Unmasking of a novel target for blocking harmful Na+ coupled acid extrusion: electrogenic Na+–HCO3− symport

See article by Kandoudi et al. [22] (pages 388–396 in this issue). Many cellular functions of myocardium, such as the sensitivity of the myofilaments to Ca2+, the activity of the rate-limiting enzyme of the glycolysis (1-phosphofructokinase) and number of functional gap junctions between adjacent myocytes, are strongly affected by lowering of pHi[1]. In fact, acidosis depresses contractility in cardiac myocytes by affecting virtually every step in the excitation–contraction coupling [2]. On the other hand, there is also evidence that cytoplasmic alkalinization is an intracellular messenger mediating growth responses of various stimuli such as stretch, neurohumoral factors and growth factors [3]. Therefore, in order to preserve proper cardiac functioning, pHi of the myocytes is maintained within narrow limits. Normally myocardial pHi is about one pH unit more alkaline than would be expected if H+ was in electrochemical equilibrium [4–7]. This indicates the existence of mechanisms actively removing acid equivalents from the cytoplasm in addition to the intracellular pHi-dependent buffering capacities (protein histidyl residues, inorganic phosphate and CO2–HCO3−) which maintain constant pHi in the cardiomyocyte [4–6,8]. In general, cells regulate their pHi on top of intracellular buffering capacity, by balancing the opposing actions of acid extruding and loading mechanisms [9]. Acid extruders are transporters that either export H+ from the cytoplasm or import a base (e.g., HCO3−) into the cytoplasm or they merely are passively diffusing molecules such as CO2. The only known transport proteins that function as acid loaders are HCO3− transporters (e.g., the electrogenic Cl−–HCO3− antiporter). It is obvious that the metabolic processes that (in)directly generate cytoplasmic H+ are the major acid loaders [9]. Historically, the principal mechanism identified for acid extrusion from cardiac cells has been the Na+–H+ antiporter (NHE family) [7]. However, over the last 15 years three different HCO3−-dependent pHi regulating mechanisms, which were known to be present in many other cell types, have been unmasked by electrophysiological studies in myocardium performed under conditions in which a physiological buffer system including CO2–HCO3− was present in the media. The following transporters have been identified: (i) the electroneutral and electrogenic Na+–HCO3− symporters (NBC family) [5,8,10–13], (ii) the Na+ dependent Cl−–HCO3− antiporter (AE family) [14] all as acid extruders and (iii) the electroneutral Cl−–HCO3−antiporter (AE family) as acid loader [6]. Furthermore, the electroneutral monocarboxylate–H+ symporter (MCT family), particularly when high rates of anaerobic glycolysis occur in the initial stage of ischemia, is a major acid extruder by removing lactic acid [15–17]. At this time most of the aforementioned myocardial acid extruders and loaders have been molecularly and structurally characterized and, moreover, some have been studied on their direct modulation by neurohumoral stimuli and their relative expression in normal and failing heart. Six isoforms of the Na+–H+ antiporter are known and they are designated NHE1–NHE6. NHE1 is the predominant isoform found in myocardial plasma membranes [18]. Its membrane-associated domain has 12 transmembrane segments and one membrane associated segment. The C-terminal hydrophilic domain is contained in the cell cytoplasm where it interacts with a variety of other proteins, including protein kinases. NHE1 becomes maximally active at low pHi (pH<6.5) and it is phosphorylated in its cytoplasmic domain in response to hormonal (e.g., ET-1) stimulation shifting its pH-dependence into a more alkaline range [18]. Hypertrophic stimuli such as stretch, endothelin-1, angiotensin II, thrombin, α1-adrenergic agonists are all able to activate NHE1 and, as expected, to induce alkalinization of cardiomyocytes (reviewed in [18]). The levels of mRNA for the Na+–H+ antiporter, and of the protein itself, have been examined in recipient hearts with chronic end-stage heart failure and unused donor hearts. Sarcolemmal NHE activity is significantly greater in the failing heart, but in contrast NHE1 protein was expressed in similar abundance in control compared to failing heart [19]. Only recently some progress in the molecular characterization of the Na+–HCO3− symporter isoforms in heart was made by cloning an electrogenic (hhNBC) [20] and electroneutral NBC (NBCn1-A or NBC3) [21] from respectively, human heart and skeletal muscle cDNA libraries. Based on electrophysiological evidence it is presumed that the electrogenic NBC in myocardium operates with stoichiometry of one Na+ for two HCO3−, whereas the renal hkNBC isoform operates by one Na+ for three HCO3−[9]. The hhNBC and the renal form hkNBC represent alternative splice products of the same gene and their hydropathy plots are consistent with the presence of at least 10 membrane spanning segments and a relatively large extracellular loop between segment 5 and 6. In one of these studies cardiac expression at the mRNA level of both NBC isoforms was confirmed, though the hybridization signal of the electrogenic NBC in normal heart was found to be very weak [20]. However, in this issue of Cardiovascular Research, the same group of investigators now report both by Northern analysis and by the more sensitive real-time quantitative reverse transcription–polymerase chain reaction (RT–PCR), on mRNA levels of hhNBC and NBCn1 in control hearts (from organ donor candidates) and in either the left or right ventricle of patients with end-stage heart failure (four had history of ischemic cardiomyopathy (ICM) and six had dilated cardiomyopathies (DCM)) [22]. Most convincing evidence for significant expression of the electrogenic NBC is that immunoblots were performed with a polyclonal antibody raised against an hhNBC-specific peptide and that mutual differences in protein expression between hearts correlated with mRNA levels, though this correlation was shown by qualitative means only for two controls and two patients with heart failure. At any rate, the results of this study demonstrate that the expression of the electrogenic NBC (shown for mRNA as well as protein) increases in heart failure whereas the expression (only shown for mRNA) of the electroneutral NBCn1 remains the same [22]. Apart from the findings in failing hearts, the observation of expression of the electrogenic NBC isoform (protein) in myocardium provides a definite proof of molecular identity of this acid extruder, of which proof for its presence in myocardium until now merely was based on data obtained in (electro)physiological studies [5,8,10–13]. That until recently this important functional evidence is ignored demonstrates the recent exhaustive review on cardiac anion transport in which the possibility of operation of electrogenic NBCs is not even considered [24]. As to the neurohumoral regulation of NBCs, another group demonstrated that angiotensin II induces alkalinization of neonatal rat ventricular myocytes by selective activation of Na+–HCO3− symport [23]. The electroneutral Cl−–HCO3− antiporters belong to a multigenic family that comprises AE1, AE2 and AE3 and these isoforms are ubiquitously expressed in vertebrate tissues. Truncated forms of AE1 and AE3 are present in the heart [24]. Topology studies suggest that the antiporter isoforms are composed of 13 transmembrane segments. In contrast to the acid extruders NHE and NBCs, the Na+-independent Cl−–HCO3− antiporters operate as a HCO3− extruder alleviating intracellular alkaline and CO2 loads. Indeed, it is demonstrated that the Cl−–HCO3− antiporter shows a steep rise in Vmax with increasing pHi above 7.0 [25]. The purinergic agonist ATP induces tyrosine phosphorylation of AE1 thereby triggering rapid activation of anion-exchange in neonatal rat cardiomyocytes [26]. The earliest functional studies on the plasma membrane monocarboxylate–H+ symporter MCT were already performed in the 1970s using isolated intestinal epithelial cells and erythrocytes [27,28]. The existence of a mammalian MCT family is now firmly established and contains at least nine members (MCT1–9) [27]. It shares the 12 transmembrane helix topology with the NHE and likely also NBC families. Western and Northern blotting has confirmed the presence of large amounts of the monocarboxylate–H+ symporter isoform MCT1 in both human and rat hearts. MCT4 which is also present in human heart, could not be detected in rat heart cells. There are no reports on post-translational regulation of cardiac MCT. However, very recently it was shown that MCT1 is strongly upregulated in a rat model of congestive heart failure [28]. Increased rates of lactate–H+ uptake (analyzed by BCECF fluorescence) were found in myocytes isolated from the failing hearts proving the functional consequence of increased MCT1 expression [28]. Where we now stand is that not only the electrogenic NBC isoform, as reported by the study of Khandoudi et al. in this issue of Cardiovascular Research, is upregulated in the failing heart but also NHE1 and MCT1 [19,22,27]. These findings implicate that the capacity of the failing heart to extrude acid must be markedly increased. As to the functions of NBCs and NHE1, increase of their activities may be unfortunate, since Na+-coupled acid extrusion is detrimental to the myocardium when it undergoes ischemia followed by reperfusion [3,18,29]. Moreover, under these conditions the electrogenic operation of NBCs might contribute to arrhythmogenesis. However, as to the function of MCT1, increase of its activity may be fortunate, since this symporter contributes largely to the Na+-independent acid extrusion during ischemia–reperfusion [15–17,30]. On the other hand, increased activities of NHE1 and NBCs promote cytoplasmic alkalinization which is known to be an intracellular messenger co-mediating growth responses of the myocardium to various stimuli such as stretch, neurohumoral factors and growth factors [3,30]. As mentioned before, there is evidence that the activities and expression of the NHE1, NBCs and AE1 respond to known hypertrophic stimuli (endothelin, angiotensin II, purinergic, etc.) [18,23,26]. All together, these findings and the possible contribution of Na+-coupled acid extruders to development of Ca2+ overload make NHE1 and NBCs potentially important candidates for targeted intervention of the remodeling processes contributing to heart failure following myocardial infarction [3,18,30,32,33]. The conditions necessary for the acute activation of NHE1 and NBCs and, thereby, development of Ca2+ overload, are present in the ischemic myocardium including a low pHi, reduced Na+–K+ ATPase activity and increased production of endocrine, paracrine and autocrine factors (endothelin, α1-adrenergic and purinergic agonists, angiotensin II, etc.) [30]. On this basis it can also be concluded that myocardial pHi will be normalized after an acid load due to ischemia by the concerted action of NHE1, NBCs, MCT1 and AE1/3 transporters and the merely passive transmembrane movement of CO2. Both NHE1 antiporter and electrogenic NBC symporter are driven by the transmembrane Na+ gradient as illustrated in the schematic of Fig. 1. In addition, depolarization of the membrane potential due to a reduction of transmembrane K+ gradient, will be in favor of action of the electrogenic Na+–HCO3− symporter. The electroneutral Cl−–HCO3− antiporter likely will not participate in acid extrusion, only when pHi sufficiently falls to raise the [HCO3−]e/[HCO3−]i ratio above the [Cl−]e/[Cl−]i[9]. However, at pHi lower than 7.0 this anion-exchanger will have a minimal activity [25] and, moreover, osmotic swelling- and agonist (endothelin-1, α1-adrenergic and purinergic agonist)-induced opening of Cl− channels [24,34,35] during ischemia will support exchange of extracellular Cl− for intracellular HCO3−. The accumulated CO2 and lactate are, when washed out during reperfusion, the major contributors to the pHi recovery (Fig. 1). Schematic of the major transport mechanisms involved in the recovery of myocardial pHi during ischemia and reperfusion. MCT1/4 stands for monocarboxylate–H+ symporter isoforms-1 and -4, NHE1 for Na+–H+ antiporter isoform-1, NBC? for electrogenic Na+–HCO3− symporter of which at present no clear-cut isoform classification is available, AE1/3 for the electroneutral Cl−–HCO3− antiporter isoform-1 and -3. Thick arrows indicate the major acid extrusion pathways (via CO2 and MCT1/4). The broken arrow indicates the passive transmembrane diffusion of CO2. The intracellular minus signs indicate the direction of the resting membrane potential. The Cl−–HCO3− antiporter could also participate in acid extrusion, but only if pHi sufficiently falls to raise [HCO3−]e/[HCO3−]i ratio above the [Cl−]e/[Cl−]i[9]. However, at pHi lower than 7.0 this transporter will have minimal activity [25] and, moreover, osmotic swelling- and agonist (endothelin-1, α1-adrenergic, purinergic)-induced opening of Cl− channels [24,34,35] during ischemia will support the exchange of extracellular Cl− for intracellular HCO3−. At present, there is abundant experimental evidence that pharmacological blockers (e.g., the amiloride analogs, and the newly developed cariporide (HOE 642) and EMD 851310) of NHE1 attenuate the detrimental consequences of myocardial ischemia and reperfusion such as contractile dysfunction, tissue necrosis and arrhythmias (extensively reviewed in Refs. [30] and [36]). There is still some controversy as to whether NHE1 inhibition is most effective when the inhibitor is administered prior to ischemia or just prior to reperfusion. The overall consensus now is that for maximal effectiveness the drug needs to be present during the ischemic period, however, some beneficial effects occur at reperfusion as well [37]. The conflict in results has been suggested to be due to differences in the experimental design (drug dose en delivery protocol) and post-ischemic protection may not always be optimal due to the vascular permeability barriers or limited collateral circulation [38]. The most commonly cited mechanism underlying the cardioprotective actions of the pharmacological blockers of the Na+–H+ antiporter during ischemia and reperfusion is the attenuation of Na+ influx-coupled H+ extrusion which reduces the harmful Ca2+ overload through the action of the Na+–Ca2+ antiporter [30,36]. Inhibition of Na+–K+–ATPase due to cellular ATP depletion, is an important prerequisite for the NHE1 involvement in ischemic and reperfusion injury and forms the basis of Na+-dependent elevation of [Ca2+]i levels resulting in cell injury [30]. Some recent studies have provided data in isolated cardiomyocytes that contradict a key role for reduced intracellular Na+ and Ca2+ accumulation during simulated ischemia in the cardioprotective mechanisms of NHE inhibitors [39,40]. But it is likely that the metabolic challenge associated with anoxia in quiescent myocytes in vitro differs from that imposed by ischemia in beating intact hearts both in vitro and in vivo [40]. For instance, in this model with a simulated ischemia, neurohumoral stimuli will not be formed, which excludes the activation of NHE. An alternate concept regarding reperfusion-induced NHE1-dependent injury through Ca2+-independent mechanisms has also been proposed. This hypothesis, termed the pH paradox suggests that during reperfusion the rapid restoration of pHi reverses the suppression of ATP resynthesis via oxidative phosphorylation and harmful actions of phospholipases and proteases [30]. In general, a major role of the Na+–H+ antiporter in the harmful Na+-coupled control of pHi in cardiomyocytes, challenged by anoxia and blocked metabolite wash-out, has barely been questioned. Eight years ago, however, Vandenbergh et al. [31] already demonstrated that the wide use of non-HCO3− containing solutions was, in fact, masking the existence of other Na+-coupled acid extruding mechanisms during ischemia and reperfusion that play a role when the more physiologically relevant HCO3−–CO2 buffer is present. By estimation of myocardial pHi from the chemical shift of 31P-nuclear magnetic resonance (NMR) signal of [PO4−]i in the perfused ferret heart, these investigators were able to demonstrate that after brief episodes of ischemia, the recovery of pHi is principally mediated by metabolite washout i.e. lactate and CO2. Na+-coupled H+ extrusion did contribute to the net H+ extrusion but with the Na+–HCO3− symport likely to be more important than the Na+–H+ antiport [31]. Around the same time Shimada et al. [41] confirmed that the cardioprotective efficacy of NHE1 inhibitors may be overestimated under HCO3−-free conditions. A year or two later Camilión de Hurtado et al. [8,10] added to these new insights experimental evidence for the existence in superfused cat papillary muscle of an electrogenic Na+–HCO3− symport mechanism that contributes almost equally to the Na+–H+ antiporter in the regulation of pHi. It should, however, be noticed that in the latter studies the pHi changes were not induced by ischemia, but by shift from HEPES- to HCO3−-buffered superfusate, through membrane depolarization by high K+e[10], or increased frequency of contraction [8]. Nevertheless, this different approach does not alter the significance of the findings, because the operation of an electroneutral Na+–HCO3− symporter was already shown before in isolated sheep Purkinje fibers [11] as well as guinea-pig ventricular myocytes [12]. The electrogenic nature of the Na+–HCO3− symporter clearly distinguishes it from the Na+–H+ antiporter, which is electrically silent. During ischemia the action of electrogenic NBC but not NHE1 will even be further promoted by the depolarization of the membrane potential due to the partial loss of the transmembrane K+ gradient (Fig. 1). Because of its electrogenic nature the Na+–HCO3− symport may also contribute to arrhythmogenesis [7]. After experimental evidence had proven that the Na+–HCO3− symport is even more important than the Na+–H+ antiport mediating the harmful Na+-dependent acid extrusion during a period of ischemia, it became interesting to apply, analogous to the studies on the role of NHE1, specific inhibitors of NBCs, to look at their effects on cardiac post-ischemic contractile recovery, tissue necrosis and/or arrhythmias. Unfortunately, the available and which are blockers of all HCO3−-dependent transporters are not because they have effects on contraction when at high for are that with many other membrane proteins [9]. This has now been by Khandoudi et al. as reported in this issue of Cardiovascular Research, by development of a specific antibody against the This polyclonal antibody was raised in against a peptide with a of the large extracellular loop between the and transmembrane segment acid of hhNBC [20]. The acid of this extracellular loop in hhNBC is very similar to the in the electrogenic rat most for the antibody to with NBC in intact rat myocardium, because it has at least not to the myocardial of the antibody to the of isolated perfused rat heart to either ischemia followed by reperfusion, or ischemia and reperfusion, and functions of the heart and reduces during reperfusion [22]. On the basis of the latter findings it is that inhibition of hhNBC a approach for ischemic heart An important now to is whether this protection is by myocardial Na+ and Ca2+ following the pHi from signal of analogous to the approach of et al. [31] would to the that by the antibody of Na+-dependent acid extrusion At this stage it is also important to if shifting from to buffer would the beneficial of the In the study of Khandoudi et al. the functional activity of the polyclonal antibody was in cells and in isolated cardiomyocytes [22]. The antibody a in both cell types, which can be to NBC In with this is that no such was detected in cells. The responses of the latter cells and cardiomyocytes to cellular were studied by to This has already been in various cell including heart cells are for about to in Na+ and HCO3− and to a During this the cells in the form and large amounts of H+ are the pHi recovery is as initial rate, The that the antibody reduces the initial of pHi recovery more than in the cells as well as in the cardiomyocytes [22]. had a lower of pHi recovery and, as would be expected, were not to the NBC However, it is not made whether cells with were with or Furthermore, these of pHi recovery have For instance, et al. demonstrated with guinea-pig ventricular myocytes that the of pHi recovery is dependent on intracellular buffering capacity and the pHi. In the presence of a constant rapid changes of pHi by may the intracellular CO2–HCO3− buffer system out of of this buffer to changes in pHi and of transporter from of pHi be made when the intracellular CO2–HCO3− buffer system is at equilibrium [6]. The that the pHi were or dependent on the of an extracellular system would have the of these The step to be is to that pHi recovery after a in the cardiomyocytes to a rise recovery and [Ca2+]i and to which of the increase in intracellular Na+ and Ca2+ is blocked by the polyclonal antibody against hhNBC and which by the NHE1 investigators cell Ca2+ during and after an acid loading in cardiomyocytes During to [Ca2+]i about whereas to a [Ca2+]i to control under these but acid conditions both Na+–Ca2+ antiport and the Na+–K+–ATPase are sufficiently active to the transmembrane Na+ and Ca2+ to pHi regulation [30].