Cardiovascular Research 52 (2001) 387–396 www.elsevier.com / locate / cardiores

Inhibition of the cardiac electrogenic sodium bicarbonate cotransporter reduces ischemic injury ´ Nassirah Khandoudi a , *, Joelle Albadine a , Philippe Robert a , Stephane Krief a , a a b Isabelle Berrebi-Bertrand , Xavier Martin , Mark O. Bevensee , Walter F. Boron b , Antoine Bril a a

´ ´ , France Cedex GlaxoSmithKline Laboratoires Pharmaceutiques, 4 Rue du Chesnay-Beauregard, BP 96205, 35762 Saint-Gregoire b Department of Cellular and Molecular Physiology, Yale University School of Medicine, New Haven, CT 06510, USA Received 16 May 2001; accepted 23 July 2001

Abstract Objective: Although it is believed that sodium-driven acid–base transport plays a central role in the development of the reperfusion injury that follows cardiac ischemia, research to date has demonstrated only a role for Na 1 / H 1 exchange (NHE). However, Na 1 -driven HCO –3 transport, which is quantitatively as important as NHE in cardiac cells, has not been examined. Methods and Results: Here the results show that a neutralizing antibody raised against the human heart electrogenic Na 1 / HCO 2 3 cotransporter (hhNBC) blocked the recovery of pH after acidic pulse both in HEK-293 cells expressing hhNBC and in rat cardiac myocytes demonstrating the presence of an electrogenic NBC in rat cardiac myocytes similar to hhNBC. Administration of anti-NBC antibody to ischemic-reperfused rat hearts markedly protects systolic and diastolic functions of the heart during reperfusion. Furthermore, using a quantitative real-time RT-PCR (TaqMan) and Western blot analysis we demonstrated that in human cardiomyopathic hearts, mRNA and protein levels of hhNBC increase, whereas mRNA levels of the electroneutral Na 1 / HCO 2 3 cotransporter (NBCn1) remain unchanged. Conclusion: Our data provide evidence that inhibition of hhNBC, whose role in cardiac pathologies could be amplified by overexpression, represents a novel therapeutic approach for ischemic heart disease.  2001 Elsevier Science B.V. All rights reserved. Keywords: Cardiomyopathy; Gene expression; Ion transport; Ischemia; Reperfusion; Ventricular function

This article is referred to in the Editorial by J.M. J. Lamers ( pages 339 – 344) in this issue.

1. Introduction Ischemic heart disease represents the principal etiology for the development of congestive heart failure. Thus, treatments that reduce post-ischemic injury could be beneficial for heart failure, as exemplified by b-adrenergic antagonists [1], angiotensin-converting enzyme inhibitors [2] or endothelin antagonists [3]. Because disturbances of

*Corresponding author. Tel.: 133-2-9928-0461; fax: 133-2-99280444. E-mail address: nassirah [email protected] (N. Khandoudi). ]

myocardial pH i have grave functional consequences, cardiac cells possess transporters that maintain pH i within narrow limits [4]. The pH-regulating systems include NHE [5] and two bicarbonate-dependent systems, Na 1 / HCO 32 cotransport [6] (NBC) and Na 1 -independent Cl 2 / HCO 2 3 exchange [7]. Although the physiological role of NHE in cardiomyocytes has been amply demonstrated [5], that of HCO –3 -dependent pH i -regulatory mechanisms had long been underestimated because most studies were carried out in HCO 2 3 -free buffers [8,9]. However, the NBC accounts for about 20–30% of total acid extrusion (i.e., sum of H 1 efflux and HCO 2 3 efflux) in cardiac Purkinje fibres [6], and for 40–50% in ventricular myocardium [10,11]. Alteration

Time for primary review 22 days.

0008-6363 / 01 / $ – see front matter  2001 Elsevier Science B.V. All rights reserved. PII: S0008-6363( 01 )00430-8

388

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

of NBC activity could thus play a major role in pathological situations such as myocardial ischemia, where intracellular proton generation is substantially elevated. Functional and molecular studies indicate the presence of at least two NBCs in the heart. An electroneutral NBC has been described in sheep cardiac Purkinje fibres [6] and in guinea-pig ventricular cells [10], whereas an electrogenic NBC has been characterized in cat papillary muscle and rat cardiac myocytes [11–13]. Following the expression cloning of an electrogenic NBC from amphibian kidney [14], other electrogenic NBCs have been cloned from mammalian kidney [15,16], pancreas [17], heart [18] and brain [19]. More recently, electroneutral NBCs (NBCn1-A–NBCn1-D) cloned from rat smooth muscle have also been shown to be expressed in heart [20]. Here, we show that, in contrast to the electroneutral cotransporter NBCn1-A [20] (also known as NBC3 [21]), the electrogenic hhNBC is overexpressed in hearts from patients with heart failure. We also demonstrate that specific inhibition of hhNBC with a polyclonal antibody raised against an hhNBC-specific peptide is effective at reducing post-ischemic dysfunction in isolated rat hearts. Thus, the specific inhibition of hhNBC could represent a major advance to the treatment of ischemic heart disease, especially in patients with cardiomyopathy.

2. Methods

2.1. Animals All animals used in the present study received human care in accordance with the principle of NIH guidelines (publication no. 85-23) for animal care.

2.2. Cell culture Experiments were performed in HEK-293 cells expressing hhNBC and in isolated rat cardiac myocytes. The complete coding region of hhNBC (GenBank Accession no. AF069510) was cloned into the eukaryotic expression vector pCDNA3.1HisC at the EcoRI site in sense and anti-sense orientation. HEK293 cells stably transfected with sense and anti-sense constructs for hhNBC were grown in Dulbecco’s modified Eagle’s medium (Life Technologies) supplemented with 10% fetal bovine serum, 10 mM Hepes and penicillin / streptomycin (100 U / ml and 100 mg / ml, respectively). When cells reached 80% of confluency, they were transfected with pCDN-hhNBC plasmid, using lipofectamin (Life Technologies, Cergy Pontoise, France). Stably transfected clones were selected by addition of G418 (1 mg / ml) to the medium. Expression of recombinant hhNBC in HEK-293 cells was tested by RT-PCR, using specific primers for hhNBC (not shown) as well as in

Western-blot experiments using anti-hhNBC specific antibodies. For cardiac myocyte experiments, rats were anesthetized and the heart was rapidly removed. Myocytes were isolated as previously described [22].

2.3. Polyclonal antibodies The deduced amino acid of hhNBC protein was analysed for highly antigenic regions using the Jameson-Wolf antigenic index. A peptide representing amino acids 629– 644 of hhNBC was synthesized (Peptide Synthesizer Model 431A, Applied Biosystems), purified and conjugated to keyhole limpet hemocyanin, using benzidine as the coupling agent. Fourteen-week-old New Zealand rabbits (two animals per peptide) were injected on day 0 with peptide-carrier conjugate (150 mg / injection) in complete Freund’s adjuvant, and every 2 weeks thereafter with peptide-carrier conjugate (50 mg / injection) in incomplete Freund’s adjuvant. Animals were bled 7 days after boosts (days 39, 69 and 95) and sera were tested at various dilutions on ELISA plates coated with unconjugated peptide.

2.4. Expression-pattern studies Ventricular samples from failing human hearts were obtained from 10 patients with NYHA class VI terminal heart failure undergoing cardiac transplantation (nine men, one woman, mean age 5063 years). Of these, four patients had history of ischemic cardiomyopathy (ICM) and six had dilated cardiomyopathy (DCM). All patients were treated with diuretics, angiotensin-converting enzyme inhibitors, antiarrhythmics, and anticoagulants until the time for transplantation. Normal RNA and protein microsomes samples were obtained from Clontech and from organ donor candidates (mean age 37.868.5 years) whose hearts could not be transplanted for non-cardiac reasons. The use of these samples was approved by the Internal Review Committees and the investigation was performed in accordance with the principles outlined in the Declaration of Helsinki. The expression level of hhNBC was measured using Northern-blot analysis, real-time quantitative RT-PCR and Western immunoblotting analysis. Total RNA from the ventricles of normal hearts, and hearts from patients with dilated cardiomyopathy (DCM) or ischemic cardiomyopathy (ICM) were extracted using the acid guanidinium thiocyanate–phenol–chloroform method [23]. Real-time quantitative RT-PCR were performed on cDNA generated from 100 ng of total RNA using murine Moloney leukemia virus reverse transcriptase (Life technologies) and random hexamers. For the PCR, we used 200 nM of both sense and antisense primers (Genset) as well as 100 nM of TAMRA-labelled primer probe (PE Applied Biosystems) in a final volume of 50 ml, the TaqMan PCR

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

core reagent kit, and a ABI PRISM 7700 Sequence Detection System (PE Applied Biosystems). Sense, antisense, and probe primers were AAGTAGAAGGCCACCATACCATTTA, CCCTGTCTTTCTCTTGTGACGTC, and CGGAGTCCATGTGCCGAAGAGTTACAG for hhNBC; GAGGTCCCTGTTCCAACCAG, ATCGTCCAATTTCATGGTACTGTG, and ATTGGGTCCAGCGGGCAAGGC for NBCn1; and GGCCATCCACAGTCTTCTGG, GGCCATCCACAGTCTTCTGG, and ACCACAGTCCATGCCATCACTGCCA for GAPDH (to account for variability in the initial quantities of cDNA). For hhNBC or GAPDH, serial dilutions of plasmid DNA or genomic DNA was amplified to generate standard curves and to assess PCR efficiency. Results are presented as cycle threshold (CT ), i.e., the PCR cycle number at which exponential PCR-generated fluorescence is first detected; the lower the CT , the higher the expression. Northern blot analysis was done on 2 mg of poly A1 RNA using standard methods [24]. In immunoblotting experiments, the polyclonal antibody typically recognized a 137-kDa band, which is believed to be glycosylated hhNBC.

2.5. Intracellular pH measurement HEK-293 cells were loaded with 1 mM of BCECF-AM [25], the acetoxymethyl ester of the pH-sensitive fluorescent dye 29,79-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF-AM, Molecular Probes, Eugene, USA) for 30 min at 378C. pH i was measured in a thermostatically controlled holding chamber (378C) in a Delta Scan dualexcitation spectrofluorometer (Photon Technology International, South Brunswick, NJ, USA). The cells were continuously superfused with Tyrode buffer solution. Fluorescence was detected in individual cells using excitation wavelengths of 440 and 490 nm at an wavelength emission of 535 nm. The pH i calibration was generated using the KCl / nigericin technique [26]. The initial rate of pH i recovery from an acid load was calculated by fitting a line to the first 2 min of the pH i recovery.

2.6. Isolated-heart perfusion Male Wistar rats (280–320 g) were anesthetized by intraperitoneal injection of sodium thiopental (50 mg kg 21 ). Hearts were then cannulated via the aorta and perfused using the Langendorff technique with Krebs– Henseleit buffer of the following composition (in mM): NaCl 118; NaHCO 3 23; KCl 4.7; KH 2 PO 4 1.2; MgCl 2 1.2; CaCl 2 1.25; glucose 11; pyruvate 2, gassed continuously with a 95%O 2 / 5%CO 2 gas mixture and maintained at 378C. A water-filled latex balloon fixed to a pressure transducer was inserted into the left ventricle for the determination of left ventricular developed pressure (LVDP) and heart rate as previously described [27]. Positive and negative dP/ dt were obtained with a differen-

389

tiator (Gould Electronics, Cleveland, OH, USA). Balloon volume was adjusted to achieve a left ventricular enddiastolic pressure (LVEDP) of 8–10 mmHg. The volume of the LV balloon was kept constant throughout the experiment, so that changes in LVEDP reflect changes in diastolic chamber distensibility [28]. The exponential time constant of left ventricular pressure decay (t ) was calculated from the linear regression of 2dP/ dt versus pressure, assuming that the asymptote of the left ventricular relaxation is zero [29]. All measurements were recorded on a thermal multichannel recorder (Gould Electronics) and transferred to computer for analysis. Coronary flow was monitored by collecting coronary sinus effluent. All hearts were initially equilibrated for 30 min after which, zeroflow or low-flow (0.5 ml / min) global ischemia was induced for either 30 or 60 min followed by a reperfusion.

2.7. Lactate dehydrogenase release Lactate dehydrogenase (LDH) leakage from the myocardium was measured, using an enzymatic assay kit (Boehringer-Mannheim, Germany), in the coronary effluent collected during reperfusion. The activity of LDH was normalized against the wet weight of heart and coronary flow rate as U / min per g wet weight. The absorbance was read at 334 nm using a Beckman DU-640 spectrophotometer.

2.8. Statistics All data are expressed as the mean6S.E.M. of n preparations. Statistical comparisons were made using either Student’s t-test or one-way analysis of variance.

3. Results

3.1. Functional evidence for hhNBC in cardiac cells Polyclonal antibodies raised against an extracellular loop of hhNBC were generated and used for both Western-blot experiments and functional studies, either in NBC-transfected HEK-293 or in isolated cardiomyocytes. The antiNBC antibody detected a single, |137-kDa band in hhNBC-transfected HEK-293 cells, but nothing in nontransfected cells (Fig. 1a). Similarly, the antibody detected a single band of about 137 kDa in microsomal fractions from rat cardiac tissue (Fig. 1b), indicating that the rat homologue of hhNBC is expressed in myocardium at the protein level and could thus be responsible for the electrogenic transport reported previously [11]. The functional activity of the polyclonal anti-NBC antibody was assessed in NBC transfected HEK-293 cells (Fig. 1c) and in isolated cardiomyocytes (Fig. 1d). In both cases, we acidified the cells with an NH 1 4 prepulse [30]. Anti-NBC slowed the pH i recovery in hhNBC-transfected

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

390

Fig. 1. Protein expression and function of hhNBC in both transfected HEK-293 cells and in rat ventricular cells. In a western blot (a), an antibody (anti-NBC) raised against one of the extracellular loops of hhNBC labelled a 137-kDa protein in hhNBC-transfected HEK-293 cells (1), but not in nontransfected HEK-293 cells (2). Similar single bands (b) were detected in the in microsomal fractions of three rat hearts (1–3). Pre-immune serum labelled no bands. Time course of pH i recovery from an acid load induced by application and withdrawal of 20 mM NH 1 4 in the presence of a solution containing 115 mM Na 1 and 25 mM HCO 2 3 (pH 7.4). The anti-NBC antibody inhibited the hhNBC in transfected HEK-293 cells (c) and in rat isolated cardiomyocytes (d).

HEK-293 cells from 0.12760.013 pH / min (pH i 5 6.4460.06; n515) to 0.04260.009 pH / min (pH i 5 6.5160.14; n55). The pH i recovery measured in cells transfected with antisense cDNA (0.06260.011 pH / min at pH i 56.7060.03; n515) was not changed by anti-NBC (0.06060.021 pH / min at pH i 56.5660.09; n55). In isolated rat cardiac myocytes, the antiserum slowed the pH i recovery from 0.09060.016 pH / min (pH i 56.7960.07; n512) to 0.04760.015 (pH i 56.7360.08; n57). The preimmune serum had no effect. Taken together, these results show that the antiserum functionally inhibits hhNBC, which is responsible for more than half of the pH i recovery in isolated cardiomyocytes.

3.2. hhNBC is involved in post-ischemic dysfunction A well-known consequence of myocardial ischemia is intracellular acidosis. During reperfusion, this low pH i activates NHE, which may overload the cell with Na 1 ,

leading via Na 1 / Ca 21 exchange to a secondary increase in [Ca 21 ] i , and thus cell damage. Several studies have reported that inhibiting NHE protects against myocardial injury following ischemia–reperfusion [5]. We previously suggested that additional mechanisms, including NBC, could also be involved [31]. To elucidate the role of hhNBC in post-ischemic contractile dysfunction, we investigated the effect of the neutralizing antibody in isolated perfused rat hearts subjected to either (i) 30-min zero-flow ischemia followed by 20-min reperfusion, or (ii) 60-min low-flow ischemia and 30-min reperfusion. NBC polyclonal antibody or pre-immune serum (1:1000 in Krebs buffer solution) was added to the perfusate 15 min before initiating ischemia, and was maintained throughout the experiment, during which serial measurements of hemodynamic parameters were performed. Basal pre-ischemic values of cardiac function were not affected by the presence of anti-NBC (Table 1).

3.2.1. Zero-flow ischemia As previously demonstrated in similar models of global ischemia, zero-flow ischemia caused left ventricular end diastolic pressure (LVEDP) to decrease in the first 10 min following the initiation of ischemia, but then to increase markedly (Fig. 2a). In anti-NBC-treated hearts, the increase in LVEDP occurred later during ischemia, although it reached a similar degree by the end of the 30-min ischemic period. Left ventricular developed pressure (LVDP), heart rate and coronary blood flow all fell to zero by the first measurement during ischemia, and were unaffected by the antiserum (Fig. 2b–d). On reperfusion, the diastolic parameter LVEDP increased markedly in control hearts (Fig. 2a), while systolic function recovered weakly, as evidenced by a slow and small increase in LVDP (Fig. 2b). Treatment with anti-NBC reduced the further increase in LVEDP during reperfusion, indicating that the antiserum limits post-ischemic contracture (Fig. 2a), and markedly increased LVDP (Fig. 2b). The other parameters measured to characterise the cardiac function recover more rapidly and more extensively in anti-NBC-treated hearts compared to that in controls. For example, dP/ dt max and dP/ dt min reached 79 and 83% of pre-ischemic value, respectively, in less than 15 min of reperfusion in the presence of antiNBC, whereas the control values remained at less than 30% of the pre-ischemic value (not shown). Heart rate (Fig. 2c) completely recovered to pre-ischemic values

Table 1 Baseline function Group

LVDP (mmHg)

LVEDP (mmHg)

dP/ dt max (mmHg / s)

dP/ dt min (mmHg / s)

HR (beats / min)

Control Anti-NBC

76.262.6 76.862.7

8.760.5 8.860.4

2266679 2140676

1701671 1568666

35169 322611

Data are shown as means6S.E.M. LVDP, left ventricular developed pressure; LVEDP, LV end-diastolic pressure; HR, heart rate.

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

391

Fig. 2. Cardioprotective effect of anti-NBC antibody during and after zero-flow ischemia. Changes in left ventricular end-diastolic pressure, LVEDP (a), left ventricular developed pressure, LVDP (b), heart rate (c), and coronary flow (d) in control (serum-treated, h) and anti-NBC-treated hearts (j) during 30-min of control perfusion, 30-min of global total ischemia and 20-min reperfusion. Both serum or polyclonal anti-NBC were added to the perfusion solution 15 min before ischemia and maintained throughout the reperfusion. Data are shown as means6S.E.M. for n58 experiments. *P,0.05, **P,0.01 anti-NBC-treated versus serum-treated groups. The ischemia period is indicated by the shaded area.

within the first 5 min of reperfusion, and coronary flow (Fig. 2d) overshot the pre-ischemic value. Such an increase in coronary flow at the time of reperfusion may be related to the protection of the cardiac function as usually observed with cardioprotective agents.

3.2.2. Low-flow ischemia During ischemia, LVEDP increased progressively with time (Fig. 3a), while LVDP, dP/ dt max and dP/ dt min rapidly fell to zero (Fig. 3b–d). In the presence of anti-NBC antibody, LVEDP increased more slowly during the ischemic period (Fig. 3a), and the complete abolition of cardiac function as determined by LVDP, dP/ dt max and dP/ dt min , tended to be delayed (Fig. 3b–d). During reperfusion, treatment with antiserum caused all cardiac function parameters to recover substantially more towards pre-ischemic values than no treatment (Fig. 3a–d). At the end of the reperfusion period, t (an index of isovolumic relaxation time) was 99626 ms in treated hearts, but 277678 ms in controls. Furthermore, the extent of lactate dehydrogenase (LDH) release into the coronary

effluent at the time of reperfusion, measured to assess the degree of cell damage following global ischemia and reperfusion, was reduced by anti-NBC treatment (Fig. 4). These data suggest that acute anti-NBC treatment improved contractile recovery and reduced the degree of contracture or irreversible myocardial injury during postischemic reperfusion.

3.3. hhNBC expression is increased in human cardiomyopathic hearts Previous work showed that hhNBC is not measurably expressed in kidney, but is highly expressed in pancreas [17,18] and, to a much lesser extent, in heart [18]. To gain additional insight into the role of the electrogenic NBC in human cardiac pathologies, the expression pattern of hhNBC was measured in normal and in cardiomyopathic hearts. Probes specific for the N-terminus unique to hhNBC were used for Northern experiments illustrated in Fig. 5a. As previously demonstrated, hhNBC mRNA was

392

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

Fig. 3. Cardioprotective effect of anti-NBC during and after low-flow ischemia. Changes in left ventricular end-diastolic pressure, LVEDP (a), left ventricular developed pressure, LVDP (b), maximum rate of contraction (dP/ dt max (c), and maximum rate of relaxation (dP/ dt min ) (d) in control (serum-treated, h) and anti-NBC-treated hearts (j) during 15-min control perfusion, 60-min global low-flow ischemia and 30-min reperfusion. Both serum and anti-NBC antibody were added to the perfusion solution 15 min before ischemia and maintained throughout the reperfusion. Data are shown as means6S.E.M. for n56 experiments. *P,0.05 anti-NBC-treated versus serum-treated groups. The ischemia period is indicated by the shaded area.

not detected in the kidney, but a very faint band at 7.5 kb was observed in normal hearts. In contrast, easily detectable bands were noted in RNA preparations in either left or right ventricle from patients with end-stage heart failure.

Fig. 4. Effect of anti-NBC on ischemia–reperfusion-induced lactate dehydrogenase (LDH) release. Myocardial effluent LDH during preischemic perfusion and at 30 s of reperfusion following 30-min global zero-flow ischemia in control and anti-NBC-treated hearts. Each value is the mean6S.E.M. for n56 experiments.

By densitometry, the signal intensity increased from an average of 1.060.3 arbitrary units in control preparations to 6.961.0 in heart failure (P,0.05). In an additional set of experiments, we used real-time quantitative RT-PCR to investigate expression of hhNBC and NBCn1 in hearts from patients with ischemic and dilated cardiomyopathies. This approach, which is more sensitive than Northern analysis, easily detected hhNBC in normal heart (Fig. 5b). Moreover, in cardiomyopathic hearts, the cycle threshold for detecting hhNBC was significantly reduced. Therefore, confirming the results obtained by Northern analysis, both dilated and ischemic cardiomyopathy increase the expression of hhNBC. The number of copies per ng of total RNA increased from 11546183 in normal to 18776200 in ischemic cardiomyopathy and to 22626156 in dilated cardiomyopathy. In contrast, cardiomyopathy did not change the cycle threshold for detecting NBCn1 (Fig. 5c). Finally, the immunoblot data shown in Fig. 5d confirm that, in human heart failure, the higher hhNBC expression observed at the RNA level was correlated with an increase at the protein level.

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

393

Fig. 5. Overexpression of hhNBC in human cardiomyopathies. (a) Northern blot experiment showing hhNBC mRNA expression in normal and ischemic cardiomyopathy. Lanes: (1) molecular marker; (2) total RNA from kidney (Clontech); (3,4) polyA1 mRNA from normal Heart (Clontech); (5–9) total RNA extracted from left ventricle specimens from human heart with terminal heart failure (donor hearts); (10–16) total RNA extracted from right ventricle specimens from human heart with terminal heart failure (donor hearts). hhNBC (b) and NBCn1 (c) mRNA expression was compared from the left ventricle of human normal, ischemic (ICM) and dilated (DCM) cardiomyopathic hearts using real time quantitative RT-PCR. Data are corrected for individual variation using GAPDH expression and results are expressed as cycle threshold. Western blot (d) showing hhNBC expression levels in normal (N) and failing (HF) human left ventricular myocardium. Representative signals from two normal and two failing hearts are shown.

4. Discussion The regulation of pH i in cardiac myocytes primarily depends on the activity of at least three membrane transporters. While activation of the Cl 2 –HCO 2 ex3

changer acidifies the cell, activation of the NHE and the NBC alkalinises it. Furthermore, Vandenberg and colleagues suggested that during myocardial ischemia metabolic washout, i.e., lactate and CO 2 , may represent a significant contributor to net pH i recovery [32]. Although

394

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

the role of NHE in cardiac pathophysiological states has been widely investigated [5], the role of NBC had been undefined. Characterization of NBC isoforms is still in its early stage, and both electroneutral [6,10] and electrogenic [11–13] Na 1 / HCO 2 3 cotransporters have been identified in heart. At the mRNA level, both an electroneutral NBC [20,21] (NBCn1) and an electrogenic NBC [18] (hhNBC) have been identified in human heart. Here, we have demonstrated, at the protein level, that the electrogenic NBC is expressed in both rat heart (Fig. 1b) and human heart (Fig. 5d), and that the specific inhibition of this electrogenic transporter using a neutralizing antibody protects the heart against ischemic injury. Antibodies to synthetic peptides based on amino acid sequences of a protein have proven useful for understanding the structure, analyzing the molecular forms and studying functional aspects of anion exchangers. Specific anti-Cl 2 / HCO 32 (AE1 isoform) raised against loops of the transporter significantly decreased the exchange function of the AE1 protein [33]. We developed polyclonal antibodies raised against an immunogenic region of hhNBC. Because the synthetic peptide had the same sequence as hhNBC between amino acids 629 and 644, the antibodies are likely to recognise an extracellular loop of the protein. Consistent with our knowledge that the amino acid sequence of this loop in hhNBC [17,18] is extremely similar to the homologous region in the electrogenic rat NBCs [15,19], the anti-NBC antibody recognised a single band in both HEK-293 cells expressing hhNBC and in rat cardiac tissue. Because there is little similarity between the hhNBC peptide used to produce the antibody and the homologous region of the electroneutral NBCs [20] (12.5% identity), and because our antibody did not recognize a second band in Western blots, it is unlikely that the antibody inhibits the electroneutral Na 1 / HCO 2 3 cotransporter. Furthermore, the anti-NBC antibody neutralized NBC activity in both hhNBC-transfected HEK-293 and isolated rat cardiac myocytes. Taken together, these results suggest that an electrogenic NBC is expressed and functional in rat cardiac cells, as previously suggested from electrophysiological approaches [12]. We further demonstrated that, in contrast to what we observed for the electroneutral NBC, the expression level of hhNBC, while rather low in normal human heart, is increased in cardiomyopathic hearts. It could be speculated that the increased expression of the electrogenic NBC would alter the electrophysiological and contractile parameters of the cardiac cells. Although this remains to be further investigated, NBC could contribute to the control of [Na 1 ] i [6,20]. Therefore, the overexpression might lead to an increase in intracellular Na 1 concentration which via activation of the Na 1 / Ca 21 exchange may trigger arrhythmias. Obviously, we cannot exclude an increase in the activity of the electroneutral NBC similar to what has been shown for the

NHE [34] but we could hypothesize that the electrogenic NBC might contribute to the higher incidence of ischemiarelated arrhythmias in the failing heart [35]. The protection afforded by inhibiting hhNBC was observed on systolic contractility and also on diastolic function in both zero- and low-flow global ischemia. Although the molecular mechanism responsible for such a pronounced protection is not defined yet, we suggest that a reduction of intracellular calcium cycling may be involved, as evidenced by the reduced contracture during both ischemia and reperfusion. During ischemia and reperfusion, intracellular acidosis leads to an activation of NBC associated to an increase Na 1 influx. In this respect, recent data clearly indicated that during no-flow ischemia, myocardial [Na] i markedly increased in perfused rat hearts [36]. Increasing [Na 1 ] i will increase reverse Na 1 / Ca 21 exchange and raise [Ca 21 ] i in cardiac cells thereby causing cardiac function, especially diastolic function, to deteriorate during ischemia and reperfusion. We showed that antiNBC exerts beneficial effects during ischemia but more profoundly during early postischemic reperfusion, especially, the heart rate returned to normal immediately at the onset of reperfusion. In this respect, it has been reported that certain interventions at the time of reperfusion or before ischemia may be expected to reduce reperfusion injury or stunning. The reperfusion in isolated ferret hearts undergoing zero-flow ischemia, with acidic perfusate prevent subsequent contractile dysfunction [37]. Moreover, using isolated perfused hearts it was reported that reperfusion with acidotic buffer (pH 6.4 or 6.6) attenuated the gain in tissue Na 1 and Ca 21 contents [38]. Finally, reoxygenation injury produced in rat neonatal cardiac myocytes has been shown to be attenuated when restoration of oxygen was carried out with acidotic medium [39]. Although further studies on [Na 1 ] i and [Ca 21 ] i changes during ischemia–reperfusion remain necessary, our results suggest that anti-NBC-induced protection seems at least in part to be mediated by reduction of the ischemia–reperfusion-induced Ca 21 overload since the time of onset and amplitude of ischemia- and reperfusion-induced contracture were significantly reduced compared to the nontreated heart. Because hhNBC is overexpressed in cardiomyopathic hearts we could hypothesize that this transporter may play a more pronounced role in patients with heart failure. In particular, upregulation of hhNBC expression could contribute to both mechanical dysfunction and arrhythmogenesis in the failing heart. Obviously this needs to be further investigated with the aim of determining whether this overexpression is cause or consequence of the disease. In conclusion, the results of the present study suggest that the electrogenic NBC, hhNBC, like the Na 1 / Ca 21 exchange [40], is overexpressed in heart failure, and that its inhibition could reduce the deleterious effect of intracellular Ca 21 overload, and therefore could be useful

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

additional therapeutic target for a variety of human cardiac diseases, such as myocardial infarction and heart failure. [18]

Acknowledgements [19]

The work was supported by SmithKline Beecham Pharmaceuticals and NIH grants (DK30344). We thank Dr. Jean Franc¸ois Faivre for computational assistance and Marie-Christine Le Boulch for technical assistance.

[20]

[21]

References [1] Metra M, Nodari S, D’Aloia A, Bontempi L, Boldi E, Cas LD. A rationale for the use of b-blockers as standard treatment for heart failure. Am Heart J 2000;139:511–521. [2] Brown NJ, Vaughan DE. Angiotensin-converting enzyme inhibitors. Circulation 1998;97:1411–1420. [3] Clozel M, Breu V, Burri K, Cassal JM, Fischli W, Gray GA, Hirth G, Loffler BM, Muller M, Neidhart W, Ramuz H. Pathophysiological role of endothelin revealed by the first orally active endothelin receptor antagonist. Nature 1993;365:759–761. [4] Frelin C, Vigne P, Ladoux A, Lazdunski M. The regulation of the intracellular pH in cells from vertebrates. Eur J Biochem 1988;174:3–14. [5] Karmazyn M, Gan TX, Humphreys RA, Yoshida H, Kusumoto K. The myocardial Na 1 / H 1 exchange. Structure, regulation, and its role in heart disease. Circ Res 1999;85:777–786. [6] Dart C, Vaughan-Jones RD. Na 1 -HCO 2 symport in the sheep 3 cardiac Purkinje fibre. J Physiol London 1992;451:365–385. [7] Puceat M, Korichneva I, Cassoly R, Vassort G. Identification of band 3-like proteins and Cl 2 / HCO 2 exchange in isolated car3 diomyocytes. J Biol Chem 1995;270:1315–1322. [8] Frelin C, Vigne P, Lazdunski M. The role of the Na 1 / H 1 exchange system in the regulation of the internal pH in cultured cardiac cells. Eur J Biochem 1985;149:1–4. [9] Shimada Y, Hearse DJ, Avkiran M. Impact of extracellular buffer composition on cardioprotective efficacy of Na 1 / H 1 exchanger inhibitors. Am J Physiol 1996;270:H692–700. [10] Lagadic-Gossmann D, Buckler KJ, Vaughan-Jones RD. Role of bicarbonate in pH recovery from intracellular acidosis in the guineapig ventricular myocyte. J Physiol London 1992;458:361–384. [11] Camilion de Hurtado MC, Perez NG, Cingolani HE. An electrogenic sodium-bicarbonate cotransport in the regulation of myocardial intracellular pH. J Mol Cell Cardiol 1995;27:231–242. [12] Aiello EA, Vila Petroff MG, Mattiazzi AR, Cingolani HE. Evidence for an electrogenic Na 1 -HCO 2 3 symport in rat cardiac myocytes. J Physiol London 1998;512:137–148. [13] Camilion de Hurtado MC, Alvarez BV, Perez NG, Cingolani HE. Role of an electrogenic Na 1 -HCO 2 cotransport in determining 3 myocardial pH i after an increase in heart rate. Circ Res 1996;79:698–704. [14] Romero MF, Hediger MA, Boulpaep EL, Boron WF. Expression cloning and characterization of a renal electrogenic Na 1 / HCO 2 3 cotransporter. Nature 1997;387:409–413. [15] Romero MF, Fong P, Berger UV, Hediger MA, Boron WF. Cloning and functional expression of rNBC, an electrogenic Na 1 / HCO 2 3 cotransporter from rat kidney. Am J Physiol 1998;274:F425–F432. [16] Burnham CE, Amlal H, Wang Z, Shull GE, Soleimani MJ. Cloning and functional expression of a human kidney Na 1 :HCO 2 co3 transporter. J Biol Chem 1997;272:19111–19114. [17] Abuladze N, Lee I, Newman D, Hwang J, Boorer K, Pushkin A,

[22]

[23]

[24]

[25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

395

Kurtz I. Molecular cloning, chromosomal localization, tissue distribution, and functional expression of the human pancreatic sodium bicarbonate cotransporter. J Biol Chem 1998;273:17689–17695. Choi I, Romero MF, Khandoudi N, Bril A, Boron WF. Cloning and characterization of a human electrogenic Na 1 -HCO 32 cotransporter isoform (hhNBC). Am J Physiol 1999;276:C576–C584. Bevensee MO, Schmitt BM, Choi I, Romero MF, Boron WF. An electrogenic Na 1 / HCO 2 3 cotransporter (NBC) with a novel COOHterminus, cloned from rat brain. Am J Physiol 2000;278:C1200– C1211. Choi I, Aalkjaer C, Boulpaep EL, Boron WF. An electroneutral sodium / bicarbonate cotransporter NBCn1 and associated sodium channel. Nature 2000;405:571–575. Pushkin A, Abuladze N, Lee I, Newman D, Hwang J, Kurtz I. Cloning, tissue distribution, genomic organization, and functional characterization of NBC3, a new member of the sodium bicarbonate cotransporter family. J Biol Chem 1999;274:16569–16575. Faivre JF, Calmels TPG, Rouanet S, Javre´ JL, Cheval B, Bril A. Characterisation of Kv4.3 in HEK293 cells: comparison with the rat ventricular transient outward potassium current. Cardiovasc Res 1999;41:188–199. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156–159. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning. In: A laboratory manual, 2nd ed., Cold Spring Harbor, NY: Cold Spring Harbour Laboratory, 1989. Rink TJ, Tsien RY, Pozzan T. Cytoplasmic pH and free Mg 12 in lymphocytes. J Cell Biol 1982;95:189–196. Thomas JA, Buchsbaum RN, Zimniak A, Racker E. Intracellular pH measurements in Ehrlich ascites tumour cells utilising spectroscopic probes generated in situ. Biochemistry 1979;18:2210–2218. Khandoudi N, Percevault-Albadine J, Bril A. Consequences of the inhibition of the sarcoplasmic reticulum calcium ATPase on cardiac function and coronary flow in rabbit isolated perfused heart: role of calcium and nitric oxide. J Mol Cell Cardiol 1998;30:1967–1977. Serizawa T, Vogel WM, Apstein CS, Grossman W. Comparison of acute alterations in left ventricular diastolic chamber stiffness induced by hypoxia and ischemia. J Clin Invest 1981;68:91–102. Weiss JL, Frederiksen JW, Weisfeldt ML. Hemodynamic determinants of the time course of fall in canine left ventricular pressure. J Clin Invest 1976;58:751–758. Boron WF, De Weer P. Intracellular pH transients in squid giant axons caused by CO 2 , NH 3 , and metabolic inhibitors. J Gen Physiol 1976;67:91–112. Khandoudi N, Bernard M, Cozzone P, Feuvray D. Mechanisms of intracellular pH regulation during postischemic reperfusion of diabetic rat hearts. Diabetes 1995;44:196–202. Vandenberg JI, Metcalfe JC, Grace AA. Mechanisms of pH i recovery after global ischemia in the perfused heart. Circ Res 1993;72:993–1003. ´ M, Roche S, Vassort G. Src family tyrosine kinase regulates Puceat intracellular pH in cardiomyocytes. J Cell Biol 1998;141:1637– 1646. Yokoyama H, Gunasegaram S, Harding SE, Avkiran M. Sarcolemmal Na 1 / H 1 exchanger activity and expression in human ventricular myocardium. J Am Coll Cardiol 2000;36:534–540. Bril A, Forest MC, Gout B. Ischemia and reperfusion-induced arrhythmias in rabbits with chronic heart failure. Am J Physiol 1991;261:H301–H307. El Banani H, Bernard M, Baetz D, Cabanes E, Cozzone P, Lucien A, Feuvray D. Changes in intracellular sodium and pH during ischaemia-reperfusion are attenuated by trimetazidine. Comparison between low- and zero-flow ischaemia. Cardiovasc Res 2000;47:688–696. Kitakaze M, Weisfeldt ML, Marban E. Acidosis during early

396

N. Khandoudi et al. / Cardiovascular Research 52 (2001) 387 – 396

reperfusion prevents myocardial stunning in perfused ferret hearts. J Clin Invest 1988;82:920–927. [38] Panagiotopoulos S, Daly MJ, Nayler WG. Effect of acidosis and alkalosis on postischemic Ca gain in isolated rat heart. Am J Physiol 1990;258:H821–H828. [39] Bond JM, Herman B, Lemasters JJ. Protection by acidotic pH

against anoxia / reoxygenation injury to rat neonatal cardiac myocytes. Biochem Biophys Res Commun 1991;179:798–803. [40] Studer R, Reinecke H, Bilger J. Gene expression of the cardiac 1 21 Na -Ca exchanger in end-stage human heart failure. Circ Res 1994;75:443–453.