Differential Expression of Kv4 K+ Channel Subunits Mediating Subthreshold Transient K+ (A-Type) Currents in Rat Brain

Paulo Serôdio, Bernardo Rudy


Serôdio, Paulo and Bernardo Rudy. Differential expression of Kv4 K+ channel subunits mediating subthreshold transient K+(A-type) currents in rat brain. J. Neurophysiol. 79: 1081–1091, 1998. The mammalian Kv4 gene subfamily and its Drosophila Shal counterpart encode proteins that form fast inactivating K+ channels that activate and inactivate at subthreshold potentials and recover from inactivation at a faster rate than other inactivating Kv channels. Taken together, the properties of Kv4 channels compare best with those of low-voltage activating “A-currents” present in the neuronal somatodendritic compartment and widely reported across several types of central and peripheral neurons, as well as the (Ca2+-independent) transient outward potassium conductance of heart cells (I to). Three distinct genes have been identified that encode mammalian Shal homologs (Kv4.1, Kv4.2, and Kv4.3), of which the latter two are abundant in rat adult brain and heart tissues. The distribution in the adult rat brain of the mRNA transcripts encoding the three known Kv4 subunits was studied by in situ hybridization histochemistry. Kv4.1 signals are very faint, suggesting that Kv4.1 mRNAs are expressed at very low levels, but Kv4.2 and Kv4.3 transcripts appear to be abundant and each produces a unique pattern of expression. Although there is overlap expression of Kv4.2 and Kv4.3 transcripts in several neuronal populations, the dominant feature is one of differential, and sometimes reciprocal expression. For example, Kv4.2 transcripts are the predominant form in the caudate-putamen, pontine nucleus and several nuclei in the medula, whereas the substantia nigra pars compacta, the restrosplenial cortex, the superior colliculus, the raphe, and the amygdala express mainly Kv4.3. Some brain structures contain both Kv4.2 and Kv4.3 mRNAs but each dominates in distinct neuronal subpopulations. For example, in the olfactory bulb Kv4.2 dominates in granule cells and Kv4.3 in periglomerular cells. In the hippocampus Kv4.2 is the most abundant isoform in CA1 pyramidal cells, whereas only Kv4.3 is expressed in interneurons. Both are abundant in CA2-CA3 pyramidal cells and in granule cells of the dentate gyrus, which also express Kv4.1. In the dorsal thalamus strong Kv4.3 signals are seen in several lateral nuclei, whereas medial nuclei express Kv4.2 and Kv4.3 at moderate to low levels. In the cerebellum Kv4.3, but not Kv4.2, is expressed in Purkinje cells and molecular layer interneurons. In the cerebellar granule cell layer, the reciprocity between Kv4.2 and Kv4.3 is observed in subregions of the same neuronal population. In fact, the distribution of Kv4 channel transcripts in the cerebellum defines a new pattern of compartmentation of the cerebellar cortex and the first one involving molecules directly involved in signal processing.


The diversity of K+ channels is responsible in large part for the variety of neuronal firing patterns and the physiological states that result from such activity (Baxter and Byrne 1991; Llinás 1988; Rudy 1988). This diversity results in part from the existence of a large repertoire of channel subunits. Among these are the members of the various subfamilies (Kv1–Kv8) of the Kv family, which are primary components of voltage-gated K+ channels (reviewed in Jan and Jan 1997; Perney and Kaczmarek 1991; Pongs 1992; Rehm and Tempel 1991; Rudy et al. 1991).

On the basis of similarities with the currents through Kv4 channels expressed in Xenopus oocytes and gene elimination methods (Baldwin et al. 1991; Pak et al. 1991; Roberds and Tamkun 1991; Serôdio et al. 1994, 1996; Tsunoda and Salkoff 1995) it is believed that Shal-related (Kv4) proteins are the key components of the classical A-type K+ channels underlying the A currents (I A) operating in the subthreshold range of action-potential generation found in many neuronal somata. Transient K+ channels that operate near action potential threshold can also be formed by Kv1.4 proteins and by other Kv1 α subunits, when complexed with β subunitsthat confer fast inactivation to otherwise noninactivating Kv1 channels (Heinemann et al. 1996; Rettig et al. 1994). However, A-type channels formed with Kv1 proteins in Xenopus oocytes are different from those formed by Kv4 subunits in at least two important features: Kv4 channels inactivate with time constants that change very little as a function of voltage, even at very negative potentials when the channels start activating, and recover very fast from inactivation as is the case with native I A in many neurons (Baldwin et al. 1991; Pak et al. 1991; Roberds and Tamkun 1991; Serôdio et al. 1994, 1996). These two features are further enhanced by coexpression of Kv4 proteins with their putative β subunits (Chabala et al. 1993; Serôdio et al. 1994, 1996). On the other hand Kv1 channels inactivate with time constants that change with voltage and recover from inactivation rather slowly (Po et al. 1993; Stühmer et al. 1989; Tseng-Crank et al. 1990). Furthermore, in this case, interaction with β subunits slows down the recovery from inactivation (Heinemann et al. 1995, 1996; Serôdio 1996).

Subthreshold-operating A-type K+ channels (I SA channels) are of particular interest because of their roles in regulating firing frequency, spike initiation, and waveform and have been found in many types of neuronal somata in mammals and other species (Byrne 1980; Connor and Stevens 1971a,b; Getting 1983; Hille 1992; Llinás 1988; McCormick and Huguenard 1992; Rudy 1988; Thompson and Aldrich 1980). Kv4 proteins are also believed to be the major components of the Ca2+-independent A current in cardiac ventricular muscle (the “transient outward current” or I to) (Dixon and McKinnon 1994; Fiset et al. 1997; Johns et al. 1997; Nakamura et al. 1997). This current plays key roles in the cardiac action potential, being responsible for early repolarization and hence the overall duration of the action potential and the length of the refractory period (Surawicz 1992; Snyders 1995). Antisense hybrid-arrest and dominant negative experiments with cardiac ventricular myocytes also support the hypothesis that the cardiac I to is mediated by Kv4 proteins (Fiset et al. 1997; Johns et al. 1997; Nakamura et al. 1997).

Three mammalian Kv4 genes, each encoding a single known protein product have been identified (Kv4.1 or mShal, Pak et al. 1991; Kv4.2 or rShal1 Baldwin et al. 1991, or RK5 Roberds and Tamkun 1991; and Kv4.3 or KShIVB Rudy et al. 1991, Serôdio et al. 1996). The formation of distinct Kv4 channels from various combinations of Kv4 subunits can contribute to the diversity of I A kinetics and voltage range of activation and inactivation observed in different cells (Serôdio et al. 1996).

To begin to understand the organization of Kv4 subunits in the mammalian brain and how they contribute to the molecular basis of A-type K+ channel diversity, we have studied the distribution of transcripts for the three known mammalian Kv4 genes in the adult rat brain by in situ hybridization. Previous studies of X-ray film autoradiograms of sagittal sections have shown that Kv4 genes are expressed in specific areas of the brain (Serôdio et al. 1996). Here we extend these preliminary observations by examining the expression of the three Kv4 genes throughout the adult rat brain and identify some of the neuronal populations by microscopic analysis of emulsion-dipped sections. The results presented here have been previously published as part of P. Serôdio's doctoral thesis (Serôdio 1996).


Preparation of cDNA probes for northern-blot analysis and in situ hybridization histochemistry

Clones for the rat Kv4.1, Kv4.2, and Kv4.3 cDNAs were obtained as previously described (Serôdio et al. 1994, 1996). Two distinct DNA fragments were prepared from each of these cDNA clones.


A 346-bp fragment spanning the sequence between nucleotides 1,077 and 1,424 of Kv4.1 was obtained by a Sac I/Hinc II restriction enzyme digestion. This sequence corresponds to the C-terminal region of an incomplete rat Kv4.1 clone previously described (Serôdio et al. 1994), available under the Genebank accession number U89873. A second, 323-bp fragment spanning the sequence between nucleotides 589 and 912 of the same Kv4.1 clone was obtained by a Nco I/Kpn I restriction enzyme digestion. This fragment corresponds to the regions between S5 and the C-terminal sequence of the previous cDNA.


A 365-bp fragment spanning the sequence between nucleotides 1,962 and 2,329 of Kv4.2 was obtained by a BamH I/BsaA I restriction enzyme digestion of a Kv4.2 cDNA previously reported (Serôdio et al. 1994). The sequence of this cDNA is identical to that reported by Baldwin et al. (1991) and is available under the Genebank accession number S64320. The probe corresponds to a region encoding the C-terminal portion of Kv4.2. A second 362-bp fragment of Kv4.2 was obtained by Acc I/BamH I restriction enzyme digestion. This fragment spans the sequence between nucleotides 1,484 and 1,846, corresponding to the regions between the S4-S5 linker and the proximal C-terminus of Kv4.2.


A 375-bp fragment of Kv4.3 spanning the sequence between nucleotides 1,763 and 2,138, corresponding to the distal C-terminal and 3′-untranslated (3′-UTR) regions of the sequence reported in Serôdio et al. (1996) was obtained by Spe I/BamH I restriction. The sequence of the cDNA is available under Genebank accession number U42975. A second 385-bp fragment, spanning the sequence between nucleotides 985 and 1,370 of Kv4.3 was obtained by Avr II/Sac I restriction enzyme digestion. This fragment corresponds to the regions between the S4-S5 linker and the proximal C-terminal region of Kv4.3.

All these cDNA fragments show <60% nucleotide identity with other sequences in the Kv family. Both probes from each Kv4 gene gave consistent results and showed no cross-reactivity with other Kv4 genes under the hybridization conditions used in these experiments.

Radioactive probes were obtained by labeling the cDNA fragments by the random hexamer primer method (Feinberg and Vogelstein 1983) with 32P-α-dCTP for Northern-blot analysis, or with 35S-α-dCTP for in situ hybridization histochemistry. The in situ hybridization experiments were repeated twice with each of the probes. The specific activities of the probes used were as follows: 5 × 108 cpm/μg and 4.8 × 108 cpm/μg for the 346 bp Kv4.1 probe; 6.5 × 108 cpm/μg and 5.7 × 108 cpm/μg for the 323 bp Kv4.1 probe; 4.4 × 108 cpm/μg and 4.2 × 108 cpm/μg for the 365 bp Kv4.2 probe; 6.2 × 108 cpm/μg and 5.9 × 108 cpm/μg for the 362 bp Kv4.2 probe; 6.5 × 108 cpm/μg and 5.1 × 108 cpm/μg for the 375 bp Kv4.3 probe; 5.1 × 108 cpm/μg and 4.8 × 108 cpm/μg for the 385 bp Kv4.3 probe.

Preparation of poly-(A) RNA and northern blot analysis

RNA was isolated from freshly dissected brains obtained from 20 day old Sprague-Dawley rats by a LiCl/Urea procedure (Dierks et al. 1981). Poly(A) RNA was prepared by oligo (dT) column chromatography (type III poly-T Sepharose from Collaborative Research) following the protocol of Sambrook et al. (1989). The RNA was ethanol-precipitated twice and resuspended in ribonuclease (RNase)-free water at a concentration of ∼1 mg/ml. mRNAs were subjected to electrophoresis in denaturing formaldehyde gels and transferred to Hybond-N membranes (Amersham) as previously described (Rudy et al. 1988). The blots were hybridized with 32P-radiolabeled DNA probes prepared from restriction fragments of the cDNAs as described. The hybridization solution consisted of 50% (vol) formamide, 5 × Denhardt's solution, 5 × SSPE (0.15 M NaCl, 1 mM EDTA, 10 mM NaH2PO4, pH 7.4), 0.3% sodium dodecyl sulfate (SDS) and 200 μg/ml denatured salmon sperm DNA. After a 4-h prehybridization at 42°C, labeled probe was added and the blots hybridized for 18 h at 42°C. The blots were washed at 68°C in 0.2 × SSC (0.15 M NaCl, 0.015 M Na citrate, pH 7) with 0.1% SDS and exposed to X-ray film at −70°C for 8–12 h with two intensifying screens.

In situ hybridization histochemistry

In situ hybridization was performed by using the methods described in Stone et al. (1990) and Weiser et al. (1994). Briefly, adult (150–250 g) male rats were perfused intracardially with 100 ml of cold saline solution (0.9% NaCl with 0.5% NaNO2 and 1,000 u Heparin), followed by 300 ml of cold 4% paraformaldehyde solution in 0.1 M phosphate buffer, pH 7.4. The brains were carefully removed, cut in blocks and postfixed for 1 h. After postfixing, the brains were washed several times in cold, 0.1 M phosphate buffer (pH 7.4) and placed in 30% sucrose (made with RNase-free water) overnight. Slices were obtained on a freezing-microtome at 40 μm thickness and prehybridized at 48–50°C on a solution containing 50% formamide, 2× SSC, 10% Dextran Sulfate, 4× Denhardt's, 50 mM dithiothreitol, and 0.5 mg/ml sonicated, denatured salmon sperm DNA. After 2 h of prehybridization, heat-denatured, 107 counts of 35S-radiolabeled probe were added and the hybridization reaction allowed to proceed overnight. After hybridization, the sections were washed in decreasing concentrations of SSC (2× to 0.1×) buffer at 48°C. After a final wash in 0.05 M phosphate buffer at room temperature, the sections were hand mounted on gelatin-coated slides and air dried. The slides were exposed to Kodak XAR-5 X-ray film for 5 days. The slides were then dipped in prewarmed Kodak NTB-2 photographic emulsion and exposed for 4 wk at 4°C, in the dark. The slides were developed in Kodak d-19 solution, fixed and counterstained with a cresyl violet solution. Data analysis and photography were performed in an Olympus BH2 photomicroscope. The atlases by Paxinos and Watson (1986) and Swanson (1992) were used to assign CNS neuronal populations.

Fig. 1.

Northern blot analysis of adult rat whole brain poly(A+) RNA (2 μg/lane) hybridized with probes specific for Kv4.1–Kv4.3 transcripts. Left: size of RNAs as obtained with Bethesda Research Labs RNA size markers is shown on left.

Fig. 2.

Differential distribution of Kv4.1 mRNAs in rat brain. X-ray autoradiograms of coronal sections of an adult rat brain at level of olfactory bulb (A), forebrain (BD), and cerebellum and medulla (EF) hybridized with 35S-labeled probes specific for Kv4.1 transcripts. AO, anterior olfactory nucleus; CA1 and CA3, CA1 and CA3 fields of Hammon's horn; DG, dentate gyrus; GrO, granule cells of olfactory bulb; P, Purkinje cells.


Northern blot analysis shows that the three known Kv4 genes are expressed in the adult rat brain, and each produces a transcript of a distinct size, however, Kv4.1 transcripts are found at extremely low levels, whereas Kv4.2 and Kv4.3 mRNAs are quite abundant (Fig. 1).

Distribution of KV4 transcripts in the rat brain

In situ hybridization histochemistry was used to study the distribution of Kv4 transcripts. Figures 2-4 depict X-ray film autoradiographs of coronal sections of an adult rat brain, hybridized with 35S-labeled probes specific for Kv4.1, Kv4.2, and Kv4.3. Kv4.1 probes produce weak hybridization signals in the auxiliary olfactory bulb, granule cells of the olfactory bulb, hippocampus, dentate gyrus, and cerebellar cortex, possibly in Purkinje cells (Fig. 2). Other brain areas show background signals (Fig. 2).

In contrast, Kv4.2 and Kv4.3 transcripts are abundant in the brain and are widely distributed (Figs. 3 and 4). Most commonly, Kv4.2 and Kv4.3 mRNAs are expressed in distinct neurons producing a reciprocal or complementary pattern of expression (Table 1). There are nuclei that express almost exclusively Kv4.2 mRNAs, such as the striatum (Fig. 3, A and C) and pontine nuclei (Fig. 4 C), whereas others express Kv4.3 transcripts strongly and have small or negligible expression of Kv4.2, such as the substantia nigra pars compacta, the deep superficial gray-optic layer of the superior colliculus, the raphe (Fig. 4, B and D), the retrosplenial cortex (Fig. 3, D and F), and the amygdala (Fig. 3 F). Kv4.2 is the predominant transcript in vestibular nuclei, the dorsal cochlear nucleus (Fig. 4 E) and the oral spinal trigeminal and facial nuclei (Fig. 4 G). A pattern of reciprocal expression is also seen in the piriform cortex and the olfactory tuberculum, two structures receiving mitral cell input (Fig. 3, A and B) and other brain regions (Table 1). In some brain areas a second type of differential expression pattern is observed. Both Kv4.2 and Kv4.3 mRNAs are expressed, but one or the other may dominate in specific subregions. For example, in the cerebellar granule cell layer, Kv4.2 is the dominant subunit in the anterior lobules, whereas Kv4.3 is expressed in increasing amounts toward the caudal regions and particularly in the flocculonodular lobe (Fig. 3, I and J; the exact lobular pattern of expression in the cerebellum is analyzed below). A similar pattern is seen in the hippocampus, as described later. In several other neuronal populations a third type of pattern is observed, where both Kv4.2 and Kv4.3 transcripts are abundant, such as in granule cells in the dentate gyrus (Fig. 3, CH) and in the anterior olfactory nucleus (Fig. 5). The distribution of Kv4.2 and Kv4.3 mRNAs in several key brain areas is further analyzed below.

Fig. 3.

Differential distribution of Kv4.2 and Kv4.3 mRNAs in adult rat brain. X-ray autoradiograms of coronal sections of an adult rat brain at level of forebrain (AF), midbrain (G and H), and cerebellum and medulla (I and J) hybridized with 35S-labeled probes specific for Kv4.2 or Kv4.3. CA1–CA3, CA1 to CA3 fields of the hippocampus; Cx, neocortex; CP, caudate-putamen; DLG, dorsal lateral geniculate nucleus; Fl, flocculus; GP, globus pallidus; Gr, granule cell layer of cerebellar cortex; Hb, habenular nucleus; LP, lateroposterior nucleus of thalamus; MG, medial geniculate nucleus; P, Purkinje cell layer of cerebellar cortex; Pfl, paraflocculus; PV, paraventricular thalamic nucleus; Py, piriform cortex; Rs, retrosplenial cortex; RT, nucleus reticularis thalamic; SnC, substantia nigra, pars compacta; Tu, olfactory tuberculum; VP, ventro-posterior complex of thalamus; VPM, ventroposteriomedial nucleus of thalamus.

Fig. 4.

Differential distribution of Kv4.2 and Kv4.3 mRNAs in brain stem. X-ray autoradiograms of coronal sections of midbrain at level of superior colliculus (A and B), midbrain at level of inferior colliculus (C and D) and medulla (EJ) of an adult rat brain hybridized with 35S-labeled probes specific for Kv4.2 or Kv4.3. 7, facial nucleus (VII); DCN, dorsal cochlear nucleus; MG, medial geniculate; Op, optic layer of superior colliculus; IC, inferior colliculus; IO, inferior olive; Pn, pontine nuclei; R, raphe; Sp5, spinal trigeminal nucleus (V); SPO, oral spinal trigeminal nucleus; Ve, vestibular nucleus.

View this table:
Table 1.

Qualitative estimates of the relative distributions of Kv4 transcripts in selected areas of the adult rat brain as determined by in situ hybridization

Fig. 5.

Distribution of Kv4.2 and Kv4.3 mRNAs in olfactory bulb. X-ray autoradiograms of sagittal (A and B) and coronal (C and D) sections and high-power dark-field photomicrographs of emulsion autoradiograms of sagittal sections (E and F), of olfactory bulb of an adult rat brain hybridized with 35S-labeled probes specific for Kv4.2 or Kv4.3. AO, anterior olfactory nucleus; Gl, periglomerular cells; GrO, granule cell layer.

Olfactory bulb

The differential expression of Kv4.2 and Kv4.3 mRNAs is clearly illustrated here. Although both genes are expressed in the periglomerular and granule cells, Kv4.3 dominates in the periglomerular area and Kv4.2 in granule cells (Fig. 5). As shown previously, Kv4.1 is also expressed in granule cells (Fig. 2), albeit at significantly lower levels than Kv4.2 and Kv4.3. Neither Kv4.2 nor Kv4.3 appear to be abundantly expressed in tufted cells.

Hippocampal formation

Strong hybridization signals were observed in several neuronal populations in the hippocampus. However, each neuronal population appears to express specific combinations of Kv4 genes (Fig. 6, A and B). Kv4.2 signals were very strong in CA1 pyramidal cells where Kv4.3 signals were negligible. Strong Kv4.2 and Kv4.3 signals were seen in CA2 and CA3 neurons and in granule cells of the dentate gyrus. As shown earlier Kv4.1 appears to be expressed throughout the hippocampal pyramidal cell layers and dentate gyrus, although at very low levels (Fig. 2). Nevertheless, in CA1 neurons, there appears to be more Kv4.1 than Kv4.3 mRNAs.

Fig. 6.

Distribution of Kv4.2 and Kv4.3 mRNAs in hippocampus, thalamus, and cerebellar cortex. Dark and bright-field photomicrographs of emulsion autoradiograms of coronal sections of an adult rat brain hybridized with 35S-labeled probes specific for Kv4.2 or Kv4.3. (A and B) medium power dark-field images of hippocampal formation showing hybridization (white signals) in pyramidal cell layers, dentate gyrus, and hippocampal interneurons of a section hybridized with a Kv4.2 (A) or Kv4.3 probe (B). CE: low-power dark-field images of thalamus of emulsion-dipped sections hybridized with a Kv4.2 (C) or Kv4.3 (D and E) probe. F and G: high-power bright-field images of ventral posterolateral (VPL) nucleus in emulsion-dipped sections counterstained with Nissl hybridized with a Kv4.2 (F) or Kv4.3 (G) probe. Note accumulation of hybridization grains on large cells, some of which are indicated with arrows. (HK) medium-power dark-field images of a rostral (HJ) or caudal (I and K) cerebellar lobule hybridized with a Kv4.2 (H and I) or Kv4.3 (J and K) probe. c. Crb, cerebellar lobule in flocculonodular lobe; Hil, hilus; DLG, dorsal lateral geniculate; Mol, molecular layer of cerebellar cortex; r. Crb, cerebellar lobule in anterior lobe; SLM, stratum lacunosum moleculare; SO, stratum oriens; SR, stratum radiatum.

Kv4.3 is also expressed in specific subsets of hippocampal interneurons (Fig. 6 B) including interneurons of the stratum oriens-alveus (OAI) and cells in the dentate hilus. A subset of hippocampal interneurons that expresses somatostatin has a similar distribution (Somogyi et al. 1984) and have been shown to have a prominent A-type current (Zhang and McBain 1995a). There is also a group of cells located in the stratum radiatum close to the hippocampal fissure and the stratum lacunosum moleculare. Interestingly, other prominent populations of hippocampal interneurons, such as the basket cells along the stratum pyramidalis do not appear to express significantly levels of any of the known Kv4 transcripts, although they also have transient K+ currents (see Discussion).


Kv4.2 transcripts are expressed moderately to weakly, but more or less homogeneously throughout dorsal thalamic nuclei. On the other hand, Kv4.3 transcripts generated strong signals in some thalamic nuclei but weak in others (Figs. 6, CE, and 3, CH). Kv4.3 mRNAs appear to be particularly abundant in several lateral nuclei, including the dorsolateral (DL), the ventral posterolateral (VPL) and posteromedial (VPM) nuclei, the dorsal lateral geniculate (DLG), and the medial geniculate (MG). Kv4.3 signals were also strong in anterior thalamic nuclei. In thalamic nuclei closer to the midline Kv4.3 signals are similar if not weaker than those observed for Kv4.2 (Fig. 6, CE, and 3, CF). Bright field images of thalamic nuclei obtained from emulsion-dipped sections show that both Kv4.2 and Kv4.3 transcripts are expressed mainly in the large cells, the thalamic relay neurons, and not in the glia, indicating that these channels are predominantly neuronal (Fig. 6, F and G).

Cerebellar cortex

Both Kv4.2 and Kv4.3 transcripts are abundant in this structure, but each displays a distinct pattern. Kv4.2 is very strongly expressed in granule cells, but in a decreasing rostro-caudal gradient (Fig. 7, A and B). The lowest concentration of Kv4.2 transcripts is present in the caudal cerebellar lobules of the floculonodular lobe (the archicerebellum) (Fig. 7, A and B, and Fig. 3 I and 6, H and I). In the granule cells, Kv4.3 is expressed in a reciprocal manner, being most abundant where Kv4.2 transcripts are at their lowest abundance (Fig. 7, A and B, 3J, and 6, J and K). The distribution of Kv4.3 in the cerebellar granule cell layer resembles the lobular distribution of calretinin positive unipolar brush cells (Abbott and Jacobowitz 1995; Yan and Garey 1996); however, Kv4.3 transcripts are present throughout the granule cells of the cerebellar lobules expressing this gene rather than in scattered clusters of a specific cell subtype (Figs. 6 K and 7 D).

Fig. 7.

Patterned distribution of Kv4.2 and Kv4.3 mRNAs in cerebellar cortex. A and B: X-ray autoradiograms of coronal (A) and sagittal (B) sections of cerebellum from an adult rat brain hybridized with 35S-labeled probes specific for Kv4.2 or Kv4.3, as indicated in figure, illustrating rostrocaudal gradients of expression of these 2 transcripts in granule cell layer. Kv4.2 signals are stronger in anterior lobules (L1–L6), whereas Kv4.3 signals are strongest in posterior lobules (L6–L10). C and D: high-power bright-field emulsion autoradiograms of coronal sections of an adult rat brain hybridized with 35S-labeled probes specific for Kv4.2 (C) or Kv4.3 (D). White arrows: Purkinje cells showing accumulation of hybridization grains when hybridized with Kv4.3 but not with Kv4.2 probes. Black arrows: molecular cell layer interneurons with Kv4.3 hybridization grains. Background signals are observed in molecular layer with Kv4.2 probes. AL, ansiform lobule; L1–10, cerebellar lobules 1–10; PFL, parafloccolus; SL, simplex lobule.

Various patterns of compartmentation of the cerebellar cortex have been defined for several neuronal markers (reviewed in Hawkes and Gravel 1991; Herrup and Kuemerle 1997) (see Discussion). Observation of coronal (Figs. 7 A and 3, I and J) and sagittal (Fig. 7 B) sections are consistent with an anteroposterior gradient with a boundary close to lobule VI similar to that observed for acetycholinesterase, tyrosine hydroxylase, and several other markers (reviewed in Herrup and Kuemerle 1997).

On the other hand Purkinje cells and molecular cell layer interneurons express mainly Kv4.3 mRNAs (Figs. 6, HK and 7, C and D). Kv4.2 is not seen in these cells types even after long exposure of the autoradiograms (Figs. 6, H andI, and 7 C). Curiously, no patterned distribution of Kv4.3 subunits, such as that seen in the granule cell layer, is observed in Purkinje cells, which display similar signals in rostral and caudal lobules (see Fig. 6, J and K).


In the neocortex, both Kv4.2 and Kv4.3 probes produced relatively weak signals throughout most cortical areas (Figs. 3 and 8). Observation of emulsion-dipped sections showed that hybridization grains were present over large cells distributed in layers II–VI and presumably include pyramidal neurons (data not shown). An exception to this is the cingulate and retrosplenial cortices, where Kv4.3 signals in layer II-III were moderate and very strong, respectively (Fig. 3, B, D, F, and H and Fig. 8).

Fig. 8.

Distribution of Kv4.2 and Kv4.3 mRNAs in cortex in dark field photomicrographs of emulsion autoradiograms. Notice relatively weak expression of Kv4.2 and Kv4.3 transcripts in layer II–VI of neocortex (Cx). Signal intensity throughout cortical layers corresponds roughly to density of pyramidal cells. On other hand, Kv4.3 transcripts are abundantly expressed in layer II-III of retrosplenial cortex (Rs).


The distribution of the mRNA transcripts from the three known mammalian Shal-homolog genes (Kv4.1, Kv4.2, and Kv4.3) in the adult rat brain was investigated by in situ hybridization histochemistry. These studies demonstrate the abundance and widespread distribution of this subfamily of channels across neural tissue. The most intriguing results are the apparent low levels of expression of Kv4.1 mRNAs (kinetically and structurally the more distinct of the three Kv4 subunits) throughout the tissue, although this transcript was discovered in a brain cDNA library (Pak et al. 1991) and the pattern of reciprocal expression of Kv4.2 and Kv4.3 transcripts across many regions of the brain.

In several brain areas, Kv4.2 and Kv4.3 mRNAs appear to be coexpressed at significant levels. Heteromultimeric channels may exist in these regions if both transcripts are present in the same neurons, because it is known that Kv proteins of the same subfamily can form heteromultimeric (as well as homomultimeric) voltage-gated K+ channels (Christie et al. 1990; Covarrubias et al. 1991; Isacoff et al. 1990, McCormack et al. 1990; Ruppersberg et al. 1990; Sheng et al. 1993; Wang et al. 1993). However, many neuronal populations express predominantly, if not exclusively, either Kv4.2 or Kv4.3. In some areas, most notably in the granule cell layer of the cerebellum, this reciprocal expression is extended to subregions of the same neuronal population, the cerebellar granule cells. Something similar, albeit less dramatic, is also seen in pyramidal cells in the hippocampus and thalamic relay neurons in the dorsal thalamus. The patterned expression of Kv4 transcripts points to a functional heterogeneity of Kv4 channels and thus of I SAs, even in populations of cells usually considered homogenous, such as the cerebellar granule cells.

Differential and reciprocal expression of Kv4.2 and Kv4.3 subunits

The most striking example of reciprocal expression is the rostrocaudal gradient of Kv4.2 and Kv4.3 transcripts in the granule-cell layer of the cerebellum. This distribution is interesting in the light of the patterning of several markers in the cerebellar cortex (reviewed in Hawkes and Gravel 1991; Herrup and Kuemerle 1997). Among these are the Zebrins, epitopes identified by antibodies isolated from a monoclonal antibodies library that display longitudinal zebralike striations across the Purkinje cells defining distinct parasagittal compartments (Hawkes 1992; Hawkes and Gravel 1991; Leclerk et al. 1990). Other patterns involving different layers and cell types of the cerebellum have been found for several enzymes, growth-factor receptors and calcium-binding proteins (Hawkes and Gravel 1991; Herrup and Kuemerle 1997). The pattern defined by the expression of Kv4.2 and Kv4.3 is similar to the anteroposterior gradient formed by several of these markers, with a boundary in or close to lobule VI (Herrup and Kuemerle 1997). This boundary may reflect distinct embryological origins as shown in developmental studies employing graft transplants between chick and quail embryos (reviewed in Le Douarin 1993; see also Herrup and Kuemerle 1997). Grafting of the metencephalic vesicle from one species to the other gives origin to a rostrocaudal gradient of specific markers across the cerebellar cortex with a boundary close to lobule VI (Herrup and Kuemerle 1997). However, this developmental gradient is seen in all cortical layers, unlike what we observe for Kv4 gene expression, indicating that factors besides early embryological origin play a role in the establishment of the Kv4 pattern. The observed layer selectivity also distinguishes the pattern found here for Kv4.2 and Kv4.3 mRNAs and that formed by other molecular markers.

The cerebellar distributions presented here are also interesting because, to our knowledge, there is no other example in which opposite patterns are produced by pairs of functionally related molecules. Moreover, despite the great interest in cerebellar patterning, its functional consequences are not well understood (Herrup and Kuemerle 1997). The fact that the pattern shown here involves molecules directly involved in signal processing might provide a unique opportunity to advance our understanding of the physiological relevance of cerebellar compartmentation and could have implications for the modular processing of cerebellar information (Hawkes and Gravel 1991).

Interestingly, Kv4.2 and Kv4.3 mRNAs also show a reciprocal pattern of expression in rat cardiac tissue, where Kv4.2 transcripts are more abundant in ventricular muscle, but Kv4.3 in the atria (Dixon and McKinnon 1994; Serôdio et al. 1996).

Functional implications

The widespread expression of Kv4 mRNAs is consistent with the prevalent presence of I SAs in mammalian neurons (Llinás 1988; Rogawsky 1985; Rudy 1988; see references below). Abundant Kv4.2 and/or Kv4.3 transcripts were found in most neurons where somatic I SAs inactivate with rates that vary little with voltage and recover fast from inactivation as is the case for Kv4 currents in Xenopus oocytes (e.g., in cerebellar granule cells, Bardoni and Belluzi 1993, Cull-Candy et al. 1989, Gorter et al. 1995; cortical neurons, Albert and Nerbonne 1995; thalamic relay neurons, Budde et al. 1992, Huguenard et al. 1991; hippocampal neurons, Ficker and Heinemann 1992; Klee et al. 1995; stratum oriens hippocampal interneurons, Zhang and McBain 1995a; dentate gyrus granule cells, Beck et al. 1992; neostriatal neurons, Surmeier et al. 1989, 1991; supraoptic neurons, Nagatomo et al. 1995; neurons from the substancia nigra pars compacta, Silva et al. 1990; and Purkinje cells, Midtgaard et al. 1993; Wang et al. 1991). This supports the suggestion that Kv4 proteins are the main components of the channels generating many of the subthreshold-operating A-type currents that are recorded in the somatodendritic compartment of CNS neurons. Interestingly, some neuronal populations that do not appear to express much of Kv4.1-Kv4.3 transcripts, but express Kv1 α and β subunits, lack somatic, subthreshold-operating I As, such as neurons in the medial nucleus of the trapezoid body (Brew and Forsythe 1995).

Nevertheless, there are some important neuronal populations in the brain where somatic A-type K+ currents have been observed electrophysiologically, which do not express Kv4.1-Kv4.3 transcripts abundantly. One such example are neurons in the globus pallidus that have prominent low voltage-activating A-type currents that also resemble the currents recorded in heterologous systems expressing Kv4 channels (Hernández-Pineda et al. 1996; Surmeier et al. 1994). However Kv4 transcripts are weakly expressed here. Similarly, an A current has been recorded from basket cells in the pyramidal cell layers of the hippocampus (Zhang and McBain 1995b), another neuronal type not expressing significant levels of Kv4.1-Kv4.3 transcripts. Further analysis of the A currents in these cells is required to clarify the components of the channels mediating these currents. Preliminary studies of the influence of external K+ on the transient K+ current of hippocampal basket interneurons suggest that this current is mediated by Kv1.4 channels (Zhang and McBain 1995b). Channels of the Kv1 subfamily, perhaps Kv1.4, may also contribute to the somatic transient K+ current in other hippocampal neurons (Eder et al. 1996; Pardo et al. 1992). However, the possibility that there are additional unidentified Kv4 genes or that there are unknown gene families that also encode for subunits of Kv4-like subthreshold-operating A channels remains open.

I SA channels are diverse in terms of their voltage dependence and kinetics (Rudy 1988). Assuming that the levels of Kv4 mRNA transcripts reflect the levels of protein, the distributions found here would indicate the presence of channels containing diverse stoichiometries of Kv4.1-Kv4.3 subunits in different neurons. The variety of channel compositions could account, at least in part, for the observed functional diversity.

Kv4.1-Kv4.3 transcripts expressed in Xenopus oocytes produce currents that display differences in voltage dependence and kinetics (Baldwin et al. 1991; Chabala et al. 1993; Pak et al. 1991; Serôdio et al. 1994, 1996). Although the differences between Kv4.2 and Kv4.3 currents are relatively small, because I SA channels operate in a voltage range where other channels are closed, even subtle changes in the voltage-range of operation or rates of activation and inactivation can have large repercussions in the overall excitability and firing properties of cells expressing these channels. However, at this stage we are unable to find any clear correlation between the type of Kv4 transcript expressed in a given neuronal population and the reported properties of their somatic I A. Several factors contribute to this. First and foremost, the characterization of the I A in different neuronal populations in the CNS has not utilized uniform methodology. In some cases the electrophysiology has been done on freshly dissociated cells, whereas cultured cells or slices have been used in other cases. The composition of the recording solution has also varied. There are also differences in the procedures used to analyze the currents. These and other technical factors can influence the reported electrophysiological properties of the currents and can obscure the differences in voltage dependence and kinetics produced by differences in channel composition. For example, three recent papers on the properties of the I A in cerebellar granule cells report midpoints of activation and inactivation that differ by >20 mV (Bardoni and Belluzzi 1993; Gorter et al. 1995; Zegarra-Moran and Moran 1994). Comparative studies using uniform recording conditions and current analysis will be required to further assess the significance of channel Kv4 subunit composition on native channel kinetics and voltage dependence.

Second, further characterization of Kv4.1-Kv4.3 channels in heterologous expression systems could reveal additional differences, which may have physiological significance. For example little is known about the differential modulation of Kv4 channels. Each Kv4 subunit has unique putative phosphorylation sites that may contribute differences in their functional and modulatory attributes in vivo. Moreover, other postranslational modifications and the degree of association of the Kv4.1-Kv4.3 channel complexes with one or more β subunits could also contribute to the properties of native, subthreshold-operating A-channels.

The Northern blot and histochemistry studies suggest that Kv4.1 transcripts are present at very low abundance in specific neuronal populations in the hippocampus, olfactory bulb, and possibly in the cerebellar Purkinje cell layer (Figs. 1 and 2). The low level expression of Kv4.1 mRNAs in rat brain, also reported for heart tissue (Dixon and McKinnon 1994) could nevertheless be physiologically significant if Kv4.1 subunits contribute important properties to heteromultimeric channels containing Kv4.2 and/or Kv4.3 proteins. On the other hand Kv4.1 proteins may be expressed only in specific subcellular regions. It is also possible that Kv4.1 transcripts are translated with higher efficiency, have a longer lifetime in the cytoplasm or that they are expressed in the brain more abundantly at earlier developmental stages.


This research was supported by National Science Foundation Grant IBN9630832 and National Institute of Neurological Disorders and Stroke Grants NS-30989 and NS-35215.


  • Address for reprint requests: B. Rudy, Dept. of Physiology and Neuroscience, New York University School of Medicine, 550 First Ave., New York, New York 10016.

  • Present address of P. Serôdio: Center for Neurobiology and Behavior, Columbia University College of Physicians and Surgeons, New York, New York 10032.


View Abstract