Differential Expression of Voltage-Gated K+ Channel Subunits in Adult Rat Heart
Relation to Functional K+ Channels?
Abstract Polyclonal antibodies against each of the K+ channel subunits (Kv1.2, Kv1.4, Kv1.5, Kv2.1, and Kv4.2) shown previously to be expressed in adult rat heart at the mRNA level were used to examine the distributions of these K+ channel subunits in adult rat atrial and ventricular membranes. Immunohistochemistry on isolated adult rat ventricular myocytes revealed strong labeling with the anti-Kv4.2 and anti-Kv1.2 antibodies. Although somewhat weaker (than with anti-Kv1.2 or anti-Kv4.2), positive staining was also observed with the anti-Kv1.5 and anti-Kv2.1 antibodies. Ventricular myocytes exposed to the anti-Kv1.4 antibody, in contrast, did not appear significantly different from background. Qualitatively similar results were obtained on isolated adult rat atrial myocytes. Western blots of atrial and ventricular membrane proteins confirmed the presence of Kv1.2, Kv1.5, Kv2.1, and Kv4.2 and revealed differences in the relative abundances of these subunits in the two membrane preparations. Kv4.2, for example, is more abundant in ventricular than in atrial membranes, whereas Kv1.2 and Kv2.1 are higher in atrial membranes; Kv1.5 levels are comparable in the two preparations. In contrast to these results, nothing was detected in Western blots of atrial or ventricular membrane proteins with the anti-Kv1.4 antibody at concentrations that revealed intense labeling of a 97-kD protein in adult rat brain membranes. A very faint band was detected at 97 kD in the atrial and ventricular preparations when the anti-Kv1.4 antibody was used at a 5- to 10-fold higher concentration. The simplest interpretation of these results is that Kv1.4 is not an abundant protein in adult rat atrial or ventricular myocytes. Therefore, it seems unlikely that Kv1.4 plays an important role in the formation of functional depolarization-activated K+ channels in these cells. The relation(s) between the (other four) K+ channel subunits and the depolarization-activated K+ channels identified electrophysiologically in adult rat atrial and ventricular myocytes is discussed in the present study.
In myocardial tissues, a variety of K+ currents with differing kinetic and voltage-dependent properties have been described.1 2 3 Similar to findings in other cells, electrophysiological studies of heart cells have revealed that K+ channels are the most numerous and diverse channels present.1 2 3 This heterogeneity has a physiological significance in the myocardium, in that different types of K+ channels function in the control of resting membrane potentials, action potential waveforms, refractoriness, and automaticity.1 Differences in K+ channels also contribute to variations in the specific electrical properties of cells in different species or from different regions of the heart in the same species.3 4 5 In addition, these channels provide multiple potential targets for the actions of endogenous neurotransmitters, neurohormones, and intracellular second messengers and a variety of exogenous drugs known to modulate cardiac function.3 6
Depolarization-activated outward K+ currents contribute importantly in determining the height and the duration of the plateau phase of the action potential in cardiac cells. In the mammalian myocardium, various types of depolarization-activated outward K+ currents have been distinguished on the basis of differences in time- and voltage-dependent properties and pharmacological sensitivities.1 2 3 7 In previous studies completed on rat cardiac myocytes, for example, we have demonstrated the presence of multiple functionally distinct types of Ca2+-independent depolarization-activated K+ currents.4 8 9 In isolated adult rat ventricular myocytes, Ito and IK, with properties similar to those described in other cells, were distinguished.8 In atrial cells, three components of the total depolarization-activated outward K+ currents (termed IKf, IKs, and Iss) with properties distinct from those of both Ito and IK in ventricular cells have been identified.9 Recently, Van Wagoner and Lamorgese10 have reported that Iss is selectively attenuated by phenylephrine, suggesting that Iss reflects the activation of a novel outward current rather than incomplete inactivation of IKf and IKs.9 Therefore, the data accumulated to date4 8 9 10 suggest the expression of at least five functionally distinct types of Ca2+-independent depolarization-activated K+ currents in the adult rat heart.
Interestingly, molecular cloning of rat cardiac K+ channels has revealed considerably more diversity than expected on the basis of electrophysiological studies, if it is assumed that functional K+ channels are homomultimeric. To date, six different K+ channel cDNAs have been cloned from rat heart: four of these belong to the Shaker (Kv1) subfamily, one to the Shab (Kv2) subfamily, and one to the Shal (Kv4) subfamily.11 12 13 When in vitro synthesized mRNA from each of these cDNAs is injected into Xenopus oocytes, depolarization-activated K+ channels are expressed; these channels display properties similar but not identical to the channels present in atrial and ventricular myocytes. For example, expression of either Kv1.411 or Kv4.214 reveals transient K+ currents similar to Ito in ventricular myocytes.8 The detailed time- and voltage-dependent properties of Ito and expressed Kv1.4 or Kv4.2, however, are different. Similarly, although expression of Kv1.212 reveals a rapidly activating current similar to the currents in atrial myocytes, the atrial K+ currents do inactivate,9 whereas expressed Kv1.2 channels do not.12 Perhaps even more interesting is that quantitative RNase protection assays reveal that five K+ channel mRNAs are expressed at appreciable levels in both atrial and ventricular tissues.15 As discussed above, the electrophysiological evidence indicates the presence of five functionally distinct K+ channel types.4 8 9 10 The RNase protection data, however, also indicate that there are tissue-specific differences in expression.15 Total Shaker subfamily (Kv1.2, Kv1.4, and Kv1.5) mRNA, for example, is higher in atria than in ventricle, whereas Kv4.2 message levels are higher in ventricle.15 As a beginning step in defining the molecular composition of the functional K+ channels, we have used K+ channel subunit–specific antibodies to determine the distributions of these subunits in adult rat heart. The results reveal differential distributions of these K+ channel subunits and provide insights into the molecular correlates of the different types of K+ channels identified electrophysiologically in adult rat atrial and ventricular myocytes.
Materials and Methods
QT-6 cells16 were cultured in medium 199 (Earle’s medium) supplemented with 5% fetal calf serum, 10% tryptose phosphate broth, antibiotics (penicillin/gentamicin), 0.15% mycostatin, and 1% dimethyl sulfoxide. Atrial and ventricular myocytes were obtained by enzymatic (0.1% to 0.2% collagenase, Worthington type II) and mechanical dissociation of adult Long Evans rat hearts, as previously described.4 8 9 Isolated cells were plated on laminin-coated coverslips and maintained in serum-free medium 199. Adult heart cells and QT-6 cells were maintained in a 95% air/5% CO2 incubator at 37°C; all media and additives were obtained from GIBCO.
Polyclonal Antibodies Against K+ Channel Subunits
Polyclonal antisera against Kv1.5 were generated by immunizing rabbits with the recombinant fusion protein pGEX-K41C, which contained 61 amino acids corresponding to residues 542 to 602 in the C terminus of Kv1.5.17 Polyclonal antisera against Kv2.1 were obtained after immunization of rabbits with the fusion protein pGEX-KC, which contained 17 amino acids corresponding to residues 853 to 857 in the C terminus of Kv2.1.18 The anti-Kv1.5 and anti-Kv2.1 antibodies were affinity-purified by binding and eluting from nitrocellulose membranes to which the pGEX-K41C or pGEX-KC fusion peptides had been previously adsorbed.17 18 Affinity-purified anti-peptide antibodies generated against unique sequences in Kv1.2,19 20 Kv1.4, and Kv4.221 were obtained from Drs Morgan Sheng and Lily Jan (University of California, San Francisco).
Expression of K+ Channel Subunits in QT-6 Cells
Kv1.2 in pGemZf(+) was obtained from Dr Ernest Peralta (Harvard University). After sequencing to confirm the presence of a critical EcoRI at the 5′ end, Kv1.2 was subcloned into the polylinker of pBK-CMV at the EcoRI and Xba I sites. The insertion of the Kv1.2 was confirmed by diagnostic restriction digests, and the junctions between the pBK vector and Kv1.2 were verified by sequencing. The Kv1.4 and Kv1.5 cDNAs in pGem7 were obtained from Dr Michael Tamkun (Vanderbilt University). Because no convenient restriction enzyme sites were available for subcloning Kv1.4, the sequence was amplified by PCR before subcloning into the pBK-CMV polylinker. PCR primers were designed to bracket the 5′ and 3′ ends of the Kv1.4 sequence. The 5′ primer contained a Sal I site, and the 3′ primer contained a BamHI site. In addition to the BamHI site, the 3′ primer was designed to introduce a silent mutation in the DNA sequence encoding the last two amino acids. This alteration still codes for the natural amino acids at these positions; however, two bases of the DNA sequence were changed such that an Aat II restriction site was introduced. The fragment obtained by PCR was digested with Sal I and BamHI and subcloned into the pBK-CMV vector. Subcloning of Kv1.4 was confirmed by diagnostic restriction digests. Because amplification by PCR can introduce errors, the entire 1.96-Kb Kv1.4 fragment was sequenced. A single base change at position 478 was found: a thymidine rather than a cytosine was present, although this change does not alter the amino acid sequence, because both GAC and GAT code for Asp. The Kv1.5 coding sequence was subcloned into the polylinker of pBK-CMV at the Sac I and HindIII sites. Proper insertion of Kv1.5 was confirmed by diagnostic restriction digests and by sequencing the 5′ and 3′ junctions between Kv1.5 and the vector.
To evaluate the cross-reactivity of antibodies against the Shaker subfamily K+ channel subunits (Kv1.2, Kv1.4, and Kv1.5), QT-6 cells were plated on glass coverslips and transiently transfected with the Kv1.2, Kv1.4, or Kv1.5 cDNAs by using the calcium phosphate precipitation method.22 Transfection efficiencies were 25% to 30%. Approximately 72 hours after transfection, cells were fixed in 2% formaldehyde and 0.2% glutaraldehyde for 10 minutes on ice. After washing with PBS (pH 7.4), cells were incubated in blocking buffer (5% normal goat serum, 0.2% Triton X-100, and 0.1% NaN3 in PBS) for 1 hour at room temperature to block nonspecific sites and to permeabilize membranes. Cells expressing each one of the K+ channel subunits were then incubated separately with antibodies against Kv1.2, Kv1.4, and Kv1.5. After washing (three times) with PBS, cells were exposed to biotinylated goat anti-rabbit IgG (Vector Laboratories) for 1 hour at room temperature and again washed (three times) with PBS. This step was followed by a 1-hour exposure to the Vector ABC reagent (avidin coupled to biotinylated alkaline phosphatase H) at room temperature. After washing (three times) with PBS, positive antibody binding was visualized by exposing the cultures to the alkaline phosphatase substrate BCIP/NBT (Bio-Rad). The appearance of the reaction product was monitored under bright-field illumination, and reactions were quenched by the addition of PBS. Cells were dehydrated in ethanol, cleared in xylene, and mounted with Krystalon.
A protocol similar to that described above was used for immunohistochemistry on isolated adult rat atrial and ventricular myocytes, except that in this case cells were fixed with 4% paraformaldehyde in PBS overnight at 4°C. Although, in general, the alkaline phosphatase method was also used, in some experiments, the Vector ABC reagent kit using avidin coupled to biotinylated horse radish peroxidase was substituted, and tetramethylbenzidine was used as the horseradish peroxidase substrate. The appearance of reaction product was also monitored under bright-field illumination, and the reaction was quenched by the addition of PBS. For generating fluorescence images of antibody-labeled cells, a Cy3-conjugated goat anti-rabbit IgG (Chemicon) was used.
Adult Long Evans rat brain and heart membranes were prepared by using modified versions of protocols described previously by others. For the preparation of rat brain membranes, the protocols of Sheng et al19 and Hartshorne and Catterall23 were adapted. Brains were isolated at room temperature and frozen at −70°C until used; in a typical preparation, two brains were used. All procedures were performed at 4°C, and all solutions contained a battery of protease inhibitors (1 mmol/L iodoacetamide, 1 mmol/L 1,10-phenanthroline, 0.5 mmol/L pefebloc, and 1 μg/mL pepstatin) to minimize proteolysis. By use of a glass homogenizer, the tissue was homogenized in a buffer containing 0.32 mol/L sucrose and 5 mmol/L Tris (pH 7.4). Nuclei and debris were pelleted by centrifugation at 1000g for 10 minutes. The supernatant was then centrifuged at 100 000g for 1 hour; the resulting pellet was resuspended in 20 mmol/L Tris and 1 mmol/L EDTA and centrifuged at 40 000g for 20 minutes. The pellet from this spin was solubilized on ice for 1 hour in 20 mmol/L HEPES, 1 mmol/L EDTA, 10% glycerol, 120 mmol/L KCl, and 2% Triton X-100 at pH 7.4. Insoluble material was pelleted by centrifugation at 78 000g for 2.5 hours. The protein content of the resulting solubilized membranes was determined with a Bio-Rad protein assay kit. Solubilized membranes were aliquoted and stored frozen at −20°C until used.
Adult Long Evans rats were anesthetized with 5% halothane/95% O2 anesthesia, and the chest cavity was opened. Atrial appendages were removed, rinsed in PBS at 4°C to remove excess blood, and immediately frozen at −70°C. The heart was then transected below the major vessels; the lower half from the apex, including both the left and right ventricles, was removed, rinsed in PBS at 4°C, and frozen at −70°C. After repeated attempts to use the rat brain membrane protocol described above to prepare cardiac membranes proved unsuccessful, a modified version of the protocols described by Hosey and colleagues24 25 in studies of cardiac muscarinic receptors was used. All procedures were performed at 4°C, and all solutions contained the following protease inhibitors: 1 mmol/L iodoacetamide, 1 mmol/L 1,10-phenanthroline, 2 μg/mL aprotinin, 1 mmol/L benzamidine, 1 μg/mL pepstatin, and 0.5 mmol/L pefebloc. Tissues were minced, diluted in 10 vol of TE at pH 7.4, and homogenized with a Brinkman Polytron. Nuclei and debris were pelleted by centrifugation at 1000g for 10 minutes. This procedure was repeated, and the supernatants from both low-speed spins were pooled and centrifuged at 40 000g for 10 minutes. The pellet was resuspended in TE containing 0.6 mol/L KI and incubated on ice for 10 minutes. After centrifuging at 40 000g for 10 minutes, the resulting pellet was resuspended in TE and centrifuged again at 40 000g. To ensure that the KI was removed, this step was repeated. The final pellet was solubilized in TE and 2% Triton X-100 on ice for 1 hour. Insoluble material was centrifuged at 13 000g for 10 minutes. This protocol was also used to prepare membranes from isolated ventricular myocytes. The protein content of all of the solubilized membrane preparations was determined by using a Bio-Rad protein assay kit. Solubilized membranes were aliquoted and stored frozen at −20°C until used.
Aliquots of solubilized QT-6, brain, ventricular, and atrial membranes were fractionated by SDS-PAGE and subjected to Western blotting for identification of K+ channel subunits. Briefly, membrane proteins (85 μg) were fractionated on 10% SDS-polyacrylamide gels and transferred to Polyscreen PVDF membranes (DuPont). PVDF membranes strips were incubated in 0.2% I-Block (Tropix) in PBS containing 0.1% Tween 20 for 1 hour at room temperature. Membranes strips were then incubated overnight at 4°C in primary antibody solutions prepared in PBS containing 5% normal goat serum, 0.2% Triton X-100, and 0.1% NaN3. After washing (twice), the strips were incubated for 1.5 hours at room temperature in alkaline phosphatase–conjugated goat anti-rabbit IgG (Tropix) diluted 1:10 000 in blocking solution. Strips were washed (three times) in I-Block and twice in assay buffer (0.1 mol/L DEA and 1 mmol/L MgCl2, Tropix). Bound antibodies were detected by using the CPSD (Tropix) chemiluminescent alkaline phosphatase substrate.
Polyclonal Antibodies Against Kv1.2, Kv1.4, Kv1.5, Kv2.1, and Kv4.2
Antibodies against Kv1.2, Kv1.4, and Kv1.5 were tested for specificity and for potential cross-reactivity with the other Shaker family members by immunohistochemistry and Western analysis on QT-6 cells transfected with DNA constructs encoding the Kv1.2, Kv1.4, or Kv1.5 proteins. To evaluate antibody cross-reactivity, cells expressing one subunit were incubated separately with primary antibodies against each of the three subunits. For example, cells expressing the Kv1.2 protein were incubated with primary antibodies to Kv1.2, Kv1.4, or Kv1.5. As illustrated in Fig 1⇓, the anti-Kv1.2 antibody specifically stains Kv1.2-transfected cells (Fig 1A⇓) but not Kv1.4-transfected (Fig 1B⇓) or Kv1.5-transfected (Fig 1C⇓) cells. The anti-Kv1.4 (Fig 1D⇓ through 1F) and the anti-Kv1.5 (Fig 1G⇓ through 1I) antibodies also display subunit specificity. As would be expected, similar experiments completed when using the anti-Kv1.x antibodies on Kv4.2-transfected cells and the anti-Kv4.2 antibody on Kv1.x-transfected cells revealed no cross-reactivity across K+ channel subfamilies (data not shown).
In parallel experiments, proteins from QT-6 cells expressing Kv1.2, Kv1.4, or Kv1.5 were extracted and fractionated directly on SDS-polyacrylamide gels. Proteins were transferred from the gels to PVDF membranes and blotted with the various K+ channel subunit-specific antibodies. For example, the anti-Kv1.4 antibody was used to blot membranes containing proteins from the Kv1.2-expressing cells, the Kv1.4-expressing cells, the Kv1.5-expressing cells, and nontransfected QT-6 cells. These experiments revealed that the anti-Kv1.4 antibody specifically recognizes a 97-kD protein that is present in Kv1.4-transfected cells but not in untransfected QT-6 cells or in Kv1.2- or Kv1.5-transfected cells. Similar specificity was seen with the anti-Kv1.2 and anti-Kv1.5 antibodies. The anti-Kv1.2 antibody recognizes a protein of 75 kD in Kv1.2-transfected cells that is not present in either Kv1.4- or Kv1.5-transfected cells; the anti-Kv1.5 antibody labels a single band at 74 kD only in Kv1.5-transfected QT-6 cells (data not shown).
Immunohistochemical Identification of K+ Channel Subunits in Myocytes
Previous work has shown that the messages for Kv1.2, Kv1.4, Kv1.5, Kv2.1, and Kv4.2 are present in the adult rat ventricles and atria.13 15 To determine if these subunits could also be detected at the protein level, two types of experiments were performed. In the first sequence of experiments, isolated myocytes were fixed and probed for the presence of K+ channel subunits immunohistochemically. In addition, adult rat atrial and ventricular membranes were isolated, and membrane proteins were separated on SDS gels and immunoblotted with the polyclonal antibodies against each of the subunits. The results of a typical immunohistochemical experiment on isolated adult rat ventricular myocytes are presented in Fig 2⇓. As is evident, several of the K+ channel subunit–specific antibodies label isolated adult rat ventricular myocytes. The label is particularly intense with the anti-Kv4.2 (Fig 2E⇓) and anti-Kv1.2 (Fig 2A⇓) antibodies. With the anti-Kv1.5 (Fig 2C⇓) and anti-Kv2.1 (Fig 2D⇓) antibodies, the labeling, although somewhat less intense and more variable than with anti-Kv4.2 or anti-Kv1.2, is significantly different from the controls (Fig 2F⇓). In marked contrast to these results, the anti-Kv1.4 (Fig 2B⇓) antibody does not label isolated adult rat ventricular myocytes; the appearance of cells exposed to anti-Kv1.4 (Fig 2B⇓) is not significantly different from background (Fig 2F⇓) staining in cells that were subjected to all the steps in the staining protocol, except that the primary antibody was omitted. Interestingly, positive antibody labeling (Fig 2A⇓, 2C⇓, 2D⇓, and 2E⇓) appears to be highly concentrated in the plasma membranes of isolated myocytes, consistent with the participation of the antigens (K+ channel subunits) recognized by these antibodies in the formation of functional depolarization-activated K+ channels (see below).
Qualitatively similar results were obtained in isolated adult rat atrial myocytes (Fig 3⇓). Labeling of atrial cells was most intense with the anti-Kv1.2 (Fig 3A⇓) and anti-Kv4.2 (Fig 3I⇓) antibodies. Although somewhat weaker, the anti-Kv1.5 antibody also consistently labeled atrial cells (Fig 3E⇓). Also similar to the results obtained in ventricular cells, the appearance of atrial cells exposed to the anti-Kv1.4 antibody (Fig 3C⇓) is not significantly different from background. However, in contrast to ventricular cells (Figs 2D⇑ and 3H⇓), the Kv2.1 antibody gives only very weak labeling of atrial cells (Fig 3G⇓).
As noted above, the labeling patterns seen with anti-Kv1.2, anti-Kv1.5, anti-Kv2.1, and anti-Kv4.2 (Figs 2⇑ and 3⇑) suggested that the antigens (K+ channel subunits) recognized by these antibodies are highly concentrated in the plasma membranes of adult rat atrial and ventricular myocytes. To explore this point further, laser scanning confocal images of isolated myocytes labeled with K+ channel subunit–specific antibodies were obtained. In these experiments, a Cy3-conjugated goat-anti-rabbit IgG secondary antibody was used. As illustrated in Fig 4⇓, anti-Kv4.2 labeling is highly concentrated in the plasma membranes of isolated adult rat ventricular myocytes. Similar results were obtained with the anti-Kv1.2, anti-Kv1.5, and anti-Kv2.1 antibodies. Interestingly, the labeling intensity appears to be highest in the intercalated disk regions of the cells (Fig 4⇓, arrows).
Western Blots of Rat Ventricular and Atrial Membrane Proteins
Western blots were conducted on rat atrial and ventricular membrane proteins after fractionation on SDS-polyacrylamide gels and transfer to PVDF membranes (see “Materials and Methods”). Typical results obtained by use of the antibodies against the Shaker subfamily subunits Kv1.2, Kv1.4, and Kv1.5 are presented in Fig 5⇓. The anti-Kv1.2 antibody recognizes a single band at 75 kD in both atrial and ventricular membranes (Fig 5A⇓), although the signal is clearly stronger in atrial than in ventricular membranes. The molecular mass (75 kD) of this band is similar to that of the protein identified by anti-Kv1.2 in Kv1.2-transfected QT-6 cells (see above). A band at approximately the same molecular mass is also seen in brain membranes (Fig 5A⇓), although as reported previously,19 20 the band identified by anti-Kv1.2 in brain appears “fuzzy,” extending over the molecular mass range of 75 to 85 kD. This observation suggests that there are multiple isoforms of Kv1.2 in the brain. If this interpretation is correct, the difference between the appearances of the bands in atrial/ventricular and brain membranes may suggest that multiple isoforms of Kv1.2 do not exist in the adult rat heart. In experiments completed on both brain and heart membrane proteins, however, the labeling of the 75-kD (in atria/ventricles) and the 75- to 85-kD (in brain) bands are specifically blocked (Fig 5A⇓) when the antibody is preincubated with the peptide against which the antibody was generated.19
A single protein band at 75 kD is recognized by the anti-Kv1.5 antibody in rat ventricular membranes, whereas two bands, at 75 kD and 60 kD, are identified with this antibody in atrial membranes (Fig 5B⇑). A 76-kD band was reportedly recognized by anti-Kv1.5 in immunoblots of GH3 cell proteins.17 Similar to the findings in the hippocampus,26 however, anti-Kv1.5 recognizes a much smaller protein (60 kD) in brain membranes (Fig 5B⇑). Both the 60- and 75-kD bands identified in immunoblots of heart membranes (and the 60-kD band in brain membranes) were eliminated (Fig 5B⇑) when the antibody was preincubated with the fusion protein against which the antibody was generated.16 Taken together, these results suggest that there may be multiple isoforms of Kv1.5 and also that these isoforms are differentially distributed in adult rat heart and brain (see “Discussion”). Alternatively, because the predicted molecular mass of the Kv1.5 protein is >60 kD,13 it is certainly also possible that the 60-kD band in adult rat brain and atria reflects breakdown of the intact 75-kD Kv1.5 protein (see “Discussion”).
Interestingly, and in contrast to the findings with anti-Kv1.2 and anti-Kv1.5 antibodies described above, nothing was detected on Western blots of adult rat atrial and ventricular membrane proteins with the anti-Kv1.4 antibody (Fig 5C⇑). However, in parallel experiments on rat brain membrane proteins, a prominent band was detected at 97 kD; this band was eliminated (Fig 5C⇑) when anti-Kv1.4 was preincubated with the N-terminal peptide against which the antibody was generated.19 20 When the anti-Kv1.4 antibody was used at 5- to 10-fold higher concentrations, Western blots of both atrial and ventricular membrane proteins revealed a very faint band at 97 kD in both samples (data not shown). Although faint, the 97-kD band in adult rat atria and ventricles in these experiments seems certain to reflect (very low) expression of Kv1.4, because the molecular mass of the protein is the same as that identified in brain (Fig 5C⇑; see also References 19 and 2119 21 ). In addition, the 97-kD bands were eliminated when the antibody was preincubated with the peptide against which it was generated.19 20 The simplest interpretation of these results is that Kv1.4 is expressed at barely detectable levels in adult rat atria and ventricles (see “Discussion”).
In parallel experiments, Western blots of rat atrial and ventricular membrane proteins revealed that expression of Kv2.1 and Kv4.2 was readily detected (Fig 6A⇓). The anti-Kv2.1 antibody, for example, specifically labels a single protein band at 130 kD in both atrial and ventricular membranes (Fig 6A⇓), although the intensity of the labeled band does appear to be higher in the atrial than in the ventricular membranes (see “Discussion”). Also of interest is the fact that a single band at 130 kD is recognized by anti-Kv2.1 in heart membranes (Fig 6A⇓), whereas multiple bands (116 to 130 kD) are identified by this antibody in rat brain (Fig 6A⇓; see also Reference 2727 ), suggesting that different isoforms of Kv2.1 are expressed in adult rat brain but not in adult rat heart (see “Discussion”).
The anti-Kv4.2 antibody specifically labels a single band at 74 kD in adult rat heart membranes (Fig 6B⇑), although this protein is more abundant in ventricular than in atrial membranes. The molecular mass of the protein identified by the anti-Kv4.2 antibody in brain membranes is the same as that in the heart, and in all cases, labeling is eliminated when the antibody is preincubated with the peptide against which the antibody was generated (Fig 6B⇑).
The experiments described above demonstrate the presence of Kv1.2, Kv1.5, Kv2.1, and Kv4.2 in membranes prepared from adult rat atria and ventricles. However, it could clearly be argued that the results obtained may not reflect the distributions of these K+ channel subunits in cardiac myocytes because of the presence of other cell types, such as smooth muscle and endothelial cells, in the tissue that was harvested, homogenized, and used for the preparation of membranes. Therefore, in separate experiments adult rat ventricular myocytes were isolated,4 8 9 membranes were prepared, and membrane proteins were fractionated on SDS gels and transferred to PVDF membranes. Western blots with the anti-Kv1.2, anti-Kv1.4, anti-Kv1.5, anti-Kv2.1, and anti-Kv4.2 antibodies revealed results (Fig 7⇓) that were indistinguishable from those obtained in proteins extracted from whole ventricles (Figs 5⇑ and 6⇑). Specifically, intense labeling was seen with the anti-Kv4.2 (lane 7D) and the anti-Kv1.2 (lane 7A) antibodies; somewhat weaker, although clearly positive, labeling was observed with the anti-Kv1.5 (lane 7B) and anti-Kv2.1 (lane 7C) antibodies. The results obtained with the anti-Kv1.4 antibody were very similar to those obtained in Western blots of membrane proteins isolated from whole adult rat atria and ventricles (Fig 5C⇑). Specifically, no labeling was detected in Western blots of adult rat ventricular myocyte membrane proteins with the anti-Kv1.4 antibody diluted at 1:500, although a very faint band at 97 kD was detected with anti-Kv1.4 at a 10-fold higher concentration (data not shown).
Differential Expression of K+ Channel Subunits in Atrial and Ventricular Membranes
The results of the present study demonstrate that four of the five K+ channel subunits shown to be expressed at the mRNA level in the adult rat heart13 15 are also readily detected at the protein level; these are Kv1.2, Kv1.5, Kv2.1, and Kv4.2. The remaining K+ channel subunit, Kv1.4, although cloned from a rat heart cDNA library13 and clearly evident in Northern blots13 and in RNase protection assays,15 was undetectable in isolated myocytes immunohistochemically and barely detectable in immunoblots of adult rat ventricular or atrial membrane proteins. It seems unlikely that the weak/absent signals with the anti-Kv1.4 antibody on cardiac cells/membranes reflect the experimental conditions used, because all of the other antibodies gave unambiguous results. In addition, the anti-Kv1.4 antibody works well in Western blots of rat brain membrane proteins (Fig 7⇑) and in immunohistochemical experiments in QT-6 cells transfected with Kv1.4 (Fig 1B⇑). Finally, it is important to emphasize that enzyme-linked immunosorbent assays revealed that the anti-Kv1.4 antibody is substantially more sensitive than either the anti-Kv1.2 or the anti-Kv4.2 antibody, both of which strongly labeled myocyte membrane proteins in immunohistochemical experiments and in Western blots. Therefore, the absence of signals in the heart samples cannot be explained by problems with the antibody. Given this situation, we are forced to conclude that Kv1.4 is not an abundant protein in adult rat atrial or ventricular myocytes and, further, that this subunit likely does not play an important role in the formation of functional depolarization-activated K+ channels in these cells (see below).
Of the four remaining K+ channel subunits found to be present in adult rat heart, differences in their distributions were evident in immunoblots. For example, Kv4.2 levels are higher in ventricular than in atrial membranes, whereas Kv1.2 is more abundant in atria. However, in contrast to the brain,19 20 there do not appear to be multiple isoforms of Kv1.2 in the adult rat heart (Fig 5A⇑). Kv1.5 appears to be similar in the two tissues, although it is of interest to note that there are two bands detected by the anti-Kv1.5 antibody in adult rat atria but not in adult rat ventricles (Fig 5B⇑). This may reflect the expression of two distinct Kv1.5 isoforms in atria. Alternatively, it is possible that the 60-kD band identified by anti-Kv1.5 in adult rat atria is a breakdown product of the intact 75-kD Kv1.5 protein. Further experiments will be necessary to distinguish between these possibilities. The relative distributions of Kv1.2, Kv1.5, and Kv4.2 revealed by the protein data presented here agree closely with the recently published RNA data.14 For Kv2.1, however, the message data show that the level of Kv2.1 mRNA in ventricle is twice that in atria, whereas the Kv2.1 protein is more abundant in atria (Fig 6A⇑).
Relation of K+ Channel Subunits to Functional K+ Channels in Ventricular Myocytes
A motivation for the proposed experiments was to probe the likely gene products underlying the depolarization-activated K+ channels previously characterized electrophysiologically in adult rat atrial and ventricular myocytes.4 8 9 10 As noted above, Northern blots13 and quantitative RNase protection assays15 have revealed the expression of Kv1.2, Kv1.4, Kv1.5, Kv2.1, and Kv4.2 in adult rat atria and ventricles. Of these, Kv1.4 and Kv4.2 seemed the most likely candidates for the protein underlying Ito in adult rat ventricular myocytes, primarily because expression of Kv1.411 or Kv4.214 in Xenopus oocytes reveals transient K+ currents that are similar (although not identical) to rat ventricular Ito.8 Alternatively, because heteromultimeric K+ channels can and do form when in vitro–synthesized mRNAs encoding K+ channel subunits belonging to the same subfamily are coinjected into oocytes,28 29 30 it also seemed possible that heteromultimeric assembly of Kv1.4 with Kv1.2 and/or Kv1.5 might underlie Ito. However, in the present study we demonstrate that Kv1.4 is barely detectable in the membranes of adult rat ventricular myocytes both by immunohistochemistry and Western blot analysis, suggesting that Kv1.4 does not contribute to rat ventricular Ito. Kv4.2, on the other hand, is abundant in ventricular membranes, leading to the obvious suggestion that Kv4.2 underlies Ito. This suggestion was made previously by Dixon and McKinnon15 on the basis of their findings that Kv4.2 message levels are higher in ventricle than atrium and vary throughout the thickness of the ventricular wall, as found for Ito (J.M. Nerbonne, unpublished data, 1994). Because Kv4.1 message levels are insignificant in adult rat ventricle,15 one can further speculate that if Kv4.2 underlies rat ventricular Ito, these channels are homomultimeric (composed of four copies of Kv4.2).31 However, it is certainly possible that there are also accessory β subunits associated with Kv4.2 in vivo and that these contribute to determining the properties of Ito channels. It has been demonstrated that there are accessory β subunits of brain voltage-gated K+ channels of the Shaker32 and Shab18 subfamilies. Recently, two Shaker subfamily β subunits were cloned from rat brain,33 and coexpression experiments revealed that the presence of a β subunit alters the kinetic properties of some members (eg, Kv1.1) of this subfamily. Further experiments will be necessary to explore the possibility that there are also β subunits that associate with Kv4.2 (or other K+ channel subunits) in the heart.
The finding that Kv1.2 is also abundant in adult rat ventricle suggests that this subunit might also be considered a reasonable candidate for Ito. However, heterologous expression of Kv1.2 yields rapidly activating noninactivating K+ currents with properties quite different from rat ventricular Ito. By analogy to the recent findings in brain noted above,33 it could certainly be suggested that there are β subunits that associate with homomultimeric Kv1.2 (or heteromultimeric Kv1.2-Kv1.5) channels to yield functional rat ventricular Ito channels. The strongest argument against a role for Kv1.2 in rat ventricular Ito is that expressed Kv1.2 currents are sensitive to nanomolar concentrations of K+ channel toxins, such as dendrotoxin,34 whereas dendrotoxin at concentrations up to 1 μmol/L has no effect on Ito (or IK) in adult rat ventricular myocytes (H. Xu and J.M. Nerbonne, unpublished data, 1994). Thus, we suggest that Kv4.2 most likely underlies Ito in adult rat ventricular myocytes.
If the hypothesis that Kv4.2 underlies functional Ito channels is correct, then the obvious questions are as follows: What are the functions of the remaining three subunits, and (specifically) what underlies IK in ventricular myocytes? Heterologous expression of either Kv1.212 or Kv1.535 yields rapidly activating K+ currents that are sensitive to 4-aminopyridine and insensitive to TEA. However, IK in rat ventricular myocytes is TEA sensitive.8 Also, as noted above, expressed Kv1.2 currents are sensitive to nanomolar concentrations of K+ channel toxins,34 whereas IK in rat ventricular myocytes is insensitive to dendrotoxin. Heterologous expression of Kv2.1, on the other hand, reveals slowly activating TEA-sensitive K+ currents36 similar to IK in adult rat ventricular myocytes.8 Therefore, of the three K+ channel subunits (other than Kv4.2) expressed in adult rat ventricular myocytes, Kv2.1 seems the likely candidate for IK. If correct, then the functional roles of Kv1.2 and Kv1.5 are unclear. If it is assumed that these subunits contribute to the formation of functional channels, it seems reasonable to suggest that there may be other (ie, in addition to Ito and IK) channels in these cells. Further experiments designed to explore each of these hypotheses directly are warranted. Recently, we have found that Ito density increases severalfold during the period of postnatal development from day 5 to adult and that during the same period IK density decreases (H. Xu and J.M. Nerbonne, unpublished data, 1994). Experiments are under way to determine if the (protein) levels of the various K+ channel subunits also change during this developmental period. These studies should provide further insights into the likely subunits that contribute to functional Ito and IK channels.
Relation of K+ Channel Subunits to Functional K+ Channels in Atrial Myocytes
Similar questions arise concerning the relation between the K+ channel subunits expressed and the K+ channels characterized electrophysiologically in adult rat atrial myocytes. Previous work has revealed the presence of three kinetically distinct current components, termed IKf, IKs, and Iss, in these cells.9 All three components are sensitive to 4-aminopyridine and insensitive to TEA.9 IKf and IKs display different rates of inactivation and rates of recovery from inactivation; Iss is noninactivating.9 Recently, Van Wagoner and Lamborgese10 demonstrated that only Iss is attenuated on application of phenylephrine, suggesting that Iss reflects the activation of a novel conductance pathway rather than incomplete inactivation of IKf and/or IKs.9 Of the heterologously expressed cloned K+ channels, IKf, IKs, and Iss most closely resemble Kv1.212 and Kv1.535 in terms of pharmacology and kinetics. Recently, it was reported that IKs, but not IKf or Iss, is sensitive to nanomolar concentrations of dendrotoxin.10 This observation is intriguing and lends further support to the suggestion that IKf, IKs, and Iss are functionally distinct populations of K+ channels.9 10 More important for the present discussion, however, this result suggests that different molecular entities likely underlie IKf, IKs, and Iss. Of the K+ channel subunits found in adult rat atrium, only Kv1.2 forms dendrotoxin-sensitive channels when expressed in Xenopus oocytes.34 Therefore, it seems reasonable to suggest that Kv1.2, either as a homomultimer or a (Kv1.2-Kv1.5) heteromultimer, underlies IKs. The time- and voltage-dependent properties of heterologously expressed Kv1.5 channels35 are similar to those of Iss.9 This observation together with the relatively low abundance of this protein in adult rat atria (Fig 5B⇑) suggests that Kv1.5 might contribute to adult rat atrial Iss.
In contrast, it seems unlikely that either Kv1.2 or Kv1.5 contributes to IKf in adult rat atrial myocytes. A more likely possibility appears to be Kv4.2, because this subunit is abundant in atrial membranes and because heterologous expression of Kv4.2 reveals rapidly activating 4-aminopyridine–sensitive K+ currents.14 As discussed above, however, expressed Kv4.2 currents also inactivate rapidly, more closely resembling ventricular Ito8 than atrial IKf.9 It is possible that coexpression of β subunits18 32 33 or posttranslational modifications37 alter the rates of inactivation of Kv4.2, thereby leading to channels with different properties in atrial (IKf) and ventricular (Ito) cells. If these various hypotheses concerning the relations between expressed subunits and functional atrial K+ channels (ie, that Kv1.2 [either alone or with Kv1.5] underlies IKs, Kv1.5 underlies Iss, and Kv4.2 underlies IKf in rat atrial myocytes) are correct, then the functional role of Kv2.1 in atrial cells is unclear. Similar to the comments above concerning our findings in ventricular cells, the simplest interpretation is that if Kv2.1 does contribute to the formation of functional channels in atrial cells, then there must be other (ie, in addition to IKf, IKs, and Iss) channels in these cells. Additional experiments aimed at exploring this possibility and further probing the molecular compositions of the functional K+ channels in atrial and ventricular myocytes will be necessary to resolve these points.
Selected Abbreviations and Acronyms
|IK||=||delayed rectifier K+ current|
|IKf||=||rapidly activating outward K+ current|
|IKs||=||slowly activating outward K+ current|
|Iss||=||steady state noninactivating atrial current|
|Ito||=||transient outward K+ current|
|PCR||=||polymerase chain reaction|
|TE||=||10 mmol/L Tris and 1 mmol/L EDTA buffer|
This study was supported by grants from the Monsanto-Searle/Washington University Biomedical Research Program, the National Institutes of Health, and the Council for Tobacco Research. Dr Trimmer is an Established Investigator of the American Heart Association. We thank Terry Der for technical assistance in the preparation of isolated adult rat atrial and ventricular myocytes and Dr Haodong Xu for assistance with atrial and ventricular isolations and for many helpful comments and discussions. We also thank Drs Morgan Sheng and Lily Y. Jan for providing us with the anti-Kv1.2, anti-Kv1.4, and anti-Kv4.2 antibodies and Dr Michael Tamkun for providing the Kv1.4 and Kv1.5 cDNAs.
This paper is dedicated to the memory of John P. Merlie, who died unexpectedly on May 27, 1995.
- Received November 28, 1994.
- Accepted April 24, 1995.
- © 1995 American Heart Association, Inc.
Cohen IS, Datyner NB, Gintant GA, Kline RP. Time-dependent outward currents in the heart. In: Fozzard HA, Jennings RB, Haber E, Katz AM, Morgan HE, eds. The Heart and Cardiovascular System. New York, NY: Raven Press Publishers; 1986:637-670.
Boyle WA, Nerbonne JM. A novel type of depolarization-activated K+ current in isolated adult rat atrial myocytes. Am J Physiol. 1991;260:H1236-H1247.
Gadsby DC. Effects of β adrenergic catecholamines on membrane currents in cardiac cells. In: Rosen MR, Janse MJ, Wit AL, eds. Cardiac Electrophysiology: A Textbook. Mt Kisco, NY: Futura Publishing Co; 1990:857-876.
Apkon M, Nerbonne JM. Characterization of two distinct depolarization-activated K+ currents in isolated adult rat ventricular myocytes. J Gen Physiol. 1991;97:973-1011.
Boyle WA, Nerbonne JM. Two functionally distinct 4-aminopyridine-sensitive outward K+ currents in adult rat atrial myocytes. J Gen Physiol. 1992;100:1047-1061.
Van Wagoner DR, Lamborgese M. Phenylephrine blocks an IK current in adult rat and human atrial myocytes. Circulation. 1994;90(suppl I):I-146. Abstract.
Paulmichl M, Nasmith P, Hellmiss R, Reed K, Boyle WA, Nerbonne JM, Peralta EG, Clapham DE. Cloning and expression of a cardiac delayed rectifier potassium channel RAK. Proc Natl Acad Sci U S A. 1991;88:7892-7895.
Roberds SL, Tamkun MM. Cloning and tissue-specific expression of five voltage gated potassium channel cDNAS expressed in rat heart. Proc Natl Acad Sci U S A. 1991;88:1798-1802.
Dixon JE, McKinnon D. Quantitative analysis of potassium channel mRNA expression in atrial and ventricular muscle of rats. Circ Res. 1994;75:252-260.
Trimmer JS. Immunological identification and characterization of a delayed rectifier K+ channel polypeptide in rat brain. Proc Natl Acad Sci U S A. 1991;88:10764-10768.
Sheng M, Tsaur M-L, Jan YN, Jan LY. Contrasting subcellular localization of the Kv1.2 K+ channel subunit in different neurons of rat brain. J Neurosci. 1994;14:2408-2417.
Hartshorne RP, Catterall WA. The sodium channel from rat brain. J Biol Chem. 1984;259:1667-1675.
Hosey MM, Fields JZ. Quantitative and qualitative differences in muscarinic cholinergic receptors in embryonic and newborn chick hearts. J Biol Chem. 1981;256:6295-6399.
Mayanil CSK, Richardson RM, Hosey MM. Subtype-specific antibodies for muscarinic cholinergic receptors, I: characterization using transfected cells and avian and mammalian cardiac membranes. Mol Pharmacol. 1991;40:900-907.
Maletic-Savatic M, Lenn NJ, Trimmer JS. Differential spatiotemporal expression of K+ channel polypeptides in rat hippocampal neurons developing in situ and in vitro. J Neurosci. 1995;15:3840-3851.
Christie MJ, North RA, Osborne PB, Douglass J, Adelman JP. Heteropolymeric potassium channels expressed in Xenopus oocytes from cloned subunits. Neuron. 1990;2:405-411.
Swanson R, Marshall J, Smith JS, Williams JB, Boyle MB, Folander K, Luneau CJ, Antanavage J, Oliva C, Buhrow SA, Bennett C, Stein RB, Kaczmarek LK. Cloning and expression of cDNA and genomic clones encoding three delayed rectifier potassium channels in rat brain. Neuron. 1990;4:929-939.