were performed as described previously [51,52]. In brief, WT and Tg2576 mice were euthanized at 3 months. Anesthesia was induced with 4% sevoflurane in a mixture of 60% N 2 O and 36% O 2 and maintained during intracardiac perfusion with 2% sevoflurane in a mixture of 60% N 2 O and 38% O 2 . Transcardial perfusion was carried out with 4% paraformaldehyde in PBS. The brains were sectioned coronally (60 µm) on a vibratome. After blocking with Rodent M block (Biocare Medical, Concord, MA, USA), sections were incubated with the following primary antibodies: rabbit polyclonal anti-K V 1.2 (1:200) and mouse monoclonal anti-NeuN (1:1000, Sigma-Aldrich, Milan, Italy). Then, sections were incubated with corresponding fluorescence-labeled secondary antibodies (Alexa488-or Alexa594-conjugated anti-mouse or anti-rabbit IgGs). DAPI was used to stain the nuclei. Images were observed using a Zeiss LSM 700 laser (Carl Zeiss, Jena (Turingia), Germany) scanning confocal microscope. Single images were taken with an optical thickness of 0.7 µm and a resolution of 1024 × 1024.

Statistical Analyses
Data are expressed as mean ± standard error of the mean (SEM). Statistical analysis was performed with unpaired t-test or one-way analysis of variance followed by the Newman-Keuls test.

The K V 2.1 Protein Expression Is Increased in the Hippocampus of 3-Month-Old Tg2576 Mice and in Tg2576 Primary Hippocampal Neurons
Starting from the previous report demonstrating that the Aβ 25-35 fragments, which are able to mimic the full-length Aβ 1-42 , induced the overexpression of K V 2.1 in the rat brain [45], we tested the hypothesis that the K V 2.1 protein expression could be modulated in the Tg2576 mouse, which overproduces the Aβ 1-42 peptide. To this aim, we performed Western blotting and immunohistochemical analyses on the hippocampus of 3-month-old Tg2576 mice, which start to display alterations in neuronal excitability and early cognitive deficits by this age. The densitometric analysis of Western blotting revealed that the hippocampus of 3-month-old Tg2576 mice displayed higher levels of K V 2.1 protein expression than the WT littermates ( Figure 1A). Accordingly, confocal double immunofluorescence analyses showed that the CA1 pyramidal neurons of the Tg2576 mice exhibited a more pronounced K V 2.1 immunofluorescence in comparison with the WT littermates ( Figure 1B). In line with previous evidence [22,40,53], the double confocal images also revealed the typical clustered K V 2.1 immunostaining on the PM of the hippocampal CA1 pyramidal neurons of both the WT and Tg2576 mice. Based on the assumption that primary neuronal cultures from Tg2576 mouse embryos, which accumulate Aβ 1-42 with the time in culture, recapitulate some of the main features of the Aβ 1-42 neurotoxicity [54], we also performed Western blotting experiments on hippocampal neurons cultured from Tg2576 and WT mouse embryos. Interestingly, we found that the Tg2576 primary hippocampal neurons also exhibited an increase in the K V 2.1 protein expression in comparison with the WT neurons ( Figure 1C).

Tg2576 Primary Neurons Display Decreased K V 2.1-Dependent I DR Density
To test the hypothesis that the K V 2.1 overexpression could be accompanied by any alterations in the channel activity, we recorded I DR in pyramidal neurons cultured from Tg2576 and WT mouse embryos by means of the whole-cell patch-clamp technique. In order to electrophysiologically isolate the I DR conductance from the total K + currents, we used a previously established protocol [48]. In particular, to record the I DR component, from a holding potential of −80 mV we applied a conditioning pulse at −40 mV lasting 1.5 s to fully inactivate the fast-inactivating current (I A ) component, followed by 250 ms steps from −80 to +40 mV (Figure 2A, bottom). Of note, the Tg2576 pyramidal neurons displayed a significantly lower I DR density than the WT neurons ( Figure 2A-C). Moreover, the analysis of the activation properties of the I DR revealed a −10 mV hyperpolarizing shift of the steady-state activation curve in the Tg2576 pyramidal neurons compared with the WT neurons ( Figure 2D).
To investigate whether the reduction of the I DR in the Tg2576 neurons was actually due to the downregulation of the K V 2.1 currents, we assessed the contribution of the K V 2.1 component to both the Tg2576 and WT neuronal I DR by recording these currents upon the intracellular diffusion of the anti-K V 2.1 monoclonal antibody (10 µg mL −1 ) through the recording pipette. Figure 3A shows a representative time-course of the effect of the anti-K V 2.1 antibody in the WT and Tg2576 neurons. As previously reported [22], the diffusion of the anti-K V 2.1 monoclonal antibody into the cells drastically reduced the I DR in the WT neurons ( Figure 3A,B), hence indicating that the K V 2.1 currents were the major components of the I DR in these neurons. By contrast, the anti-K V 2.1 antibody produced a markedly lower effect in reducing the I DR at 6 min in the Tg2576 neurons than in the WT neurons ( Figure 3A-C), hence suggesting that the K V 2.1 current component in the Tg2576 neurons was lower than in the WT neurons. The subtraction of the antibody-resistant currents at 6 min from the initial current illustrated in Figure 3C clearly represents this result. The representative traces of the I DR and relative current-voltage relationship recorded by 6 min, without (top) and with (bottom) the anti-K V 2.1 monoclonal antibody in the recording pipette, are illustrated in Figure 3D and 3E.
In agreement with the major contribution of the K V 2.1 component to the I DR in WT neurons, we found that the steady-state parameters of the I DR activation recorded in the WT neurons upon the intracellular diffusion of the anti-K V 2.1 antibody were significantly different from those recorded in the WT neurons in the absence of the antibody ( Figure 3F). This data likely suggested that upon the blockade of the K V 2.1 currents by the anti-K V 2.1 antibody, the contribution of non-K V 2.1 components characterized by different activation properties were unmasked. Interestingly, the left-shift of the voltage dependence of the steady-state activation curve of the I DR observed in the WT neurons upon the K V 2.1 blockade was similar to that observed in the Tg2576 neurons ( Figure 3F), thereby suggesting that a loss of the K V 2.1 currents might underlie the hyperpolarizing shift of the steady-state activation curve observed in the Tg2576 neurons.

Glutamate Recovers the Reduction of the K V 2.1-Mediated I DR Caused by Increased Channel Clustering
As mentioned above, the majority of the K V 2.1 channels expressed in pyramidal neurons reside within high-density cell surface clusters. Clustered channels are held in a non-conducting state, while non-clustered channels, which are spread diffusely over the cell surface, mediate the majority of the whole I DR . Based on this aspect, we hypothesized that the reduction of the K V 2.1-mediated I DR in the Tg2576 neurons could be due to an increase in channel clustering. Further prompted by the evidence that the overexpression of K V 2.1 results in increased channel clustering [55], we performed confocal experiments to assess possible changes in channel clustering in the Tg2576 neurons. To this aim, both the Tg2576 and WT hippocampal neurons were fixed and stained with the anti-K V 2.1 rabbit polyclonal antibody (1:1000, Alomone Labs, Jerusalem, Israel) and subjected to imaging on a confocal microscope. Interestingly, if compared with the WT neurons, the Tg2576 neurons displayed an evident increase in channel clustering. In particular, while microclusters of K V 2.1 channels were present in both the WT and Tg2576 neurons, markedly larger clusters of K V 2.1, usually defined as "macroclusters", were clearly detectable only in the Tg2576 neurons ( Figure 4A).
To functionally investigate whether the increase of channel clustering was actually responsible for the reduction of the K V 2.1 currents previously observed in the Tg2576 neurons, we tested the effect of a declustering stimulus on the K V 2.1 currents in both the WT and Tg2576 neuronal cultures. In particular, based on the evidence provided by the Trimmer group that glutamate is able to induce a calcineurin-dependent K V 2.1 dephosphorylation and the subsequent dispersal of channel clusters [40], we exposed both the WT and Tg2576 neurons to glutamate (10 µM, 10 min). In line with the previous results provided by Mohapatra and coworkers [33], we found that the exposure to 10 µM glutamate for 10 min significantly increased the I DR in the WT neurons ( Figure 4B-E). Notably, glutamate exerted a stimulatory effect on the I DR also in the Tg2576 neurons, restoring the I DR density at 10 min to levels similar to those recorded in the WT neurons ( Figure 4B-E). Consistent with the higher levels of K V 2.1 channels expressed in the Tg2576 neurons and become available for conduction upon the glutamate-induced declustering, the stimulatory effect of glutamate on the I DR seemed to be greater in the Tg2576 neurons than in the WT neurons ( Figure 4B,C). Interestingly, we also found that the steady-state parameters of the I DR activation in the Tg2576 neurons upon glutamate exposure were significantly different if compared with those in the Tg2576 neurons under control conditions ( Figure 4F). Of note, the shift in the voltage-dependence of activation induced by glutamate restored the steady-state activation curve to values comparable to those of the WT neurons ( Figure 4F). To exclude the possibility that the stimulatory effect of glutamate on the I DR could be unrelated to K V 2.1 declustering, we exposed both the WT and Tg2576 neurons to glutamate (10 µM, 10 min) in the presence of the anti-K V 2.1 antibody in the recording pipette. Remarkably, when K V 2.1 was blocked by the intracellular diffusion of the anti-K V 2.1 antibody, the glutamate-induced upregulation of the I DR was not observed in the WT or in the Tg2576 neurons ( Figure 5A,B). This result clearly shows that the upregulation of the I DR induced by glutamate was mediated by the specific upregulation of the K V 2.1 currents resulting from the glutamate-induced channel declustering.

K V 2.1 Clustering Is Increased in Cortical Layer V Pyramidal Neurons of the Somatosensory Cortex of 3-Month-Old Tg2576 Mice
To investigate whether K V 2.1 clustering was increased in the Tg2576 model also in vivo, we performed confocal double immunofluorescence analyses on the somatosensory cortex of both 3-month-old WT and Tg2576 mice, where the K V 2.1 protein expression and function have been extensively characterized [23,56]. Interestingly, the confocal double immunofluorescence analysis revealed that the anti-K V 2.1 antibody depicted a clear and pronounced clustered PM labeling of NeuN-positive neurons in layer V of the somatosensory cortex of both 3-month-old WT and Tg2576 mice ( Figure 6A). More important, quantitative analyses showed a significantly higher number of cortical neurons displaying K V 2.1 clustering in the layer V of the Tg2576 mice in comparison with the WT littermates ( Figure 6B).

Discussion
In the present study, we demonstrated that K V 2.1 protein expression was significantly increased in primary neurons from the Tg2576 mouse, which, carrying the human Swedish double mutation of the APP, accumulates the Aβ 1-42 peptide overtime in culture [54,57]. However, as a consequence of the increased channel clustering, the K V 2.1-mediated currents were markedly reduced in the Tg2576 neurons, resulting in a decrease in the whole I DR along with clear changes in their voltage-dependence of activation. Moreover, we showed that the K V 2.1 expression and clustering levels were modulated also in vivo in Tg2576 mice.
First, Western blotting analyses on the hippocampal lysates from 3-month-old Tg2576 mice revealed a significant increase in the K V 2.1 protein levels in comparison with the WT hippocampal lysates, hence providing the first evidence, to our knowledge, that the K V 2.1 protein expression is modulated in a transgenic in vivo model of AD. In agreement, immunohistochemical analyses on WT and Tg2576 brain sections showed a more pronounced K V 2.1 immunostaining in the CA1 pyramidal neurons of the Tg2576 mice in comparison with the WT littermates. Of note, Tg2576 mice, a widely used transgenic model of the AD pathology, start by 3 months of age to intracellularly accumulate Aβ 1-42 trimers, which have been implicated in the appearance of precocious cognitive deficits and high seizure susceptibility [58][59][60]. In line with the previous evidence demonstrating that the neuronal cultures from Tg2576 mouse embryos recapitulate some of the main features of the Aβ 1-42 neurotoxicity [54], we found that the K V 2.1 protein expression was also increased in the hippocampal neurons cultured from Tg2576 embryos.
Based on the evidence that K V 2.1 is the main channel α subunit mediating I DR in hippocampal neurons [16], we investigated whether its overexpression could affect the whole I DR density in the Tg2576 primary hippocampal neurons. Consistent with the aforementioned literature, our electrophysiological experiments revealed that K V 2.1 was the main channel α subunit mediating I DR in the WT pyramidal neurons, since the intracellular diffusion of the anti-K V 2.1 monoclonal antibody induced a significant reduction of the I DR in these neurons. Interestingly, we noticed that the I DR density was significantly reduced in the Tg2576 neurons compared with the WT neurons. More important, the intracellular diffusion of the anti-K V 2.1 antibody produced a markedly lower effect in promoting the decrease of the I DR density in the Tg2576 neurons than in the WT neurons, hence suggesting that the percentage of the conducting K V 2.1 channels in the Tg2576 neurons was reduced or functionally compromised. In addition, the analysis of the voltage-dependence of activation revealed significant differences in the steady-state parameters of the I DR activation in the Tg2576 neurons compared with the WT neurons. Interestingly, the shift of the voltagedependence of activation to more negative values observed in the Tg2576 neurons was similar to that produced by the intracellular diffusion of the anti-K V 2.1 antibody in the WT neurons. This observation suggested that the left-shift in the voltage-dependence of activation occurring in the Tg2576 neurons could reflect a lower contribution of the K V 2.1 current component to the whole I DR .
Although apparently divergent from the higher expression levels observed in the Tg2576 neurons, our functional results are in line with previous evidence clearly indicating that the K V 2.1 channel conductance decreases as the expression density on the cell surface membrane increases. Indeed, the Tamkun group provided solid data demonstrating that the percentage of non-conducting K V 2.1 channels strictly depends on their expression levels and consequent channel density, whereas it appeared to be independent on cluster localization per se [61]. On the other hand, it has been extensively reported that the vast majority of the K V 2.1 channels residing within clusters are held in a non-conducting state [20,41,42], in agreement with the fact that the high concentration of channels within a cluster results in a 5-fold to 10-fold greater density in this restricted space than outside [61]. Prompted by the evidence that the overexpression of K V 2.1 promotes its clusterization, probably by inducing the fusion of smaller clusters to form macroclusters [55], we also hypothesized that the lower K V 2.1-mediated I DR density observed in the Tg2576 neurons could be due to an increase in channel clustering. In confirmation of this hypothesis, immunocytochemical analyses showed a marked change in the K V 2.1 clustering in the Tg2576 pyramidal neurons. In particular, the K V 2.1 immunostaining revealed that, along with the typical clustered expression detected in both the WT and Tg2576 neurons, a massive presence of macroclusters was detectable only in the Tg2576 neurons.
While the causal relationship between the K V 2.1 overexpression and its clustering has not been fully defined, it was instead clearly demonstrated that K V 2.1 channel clustering strictly depends on its phosphorylation state. Indeed, it was reported that the phosphorylation of specific residues of K V 2.1, identified in mutagenesis studies, drastically affects both the clustered localization and the voltage-dependent gating of the channel [32,34,35,40,62]. Intriguingly, the phosphorylation of the proximal restriction and clustering domain is propaedeutic to the binding of K V 2.1 to the endoplasmic reticulum (ER)-resident VAPA and VAPB (vesicle-associated membrane and protein-associated protein A and B) and the subsequent formation of specialized ER-PM junctions. The function of the K V 2.1induced ER-PM junctions appeared to be unrelated to K + conductance, hence providing an explanation for the non-conducting state of the K V 2.1 channels residing within clusters [20,42,[63][64][65]. Physiologically, K V 2.1 is highly phosphorylated in resting neurons and undergoes a neuronal activity-induced dephosphorylation through a Ca 2+ /calcineurindependent mechanism [32,40]. Consistently, stimuli that increase neuronal activity and/or cytoplasmic Ca 2+ levels induce the dispersal of K V 2.1 clusters and enhance the voltagesensitivity [32,35,40].
Therefore, we tested the possibility that a declustering stimulus could restore the K V 2.1 currents in the Tg2576 neurons. In particular, based on the results provided by the Trimmer group demonstrating that glutamate stimulation rapidly induces a calcineurindependent dephosphorylation of K V 2.1 and its subsequent translocation from clusters to a more diffuse localization [40,63], we measured the I DR in the Tg2576 neurons upon glutamate application. Interestingly, we found that a brief application of 10 µM glutamate was actually able to increase the I DR in the Tg2576 neurons. In line with the results provided by the Trimmer group [33,35,66], the same treatment induced an increase in I DR also in the WT neurons, where the majority of K V 2.1 channels are physiologically clustered. However, the increase of the I DR induced by glutamate exposure in the Tg2576 neurons was greater than that observed in the WT neurons, arguably due to the larger amount of clustered K V 2.1 channels present in the Tg2576 neurons and become available for conduction upon glutamate-induced declustering. In this regard, it is worth noting that, unlike the results provided by Misonou and co-workers, a study by the Tamkun group showed that the K V 2.1 declustering induced by swinolide A or alkaline phosphatase resulted in no change in current density or voltage-dependence of the channel [41]. Nonetheless, the same authors stated that it cannot be ruled out that distinct declustering stimuli may have different results or that the cellular model they used in their study, namely HEK cells, does not completely mimic a neuronal model.
To exclude any possible non-specific effects of glutamate on I DR , we recorded the I DR in both the WT and Tg2576 neurons upon glutamate exposure in the presence of the anti-K V 2.1 antibody in the recording pipette. Importantly, when blocking K V 2.1 with the anti-K V 2.1 antibody, the glutamate-induced upregulation of the I DR was not observed in the WT or in the Tg2576 neurons, hence indicating that the positive effect of glutamate on the I DR density was related to K V 2.1 declustering. Interestingly, glutamate application also produced a significant depolarizing shift in the voltage-dependence of activation in the Tg2576 neurons. Of note, this depolarizing shift restored the conductance-voltage relationship to values near to those of the WT neurons. In this respect, it should be noted that molecular studies by the Trimmer group clearly demonstrated that the calcineurindependent dephosphorylation of K V 2.1 channels yields hyperpolarizing shifts in their voltage-dependent activation, which seems to be not in line with our results [34,35,40]. However, it should be considered that, while the treatment with glutamate made the K V 2.1 current component newly available, thus significantly increasing its contribution to the whole I DR , the I DR registered in the Tg2576 neurons in control conditions had to arguably be attributed to non-K V 2.1 components with different activation properties. In agreement, we observed that the blockade of K V 2.1 in the WT neurons produced a left shift in the steady-state I DR activation curve, likely suggesting that the loss of the K V 2.1 current component brought out a major contribution of non-K V 2.1 components to the voltage-dependence of activation. Thereby, it is conceivable that the functional effect of K V 2.1 modulation by glutamate on the I DR activation properties could reflect the restoring of the K V 2.1 current contribution to the whole I DR in the Tg2576 neurons, with no change in the kinetic properties of the channel. In this respect, it is worth noting that this is the first report, to our knowledge, on the effect of glutamate on the K V 2.1-mediated I DR in a cellular model where the K V 2.1 channels are both overexpressed and more clustered.
Of note, we have previously demonstrated that primary hippocampal neurons from the Tg2576 mouse display increased excitability if compared with WT neurons [13]. In the present study, we found that the overexpression of K V 2.1 and its increased clusterization occurred in Tg2576 pyramidal neurons both in vitro and in vivo. Our in vitro experiments also revealed that the increased K V 2.1 channel clustering caused a significant reduction of the K V 2.1-mediated currents and considerable changes in the voltage-dependence of activation of the I DR in Tg2576 neurons. Therefore, it is likely that similar functional changes may occur in the Tg2576 pyramidal neurons in vivo. More importantly, a downregulation of the K V 2.1-mediated currents might be expected to have a crucial impact on hippocampal and cortical excitability in the Tg2576 mouse. In this regard, a previous study by Frazzini and colleagues [43] revealed that the reduction in the K V 2.1 current density due to channel oxidation was associated with the spontaneous hyperexcitability observed in the neurons from the 3xTg-AD mouse, a transgenic model characterized by elevated levels of oxidative stress in the early phases of the pathology [67]. In particular, they observed increased [Ca 2+ ] i spike frequency in the 3xTg-AD neurons, where the K V 2.1-mediated currents were downregulated, and obtained similar results in the non-transgenic neurons upon the pharmacological blockade of K V 2.1 [43].
Tg2576 mice have been reported to display an ectopic expression of the Neuropeptide Y, a marker of chronic seizures, and an aberrant network activity even before the onset of memory deficits [60,[68][69][70], hence emerging as an animal model particularly suitable to study neuronal hyperactivity in association with Aβ 1-42 accumulation. However, I DR had not been yet investigated in this transgenic model. Nonetheless, in recent studies, the D'Amelio group reported that pyramidal neurons of both dorsal and ventral CA1 of the Tg2576 mouse displayed altered firing and synaptic transmission in the pre-plaque stages of AD [59,71,72]. Intriguingly, based on the results of computational models, they suggested that the persistent perturbation of the I DR conductance and kinetic properties could be a critical determinant of the firing alterations observed in the Tg2576 ventral CA1 neurons [72].
Finally, considered the non-conducting and structural functions of K V 2.1, further implications unrelated to neuronal excitability could arise from its overexpression in AD. Indeed, immediately facing astrocytic processes and lying on neuronal cell membranes over the ER subsurface cisternae [73], the K V 2.1 clusters play a key role in inducing ER-PM junctions and provide specialized molecular platforms to regulate both somatodendritic Ca 2+ signaling and ion channel trafficking. In particular, it has recently been shown that K V 2.1 clustering at neuronal ER-PM junctions promotes the clustering of PM L-type Ca 2+ channels and their coupling to ryanodine receptor ER Ca 2+ release channels [74], hence allowing the generation of localized Ca 2+ release events. Moreover, the fact that K V 2.1 clusters induce the formation of -and localize at-ER-PM junctions confers to K V 2.1 the ability not only to trigger a structural remodeling of the ER but also to affect neuronal Ca 2+ handling by modulating ER Ca 2+ uptake [75]. Of note, we have recently reported about the occurrence of an ER Ca 2+ remodeling in the Tg2576 primary neurons. In particular, we showed that the enhanced activity of the Na + /Ca 2+ exchanger 3 promotes an increased ER Ca 2+ refilling in Tg2576 neurons in comparison with WT neurons [47]. Moreover, we demonstrated that the I A -mediating K V 3.4 channel subunit, which is overexpressed and functionally upregulated in Tg2576 reactive astrocytes, regulates astrocytic [Ca 2+ ] i transients and ER Ca 2+ levels [52,76]. The findings reported in the present study bring further information about the ionic dysregulation occurring in the Tg2576 mouse brain, hence suggesting new pathways to explore in studying the alterations of Ca 2+ signaling and homeostasis in AD and the involvement of potassium channels in these processes.

Conclusions
Undoubtedly, the involvement of I DR, and in particular of the K V 2.1 currents, as well as of many other ionic mechanisms, in the AD-related alterations of neuronal excitability deserves much attention. In fact, how AD brains develop hyperexcitability is still largely unknown. Changes in the excitatory/inhibitory balance [77][78][79] and the increase in intrinsic membrane excitability properties [80] have been suggested as possible explanations, with the Aβ peptide proposed as the main determining factor. In this context, the Aβ-induced alteration of protein expression and function of different K + channels could represent a crucial mechanism promoting hyperexcitability and altered neuronal activity. Considering the pivotal role of the K V 2.1-mediated I DR in regulating the intrinsic excitability of pyramidal neurons, further studies should be performed to address this issue in AD models. In addition, since the alteration of cytosolic and organellar Ca 2+ homeostasis is considered a critical event in the AD pathogenesis [81], the impact of the K V 2.1 modulation on these processes should be further investigated in AD models.

Conflicts of Interest:
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.