The Importance of Immunohistochemical Analyses in Evaluating the Phenotype of Kv Channel Knockout Mice

Summary

To gain insights into the phenotype of Kv1.1 and Kv4.2 knockout mice, we used immunohistochemistry to analyze expression of component principal or α subunits and auxiliary subunits of neuronal Kv channels in knockout mouse brains. Genetic ablation of the Kv1.1 α subunit did not result in compensatory changes in the expression levels or subcellular distribution of related ion channel subunits in hippocampal medial perforant path and mossy fiber nerve terminals, where high levels of Kv1.1 are normally expressed. Genetic ablation of the Kv4.2 α subunit did not result in altered neuronal cytoarchitecture of the hippocampus. While Kv4.2 knockout mice did not exhibit compensatory changes in the expression levels or subcellular distribution of the related Kv4.3 α subunit, we found dramatic decreases in the cellular and subcellular expression of specific KChIPs that reflected their degree of association and colocalization with Kv4.2 in wild-type mouse and rat brains. These studies highlight the insights that can be gained by performing detailed immunohistochemical analyses of Kv channel knockout mouse brains.

Advances in mouse genetic techniques over the past two decades have led to an increasing availability of constitutive and inducible knockout (KO) and transgenic mouse lines. These lines are crucial to our understanding of the role of individual molecules in normal physiology, and many serve as important models of human disease. This has been especially relevant to the study of epilepsy, with a number of mouse models that have proven useful not only in studies of underlying mechanisms, but also for evaluation of novel therapeutic approaches (Frankel 2009). Studies on how the altered genotype of these mice affects phenotype typically employ a battery of behavioral tests to determine the impact on the whole animal. An array of electrophysiological, metabolic, neurochemical and biochemical assays typically follow. Investigators often perform histochemical analyses to determine whether brain structure is altered. Here we focus on the use of immunohistochemistry-based neuroanatomical approaches to gain additional insights into the impact of the genetic manipulation of a particular epilepsy-associated target gene on the expression and localization of other brain proteins as a powerful approach to gain insights into the molecular and cellular bases of the observed phenotype.

Here, we focus on mice with genetic manipulations in voltage-gated potassium or Kv channels. Kv channels are members of a large gene family whose products co-assemble to form multisubunit complexes in mammalian neurons. These complexes contain four principal, voltage-sensing and pore-forming α subunits, together with cytoplasmic and/or transmembrane auxiliary subunits (Vacher, et al. 2008). As such, the straightforward evaluation of the impact of a genetic manipulation in one component subunit of a Kv channel complex may be confounded by subsequent effects on interacting subunits. Upregulation of similar subunits to compensate for the loss of a single component, altered stoichiometry of the remaining channel complexes, or elimination of associated subunits due to the aberrant nature of the resultant complexes are all possible confounding outcomes of single gene Kv channel subunit KOs. Careful evaluation of the levels of expression of related and/or associated subunits is therefore a necessary step in evaluating the phenotype of these mice. Immunoblot analyses can be a sensitive and fairly quantitative approach to measure steady-state expression level of individual proteins in samples prepared from the entire brain, from specific brain regions, or from neuronal cultures. However, many Kv channels are localized to specific neuronal populations, and within these neurons to precise subcellular domains, which impacts or even defines their contribution to neuronal function (Vacher, et al. 2008). As such immunohistochemistry can provide important insights into the cellular and subcellular localization of the specific proteins that are not obtained by bulk biochemical procedures such as immunoblots. Here we present work from our laboratory to determine the impact of genetic manipulation of Kv channel subunit expression in mouse brain on the expression and localization of related and/or associated Kv channel subunits.

Methods

Antibodies

Antibodies used for immunohistochemistry were: mouse monoclonal antibodies (mAbs): anti-Kv1.1 K36/15 (IgG2b), anti-Kv1.4 K13/31 (IgG1), anti-Kv2.1 K89/34 (IgG1), anti-Slo1 L6/60 (IgG2a), anti-Kv4.2 K57/1 (IgG1), anti-Kv4.3 K75/41 (IgG1), anti-KChIP2 K60/73 (IgG1), and anti-KChIP3 K66/38 (IgG2a). These mAbs are available from the UC Davis/NIH NeuroMab Facility (neuromab.ucdavis.edu). Rabbit polyclonal anti-Kv2.1 KC antibody (Trimmer 1991) and mouse mAb anti-calbindin CB-955 (Sigma, St. Louis, MO) were also used.

Immunohistochemistry

Mice were deeply anesthetized with sodium pentobarbital (60 mg/kg, i.p.) and perfused through the ascending aorta with 20 ml of 0.9% saline followed by 80 ml of fixative containing 4% formaldehyde, prepared from freshly depolymerized paraformaldehyde, in 0.1 M sodium phosphate buffer (PB), pH 7.4. The brains were removed, cryoprotected for 18–48 hr in 20% sucrose, frozen in a bed of pulverized dry ice, and then cut into 40 µm sagittal sections on a sliding microtome (Michrom HM 450, Richard-Allan Scientific, Kalamazoo, MI) equipped with a freezing stage (Physitemp BFS-30TC and PTU-3, Clifton, NJ). Sections were collected in 0.1M PB and processed immediately for immunohistochemistry as described below.

For immunohistochemistry, sections were incubated for 1 hr in 10 mM PB (pH 7.4) containing 0.9% NaCl (PBS) and 0.5% Triton X-100. The sections were then washed in PBS and incubated for 1 hr in antibody vehicle (10% goat serum 0.3% Triton X-100 in PBS). After additional washes in PBS, sections were incubated overnight at 4°C in vehicle containing mouse mAbs and affinity-purified rabbit polyclonal antibodies. Following overnight incubation in the primary antibodies, sections were washed 3 times for 10 min each in PBS, incubated for 1 hr in vehicle containing affinity-purified species and/or IgG-subclass-specific goat secondary antibodies conjugated to Alexa fluors (Life Technologies, Carlsbad, CA). Sections were then washed again 3 times for 10 min each in PBS, dried, and cover-slipped using ProLong Gold mounting medium (Life Technologies).

Fluorescence images were taken using the same exposure time for each set of samples to directly compare the signal intensity. Images were acquired with a CCD camera installed on an Axiovert 200M microscope with 63X, 1.3 numerical aperture (NA) lens or 20X, 0.8 NA lens, and an ApoTome coupled to Axiovision software (Zeiss, Oberkochen, Germany). Images were exported from Axiovision as TIFF or JPG files, and imported into Adobe Photoshop.

Thionin Staining of Nissl Substance

Sagittal mouse brain sections prepared as described above were defatted in 1:1 chloroform:ethanol solution for 3 h at RT, and rehydrated through an ethanol:water series. Sections were stained for 3 min at RT in a solution of 0.05% thionin in 0.5 M sodium acetate buffer pH 4.2, rinsed with water, and destained in 70, 95 and 100% ethanol. Sections were cleared in xylene and mounted in Permount.

Results

Insights from immunohistochemical studies of Kv1.1 KO mice

Constitutive Kv1.1 KO mice display a prominent phenotype, in exhibiting spontaneous seizures, increased excitability in CA3 pyramidal neurons, and epileptiform activity in slices prepared from their hippocampi (Smart, et al. 1998). In their initial analyses of the structural features of the brains of Kv1.1 KO mice, Dr. Jürgen Wenzel and his colleagues used hematoxylin-eosin staining to demonstrate that these mice had relatively normal cytoarchitecture in cerebral cortex, cerebellar cortex and hippocampus (Smart, et al. 1998). They also showed that cerebellar basket cell terminals, which form large pinceau structures enwrapping the initial segment of Purkinje cell axons and which typically express high levels of Kv1.1 and Kv1.2 (Wang, et al. 1993), had normal morphology, and had levels of Kv1.2 expression and localization similar to that seen in wild-type (WT) mice (Smart, et al. 1998). These results were important in interpreting data from subsequent studies showing that these nerve terminals have enhanced GABA release (Zhang, et al. 1999), leading to the ataxic phenotype of the Kv1.1 KO mice. The hyperexcitability of these nerve terminals is presumably due to an inability of the Kv1.2 homotetrameric channels present in the pinceau of the Kv1.1 KO mice to support the function typically carried out by heteromeric Kv1.1:Kv1.2 channels found at these sites in WT mice, either due to differences in the biophysical properties of the homo-versus hetero-tetrameric channels, or overall reduced Kv1 channel expression, given the decreased number of component subunits available to assemble these channels.

A more detailed analysis of the hippocampi of the Kv1.1 KO mice was undertaken by Wenzel and his colleagues (including some of us), using a battery of histochemical and immunohistochemical analyses (Wenzel, et al. 2007b). These immunohistochemical analyses were focused on determining whether the genetic ablation of Kv1.1 expression impacted expression of other Kv channel subunits known to associate and colocalize with Kv1.1 at specific sites in WT hippocampus (Vacher, et al. 2008). Within the hippocampus, Kv1 channels in medial perforant path nerve terminals are formed as heterotetrameric assemblies of Kv1.1, Kv1.2 and Kv1.4 α subunits, while those in hippocampal mossy fiber axons and nerve terminals are formed as heterotetramers of Kv1.1 and Kv1.4 (Sheng, et al. 1993, Wang, et al. 1993, Wang, et al. 1994, Rhodes, et al. 1997, Monaghan, et al. 2001). Kv1.4 is unique among Kv1 channel α subunits in having an inactivation particle (Stuhmer, et al. 1989), a specialized structural domain on its cytoplasmic amino terminus that can confer rapid inactivation on Kv1.4-containing Kv1 channels. The stoichiometry of Kv1.4 subunits in a heteromeric channel complexes is crucial to channel function, as the kinetics of inactivation become more rapid with each additional Kv1.4 subunit incorporated into the channel tetramer (Ruppersberg, et al. 1990).

In our investigations of the impact of eliminating Kv1.1 expression on the expression of other Kv channel subunits, we performed immunoblots on whole brain preparations and found no differences in the overall expression level of Kv1.2 and Kv1.4 in WT and Kv1.1 KO brains (Wenzel, et al. 2007b). We used immunofluorescence staining with mouse mAb specific for Kv1.1 (K36/15) to first show that the localization of Kv1.1 in the hippocampus of our WT mice was similar to that which had been published previously for both rat (Sheng, et al. 1993, Veh, et al. 1995, Rhodes, et al. 1997) and mouse (Wang, et al. 1993, Smart, et al. 1998, Wenzel, et al. 2007b), and that the Kv1.1 staining was completely eliminated in the sections prepared from Kv1.1 KO mice (Figure 1). As also shown in Figure 1, the expression and localization of Kv1.2 (determined by staining with mouse mAb K14/16) and Kv1.4 (determined by staining with mouse mAb K13/31) in the middle molecular of the dentate gyrus, the location of medial perforant path nerve terminals and where these subunits normally colocalize with Kv1.1, was not noticeably different between WT and KO mouse hippocampi. Similarly, the expression of Kv1.4 in mossy fiber terminals in stratum (s.) lucidum of CA3, where it normally colocalizes with Kv1.1, was also not affected by the elimination of Kv1.1 expression (Figure 1). These studies suggest that elimination of Kv1.1 must impact Kv1 channels at these sites by either yielding a lower density of channels, due to a smaller overall pool of subunits with which to assemble tetrameric channels, as well as altered inactivation gating due to a higher relative representation of Kv1.4 in the remaining channels, as the remaining channels presumably now comprise Kv1.2 and Kv1.4 (in medial perforant path nerve terminals) or Kv1.4 alone (in mossy fiber terminals). This change in channel number and the open probability would affect the overall amplitude of Kv1-based currents that regulate glutamate release from these nerve terminals, due to fewer overall channels, and a enhanced propensity of those that remain to enter an inactivated state shortly after activation. The reduction in presynaptic Kv channels at these sites could contribute to the epileptic phenotype exhibited by the Kv1.1 KO mice.

Figure 1.

Staining for potassium channel subunits and calbindin in WT and Kv1.1 KO mouse hippocampus. Sagittal brain sections prepared from age- and sex-matched WT and Kv1.1 KO mice were used in immunofluorescence staining with mAbs specific for the target proteins as indicated. *: middle molecular layer of the dentate gyrus. **: mossy fiber pathway. Scale bar = 200 µm.

We also investigated the expression of two other potassium channels expressed at high levels in hippocampal neurons, the Kv2.1 delayed rectifier Kv channel (determined by staining with the mouse mAb K89/34), and the Slo1 large conductance calcium- and voltage-activated or BK potassium channel (determined by staining with the mouse mAb L6/60). As shown in Figure 1, expression of the somatodendritic Kv2.1 channel is not altered in the Kv1.1 KO mouse hippocampus. The Slo1 BK channel, which is expressed in the same nerve terminals in the medial perforant path and mossy fiber pathways (Misonou, et al. 2006a) that normally express high levels of Kv1.1, is also not visibly altered by genetic elimination of Kv1.1 (Figure 1). We also used immunohistochemistry for the calcium binding protein calbindin (determined by staining with the mouse mAb CB-955), which is expressed at high levels in dentate granule cells and their processes, to show that the overall morphology of these cells is not dramatically altered by Kv1.1 ablation (Figure 1). Together, these immunohistochemical studies reveal a lack of any obvious upregulation of other potassium channel subunits in the hippocampi of mice lacking the prominent Kv1.1 subunit.

Immunohistochemical analyses of Kv4.2 KO mice reveal a similar lack of compensatory upregulation of related and associated α subunits

Kv4.2 is a voltage-gated potassium channel subunit that is a prominent component of low threshold, rapidly inactivating “A-type” channels in mammalian neurons. Kv4.2-containing channels are highly expressed on dendrites of most principal brain neurons (Vacher, et al. 2008), where they play a crucial yet dynamic role in dendritic integration through their regulation of dendritic excitability and backpropagating action potentials (Jerng, et al. 2004). Altered Kv4.2 expression and/or function has also been implicated in contributing to epileptogenesis (Bernard, et al. 2004, Singh, et al. 2006, Monaghan, et al. 2008). As such, it was somewhat surprising that in most aspects the constitutive Kv4.2 KO mice were grossly normal (Jung 2002). The relatively subtle phenotype of the Kv4.2 KO mice suggested compensatory mechanisms were in place that masked the full effect of eliminating Kv4.2 expression. Arguably the simplest mechanism for compensation would be the upregulation of other dendritic Kv channel subunits. Following from the work of Wenzel and colleagues in Kv1.1 KO mice (Smart, et al. 1998, Wenzel, et al. 2007b) we undertook an immunohistochemical analysis of the expression and location of such candidate Kv channel subunits, employing monoclonal and polyclonal antibodies specific for individual channel subunits in multiple color immunofluorescence labeling studies in brain sections prepared from WT and Kv4.2 KO mice (Menegola & Trimmer 2006). Before undertaking this immunohistochemical analysis, we first investigated the gross anatomical characteristics of these brains by thionin staining for Nissl substance. As shown in Figure 2, the gross anatomical characteristics of the hippocampus of WT (top) and Kv4.2 KO (bottom) mice are indistinguishable.

Figure 2.

Nissl staining in WT and Kv4.2 KO mouse hippocampus. Sagittal brain sections prepared from age- and sex-matched WT (top panel) and Kv4.2 KO (bottom panel) mice were stained with thionin to reveal the Nissl substance. Scale bar = 200 µm.

We also verified the lack of Kv4.2 expression in Kv4.2 KO mice. Double label immunofluorescence staining was performed on brain sections from WT and Kv4.2 KO mice using the Kv4.2-specific mAb K57/1 (Rhodes, et al. 2004) in red, and the rabbit polyclonal antibody KC (Trimmer 1991) against the Kv2.1 potassium channel in green. As shown in Figure 3, WT mice show intense Kv4.2 immunofluorescence staining in the molecular layer of the dentate gyrus, corresponding to dendrites of dentate granule cells, in s. oriens and s. radiatum but not s. lucidum of CA3, corresponding to the apical and basal dendrites of CA3 pyramidal cells, and in s. radiatum and s. oriens of CA1, corresponding to apical and basal dendrites of CA1 pyramidal cells. Much lower levels of Kv4.2 staining are observed in s. lacunosum moleculare of CA1, one example of the non-uniform expression of Kv4.2 within neuronal dendrites (Figure 3). The pattern of Kv4.2 staining in WT mouse hippocampus is very similar to that observed for Kv4.2 in rat hippocampus (Sheng, et al. 1992, Rhodes, et al. 2004). All of this staining is eliminated in brain sections from Kv4.2 KO mice (Figure 3), incubated in the same primary and secondary antibodies and imaged under identical conditions.

Figure 3.

Staining for potassium channel α subunits in WT and Kv4.2 KO mouse hippocampus. Sagittal brain sections prepared from age- and sex-matched WT and Kv4.2 KO mice were used in double label immunofluorescence staining with mAbs specific for Kv4.2 (left panels) or Kv4.3 (right panels) in red, and polyclonal antibodies against Kv2.1 in green, proteins as indicated. *: molecular layer of the dentate gyrus. **: s. radiatum of CA3. ***: s. radiatum of CA1. Scale bar = 200 µm.

Kv4.3 is the other member of the dendritic A-type potassium channel family expressed at high levels in mammalian brain (Serodio & Rudy 1998). While the biophysical properties of Kv4.2 and Kv4.3 have some distinctions (Guo, et al. 1999, Jerng, et al. 2004), and heteromeric assembly of Kv4.2 and Kv4.3 yields channels with characteristics intermediate between those formed as homotetramers of either Kv4.2 or Kv4.3 (Guo, et al. 1999), overall they are quite similar. Both Kv4.2 and Kv4.3 are localized to dendrites, and to a lesser extent, somata, in expressing neurons (Rhodes, et al. 2004). Moreover, both Kv4.2 and Kv4.3 interact with the same set of auxiliary subunits, including KChIPs and DPPs (An, et al. 2000, Nadal, et al. 2003, Jerng, et al. 2004, Rhodes, et al. 2004, Maffie & Rudy 2008, Vacher & Trimmer 2011). Intriguingly, while in each of these cases their characteristics do not appear to be so different, some neurons express Kv4.2 alone, some Kv4.3 alone, and others both subunits (Rhodes, et al. 2004), suggesting the requirement for homotetramers of each channel type, and heterotetramers of Kv4.2 and Kv4.3, in specific neuronal populations. However, given the overall similarities between Kv4.2 and Kv4.3, it remains that one might expect that a consequence of the elimination of Kv4.2 would be a compensatory upregulation of Kv4.3.

Double label immunofluorescence staining was performed on brain sections from WT and Kv4.2 KO mice using the Kv4.3-specific mAb K75/41 (Rhodes, et al. 2004) in red, and the rabbit polyclonal antibody KC against the Kv2.1 potassium channel (Trimmer 1991) in green (Figure 3). Note that the specificity of the anti-Kv4.3 mAb was also validated independently in experiments showing elimination of immunohistochemical staining in the brains of Kv4.3 KO mice (Burkhalter, et al. 2006). Figure 3 shows images of the hippocampal formation from WT and KO mice stained with this mouse mAb. Strong Kv4.3 signal is observed in the molecular layer of the dentate gyrus in sections from both WT and Kv4.2 KO mice, corresponding to high levels of expression of Kv4.3 in the dendrites of dentate granule cells. Kv4.3 is also observed in s. radiatum of CA3 and s. oriens of CA3 in both WT and Kv4.2 KO sections, corresponding to dendrites of CA3 pyramidal neurons. These patterns of Kv4.3 expression in WT mouse hippocampus are similar to that observed in rat (Rhodes, et al. 2004). Remarkably, in both WT and Kv4.2 KO sections, Kv4.3 staining is largely absent from s. radiatum, s. lacunosum moleculare and s. oriens of CA1. Kv4.2 is normally expressed at high levels in s. radiatum of CA1 (Figure 3), and the lack of any obvious change in Kv4.3 staining here, as well as other sites where Kv4.2 is normally expressed (Menegola & Trimmer 2006), suggests that upregulation of Kv4.3, and maintenance of dendritic A-type potassium currents, as a simple compensatory mechanism for maintaining neuronal function, has not occurred in response to the absence of Kv4.2 (Menegola & Trimmer 2006). Independent immunoblot analyses performed with a commercially available anti-Kv4.3 polyclonal antibody provided similar results, in that no differences were observed between crude membrane fractions from WT and Kv4.2 KO hippocampi (Chen, et al. 2006).

Kv2.1 is a major somatic and dendritic voltage-gated potassium channel in mammalian neurons (Trimmer 1991, Vacher, et al. 2008). Kv2.1 is broadly expressed in most brain neurons, including the dentate granule cell and CA pyramidal cells in hippocampus that express high levels of Kv4.2 (Rhodes, et al. 2004). Given the cellular and subcellular pattern of Kv2.1 expression, and the prominent role that Kv2.1 plays in determining neuronal excitability (Misonou, et al. 2004, Misonou, et al. 2005, Misonou, et al. 2006b, Mohapatra, et al. 2009, Ikematsu, et al. 2011, Plant, et al. 2011), one might expect that Kv2.1 expression might be upregulated in these cells to compensate for the lack of Kv4.2 in their dendrites. However, no obvious differences in Kv2.1 staining were observed between the hippocampi from WT and Kv4.2 KO mice (green staining in Figure 3).

Immunohistochemical analyses of auxiliary subunit expression in Kv4.2 KO mice reveals unexpected changes

Dendritic Kv4 channels in mammalian brain are multiprotein complexes containing principal and auxiliary subunits (Jerng, et al. 2004, Maffie & Rudy 2008, Vacher & Trimmer 2011). The first of these discovered were a family of cytoplasmic auxiliary subunits termed KChIPs (An, et al. 2000). KChIPs exert pronounced effects on diverse functional features of Kv4 channel complexes (An, et al. 2000, Bahring, et al. 2001, Shibata, et al. 2003, Jerng, et al. 2004, Vacher & Trimmer 2011). KChIPs are also extensively associated and colocalized with Kv4 subunits in mammalian brain (An, et al. 2000, Rhodes, et al. 2004, Strassle, et al. 2005, Vacher, et al. 2008). Interestingly, different KChIPs are found in different neuronal populations, and exhibit distinct patterns of association and colocalization with Kv4.2 and Kv4.3, suggesting cell-specific Kv4:KChIP combinations (Rhodes, et al. 2004).

We performed immunofluorescence staining using mAbs specific for each of the KChIPs 1–3 to determine whether the loss of Kv4.2 alters expression and localization of the auxiliary KChIP subunits (Menegola & Trimmer 2006). We found that the expression of KChIP1, which is primarily expressed in interneurons throughout the hippocampus (Rhodes, et al. 2004, Menegola, et al. 2008), is for the most part unaltered in the Kv4.2 KO mice (Menegola & Trimmer 2006). However, expression of KChIP2 and KChIP3 is fundamentally changed, in that their expression is greatly reduced in the hippocampus of Kv4.2 KO mice. While a decrease in KChIP expression had been observed on immunoblots using the same mAbs (Chen, et al. 2006), it was the immunohistochemical analysis of brain sections that revealed the remarkable nature of these decreases. These analyses revealed that expression of these KChIPs was reduced in those hippocampal subfields and cells in which they would normally be found associated and colocalized with Kv4.2, but were spared in subfields and cells in which they are normally found associated and colocalized with Kv4.3. As mentioned above, the expression of KChIP1, which is predominantly found associated and colocalized with Kv4.3 in interneurons, is relatively unchanged in Kv4.2 KO mice (Menegola & Trimmer 2006), although immunoblot analyses suggested a reduction in overall expression (Chen, et al. 2006). In rat hippocampus (Rhodes, et al. 2004), and in the hippocampus of WT mice (Figure 4), KChIP2 expression (as determined here by staining with the mouse mAb K60/73 in red) is found associated and colocalized with both Kv4.2 and Kv4.3 in the dendrites of dentate granule cells in the dentate gyrus molecular layer, and in s. oriens and s. radiatum of CA3. In these regions of KChIP2, Kv4.2 and Kv4.3 association and colocalization, the elimination of Kv4.2 led to a partial reduction in KChIP2 expression (Figure 4). However, in s. oriens and s. radiatum of CA1, where Kv4.2 is present in the absence of Kv4.3, the elimination of Kv4.2 led to a corresponding elimination of KChIP2 expression (Figure 4).

Figure 4.

Staining for potassium channel auxiliary subunits in WT and Kv4.2 KO mouse hippocampus. Sagittal brain sections prepared from age- and sex-matched WT and Kv4.2 KO mice were used in double label immunofluorescence staining with mAbs specific for KChIP2 (left panels) or KChIP3 (right panels) in red, and polyclonal antibodies against Kv2.1 in green, proteins as indicated. *: molecular layer of the dentate gyrus. **: s. radiatum of CA3. ***: s. radiatum of CA1. Scale bar = 200 µm.

KChIP3 expression, as defined here by staining with the mouse mAb K66/38 in red, is found at sites in dentate gyrus and CA3 that typically contain KChIP2, Kv4.2 and Kv4.3. The expression of KChIP3 was markedly reduced in the molecular layer of the dentate gyrus, and in s. radiatum of CA3 in the Kv4.2 KO mice (Figure 4). These results show that the reduction of KChIP expression upon elimination of Kv4.2 corresponds precisely with their extent of association and colocalization with Kv4.2 in WT brains. In neurons expressing Kv4.3 alone, little or no KChIP reduction is observed, in neurons expressing Kv4.2 and Kv4.3, a partial reduction is observed, while in neurons expressing Kv4.2 alone, an almost complete elimination is observed. This suggests that association of KChIPs with Kv4 channels is a primary determinant of their expression levels.

Discussion

Here we present results obtained by performing immunohistochemical analyses comparing the brains of WT and Kv channel KO mice. These results reinforce those obtained in previously published studies, and highlight the novel and important insights that can be obtained using this approach. Biochemical analyses, such as those provided by immunoblots, can provide valuable insights into the expression levels of specific proteins in samples prepared from WT and Kv channel KO brains. Immunoblot analyses have been performed previously on Kv1.1 (Wenzel, et al. 2007b) and Kv4.2 (Chen, et al. 2006) KO mice, and have provided important data on lack of compensatory changes in other Kv channel α subunits, and, in the case of the Kv4.2 KO mice, significant down-regulation of KChIP expression levels (Chen, et al. 2006). While immunoblot analyses provide a reliable view of changes in overall protein expression within an entire brain region, the nature of sample preparation, which involves generating a single sample from a piece of brain tissue that contains numerous and diverse neuronal cell types, obscures changes in specific neuronal populations of their subcellular domains. For example, immunoblot analyses would fail to detect changes in a situation whereby increased expression levels occur in one neuronal population simultaneous with decreased expression in another population. Such a situation could occur in neuronal networks whose overall activity could be stabilized by opposing changes in excitatory and inhibitory neurons. Moreover, as trafficking of Kv channels to precise subcellular domains is a key component of their functional role (Vacher, et al. 2008), it is also possible that Kv channel KO mice could exhibit similar levels of overall protein expression as measured by immunoblots, but have aberrant neuronal function due to altered subcellular targeting of the remaining channels. As such, the utility of using immunohistochemistry to complement, or even supersede immunoblot analyses, is paramount.

This approach has been used by Wenzel and colleagues in their studies of Kv1.1 KO mice (Smart, et al. 1998, Wenzel, et al. 2007b), as well as in studies of zinc transporter 3 KO mice (Cole, et al. 1999, Cole, et al. 2000, Lopantsev, et al. 2003), p35 KO mice (Wenzel, et al. 2001, Patel, et al. 2004, Wenzel, et al. 2007a) and mouse models of the fragile X premutation (Hunsaker, et al. 2009, Wenzel, et al. 2010). In the case of the studies on the mice lacking expression of the Kv1.1 and Kv4.2 potassium channel α subunits, immunohistochemical analyses were critical to demonstrating that the overall expression levels and subcellular distributions of associated α subunits were relatively unaffected by the elimination of their binding partners. In the case of the Kv4.2 KO mice, the finding that loss of Kv4.2 expression led to precise loss of expression of specific auxiliary KChIP subunits in a pattern that accurately reflected their normal association and colocalization with Kv4.2 led to important insights. First, this finding provided strong evidence that the primary role of KChIPs is as auxiliary subunits of Kv4 channels, in spite of the other functions that have been attributed to KChIPs. Second, while it had been previously shown that KChIPs fundamentally impact the properties of Kv4 α subunits, including their stability, these results showed that Kv4 α subunits also reciprocally impact KChIPs. In fact, a recent study provides a potential mechanism for the observed loss of Kv4.2-associated KChIPs upon elimination of Kv4.2, in showing that Kv4.2 co-expression stabilizes KChIP protein expression (Foeger, et al. 2010). Together, the immunohistochemical approaches to the study of KO mice championed and pursued with a high level of ingenuity and expertise by Dr. Jürgen Wenzel have proven to be crucial to the analysis of Kv channel KO mice, and will continue to be of great value as additional KO mice, both constitutive and inducible, become available in future years.