OUP user menu

Disproportionate enhancement of myocardial contractility by the xanthine oxidase inhibitor oxypurinol in failing rat myocardium

Harald Kögler, Heather Fraser, Sylvia McCune, Ruth Altschuld, Eduardo Marbán
DOI: http://dx.doi.org/10.1016/S0008-6363(03)00512-1 582-592 First published online: 1 September 2003


Objective: Xanthine oxidase (XO) inhibitors enhance myofilament Ca2+ responsiveness of normal rat myocardium. We examined whether this inotropic action is preserved or magnified in failing rat myocardium and whether the magnitude of this effect correlates with tissue xanthine-oxidoreductase (XOR) activity. Methods: Hearts of 18–20 month-old SHHF (spontaneous hypertensive/heart failure) rats with end-stage heart failure, as well as of normal control rats, were perfused with the XO inhibitor oxypurinol. Afterwards, [Ca2+]i and tension were measured simultaneously in fura-2-loaded intact isolated right ventricular trabeculae. XOR activity was determined fluorometrically in myocardial homogenates. Results: In failing myocardium, 100 μM oxypurinol significantly increased systolic twitch tension (by 87 and 92% at 1.0 and 1.5 mM extracellular [Ca2+], respectively), without altering [Ca2+]i transient amplitude. Oxypurinol did not alter the midpoint or cooperativity of the steady-state tension-[Ca2+]i relationship, but significantly enhanced maximum Ca2+-activated tension by 75% in failing myocardium. Oxypurinol also exerted a positive inotropic effect in failing myocardium, which was, however, of significantly smaller relative magnitude. Failing rat myocardium exhibited higher XOR activity than nonfailing myocardium, and this activity was largely suppressed in oxypurinol-treated preparations. Conclusions: The magnitude of functional improvement with XOR inhibitors depends on the initial level of XOR activity. Specifically, the inotropic actions of oxypurinol are more pronounced in failing rat myocardium, a tissue that exhibits enhanced XOR activity. Our findings rationalize how XO inhibitors boost cardiac contractility and improve mechanoenergetic coupling, and why the effects might be relatively ‘selective’ for heart failure.

  • Heart failure
  • Contractile function
  • Calcium
  • Inotropic agents
  • Free radicals

Time for primary review 41 days.

This article is referred to in the Editorial by J. Neumann (pages 534–535) in this issue.

1 Introduction

The formation of oxygen-derived free radicals (OFR) is enhanced in congestive heart failure (CHF) [1–3]. An important source of OFR in CHF is the generation of superoxide radical (⋅O2) by myocardial mitochondria [4]. Xanthine oxidase (XO) is also likely to contribute to ⋅O2 generation in CHF; XO and xanthine dehydrogenase (XDH) are isoenzymes of xanthine oxidoreductase (XOR), which catalyzes the oxidation of hypoxanthine and xanthine to urate during purine catabolism in mammals. While XDH preferentially transfers the electrons released during the oxidation process to NAD+, XO exclusively utilizes molecular oxygen, thereby generating ⋅O2 [5]. ⋅O2 may exert detrimental effects on cardiac function either by transformation to the highly-reactive hydroxyl radical (⋅OH) via the Haber–Weiss reaction [6] or by reacting with NO to form peroxynitrite, which can cause protein nitration [7]. The latter mechanism reduces NO bioavailability, thereby limiting cardiovascular effects of this compound.

Pérez et al. reported that treatment of trabeculae isolated from nonfailing or stunned rat hearts with a cocktail of the XO-inhibitor allopurinol and the ⋅OH scavenger N-(2-mercaptopropionyl)-glycine (MPG) substantially improved contractility by enhancing Ca2+ responsiveness [8]. Replacing allopurinol with oxypurinol (synonym: alloxanthine), the active metabolite of allopurinol that blocks XO activity by binding tightly at its active site [9], caused similar effects. However, in vivo studies in dogs indicated that hemodynamic function was only slightly [10], if at all [11], improved after i.v. administration of allopurinol to nonfailing animals. In dogs with rapid ventricular pacing-induced heart failure, however, the inotropic effects afforded by allopurinol were unambiguous, and contractile efficiency was enhanced [10,11]. The observed improvements of contractile function and myocardial efficiency were associated with enhanced XO activity in failing dog myocardium, compared to nonfailing controls [10]. In human patients with idiopathic dilated cardiomyopathy, intracoronary infusion of allopurinol likewise increased myocardial efficiency [12], an effect which consisted of markedly reduced oxygen consumption in the face of a modest (statistically insignificant) rise in contractile function. Taken together, the available data hint that the effect of XO-inhibitors may be magnified in the failing heart, but a direct comparison has not been made at the muscle level.

Here, we used spontaneously hypertensive/heart failure (SHHF/Mccfacp) rats (henceforth referred to as SHHF rats) as a CHF model and tested whether treatment with an XO-inhibitor exerts more pronounced inotropic effects in failing than in nonfailing myocardium, and whether this is associated with different levels of tissue XO/XOR activity. Lean male SHHF rats heterozygous for the facp (corpulent) gene, which encodes a defective leptin receptor, exhibit arterial hypertension and eventually develop CHF as a consequence of chronic pressure overload at about 18 months of age [13]. SHHF-rat myocardium exhibits defective Ca2+ cycling and contractile dysfunction [14] as well as excitation–contraction coupling abnormalities [15] resembling those present in human CHF. A preliminary report of this work has appeared [16].

2 Methods

2.1 Muscle preparation and force measurement

All experimental procedures conformed the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996) and followed previous descriptions [17]. Lean male SHHF rats were sacrificed at age 18–20 months as soon as physical examination revealed clear clinical signs of CHF (tachypnea/dyspnea, cyanosis, lethargy, peripheral edema). Age-matched Wistar–Furth rats and 10-week-old LBN-F1 rats served as controls. Since none of the functional parameters determined exhibited any significant difference between the two control rat groups, data from both groups were pooled for statistical analysis. Animals were anesthetized by intraperitoneal injection of sodium pentobarbital (∼50 mg) and hearts quickly removed and mounted for retrograde perfusion with a high-potassium (20 mM) modified Krebs–Henseleit (K–H) solution, equilibrated with 95% O2/5% CO2, containing (mM): Na+ 142, K+ 20, Mg2+ 1.2, Ca2+ 0.5, Cl 142, PO43− 2, SO42− 1.2, HCO3 20, glucose 10, pH 7.37–7.43. Trabeculae were microdissected from the right ventricles and mounted in a superfusion bath between a force transducer (type AE 801, SensoNor, Horten, Norway) and a hook connected to a micromanipulator for length adjustment. All measurements were carried out at room temperature (22–23°C). Preparations were superfused at 10 ml min−1 with the same K–H solution except for [K+] 5 mM and [Cl] 127 mM and excited by bipolar electrical field stimulation, amplitude 5–10 V, pulse width 5 ms (Grass SD9 Stimulator; Grass Instruments, Quincy, MA, USA), at 0.5 Hz. After equilibration, muscles were stretched to the length at which developed twitch tension was maximal (Lmax). To obtain a steady-state tension-[Ca2+]i relation, tetanic contractions were elicited by intermittent 10-Hz stimulation in the presence of 5 μM ryanodine at various extracellular Ca2+ concentrations ([Ca2+]o; range: 0.25–18.0 mM), thereby achieving an equilibrium between the intracellular free Ca2+ concentration ([Ca2+]i) and the level of myofilament activation.

2.2 Measurement of [Ca2+]i

Fura-2 pentapotassium salt (Molecular Probes, Eugene, OR, USA) was microinjected iontophoretically into 3–5 cells of each preparation and allowed to spread evenly throughout the preparation via gap junctions [17]. Fura-2 fluorescence was excited at 340 and 380 nm, and the emitted light signal was collected at 510 nm by a photomultiplier tube (R2693, Hamamatsu, Bridgewater, NJ, USA), filtered at 100 Hz, collected by an A/D converter, and stored on a computer for subsequent analysis. Raw fluorescence data were corrected for autofluorescence and [Ca2+]i was calculated using the following equation [18]: [Ca2+]i=Kd (RRmin)/RmaxR), where R is fluorescence ratio (340 nm/380 nm), Kd is the apparent dissociation constant of fura-2, Rmax is the fluorescence ratio (340 nm/380 nm) at saturating [Ca2+], and Rmin is the ratio (340 nm/380 nm) at zero [Ca2+]. Values for Kd, Rmax, and Rmin were obtained as reported previously [19].

2.3 Administration of oxypurinol

Oxypurinol (Sigma, St. Louis, MO, USA) was dissolved in 1 M NaOH (300 mM stock) and diluted into the K–H solution to a final concentration of 100 μM. The resulting minor alkalinization of the solution was fully compensated by bubbling with 95% O2/5% CO2. In the oxypurinol-treated groups, oxypurinol was present from onset of perfusion of the excised heart, which was maintained for 15 min before starting to dissect a trabecula, and throughout the entire experiment. In the drug-free groups, an aliquot of solvent (1 M NaOH) was added at the same time.

2.4 Measurement of xanthine oxidase/xanthine oxidoreductase activity

XO/XOR activity in tissue homogenates was determined using a fluorometric assay measuring the conversion of pterin to isoxanthopterin [20]. At 345 nm excitation, formation of isoxanthopterin can be monitored by the increase in fluorescence emitted at 390 nm. Tissue samples of 25–50 mg were obtained from the free right ventricular wall immediately prior to dissection of a trabecula for force measurement, shock-frozen in liquid N2, and stored at −80°C until use. In oxypurinol-treated preparations, hearts had been exposed to 100 μM oxypurinol for 15 min prior to collection of the samples. Frozen samples were homogenized in 1:4 (w/v) of ice-cold HEPES buffer (10 mM) containing sucrose (320 mM), DTT (1 mM), leupeptin (10 μg/ml), soybean trypsin inhibitor (10 μg/ml), aprotinin (2 μg/ml), EDTA (0.1 mM) and PMSF (0.1 mM), and centrifuged at 13 000 rpm for 35 min at 4°C. Reducing conditions and the presence of protease inhibitors are required to prevent spontaneous conversion of XDH to XO. Baseline was measured after mixing the supernatant (25 μl) with KH2PO4 buffer (50 mM, pH 7.2) in a quartz cuvette. XO activity was calculated from the slope of the linear increase in fluorescence obtained during 2 min following the addition of the substrate, pterin (10 μM). Then, methylene blue (10 μM), which is used as an electron acceptor by XDH, was added and the subsequent increase in fluorescence over time was used to calculate the combined activities of XO and xanthine dehydrogenase (XDH) (total XOR activity). The reaction was stopped by the addition of allopurinol (50 μM). For calibration, the step change in fluorescence intensity upon addition of a standard concentration of isoxanthopterin was measured. Values are expressed as pmol min−1 mg protein−1, with protein concentrations determined using the Lowry protein assay (Bio-Rad).

2.5 Data analysis and statistics

All force values were normalized to the cross-sectional area of each muscle (width×thickness×π/4) and are expressed as tension in mN/mm2. To characterize the steady-state tension-[Ca2+]i relationship, tension and [Ca2+]i data obtained during tetanic contractions were fitted to the Hill equation: FXFmin=(FmaxFmin) [CaX]n/(Ca50n+[CaX]n), where FX is actual tension, Fmin is resting tension, Fmax is maximal Ca2+-activated tension, [CaX] is actual [Ca2+]i, Ca50 is the [Ca2+]i required for half-maximal tension development, and n is the Hill coefficient. The relative magnitudes of inotropic effects in SHHF and failing preparations were compared after normalizing the tension data from oxypurinol-treated preparations with respect to the mean absolute tension of the respective drug-free group under the same conditions. Statistical significance was assessed using unpaired Student’s t-test. P<0.05 was considered statistically significant. Pooled data are expressed as means±S.E.M.

3 Results

3.1 Effects of oxypurinol on contractility in failing myocardium

Hearts of SHHF rats (n=13) were enlarged, and all but one animal showed pleural effusions, liver congestion, and ascites, while such clinical signs of heart failure were completely absent in control rats (n=16). Body weight showed a tendency to decrease in SHHF compared to age-matched control rats (461±44 vs. 682±113 g, respectively, P=n.s.). Both heart weight:body weight and lung weight:body weight ratios were significantly increased in SHHF, compared to age-matched control rats (Fig. 1), indicating severe myocardial hypertrophy and lung congestion. These parameters were similar in SHHF rats assigned to the oxypurinol-treated or drug-free groups, indicating comparable severity of heart failure.

Fig. 1

Heart weight:body weight and lung weight:body weight ratios. Black bars: Age-matched controls (n=16); grey bars: SHHF rats (n=13). Both hypertrophy/heart failure markers are significantly enhanced in SHHF rats, compared to age-matched control rats.

We simultaneously examined calcium cycling and tension development during twitches of intact right ventricular trabeculae from drug-free and oxypurinol-treated SHHF rat hearts at various [Ca2+]o (range 0.5–1.5 mM). Fig. 2 shows representative original recordings. Due to the hypertrophy and ventricular remodeling, not every SHHF heart yielded long, thin trabeculae suitable for microinjection and tension measurement, but such preparations could be obtained from five drug-free and four oxypurinol-treated failing hearts. In these preparations, 100 μM oxypurinol did not alter the amplitude of [Ca2+]i transients or diastolic [Ca2+]i (Fig. 3A). Nevertheless, twitch tension was significantly enhanced in oxypurinol-treated preparations (n=4) at [Ca2+]o of 1.0 and 1.5 mM by 87% and 92%, respectively, compared to drug-free failing trabeculae (n=5; Fig. 3B). Neither the time from peak tension to half relaxation of the twitch (186±24 drug free, n=5, vs. 223±20 ms oxypurinol treated, n=4) nor the time for [Ca2+]ito decrease from peak to 50% (265±38 drug free, n=5, vs. 184±11 ms oxypurinol treated, n=4) were significantly different between groups. Oxypurinol did not affect resting tension. During tetanic contractions induced by 10-Hz stimulation in the presence of 5 μM ryanodine, steady-state tension was enhanced in oxypurinol-treated preparations despite similar steady-state [Ca2+]i levels (Fig. 4A). Both oxypurinol-treated and drug-free failing rat heart preparations exhibited a sigmoidal steady-state tension-[Ca2+]i relationship that was well-fit by the Hill equation (Fig. 4B). While oxypurinol did not affect resting tension, EC50, and Hill coefficient, maximum Ca2+-activated tension nearly doubled from 46.6±3.9 (n=5) to 81.6±6.7 mN/mm2 (n=4; P=0.006), indicating that a markedly elevated maximum force-generating capacity, rather than enhanced Ca2+ sensitivity, underlies the increase in twitch tension afforded by oxypurinol.

Fig. 4

Effect of oxypurinol on [Ca2+]i and tension during tetani in failing rat myocardium. (A) Representative original [Ca2+]i (top) and tension (bottom) recordings from RV trabeculae obtained from untreated (left panels) and oxypurinol-treated (right panels) SHHF hearts at the designated [Ca2+]o. While the [Ca2+]i levels attained during tetani are similar, the oxypurinol-treated preparation develops higher tension. (B) Steady-state tension-[Ca2+]i relationship. Filled symbols, drug-free failing myocardium (n=5); open symbols, oxypurinol-treated failing myocardium (n=4). Maximum Ca2+-activated tension is increased without alterations in Ca2+ sensitivity. Invisible error bars indicate that error is smaller than the symbol. P value indicates difference between the mean maximum isometric tension values.

Fig. 3

Effect of oxypurinol on Ca2+ cycling and tension in failing rat myocardium; group data. (A) Peak systolic (squares) and resting (circles) [Ca2+]i during twitches. (B) Peak systolic (squares) and resting (circles) tension during twitches. (A,B) Filled symbols, drug-free failing myocardium (n=5); open symbols, oxypurinol-treated failing myocardium (n=4). Oxypurinol has no effect on Ca2+ cycling but significantly increases force development. Invisible error bars indicate that error is smaller than the symbol; n.s., not significant.

Fig. 2

Effect of oxypurinol on Ca2+ cycling and tension in failing rat myocardium. Original [Ca2+]i (top) and tension (bottom) recordings from RV trabeculae obtained from untreated (left panels) and oxypurinol-treated (right panels) SHHF hearts at 1.0 mM [Ca2+]o. Both [Ca2+] and tension traces exhibit the slow kinetics typically observed in failing myocardium. While the Ca2+-transient amplitude is similar, developed twitch tension is substantially higher in the oxypurinol-treated preparation.

3.2 Effects of oxypurinol on contractility in normal rat myocardium

We carried out the same experimental protocol in isolated trabeculae from nonfailing rat hearts. Original recordings are shown in Fig. 5. Also in these preparations (n=8 in both the drug-free and oxypurinol-treated groups), treatment with 100 μM oxypurinol did not significantly alter [Ca2+]i transient amplitudes or resting [Ca2+]i levels (Fig. 6A). Twitch tension, however, was significantly enhanced by 35% due to oxypurinol treatment at 1.5 mM [Ca2+]o (Fig. 6B). Neither the time from peak tension to half relaxation of the twitch (196±27 drug free, n=8, vs. 203±27 ms oxypurinol treated, n=8) nor the time for [Ca2+]ito decrease from peak to 50% (128±11 drug free, n=8, vs. 154±16 ms oxypurinol treated, n=8) were significantly different between groups. During tetanic contractions (representative examples shown in Fig. 7A) steady-state tension development was enhanced in oxypurinol-treated preparations, while steady-state [Ca2+]i levels were unaltered. Assessment of the steady-state tension-[Ca2+]i relationship in nonfailing rat preparations revealed an increase in maximum Ca2+-activated tension from 114.4±5.2 (n=8) to 143.3±4.0 mN/mm2 (n=8, P<0.001). As in failing rat heart preparations, resting tension, EC50, and Hill coefficient were not affected by oxypurinol (Fig. 7B). Thus, also in nonfailing rat myocardium oxypurinol exerts a positive inotropic effect by enhancing the maximum force-generating capacity without affecting Ca2+ sensitivity, but the effects of oxypurinol in control muscles are much less striking than in SHHF muscles. In the next section, we compare the effects quantitatively.

Fig. 7

Effect of oxypurinol on [Ca2+]i and tension during tetani in nonfailing rat myocardium. (A) Representative original [Ca2+]i (top) and tension (bottom) recordings from RV trabeculae obtained from untreated (left panels) and oxypurinol-treated (right panels) SHHF hearts at the designated [Ca2+]o. While the [Ca2+]i levels attained during tetani are similar, the oxypurinol-treated preparation develops higher tension. (B) Steady-state tension-[Ca2+]i relationship. Filled symbols, drug-free nonfailing myocardium (n=8); open symbols, oxypurinol-treated nonfailing myocardium (n=8). Like in failing myocardium, also in nonfailing control myocardium an increase in maximum Ca2+-activated tension accounts for the positive inotropic effect, without alterations in Ca2+ sensitivity. Invisible error bars indicate that error is smaller than symbol. P value indicates difference between the mean maximum isometric tension values.

Fig. 6

Effect of oxypurinol on Ca2+ cycling and tension in nonfailing control rat myocardium; group data. (A) Peak systolic (squares) and resting (circles) [Ca2+]i during twitches. (B) Peak systolic (squares) and diastolic (circles) tension during twitches. (A,B) Filled symbols, drug-free myocardium (n=8); open symbols, oxypurinol-treated myocardium (n=8). Oxypurinol has no effect on Ca2+ cycling but significantly increases force development. Invisible error bars indicate that error is smaller than symbol; n.s., not significant.

Fig. 5

Effect of oxypurinol on [Ca2+]i and tension during twitches in nonfailing rat myocardium. Original [Ca2+]i (top) and tension (bottom) recordings from RV trabeculae obtained from untreated (left panels) and oxypurinol-treated (right panels) age-matched control hearts at 1.0 mM [Ca2+]o. While the Ca2+ transient amplitude is similar, the oxypurinol-treated preparation develops higher twitch tension. However, the oxypurinol-induced enhancement of contractility is less pronounced than in the failing group.

3.3 Comparison between failing and nonfailing rat myocardium

In nonfailing preparations treated with 100 μM oxypurinol, maximum Ca2+-activated tension was 25.2±3.5% higher than in drug-free muscles (n=8 in each group), while in oxypurinol-treated trabeculae from SHHF rats (n=4) the increase in maximum Ca2+-activated tension was 75.0±14.3% compared to drug-free preparations (n=5). Thus, the fractional increase in tension was 3-fold greater in failing than in nonfailing myocardium (P=0.036). Maximum Ca2+-activated tension in drug-free SHHF preparations (n=5) was 41±8% of that developed by drug-free nonfailing preparations (n=8), indicating severe contractile dysfunction in the failing group. Oxypurinol-treated preparations (n=4), on the other hand, exhibited a maximum Ca2+-activated tension equaling 72±12% of that developed on average by drug-free nonfailing control trabeculae (n=8). Thus, although oxypurinol treatment resulted in substantial improvement of contractility in failing myocardium, contractile function still remained impaired compared to drug-free nonfailing myocardium (P=0.007).

3.4 Xanthine oxidase/xanthine oxidoreductase activity

We determined tissue XO and XOR activity in right ventricular homogenates from drug-free failing and nonfailing hearts as well as failing hearts treated with oxypurinol (Fig. 8). The total XOR activity (i.e. combined XO and XDH activities) in RV from drug-free failing SHHF rat hearts (n=8) was significantly enhanced compared to control myocardium (n=5) (P=0.044; Fig. 8A). XO activity tended to be almost 2-fold higher in failing (n=8) than in nonfailing myocardium (n=5) (P=0.07; Fig. 8B). Both XOR and XO activities were largely suppressed in oxypurinol-treated SHHF rat myocardium (n=5) (Fig. 8A and B).

Fig. 8

Xanthine oxidase/xanthine oxidoreductase activities in tissue homogenates. (A) XO activities. (B) XOR activities. Black bar: drug-free control myocardium (n=5); grey bar: drug-free SHHF rat myocardium (n=8); empty bar: oxypurinol-treated SHHF rat myocardium (n=5). XOR activity is significantly higher in failing than in nonfailing myocardium. Oxypurinol treatment largely suppresses both XO and XOR activity in failing rat myocardium.

4 Discussion

In the present study, treatment of rat failing and nonfailing isolated myocardium with the XO inhibitor oxypurinol increased twitch force without altering [Ca2+]i transient amplitude. The steady-state tension-[Ca2+]i relationship revealed that this inotropic effect was ascribable to an increase in maximum Ca2+-activated tension, without alterations in Ca2+ sensitivity. Pérez et al. first reported that intact cardiac trabeculae isolated from nonfailing rat hearts, when treated with the XO inhibitor allopurinol in combination with the hydroxyl radical scavenger MPG, exhibited greatly improved contractile function due to enhancement of Ca2+ responsiveness [8]. In that study, in addition to the inotropic effect, a reduction in [Ca2+]i transient amplitude resulted from treatment with the allopurinol/MPG cocktail. In the present investigation, no such effect on Ca2+ handling was observed after treatment with oxypurinol alone. The differential effects on Ca2+ cycling may arise from the absence of MPG in the present study, or from the use of oxypurinol rather than the parent compound allopurinol, but this was not tested here. Nevertheless, the central finding remains an enhancement in myofilament Ca2+ responsiveness with XO-inhibitor treatment, resulting in a positive inotropic effect unaccompanied by a rise in activator calcium. These inotropic effects of oxypurinol were qualitatively similar in failing and nonfailing myocardium. We found, however, contractility to be improved to a markedly greater extent in preparations from failing hearts.

In order to understand this apparent ‘selectivity’ for failing myocardium one needs to consider potential mechanisms mediating the inotropic effects exerted by oxypurinol: Apart from their XO-inhibitory properties, allopurinol and oxypurinol are direct (albeit weak) ⋅OH scavengers [21]. Furthermore, other presently unidentified direct actions of the drug on the myofibrillar proteins could account for the enhancement of myofilament Ca2+ responsiveness observed here and previously [8]. Elevated ⋅O2 levels are found in failing myocardium, and both mitochondria and xanthine oxidase have been implicated in their generation [4]. ⋅O2 and NO react with each other, thereby forming peroxynitrite. Therefore, inotropic effects of XO inhibition could either result from reduced generation of ⋅O2 or peroxynitrite, or from increased bioavailability of NO. In dogs with pacing-induced heart failure, blockade of myocardial iNOS-dependent NO formation increased cardiac output [22], which argues against the notion that XO-inhibition exerts positive inotropic effects by increasing NO bioavailability. Pretreatment of dogs with the nonisoform specific NOS inhibitor, l-NMMA, on the other hand, completely abolished the inotropic and energy-sparing effects of allopurinol [23], indicating that reducing ⋅O2 formation alone was not sufficient to improve contractile function. Furthermore, exogenously applied peroxynitrite reduced contractility in isolated rat cardiac myocytes [24]. Taken together, these studies point to peroxynitrite as a potential culprit in the impairment of force generation and energetic efficiency in CHF models.

Thus, it is conceivable that improvement of contractile function by oxypurinol is due to its property of inhibiting XO. If this is the case, the magnitude of the inotropic effect will be affected by the level of XO activity accessible to blockade by XO-inhibitors. Consistent with this hypothesis, intravenous administration of allopurinol substantially improved hemodynamic function and energetic efficiency in dogs suffering from rapid ventricular pacing-induced CHF, while in nonfailing control dogs an only minor inotropic effect was detectable [10,11]. These differential effects were associated with enhanced XO activity in failing dog myocardium, compared to nonfailing controls [10]. Elevation of myocardial xanthine oxidoreductase activity has been described in several rat CHF models, e.g. monocrotaline-induced right ventricular failure and postmyocardial infarction failure [25]. In the present study, we used a different model of heart failure in the rat, the spontaneous hypertensive/heart failure rat. Failing animals had significantly higher levels of XOR activity than nonfailing controls, and XO isoenzyme activity tended to be considerably enhanced. The inotropic action of oxypurinol was more pronounced in failing than in nonfailing myocardium, and this improvement of contractile function was associated with nearly complete suppression of the elevated XO/XOR activity by oxypurinol in the failing group. We have no reason to assume that the XO/XOR-inhibitory action of oxypurinol might differ between groups, although we did not test whether XO/XOR activity was suppressed to a similar extent by oxypurinol in the nonfailing group.

An interesting question remains as to why in our study nonfailing myocardium showed a clear inotropic response to XO inhibition (albeit weaker than in the failing myocardium), while other studies testing the effects of allopurinol in vivo found only subtle effects in nonfailing controls. A critical aspect when considering these differences may be the species used. In failing dog myocardium XO activity was about 5-fold that exhibited by failing myocardium [10], while in our study using rat the difference was only 2-fold. Higher intrinsic XO activity in the nonfailing rat myocardium, compared with dog, could be an explanation. The presence of higher XO activity in nonfailing rat myocardium, in turn, could explain why XO inhibition results in greater improvement of contractility in control rat myocardium. This was, however, not tested here, and a study directly comparing XO expression in rat and dog to our knowledge has not yet been published.

Another substantial difference between this and previous studies is that we, for the first time, report a difference in the extent of inotropic effects of XO-inhibition between failing and nonfailing hearts at the level of the isolated myocardium. When allopurinol is applied in vivo [10,11], vascular effects of XO-inhibition may have interfered with the observed improvement of hemodynamic parameters. XO inhibitors enhance nitric oxide (NO)-dependent vasodilation, presumably due to an increase in NO bioavailability subsequent to reduction in ⋅O2 levels [26–28]. Thus, when hemodynamic parameters are being assessed as a measure of cardiac contractile function, changes in intrinsic myocardial contractility on the one hand and in pre- and/or afterload on the other hand may have offset each other. Moreover, the vascular response to oxypurinol is augmented under conditions of increased oxidative stress, which was discussed to be due to increased endogenous formation of peroxynitrite subsequent to ⋅O2formation under these circumstances [29]. This could certainly apply in heart failure.

The present study identifies an important action of oxypurinol, which is magnified in the failing myocardium, but information regarding the biochemical basis of these effects remains to be obtained. Release of endogenous catecholamines from intracardiac stores does not seem to play a role, since in this case an increase in tension development would be accompanied by a considerable rise in [Ca2+]i transient amplitude and enhanced twitch and [Ca2+]i transient kinetics, which we did not observe in this study. It is clear from previous work that the effects of XO inhibition are tied to NO metabolism, since nonisoform specific inhibition of NOS abolished the inotropic effects of allopurinol in vivo [23]. We selected a concentration of oxypurinol (100 μM) readily achieved in human patients treated for gout with 300 mg of allopurinol [30]. Since the effects of NO are known to be biphasic, it would certainly be logical and worthwhile to test whether lower concentrations of the XO inhibitor (e.g. 10 and 30 μM) exert a qualitatively similar and graded response. However, such experiments are beyond the scope of the present study, whose major purpose was to establish the preferential response in the failing heart and to investigate its basis at the level of E–C coupling.

In evaluating the therapeutic prospects for XO inhibitors in heart failure, three considerations are noteworthy. First, the inotropic effects are much enhanced in failing myocardium, as is basal XOR activity. This means that the potential benefits may be rather ‘selective’ for heart failure [10,11], a property unlike that of any other inotropic agent: Phosphodiesterase inhibitors, for example, boost contractility less in human and canine heart failure than in nonfailing controls [31,32]. The unique selectivity of XO inhibitors emphasizes that such therapy is targeted at a mechanism specifically associated with heart failure, rather than the symptomatic therapy which is achieved with conventional inotropic agents. Second, the increase in myofilament responsiveness is not accompanied by an increase in calcium transients (this study and Ref. [8]). This finding rationalizes the observed increase of myocardial contractile efficiency in vivo [10]: if less calcium cycling suffices to produce more force, less total energy will be expended during contractile activation. Such an effect explains how XO inhibitors ameliorate the mechanoenergetic uncoupling seen in heart failure [10]. Third, the boost of myofilament function involves neither a rise in resting tension nor a sensitizing shift in [Ca2+]50; thus, diastolic tone is not increased, nor is relaxation impaired. Overall, this study suggests that XO inhibition represents an especially promising therapeutic approach to improve cardiac contractile function in heart failure.


This work was supported by a fellowship from the Deutsche Forschungsgemeinschaft (DFG) to HK (KO 1873/1-1), by the Alberta Heritage Foundation (fellowship to HF), and by grants of the National Heart, Lung and Blood Institute of the NIH, R01 HL48835 (to RA) and R01 HL44065 (to EM). EM holds the Michel Mirowski, M.D. Professorship of Cardiology of the Johns Hopkins University.


  • 1 Present address: CV Therapeutics, Palo Alto, CA, USA.

  • 2 Present address: Myogen, Westminster, CO, USA.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
View Abstract