Isoform-specific regulation of the sodium pump by α- and β-adrenergic agonists in the guinea-pig ventricle

Abstract

1. Guinea-pig ventricle was used in the RNase protection assays to determine which α-isoforms of the Na⁺-K⁺ pumps are present, and ventricular myocytes were used in whole cell patch clamp studies to investigate the actions of α- and β-adrenergic agonists on Na⁺-K⁺ pump current.

2. RNase protection assays showed that two isoforms of the α-subunit of the Na⁺-K⁺-ATPase are present in guinea-pig ventricle. The mRNA for the α₁-isoform comprises 82 % of the total pump message, the rest being the α₂-isoform.

3. We have previously shown that β-adrenergic agonists affect Na⁺-K⁺ pump current (I_p) through a protein kinase A (PKA)-dependent pathway. We now show that these β-effects are targeted to the α₁-isoform of the Na⁺-K⁺ pumps.

4. We have also previously shown that α-adrenergic agonists increase I_p through a protein kinase C (PKC)-dependent pathway. We now show that these α-isoform effects are targeted to the α₂-isoform of the Na⁺-K⁺ pumps.

5. These results suggest the effects of adrenergic activation on Na⁺-K⁺ pump activity in the heart can be regionally specific, depending on which α-isoform of the Na⁺-K⁺ pump is expressed.

In cardiac myocytes Na⁺ enters and K⁺ leaves the cell by moving down their electrochemical gradients during each action potential. An active transport process, the Na⁺-K⁺ exchange pump, helps restore these ion gradients by transporting three Na⁺ ions out of the cell and returning two K⁺ ions to the cell interior at the cost of one ATP molecule. Molecular studies have revealed that the pump is composed of one α- and one β-subunit which combine to form the functional transporter. The α-subunit carries out the ATPase activity and is responsible for binding cardiac glycosides which can prevent ion transport. Three different α-subunit isoforms have been discovered, one of which, the α₁, has a distinctly lower affinity for glycosides than do the other two (Sweadner, 1989).

We previously reported that two functionally distinct types of Na⁺-K⁺ pumps co-exist in guinea-pig ventricular myocytes (Gao et al. 1995). Thus, we define total pump current (I_T) as the sum of the low dihydro-ouabain (DHO)-affinity pump current (I_l) and the high DHO-affinity pump current (I_h): I_T=I_l+I_h. These two types had about a 100-fold difference in affinity (0·75 μM and 72 μM) for DHO and differed in their responses to external pH and K⁺. The difference in DHO affinity suggested that the low-affinity pumps might be the α₁-isoform and the high-affinity pumps the α₂- or α₃-isoform. Berrebi-Bertrand et al. (1991) reported that the α₁- and α₂-isoforms are present in these cells, based on separation by glycoside affinity and molecular weight. However, Sweadner et al. (1994) were unable to detect any α₂-isoform, using an antibody raised against rat axolemma that was able to detect α₂ in guinea-pig brain. McDonough et al. (1996) demonstrated the presence of α-subunits of Na⁺-K⁺ pumps in both the sarcolemma and T-tubule of guinea-pig myocytes, although their antibody did not attempt to distinguish among isoforms. Thus, the α-subunits are distributed on both sarcolemma and T-tubules but the molecular basis of the two types of pumps remains in question. The patch clamp data in Gao et al. (1995) were consistent with the high DHO-affinity-type pumps being a small fraction of the total pump protein, even though they represented a significant fraction of total pump current. If so, the antibody used by Sweadner et al. (1994) might not have been sufficiently specific to identify the α₂-isoform over non-specific binding. In the present study, we used RNase protection assays to identify isoforms of the Na⁺-K⁺ pump present in these myocytes. This assay is highly specific and quite sensitive, although RNA levels do not necessarily reflect the amount of protein. Nevertheless, the absence of mRNA for the α₃-isoform strongly suggests that this protein is not expressed, whereas the amounts of mRNA for the α₁- and α₂-isoforms correlates well with the functional data on pump current.

Previous studies (Shah et al. 1988; Wilde & Kleber, 1991; Gao et al. 1992; Wang et al. 1998) have demonstrated that catecholamines, which activate either α- or β-adrenergic receptors, alter Na⁺-K⁺ pump current. Gao et al. (1994, 1995) showed that all of the effects of β-adrenergic activation on I_T require protein kinase A (PKA)-mediated phosphorylation. In contrast, Wang et al. (1998) showed that all of the effects of α-adrenergic activation on I_T required protein kinase C (PKC)-mediated phosphorylation. In these earlier studies, we recorded the effects of α- or β-adrenergic activation on total pump current. However, given the presence of two types of Na⁺-K⁺ pumps in these cells, the PKA and PKC pathways of modulating total pump current may have different target proteins. The purpose of the current study was to test this possibility and determine the molecular identity of the two types of pumps.

METHODS

Patch clamp experiments

Electrophysiology

Guinea-pig ventricular myocytes were isolated essentially as described in our previous studies (Gao et al. 1992). Male guinea-pigs weighing 300-500 g (1-2 months old) were killed with sodium pentobarbitone (1 ml of 360 mg ml⁻¹) solution by peritoneal injection. The heart was then removed. All of the procedures are consistent with the guidelines laid down by the institution's animal welfare committee. The isolated cells were studied using the whole cell patch clamp technique and an Axopatch 1A amplifier (Axon Instruments). Initial resistances of patch pipettes were 1-2 MΩ and seal resistances were 10-20 GΩ. After achieving a seal and rupturing the underlying cell membrane, we waited at least 5 min for the pipette and cell solutions to come to steady state. The recording temperature was 25-32°C and constant to within 0·5°C within any experiment. We have previously reported that the inhibitory effect of β-adrenergic activation on I_T is not voltage dependent at low [Ca²⁺]_i (15 nM). However, at high [Ca²⁺]_i (1·4 μM), β-adrenergic activation increases I_T at negative voltages while leaving I_T unchanged at positive potentials (Gao et al. 1996). Therefore, in the present study, we voltage clamped the cells to -60 mV to study the inhibitory (low [Ca²⁺]_i) and stimulatory (high [Ca²⁺]_i) effects of β-adrenergic activation on the two different isoforms of the Na⁺-K⁺ pump. Wang et al. (1998) reported that α-adrenergic activation increased I_T and this stimulation was not voltage dependent at either low or high [Ca²⁺]_i. Thus, we clamped the cells to 0 mV to investigate which isoform of Na⁺-K⁺ pump was stimulated by α-adrenergic activation. Because pump current is larger at 0 mV than -60 mV, voltage clamping to 0 mV increased our signal-to-noise ratio and gave us better resolution. Na⁺-K⁺ pump current was measured as the decrease in outward current produced by extracellular application of the cardiac glycoside dihydro-ouabain (DHO, 1 mM or 5 μM) or strophanthidin (0·5 mM), two specific reversible inhibitors of Na⁺-K⁺ pump activity in these cells. DHO at 1 mM or strophanthidin at 0·5 mM (Gadsby & Nakao, 1989) could block roughly 96 % of I_T, and 5 μM DHO blocked 89 % of I_h while blocking only 9 % of I_l (Gao et al. 1995). We equate effects on I_T with the 1 mM DHO blockade and effects on I_h with the 5 μM DHO blockade. The properties of I_l are then proportional to the difference, I_l=I_T - I_h. All patch clamp data were displayed on a digital storage oscilloscope and recorded on computer disk for later analysis. For each experimental treatment, both control and test pump currents were obtained from the same cell. The ratio of test to control current was calculated to normalize out cell-to-cell variability in size and pump density. All values are given as means ± standard deviation. Statistical significance between control and test groups was tested with Student's t test for paired comparisons.

Solutions

Our standard pipette solution contained (mM): 5 Na₂-ATP, 0·2 GTP, 50 sodium aspartate, 20 potassium aspartate, 30 CsOH, 20 TEACl, 5 Hepes, 5 MgSO₄, 11 EGTA, 10 glucose, 1 or 10 CaCl₂, pH = 7·2. The 60 mM [Na⁺] in the pipette solution largely saturates the internal Na⁺ site of the Na⁺-K⁺ pump, and so guarantees that small increases in [Na⁺]_i that occur during partial pump blockade will not significantly change I_p. Free [Ca²⁺] in the pipette solution containing 11 mM EGTA with 1 and 10 mM CaCl₂ were 15 nM and 1·4 μM, respectively, according to calculations using the SPECS program (Fabiato, 1988) and measurements using BAPTA as a calcium indicator. Free [Mg²⁺] in the pipette was 52 μM given by calculation using the above program. The cells were superfused with an external solution containing (mM): 137·7 NaCl, 2·3 NaOH, 1 MgCl₂, 10 glucose, 5 Hepes, 4·6 KCl, 1·8 CaCl₂, 0·5 BaCl₂, 0·2 CdCl₂, pH = 7·4. Various agents added to the external solution were: 1 mM or 5 μM DHO, or 0·5 mM strophanthidin, 0·5 μM isoprenaline and 10 μM noradrenaline. Stock solutions of DHO (85 mM) or strophanthidin (0·5 M), isoprenaline (1 mM) and noradrenaline (10 mM) were diluted to the final concentrations just before the experiments. The external solution was changed by switching to a new superfusate through an hydraulic switch.

RNase protection assay

Cloning of probes

In order to perform the RNase protection assays it was necessary to clone probes that were specific not only for the three Na⁺-K⁺ pump isoforms (α_1,α₂ and α₃), but also specific for guinea-pig. Since sequence information for guinea-pig Na⁺-K⁺ pumps was not available, a degenerate PCR cloning strategy was employed, utilizing sequence information for rat and human Na⁺-K⁺ pump isoforms from various databases. Degenerate PCR primers were synthesized and the same primer set generated both the α₁- and α₃-isoform probes. The upstream primer was the amino acid sequence, KNCLVK and the oligonucleotide sequence was (FASTA format) 5′-AARAAYGYYTNGTNAAR-3′. The downstream primer was generated from the amino acid sequence FNTTNKY, and the oligonucleotide sequence was 5′-RTAYTTRTTNGTRTTRAA-3′. A separate primer set was used to make the α₂-isoform probe. The upstream primer was the amino acid sequence, PEWVKF and the oligonucleotide sequence was 5′-CCNGARTGGGTNAARTTY-3′. The downstream primer was generated from the amino acid sequence HGCKVDN, and the oligonucleotide sequence was 5′-RTTRTCNACYTTRCANCCRTG-3′.

Guinea-pig left ventricle mRNA was reverse transcribed, and the subsequent cDNA was used as a template for the PCR reactions. The above four primers were used successfully for PCR amplification of the three isoforms of the α-subunit of the Na⁺-K⁺ pumps from guinea-pig, using TAQ polymerase in DMSO-containing buffer (Perkin-Elmer Corp., Norwalk, CT, USA). The -380 bp PCR products were isolated by agarose gel electrophoresis, purified using the Glass-Max system (Gibco BRL, Gaithersburg, MD, USA), then cloned into pBluescript (Stratagene, La Jolla, CA, USA), and transformed into X1-1 Blue E. coli (Stratagene). Insert-bearing individual clones were identified, the plasmid DNA was purified using the Wizard mini-prep system (Promega, Madison, WI, USA), and then sequenced with Sequenase (USB, Cleveland, OH, USA).

Preparation of RNA

Poly-A⁺ RNA was isolated from adult animals using paramagnetic poly-dT beads (Dynal Inc., Lake Success, NY, USA). After isolation, tissue from the apex of the left ventricle and mixed samples from various regions of the brain were homogenized fresh, in reagents supplied with the Dynal kit and the poly-A⁺ RNA extracted as described using the Dynal mRNA extraction kit.

Assays

RNA probes were prepared as described previously (Dixon & McKinnon, 1994). In all experiments a significant stretch of non-hybridizing sequence of -50 bp was included with the probe to allow easy distinction between the probe and the specific protected band. The specificity of the assay is such that there was no evidence for cross-reactivity between the probes and any other transcripts.

RNase protection assays were performed essentially as described previously (Dixon & McKinnon, 1994). For each experiment, 1 μg of poly A⁺ RNA was used. Cyclophilin probes were included in the hybridization reaction to confirm that sample was not lost during the course of the experiment. A guinea-pig cyclophilin probe was not available, so a rat cyclophilin probe was used. As expected, the probe protected smaller fragments than the full-length band found in rat RNA. Nonetheless, the same protected fragments appeared replicably, regardless of the tissue used, and hence this was an acceptable internal control. Yeast tRNA (5 μg) was used as a negative control for probe self-protection bands. The RNase protection assays shown in Fig. 1 are all 2 day exposures.

Figure 1. RNase protection assays of the three α-isoforms of the Na⁺-K⁺ pump in brain and left ventricle.

Messenger RNA for all three α-isoforms of the Na⁺-K⁺ pump is abundantly expressed in guinea-pig brain. In contrast, the α₁-isoform is the dominant transcript in guinea-pig left ventricle, comprising 82 ± 10 % (mean ±s.e.m.) of the total. Message for the α₂-isoform is also expressed in guinea-pig left ventricle, being 18 ± 8 % of the total mRNA, based on quantification from three different samples. No α₃-mRNA was detected in left ventricle, even after long exposure times.

For the comparison of the α₁- and α₂-isoform mRNA levels in ventricle, the specific protected signals were quantified directly from dried RNase protection gels using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA, USA). In this experiment three independent samples of RNA were used. The intensity of the cyclophilin signal was used to normalize for the amount of material present and the time of exposure.

RESULTS

RNase protection assays

Our first experiments aimed at relating our measurements of pump current to molecular correlates. We therefore quantified the level of message for each of the pump α-isoforms in guinea-pig tissue. As described in the Methods, we designed probes specific for the three α-isoforms of the Na⁺-K⁺ pump in guinea-pig. We then performed RNase protection assays on guinea-pig ventricle and brain. Figure 1 illustrates a sample set of results. Assays are included for each of the α-isoforms along with a lane for probe and tRNA. Cyclophilin was included as an internal control for the amount of message loaded in each lane. The results show the presence of a large band in brain tissue for all three α-isoforms, while in left ventricle only α₁- and α₂-isoform message are present. The α₁-isoform mRNA is far more prevalent than the α₂. Based on quantification from three tissue samples, the α₂-isoform mRNA is present at levels of only about 18 % of the total. These results suggest that the high- and low-affinity components of the glycoside inhibition curve we previously reported correspond specifically to α₂- and α₁-isoforms, respectively (see Discussion for quantitative comparison of the fraction of high-affinity isoform estimated from biophysical and molecular approaches).

β-Adrenergic effects

Our previous studies using guinea-pig ventricular myocytes have demonstrated that β-adrenergic activation alters total pump current (I_T), with the effect depending on intracellular calcium ([Ca²⁺]_i). At low [Ca²⁺]_i (15 nM), the β-agonist isoprenaline caused about a 20 % inhibition of I_T at all membrane voltages in the physiological range. In contrast, at high [Ca²⁺]_i (1·4 μM), isoprenaline did not inhibit I_T, but caused a negative voltage shift in the I_T current-voltage relationship. As a consequence, with a normal physiological extracellular solution, isoprenaline increased I_T by about 25 % at an intracellular voltage (V_m) of -60 mV but had no effect when V_m was 0 mV. Given the presence of two Na⁺-K⁺ pump α-isoforms, we decided to reinvestigate these actions of β-activation to determine if one or both α-isoforms are the targets for these β-adrenergic effects.

We first investigated the actions of isoprenaline at a low [Ca²⁺]_i of 15 nM. A sample set of results are provided in Fig. 2. In the experiments illustrated in Fig. 2A, 5 μM DHO was applied to block high DHO-affinity Na⁺-K⁺ pumps (this concentration of DHO should block 89 % of high-affinity Na⁺-K⁺ pumps, while blocking only 9 % of low-affinity pumps). I_h was 42 pA in control conditions and 44 pA in the presence of 0·5 μM isoprenaline. The results for the six cells studied with this protocol gave an average ratio of the pump current in the presence of isoprenaline to that in control of 0·96 ± 0·05 (P > 0·1). This suggests that the effects of β-adrenergic activation in low [Ca²⁺]_i are not targeted to the α₂-isoform of the Na⁺-K⁺ pumps. We next examined the effects of isoprenaline in low [Ca²⁺]_i on the lower affinity Na⁺-K⁺ pumps. A sample result is provided in Fig. 2B. The entire experiment was performed in the presence of 5 μM DHO to block I_h. In control conditions, I_l (the current blocked by 1 mM DHO) was 153 pA. Isoprenaline (0·5 μM) was then added to the perfusate and 1 mM DHO was applied again. I_l was 120 pA under these conditions. The results for the five cells studied gave a ratio of I_l in the presence of isoprenaline to that in control of 0·77 ± 0·06 (P < 0·001). These results suggest the effects of β-activation in low [Ca²⁺]_i are targeted to the α₁-isoform of the Na⁺-K⁺ pumps.

Figure 2. β-Adrenergic activation at low [Ca²⁺]_i inhibits only the low DHO-affinity (α₁) isoform of the Na⁺-K⁺ pump.

Cells were held at -60 mV. A, application of 5 μM DHO was used to demonstrate that isoprenaline (ISO) (0·5 μM) has no effect on the high DHO-affinity (α₂) isoform of the Na⁺-K⁺ pumps. B, in the presence of 5 μM DHO to block the high DHO-affinity Na⁺-K⁺ pumps, 1 mM DHO was used to demonstrate that isoprenaline (0·5 μM) inhibits the low DHO-affinity (α₁) isoform of the Na⁺-K⁺ pumps. Note application of isoprenaline induces an inward shift in holding current due to activation of a chloride conductance (Harvey & Hume, 1989). The vertical bars in A and B illustrate the measured I_p amplitude.

We performed a similar set of experiments at a high [Ca²⁺]_i of 1·4 μM and V_m of -60 mV. We first examined the high-affinity pumps. A sample result is provided in Fig. 3A. I_h in control was 44 pA, while that in the presence of 0·5 μM isoprenaline was 46 pA. The ratio of I_h in the presence of isoprenaline to that in control conditions for the five cells studied was 1·05 ± 0·01 (P < 0·001). Part of this small difference was due to block by 5 μM DHO of a small fraction (≡ 9 %) of low DHO-affinity pumps. A much larger effect was observed when examining specifically the effects on I_l. A sample result is illustrated in Fig. 3B. Again 5 μM DHO was present throughout to block I_h. I_l in control was 122 pA, and I_l in the presence of isoprenaline was 147 pA. The results from all seven cells studied indicate the ratio of I_l in isoprenaline to that in control was 1·22 ± 0·03 (P < 0·001). These results, in comparison with the data in low [Ca²⁺]_i, suggest all of the β-adrenergic effects on pump current are specially targeted to the α₁-isoform of the Na⁺-K⁺ pumps.

Figure 3. β-Adrenergic activation at high [Ca²⁺]_i stimulates only the low DHO-affinity (α₁) isoform of the Na⁺-K⁺ pumps.

Cells were held at -60 mV. A, application of 5 μM DHO was used to demonstrate that isoprenaline (ISO) (0·5 μM) has no effect on the high DHO-affinity (α₂) isoform of the Na⁺-K⁺ pumps_.B, in the presence of 5 μM DHO to block the high DHO-affinity (α₂) isoform of the Na⁺-K⁺ pumps, 1 mM DHO was used to demonstrate that isoprenaline (0·5 μM) stimulates the low DHO-affinity (α₁) isoform of the Na⁺-K⁺ pumps.

α-Adrenergic effects

Our previous studies of the effects of α-adrenergic agonists have shown that I_T is increased by α-adrenergic activation with a maximum increase of about 15 %. The sensitivity of the increase depends on [Ca²⁺]_i but the maximum increase does not. We next investigated whether the maximal increase is mediated through the high or low DHO-affinity Na⁺-K⁺ pumps.

We first investigated the effects of noradrenaline (NA) on I_h. These experiments are performed in the presence of 10 μM propranolol to block β-receptors. A sample set of results is provided in Fig. 4A. Upon application of DHO there was an initial increase in outward current (due to a stimulation of the Na⁺-K⁺ pump; see small print section below) and then an inward movement due to pump blockade. The amplitude of the pump current in control was 20 pA and in the presence of 10 μM NA was 28 pA. In a total of seven cells, the ratio of I_h in the presence of NA to control was 1·38 ± 0·04 (P < 0·001).

Figure 4. α-Adrenergic activation stimulates only the high DHO-affinity (α₂) isoform of the Na⁺-K⁺ pumps.

[Ca²⁺]_i was 15 nM and propranolol was always present to block β-adrenergic activation. Cells were voltage clamped to 0 mV. A, 5 μM DHO was used to study the effects of α-adrenergic activation by noradrenaline (NA) on the high DHO-affinity (α₂) isoform of the Na⁺-K⁺ pumps. NA increased I_h. B, effect of NA with 5 μM DHO present throughout the experiment to block the high DHO-affinity pumps, 0·5 mM strophanthidin (STR) was used to block the low-affinity pumps. NA had no effect on I_l. The vertical bar in A illustrates the measured I_p amplitude.

In our studies of the actions of 5 μM DHO we frequently observed an outward initial shift in holding current before the inward shift was observed. We found this outward shift never occurred on application of 1 mM DHO and was present only at slow perfusion rates. Stimulation of the Na⁺-K⁺ pump had previously been reported at low concentrations of cardiotonic steroids. Preliminary results using concentrations of DHO between 10⁻⁹ and 10⁻⁷ M demonstrated a sustained outward shift in holding current that was eliminated on removal of [K⁺]_o. This apparent stimulation of I_p is currently under investigation to determine its role in glycoside action as well as its molecular origin (one or both Na⁺-K⁺ pump α-isoforms).

We next examined the effect of NA on I_l by blocking I_h with 5 μM DHO throughout the experiment, and then applying 1 mM DHO to determine I_l. A sample set of results is shown in Fig. 4B. I_l was 97 pA in control solution and 98 pA after application of 10 μM NA. The ratio of the pump current in NA to that in control solution for eight cells was 1·01 ± 0·02 (P > 0·2). These results indicate that the effects of α-adrenergic activation on Na⁺-K⁺ pump current are targeted to the α₂-isoform only.

DISCUSSION

Previous functional studies have demonstrated the presence of two glycoside binding sites of different affinity in cardiac myocytes (Mogul et al. 1989; Berlin et al. 1992; Gao et al. 1995). Our results have demonstrated the existence of mRNA for two isoforms of the α-subunit of the Na⁺-K⁺ pump in guinea-pig ventricular myocytes. The α₁-isoform (low glycoside-affinity isoform) is more prevalent, comprising 82 ± 10 % of the total Na⁺-K⁺ pump message in guinea-pig ventricle. Our previous studies of DHO inhibition of Na⁺-K⁺ pump current (Gao et al. 1995) demonstrated that in physiological conditions, the high-affinity binding site contributed about 40 % of total pump current. However, to estimate the fraction of high-affinity pump protein, corrections had to be made for both extracellular K⁺ activation (K_0·5= 0·4 mM for the I_h, K_0·5= 3·7 mM for the I_l) and pH_o (I_h is fully activated at pH_o= 7·4, whereas I_l is just 32 % activated at the same pH_o). If one assumes that both pump types have the same maximal turnover rate then about 10 % of total pump protein is of the high-affinity type (Gao et al. 1995). Given the number of assumptions in this estimate, it is in reasonable agreement with the 18 ± 8 % value for the α₂-isoform mRNA fraction. This relatively small fraction is probably the reason that the presence of the α₂-isoform in guinea-pig ventricle could not previously be demonstrated using Western blots (Sweadner et al. 1994).

Previous studies of catecholamine action on Na⁺-K⁺ pump current recorded effects only on I_T (Desilets & Baumgarten, 1986; Shah et al. 1988; Williamson, et al. 1993; Gao et al. 1994; Wang et al. 1998). Activation of α- and β-adrenergic receptors initiated distinct second messenger pathways leading to PKC- and PKA-dependent phosphorylation, respectively. We did not determine whether (1) both receptors and their associated kinases lead to changes in both high and low DHO-affinity pumps, or (2) one or both adrenergic receptors and the associated kinases are specifically coupled to one type of pump. Our results here demonstrate that specificity does exist. Synthesizing our molecular and biophysical studies, our results demonstrate that β-adrenergic activation initiates the cAMP-PKA pathway and alters activity of the α₁ (low DHO-affinity) isoform of the Na⁺-K⁺-ATPase. In contrast, α-adrenergic activation initiates the PKC pathway and alters activity of the α₂ (high DHO-affinity) isoform of the Na⁺-K⁺-ATPase.

The functional differences in I_h and I_l previously described (Gao et al. 1995) can now be ascribed to functional differences in the α₂- and α₁-isoforms, respectively. The present study demonstrates isoform-specific coupling to autonomic imput. Various tissues within the heart, and indeed different organs of the body, express specific isoforms of the Na⁺-K⁺ pumps as well as specific signal transduction proteins. Our results from cardiac ventricular cells reveal some of the reasons these different isoforms are expressed. Each particular mix of isoforms will have a unique response to a variety of physiological stimuli. Thus, the selection of which Na⁺-K⁺ pump isoform(s) to express is another dimension in a cell's ability to respond optimally to its environment.