Nitric oxide can acutely modulate its biosynthesis through a negative feedback mechanism on l-arginine transport in cardiac myocytes

Abstract

Nitric oxide (NO) plays a central role as a cellular signaling molecule in health and disease. In the heart, NO decreases the rate of spontaneous beating and the velocity and extent of shortening and accelerates the velocity of relengthening. Since the cationic amino acid l-arginine (l-Arg) is the substrate for NO production by NO synthases (NOS), we tested whether the transporters that mediate l-Arg import in cardiac muscle cells represent an intervention point in the regulation of NO synthesis. Electrical currents activated by l-Arg with low apparent affinity in whole cell voltage-clamped rat cardiomyocytes were found to be rapidly and reversibly inhibited by NO donors. Radiotracer uptake studies performed on cardiac sarcolemmal vesicles revealed the presence of high-affinity/low-capacity and low-affinity/high-capacity components of cationic amino acid transport that were inhibited by the NO donor S-nitroso-N-acetyl-dl-penicillamine. NO inhibited uptake in a noncompetitive manner with K_i values of 275 and 827 nM for the high- and low-affinity component, respectively. Fluorescence spectroscopy experiments showed that millimolar concentrations of l-Arg initially promoted and then inhibited the release of endogenous NO in cardiomyocytes. Likewise, l-Arg currents measured in cardiac myocytes voltage clamped in the presence of 460 nM free intracellular Ca²⁺, a condition in which a Ca-CaM complex should activate endogenous NO production, showed fast activation followed by inhibition of l-Arg transport. The NOS inhibitor N-nitro-l-arginine methyl ester, but not blockers of downstream reactions, specifically removed this inhibitory component. These results demonstrate that NO acutely regulates its own biosynthesis by modulating the availability of l-Arg via cationic amino acid transporters.

nitric oxide (NO) is a signaling molecule that plays major regulatory roles in the cardiovascular system (6). NO biosynthesis is mediated by NO synthase (NOS), a dioxygenase that uses 1.5 mol of NADPH and 2 mol of O₂ in the oxidation of a guanidino nitrogen from its substrate l-arginine (l-Arg) to produce NO and citrulline (10). This reaction requires flavin mononucleotide (FMN), FAD, tetrahydrobiopterin (BH₄), and, in the constitutively expressed endothelial (eNOS) and neuronal (nNOS) isoforms of the enzyme, a Ca-CaM complex (6, 26).

The main molecular target of NO is a soluble guanylyl cyclase (sGC), a heme-containing enzyme that, upon NO binding, increases several hundredfold the rate of synthesis of cGMP from GTP (24). NO synthases are among the most highly regulated enzymes in biology (5). In particular, endogenous NO as well as NO donors reversibly inhibit NOS activity (1, 29, 32) in a reaction that involves direct NO modification of −SH residues in the enzyme (28). This inhibition, referred to as feedback autoregulation in some of these reports, represents the classical mechanism of enzyme inhibition by a reaction product (30).

In the heart, ventricular myocyte-generated NO has divergent effects on cardiac function, depending on the NOS isoform and its subcellular location (5). NO produced by caveolar eNOS promotes cGMP synthesis via sGC, which is used by cGMP-dependent protein kinases to phosphorylate L-type calcium channels. Phosphorylation has an inhibitory effect on these channels thus decreasing β-adrenergic-induced myocardial contractile performance (2, 3). On the other hand, nNOS appears to localize to the sarcoplasmic reticulum in cardiomyocytes and interacts with ryanodine receptors. nNOS-generated NO stimulates calcium release from the sarcoplasmic reticulum via direct activation of these receptors, an effect that increases myocardial contractility (5). Thus, a fine-tuned modulation of NO synthesis is crucial for proper cardiac function.

Cellular NO production is absolutely dependent on l-Arg availability. A pertinent question then relates to the sources of this amino acid in cardiac myocytes. Endogenous de novo synthesis of l-Arg only takes place in liver and kidneys (23). Nonetheless, some NO-producing cell types such as vascular endothelial cells (12) and macrophages (33) are capable of recycling l-Arg from citrulline pools. Cardiac muscle cells, on the other hand, lack some of the enzymes required for l-Arg synthesis or recycling from citrulline (11, 25, 31). Instead, these cells must import l-Arg from the circulation to ensure adequate intracellular levels of this amino acid. Cationic amino acids appear to enter cardiomyocytes through the activity of a family of transporters known as system y⁺ (17, 27, 31). These plasma membrane-bound carriers transport the l-enantiomers of arginine, lysine (l-Lys), and ornithine (l-Orn) with similar efficiency and in a Na⁺-independent manner (8). We have recently found two l-Lys uptake components in cardiac sarcolemmal vesicles, with functional properties that are consistent with the activity of the high-affinity, low-capacity CAT-1, and the low-affinity, high-capacity CAT-2A members of system y⁺ (17). These transporters are likely to play a key role in the various metabolic pathways that involve l-Arg and l-Lys (18). In particular, the activity of these carriers might represent a critical regulatory point for cardiac NO synthesis and metabolism.

The present work tested whether NO synthesis is acutely self-regulated by an inhibitory mechanism of this gas on l-Arg transport. We found two components of cationic amino acid transport that are noncompetitively inhibited by S-nitroso-N-acetyl-dl-penicillamine (SNAP)-derived NO in cardiac sarcolemmal vesicles. Furthermore, l-Arg transport is shown to be blocked in real time by NOS-derived NO in cardiac muscle cells. This novel mechanism is unique in that, as the product of a downstream enzymatic reaction, NO can acutely modulate the activity of the transporters that supply the substrate for its biosynthesis.

MATERIALS AND METHODS

Adult male or female Sprague-Dawley rats were injected with pentobarbital sodium (Nembutal, 100 mg/kg ip), and hearts were removed under complete anesthesia. Animal protocols were reviewed and approved by the Institutional Animal Care and Use Committee of the New Jersey Medical School. For voltage-clamp and fluorescence studies, ventricular myocytes were enzymatically isolated from a single heart according to published methods (21). For uptake experiments, ventricles from three hearts (1.2–1.4 g) were used to prepare fresh giant sarcolemmal vesicles.

Voltage-clamp experiments.

Freshly isolated ventricular cardiomyocytes were placed in a superfusion chamber on the stage of an inverted microscope and superfused at 36 ± 1°C with a Tyrode solution containing (in mM) 145 NaCl, 5 KCl, 10 dextrose, 2 CaCl₂, 1 MgCl₂, and 10 HEPES-NaOH, pH 7.40 (23°C). Myocytes were whole cell voltage clamped with low-resistance (0.9–1.3 MΩ) patch electrodes back-filled with an intracellular salt solution containing (in mM) 110 potassium aspartate, 20 tetraethylammonium chloride (TEACl), 4 MgCl₂, 0.7 MgATP, 10 EGTA-Tris, 5 glucose, and 10 HEPES-KOH, pH 7.30 (23°C). After establishment of a gigaohm seal, the superfusion solution was switched to a Na⁺- and K⁺-free solution containing (in mM) 145 tetramethylammonium chloride (TMACl), 2.3 MgCl₂, 0.2 CdCl₂, 5.5 dextrose, and 10 HEPES-Tris, pH 7.40 (23°C). l-Arg- and l-Lys-containing solutions were prepared by equimolar substitution of TMA. Extracellular TMA and Cd²⁺ as well as intracellular TEA were added to block contaminating ionic currents. Cells were exposed to these blockers for 5 min before further manipulations.

The release of NO from sodium pentacyanonitrosyl ferrate(III) dihydrate (SNP) is triggered by visible light and decays with a half-time of ∼60 min (14). These characteristics make SNP a convenient NO donor for voltage-clamp experiments. Superfusion solutions containing 1 mM SNP were protected from light and prepared fresh several times in any given experimental day. SNAP, on the other hand, produces NO spontaneously at physiological pH and decays faster than SNP (14). Therefore, for those electrophysiological experiments involving this NO donor, SNAP powder was dissolved as needed in the superfusion solution and added directly to the bath containing the voltage-clamped myocyte.

Voltage-clamp experiments were performed with a low-noise Axopatch 200B patch-clamp amplifier using a Digidata 1400 and pCLAMP 10 software for data acquisition and analysis (Molecular Devices, Sunnyvale, CA).

Voltage-clamp protocol.

The 100-ms-long step changes in membrane potential (V_m) were produced from a holding potential of −40 mV to various V_m in the range −100 to +40 mV at 2 Hz. These V_m jumps were applied before and during exposure to l-Arg-containing solution, and again after l-Arg removal, to obtain the respective current [membrane ionic current (I_m)]-V_m relationships. l-Arg-activated currents are defined as the difference between I_m levels measured in the absence and presence of the amino acid.

Isolation of giant vesicles from cardiac ventricles.

Cardiac sarcolemmal vesicles were prepared according to published procedures (16, 17).

l-[¹⁴C]Lys uptake.

Sarcolemmal vesicles loaded with 140 mM KCl and 20 mM MOPS-KOH, pH = 7.4 (KCl-MOPS solution), were suspended in 80 μl of the same solution and incubated at 36 ± 1°C in the absence (10 μl KCl-MOPS) or presence of 0.44 mM N-ethylmaleimide (NEM) (10 μl of freshly prepared 4 mM NEM in KCl-MOPS). After at least 10 min, 10 μl of SNAP solution were added to the vesicles. Since half the concentration of SNAP releases tenfold larger NO levels with a faster kinetics compared with SNP (14), SNAP was the NO donor of choice for these assays that required high but brief exposure of vesicles to NO. SNAP is light sensitive, so that it was dissolved with KCl-MOPS in aluminum foil-wrapped tubes 2 min before use. For those uptake experiments that used SNP, vesicles were incubated with this NO donor for 20 min before adding NEM. The uptake reaction was started by adding 100 μl of a solution containing twice the final concentrations of l-[¹⁴C]Lys and unlabeled l-Lys, in KCl-MOPS. Uptake was stopped at the desired times and vesicles were washed, radioactivity counted, and the amount of vesicle protein determined as previously described (17).

Nitric oxide release.

Immediately before measuring NO release, appropriate amounts of SNAP were dissolved in the following bicarbonate-buffered salt solution (in mM): 137 NaCl, 4.7 KCl, 1.2 MgSO₄, 2.0 CaCl₂, and 18 NaHCO₃, pH 7.4 at 35°C. This solution was selected because the KCl-MOPS solution used for uptake experiments strongly interfered with NO-electrode measurements. The bicarbonate-buffered solution, which had a pH and ionic strength similar to those of the KCl-MOPS solution, was equilibrated with 5% CO₂ and 95% N₂ to maintain constant pH levels. NO production was measured at 35–37°C with Nafion-coated recessed-tip microelectrodes (4), using known amounts of NO in N₂ for equipment calibration. SNAP solutions (contained in syringes wrapped with aluminum foil) were delivered at a rate of 1 ml/min into the bath containing the NO-sensitive electrode.

Fluorescence measurements.

Freshly isolated myocytes were suspended in 1.5 ml of Langendorf solution containing (in mM) 135 NaCl, 5.4 KCl, 1 MgCl₂, 0.2 CaCl₂, 0.33 NaH₂PO₄, and 10 HEPES-Na, pH = 7.2 at 23°C, and were incubated for 1 h at 23°C with 10 μM of the dye 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate (DAF-FM diacetate) in aluminum foil-wrapped glass tubes. The supernatant was removed and cells were incubated in 1.5 ml of Langendorf solution for 15 min at 37°C. After supernatant removal, cells were resuspended in 2 ml of Langendorf solution, and 225-μl aliquots were distributed in 96-well plates. The time course of DAF-FM fluorescence changes was followed with a Cary Eclipse spectrofluorometer. Results similar to those reported in this work were obtained in myocytes loaded with 1 or 50 μM DAF-FM diacetate.

Data analysis.

Current traces were sampled at 25 Hz and low-pass filtered at 6 Hz. Linear cell capacitance was calculated with Clampex using 5-mV depolarizing pulses, and current analysis was performed with Clampfit. Clampex and Clampfit routines are included in pClamp 10 (Molecular Devices). Data are displayed as means ± SE for the indicated number of experiments. Statistical significance was determined using Student's t-tests (P < 0.05). Curve fitting was performed with nonlinear least-squares routines included in SigmaPlot v10.0 (Systat Software) using statistical weights proportional to (SE)⁻¹.

Reagents.

[¹⁴C]-UL-l-lysine hydrochloride, specific activity 228 Ci·mol⁻¹, SNAP, and 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) were purchased from Sigma. The hydrochloride salts of l-Lys and l-Arg, as well as ATP (Mg salt) and SNP, were purchased from Sigma-Aldrich. NEM and N-nitro-l-arginine methyl ester (l-NAME) were from Fluka. 7-Nitro indazole sodium (7-NINA) and (9S,10R,12R)-2,3,9,10,11,12-hexahydro-10-methoxy-2,9-dimethyl-1-oxo-9,12-epoxy-1H-diindolo[1,2,3-fg:3′,2′,1′-kl]pyrrolo[3,4-i][1,6]benzodiazocine-10-carboxylic acid, methyl ester (KT-5823) were from Tocris Bioscience. (Rp)-8-(para-chlorophenylthio)guanosine-3′,5′-cyclic monophosphate (Rp-8-pCPT-cGMP), a kind gift of Dr. Annie Beuve, was originally from Biolog Life Science Institute. DAF-FM diacetate was from Invitrogen. Collagenase type II was obtained from Worthington Biochemical. Salts and reagents were of analytical reagent grade.

RESULTS

NO inhibits l-Arg currents in ventricular cardiomyocytes.

The effect of NO on cationic amino acid transport has not been previously evaluated. To study this effect, inward currents reported to be activated by millimolar concentrations of l-Arg (27) were measured in cardiac muscle cells exposed to the NO donor SNP. Myocytes were whole cell voltage clamped at −40 mV with patch electrodes filled with a high-K⁺ intracellular salt solution and superfused with a 145 mM TMA-containing solution. Activation with 10 mM l-Arg in the absence or presence of 1 mM SNP yielded steady-state current levels of −1.1 and −0.6 pA/pF, respectively, for the traces shown in Fig. 1A. Therefore, simultaneous incubation with the NO donor produced a 45% inhibition at this V_m. Current levels returned to >90% of control values 3–4 min after SNP removal (Fig. 1A), showing that NO inhibition was reversible. Endogenous NO production mediated by constitutive Ca-CaM-dependent NOS (6) was prevented by including 10 mM EGTA in the electrode solution. Accordingly, similar results were obtained when 10 mM l-Lys, which is not a NOS substrate, replaced l-Arg in the bath solution (not shown).

Fig. 1.

Effect of nitric oxide (NO) on l-arginine (l-Arg) currents in voltage-clamped cardiomyocytes. A: inward current activated by 10 mM l-Arg in a cell held at −40 mV before (left), during simultaneous exposure to 1 mM sodium pentacyanonitrosyl ferrate(III) dihydrate (SNP; middle), and 4 min after SNP removal (right). B: membrane ionic current (I_m)-membrane potential (V_m) relationships for measurements performed in the absence (●) and presence (○) of SNP (n = 5, each). Curves represent fitting of Eq. 1. Notice that, although l-Arg transport is a passive process (27), a reversal potential approaching +∞ is consistent with these “zero-trans” experiments provided that there is no significant subsarcolemmal accumulation of l-Arg. Measurements from three myocytes exposed to 100 μM S-nitroso-N-acetyl-dl-penicillamine (SNAP; ▼) are also shown.

The inhibitory effect of NO on l-Arg currents was studied in the range of V_m −100 to +40 mV by applying a voltage-clamp protocol (see materials and methods). The corresponding I_m-V_m relationships collected from five cells are shown in Fig. 1B for control and SNP-treated myocytes. NO blocked l-Arg-activated currents by 25–50% depending on V_m, an observation that suggests V_m-dependent inhibition. Accordingly, I_m-V_m curves were analyzed with the equation:

$$I_m=I_{\text{o}}\times \exp (-\text{λ}\times zF{V}_{\text{m}}/RT)+I_{\infty }$$ (1)

where I_o represents (I_m − I_∞) at V_m = 0; λ accounts for the steepness of the I_m-V_m curve, I_∞ is the value of I_m at V_m → +∞, z is the valence of the transported species, F is Faraday's constant, R is the universal gas constant, and T is absolute temperature. Assuming z = 1, λ was found to be 0.56 ± 0.02 in the absence and 0.70 ± 0.03 in the presence of NO, a significantly larger value. Thus, l-Arg transport was rapidly and reversibly inhibited by exogenous NO in a manner that increased the V_m dependence of the transport process in cardiac myocytes.

The effect of 100 μM SNAP, an NO donor chemically unrelated to SNP, was also tested on 10 mM l-Arg-activated currents. The degree of inhibition was similar to that obtained with a tenfold larger SNP concentration (filled triangles in Fig. 1B). Therefore, SNAP is a more potent inhibitor, likely owing to different mechanisms and kinetics of NO release by these two compounds (14). Potencies aside, these results imply that the released NO is responsible for l-Arg current inhibition independently of the chemical structure of the NO donor.

NO selectively inhibits two components of l-Lys uptake in cardiac sarcolemmal vesicles.

The rapid onset of inhibition suggests either a direct interaction between NO and protein residues involved in l-Arg binding/transport or transporter phosphorylation via NO-sensitive cGMP-dependent protein kinases (13, 34). To distinguish between these possibilities, the effect of NO donors on cationic amino acid uptake was studied in giant sarcolemmal vesicles prepared from rat heart ventricles (16, 17). The rate of NEM-sensitive l-[¹⁴C]Lys uptake was measured in vesicles incubated with 0.05–50 mM l-Lys in the presence of 0, 10, 20, and 100 μM SNAP. Radiolabeled l-Lys instead of l-Arg was chosen for these experiments as previously discussed (17). Giant sarcolemmal vesicles, loaded with and suspended in 140 mM KCl and 20 mM MOPS-KOH, pH = 7.4, were incubated in the absence or presence of NEM for at least 10 min. Freshly prepared SNAP solution (10×) was then added to the vesicles, and the uptake reaction was initiated 7–13 min later depending on the tested SNAP concentration ([SNAP]) (see online Supplemental Material, available at the Journal website). Vesicles were treated with SNP in some experiments to assess the potential effect of the NO-donor chemical structure on l-Lys uptake. Results in Fig. 2 show the presence of two l-Lys uptake components that were inhibited by SNAP in a concentration-dependent manner. Both uptake components were also sensitive to 1 mM SNP (filled squares in Fig. 2A for 10 mM l-Lys and Fig. 2B for 0.2 mM l-Lys). Inhibition was found to be noncompetitive and, thus, the entire data set was simultaneously fitted in two variables with the following expression that describes the effect of a noncompetitive blocker on two hyperbolic uptake components:

$$v=\frac{{\text{V}}_{\max ,\text{h}}}{\left(1+\frac{{\text{K}}_{\mathrm{\text{m,h}}}}{[\text{l}-\text{Lys}]}\right)\left(1+\frac{[\text{SNAP}]}{K_{\text{i},\text{h}}}\right)}+\frac{V_{\max ,\text{l}}}{\left(1+\frac{K_{\text{m},\text{l}}}{\text{l}-\text{Lys}}\right)\left(1+\frac{[\text{SNAP}]}{K_{\text{i},\text{l}}}\right)}$$ (2)

where the subscripts “h” and “l” represent high- and low-affinity components, respectively. Best-fit values for all six parameters were as follows: K_m,h = 0.198 ± 0.142 mM, K_i,h = 7.3 ± 3.3 μM, V_max,h = 0.096 ± 0.041 nmol·mg vesicle protein⁻¹·min⁻¹, K_m,l = 15.7 ± 1.7 mM, K_i,l = 22.0 ± 1.5 μM, and V_max,l = 3.80 ± 0.15 nmol·mg vesicle protein⁻¹·min⁻¹. Curves through the data points in Fig. 2, A and B, were generated with Eq. 2 and these best-fit values. SNAP inhibited both l-Lys uptake components with a threefold difference in K_i values. This finding has two important implications. First, exogenous NO inhibits uptake in vesicles lacking intracellular substrates, cofactors, and soluble enzymes, i.e., inhibition likely results from direct interaction between NO and the transporters. Second, NO selectively regulates each component of cationic amino acid transport in cardiomyocytes. Apparent affinities and maximal turnover rates for both uptake components, as well as their respective ratios (K_m,l/K_m,h ≅ 79; V_max,l/V_max,h ≅ 40), were in agreement with those previously reported (17).

Fig. 2.

Effect of NO on l-lysine (l-Lys) uptake in giant cardiac sarcolemmal vesicles. A: extravesicular l-Lys concentration ([l-Lys]) dependence of uptake measured at 37°C for the range 0.05–50 mM l-Lys in the presence of 0 (●), 20 (○), and 100 μM SNAP (▼). Experiments performed with 10 μM SNAP, although included in data analysis, are not shown for the sake of clarity. Vesicles were exposed to SNAP for the appropriate time to ensure steady NO release before uptake measurements (Supplemental Fig. S1). Symbols are the mean ± SE of 0.2 mM N-ethylmaleimide (NEM)-sensitive uptake from 3–5 experiments, each performed in triplicate (n = 45). Curves represent the solution of Eq. 2 simultaneously fitted to the entire data set. The filled square is the mean ± SE of 3 experiments performed with 1 mM SNP. Notice that the level of inhibition is similar to that obtained with 20 μM SNAP. B: detail of curves in the range 0.05–0.5 mM l-Lys to show the effect of NO on the high-affinity uptake component. Symbol code as in A. The filled square represents the mean ± SE of 3 experiments performed with 1 mM SNP. C: calibration of steady-state NO levels against [SNAP] (n = 3).

The amount of NO released by SNAP was measured with a NO-selective electrode (Supplemental Fig. S1). Figure 2C shows that NO levels were a linearly increasing function of [SNAP] with slope 37.6 ± 4.5 nM NO/μM SNAP. Interpolation of SNAP K_i values, 7.3 and 22.0 μM, yielded NO values of 275 and 827 nM, respectively. Basal NO levels of 200–1,000 nM have been measured in vivo with NO-selective electrodes in the intestine perivascular region of Sprague-Dawley rats (7). Therefore, NO modulation of cardiac cationic amino acid import appears to occur at physiologic NO levels.

Endogenous NO production rate shows a biphasic [l-Arg] dependence in DAF-FM-loaded myocytes.

It has been shown that, because of its high capacity, the low-affinity transporter accounts for roughly half of total cationic amino acid transport at physiologic plasma levels of l-Arg, l-Lys, and l-Orn (17). These recent findings challenge the common notion that only the high-affinity transporter is physiologically relevant. High-affinity l-Arg transport in cardiac myocytes can only be studied with radioisotope techniques because of the scarce capacity of this transporter (17), which escapes detection with electrophysiological or fluorescence methods (27). Therefore, once established that exogenously produced NO inhibits both uptake components, we took advantage of the prominent capacity of the low-affinity transporter to study the involvement of endogenous NO on l-Arg transport regulation.

Fluorescence spectroscopic studies were performed in cardiomyocytes loaded with the dye DAF-FM, which greatly increases its fluorescence quantum yield in the presence of NO (15). After recording baseline fluorescence for ∼15 min, the effect of 1, 2, 5, 10, 20, and 50 mM l-Arg on fluorescence intensity changes was measured. Results in Fig. 3A show a linear increase in NO production upon addition of 1 mM l-Arg that lasted up to 40 min. Indeed, this increase was preceded by a 15-min lag, which was subtracted from the records. The lag decreased to 10 min with 2 mM, 3 min with 5 mM, and disappeared at 10 mM l-Arg. Since these experiments were performed at 22–23°C, a slower l-Arg-transport rate that delays NO synthesis and/or a slower NOS activity and/or slow reaction kinetics between DAF-FM and NO at [l-Arg] < 10 mM may explain this behavior. Starting at 2 mM l-Arg, the increase in NO-produced fluorescence became biphasic, with an initial fast component followed by a slower phase, as displayed in Fig. 3, B and C, for 10 and 50 mM l-Arg, respectively. Figure 3B also shows that fluorescence increases were not mimicked by exposing myocytes to 10 mM l-Lys (see discussion). Both fluorescence components were analyzed by linear regression, yielding slopes (dF/dt) equal to NO production rates. Larger [l-Arg] increased the initial rate of NO production and reduced the slope of the slower component. The initial rate of NO production increased hyperbolically as a function of [l-Arg] with a K_0.5 = 2.9 ± 0.8 mM (Fig. 3D). The subsequent, slower NO production rate decreased hyperbolically as a function of [l-Arg] with a K_i = 1.9 ± 1.0 mM (Fig. 3E). These K_0.5 and K_i values were found to be not statistically significantly different.

Fig. 3.

Effect of l-Arg on NO-induced fluorescence changes in 4-amino-5-methylamino-2′,7′-difluorofluorescein (DAF-FM)-loaded cells. A: time course of endogenous NO production in the presence of 1 mM l-Arg (n = 3). A 15-min lag was subtracted from these records. Linear regression analysis yielded an NO production rate (dF/dt) equal to 23.1 ± 0.8 fluorescence units (FU)/min. AU, arbitrary units. B: time course of NO release in the presence of 10 mM l-Arg (●). Linear regression yielded an initial rate of NO production of 49.3 ± 2.6 FU/min and a subsequent rate of 10.6 ± 0.6 FU/min. No lag was apparent in this or in 6 other similar experiments. Also shown is the time course of fluorescence changes when 10 mM l-Lys replaced l-Arg in the extracellular medium (○; average of 2 experiments). The lack of fluorescence changes observed with l-Lys serves as a control for nonspecific DAF-FM photoactivation during prolonged exposure to the excitation light. C: time course of NO production with 50 mM l-Arg. Initial NO production rate = 102.7 ± 11.4 FU/min; subsequent rate = 6.4 ± 1.1 FU/min (n = 5). D: initial NO production rates for 1–50 mM l-Arg. The curve through the data points is the best-fit hyperbola with K_0.5 = 2.92 ± 0.75 mM and V_max = 100.7 ± 7.2 FU/min. E: secondary rate of NO production, as a percentage of that measured with 1 mM l-Arg, vs. [l-Arg]. The curve represents an inhibitory hyperbola with V₀ = 122.8 ± 22.5%, K_i = 1.89 ± 0.99 mM, and V_∞ = 34.9 ± 8.9%.

Thus, millimolar concentrations of l-Arg promoted the initial synthesis of endogenous NO in ventricular cardiomyocytes. This effect was followed by a decrease in the rate of additional NO production.

Direct inhibition of l-Arg currents by endogenous NO in voltage-clamped myocytes.

An alternative interpretation of the results in Fig. 3 is that NO inhibits NOS activity rather than (or in addition to) l-Arg transport. Therefore, experiments were designed to directly test the effect of endogenously generated NO on l-Arg transport by including free Ca²⁺ in the intracellular medium. CaM activation of constitutive NOS has been reported to occur with a K_0.5 for Ca²⁺ in the range 0.2–0.4 μM (26). Thus, cardiomyocytes were voltage clamped with the electrode solution described in materials and methods modified to include (in mM) 9 EGTA, 1 K-HEDTA (potassium N-hydroxyethyl ethylenediamine triacetic acid), and 7.2 CaCl₂. This combination of chelating agents and Ca²⁺, together with 0.7 mM MgATP and 4.0 mM MgCl₂, resulted in a free [Ca²⁺] = 460 nM, as calculated with the program MaxChelator (37°C, pH = 7.3, ionic strength = 0.15 N). Under these conditions, application of 10 mM l-Lys produced an inward current of expected aspect and magnitude (Fig. 4A, left). Two minutes after l-Lys removal, the cell was exposed to 10 mM l-Arg. The inward current initially increased, approaching a maximum at ∼6 s and then decayed exponentially to a steady level that was 48% of the peak value (Fig. 4A, middle). Subsequent application of 10 mM l-Lys resulted in a monotonically increasing current that reached a steady-state level ∼44% of that obtained when l-Lys was first applied (Fig. 4A, right). An average additional 2 min were required for l-Lys currents to display an 80–90% recovery (not shown). The biphasic behavior of l-Arg currents in the presence of intracellular Ca²⁺ was consistently observed in 7 myocytes, yielding a 62 ± 11% inhibition and an average rate constant for current decay of 0.25 ± 0.05 s⁻¹. Although BH₄, FAD, and FMN were not included in the electrode solution, the presence of NOS activity can be explained by tight binding of these endogenous cofactors to the enzyme, some of which copurify with NOS (6).

Fig. 4.

Effect of endogenously produced NO on cationic amino acid-activated currents in cardiomyocytes. A: cells were voltage clamped at −40 mV with an electrode solution that included a calculated [Ca²⁺]_free = 460 nM and superfused with 10 mM l-Lys (left), 10 mM l-Arg (middle), and again 10 mM l-Lys after 2-min washout of l-Arg (right). l-Arg-activated current decayed exponentially toward a minimum that was 48 ± 4% of the peak value, with a rate constant of 0.21 ± 0.08 s⁻¹ (fitting shown superimposed). B: a myocyte voltage clamped as described in A was exposed to 1 mM N-nitro-l-arginine methyl ester (l-NAME) for 5 min before application of 10 mM l-Arg. C: inward current activated by 10 mM l-Arg in cardiomyocytes voltage clamped at −40 mV with 460 nM free Ca²⁺ in the absence (a) and presence (b) of 20 μM 7-nitro indazole sodium (7-NINA). Current activated by 10 mM l-Arg was also measured in cells voltage clamped with a Ca²⁺-free, 10 mM EGTA-containing electrode solution and exposed to 20 μM 7-NINA to show the lack of effect of this compound on l-Arg transport (c). Current density levels are representative of 3–5 experiments for each condition. D: steady-state I_m-V_m curves for the following current traces: A, left (●); A, middle (▼); and B (○). Lines through the data points represent fitting of Eq. 1. E: a myocyte voltage clamped as described in A was exposed to 10 mM l-Arg, and current was continuously recorded at the holding potential for ∼5 min. Gap duration: 160 s.

Incubation of myocytes with 1 mM of the NOS inhibitor l-NAME (n = 5) removed the inhibitory component of l-Arg currents (Fig. 4B). On the other hand, inclusion of 5 μM of the sGC inhibitor ODQ (9) in the electrode solution did not modify the biphasic response elicited by 10 mM l-Arg in the presence of intracellular Ca²⁺ (n = 3; not shown). Likewise, inhibitors of cGMP-dependent protein kinases, Rp-8-pCPT-cGMP [15 μM (9)] and KT-5823 [5 μM (34)], when added to the electrode solution, had no effect on the biphasic pattern displayed by l-Arg currents (n = 3 each; not shown). These results demonstrate the direct modulation of l-Arg transport by NOS-produced NO.

The fairly selective nNOS inhibitor 7-NINA, which has been reported to block the activity of this NOS isoform at submicromolar concentrations (22), was also tested. Results in Fig. 4C show that addition of 20 μM 7-NINA did not prevent the appearance of an inhibitory component in 10 mM l-Arg-activated currents when cardiomyocytes were voltage clamped in the presence of free intracellular Ca²⁺ (trace b). In fact, this behavior was similar to that of control, untreated cells (trace a). Exposure to 7-NINA had no effect per se on 10 mM l-Arg-activated currents produced in the absence of free intracellular Ca²⁺ (trace c). Together with the effect of l-NAME (Fig. 4B), these data suggest that eNOS-generated NO is responsible for the regulation of l-Arg transport in cardiac muscle cells.

The V_m dependence of l-Arg currents elicited in the presence of intracellular Ca²⁺ (Fig. 4D) was similar to that found when testing exogenous NO donors (Fig. 1B). Analysis with Eq. 1 of I_m-V_m curves obtained in the presence of l-Lys, l-Arg, and l-Arg + l-NAME yielded best-fit λ-values of 0.55 ± 0.04, 0.73 ± 0.06, and 0.59 ± 0.04, respectively. The value of λ obtained from the l-Arg I_m-V_m curve was statistically significantly larger than those obtained with l-Lys and l-Arg + l-NAME. Therefore, current block by endogenous NO also increased the V_m dependence of cationic amino acid transport.

To test whether reduced l-Arg transport that results from increased NO levels further affects NO synthesis, cardiac myocytes voltage clamped in the presence of intracellular Ca²⁺ were exposed to 10 mM l-Arg and the time course of current production was continuously recorded for ∼5 min. The current record displayed in Fig. 4E, which is representative of three experiments, shows that current inhibition lasted ∼3 min (72 ± 14% inhibition relative to the peak), after which, inward current slowly began to rise, peaked (75 ± 8% of the initial peak), and was once again inhibited (64 ± 12% inhibition relative to the second peak). Altogether, these results strongly suggest that NOS-generated NO blocks l-Arg transport, and reduced levels of l-Arg, in turn, decrease NO synthesis thus defining a negative feedback mechanism.

l-Arg concentration dependence of current inhibition.

Currents were measured as a function of [l-Arg] to quantitatively correlate NOS-produced NO and the degree of current inhibition. Figure 5A shows superimposed normalized current traces elicited by 1, 5, and 20 mM l-Arg. Inhibition, relative to peak values, clearly increased with [l-Arg]. This effect was analyzed by fitting exponential functions to the decaying portion of current traces obtained with 0.5–20 mM l-Arg. Current inhibition displayed a hyperbolic dependence on [l-Arg] with a maximal value of 84.3 ± 5.8% and a K_0.5 = 2.2 ± 0.5 mM (Fig. 5B). Likewise, the rate constant for current decay increased hyperbolically with [l-Arg] with parameter values k_max = 0.30 ± 0.02 s⁻¹ and K_0.5 = 2.5 ± 0.4 mM (Fig. 5C). These l-Arg K_0.5 values are similar to those found for NO-mediated fluorescence changes in DAF-FM-loaded cells. Together with the lack of an inhibitory component in currents activated by l-Lys and l-Arg + l-NAME, these results establish a direct relationship between regulation of l-Arg transport, NOS activity, and NO.

Fig. 5.

[l-Arg] dependence of current inhibition by endogenous NO. A: superimposed current traces elicited by the shown [l-Arg] in cardiomyocytes whole cell voltage clamped in the presence of free intracellular Ca²⁺. Traces are displayed as percentage of peak values. Exponential functions were fitted to the decaying portion of current traces obtained at various [l-Arg]. B: current inhibition as a function of [l-Arg]. C: rate constant vs. [l-Arg]. Symbols represent the mean ± SE of 3–7 experiments. Curves represent hyperbolic functions with best-fit parameters reported in the text.

DISCUSSION

This work reports two novel findings in cardiac ventricular myocytes. First, NO acutely, reversibly, and selectively inhibits the activity of high- and low-affinity cationic amino acid transporters. Second, by modulating the transport kinetics of l-Arg, which is the substrate for its biosynthesis, NO participates in a negative feedback self-regulatory mechanism. This mechanism is different from the classical enzyme inhibition by a reaction product (30).

l-Arg currents were significantly inhibited by different NO donors. Although an effect of NO donor by-products cannot be conclusively excluded, the unrelated chemical structures of SNP and SNAP indicate that current inhibition was mediated by the released NO. The rapid onset of inhibition allowed us to rule out “long-term” effects such as regulation of protein synthesis or trafficking of transporter molecules as an explanation for this observation. NO donors also inhibited high- and low-affinity components of l-Lys uptake in cardiac sarcolemmal vesicles loaded only with a KCl-MOPS solution. Thus, uptake inhibition appears to be the result of direct NO-transporter interactions. Both uptake components were noncompetitively inhibited by exogenous NO with a threefold difference in K_i values. Since block of uptake by NO was found to be reversible, noncompetitive inhibition is likely due to interactions between NO and transporter domains that participate in conformational transitions associated with amino acid translocation (rather than binding/release reactions). Changes in V_max due to complete inhibition of some carrier molecules while others are located in unreachable pools seem unlikely given the highly diffusible nature of NO.

Isolated rat ventricular myocytes have background NO levels of ∼100 nM (2). On the basis of Eq. 2 and our best-fit parameters, this NO concentration will block only 16% of total cationic amino acid transport, considering a combined plasma concentration of 0.6 mM for l-Arg, l-Lys, and l-Orn (17, and references therein). This more realistic concentration, rather than that of l-Arg alone (∼0.25 mM), was used in these calculations because l-Lys (∼0.25 mM) and l-Orn (∼0.1 mM) are equally effective competitors for transport. Thus, roughly two out of five binding/translocation events will result in l-Arg import and, eventually, in NO production. NO K_i values of 275 and 827 nM, which fall within physiologic levels (7), will inhibit total cationic amino acid transport by 33.5 and 58.5%, respectively. Thus, a simultaneous array of high- and low-affinity transporters is advantageous because total uptake will be just 33% inhibited at [NO] ∼275 nM (K_i,h). In fact, the low-affinity carrier, which was calculated to account for ∼60% of total cardiac cationic amino acid transport at physiologic plasma levels of l-Arg, l-Lys, and l-Orn (17), will be the predominant functional transporter at intermediate [NO] because of its lower sensitivity to inhibition.

l-Arg increased the initial rate of NO release and decreased the rate of subsequent NO production, in both cases with a K_0.5 ∼2 mM, in DAF-FM-loaded myocytes. Since l-Arg activation of NO release preceded inhibition, similar K_0.5 values suggest that the released NO was stoichiometrically responsible for the subsequent decrease in the rate of NO synthesis. These effects were not mimicked by l-Lys, showing that cell membrane depolarization induced by the inward movement of positive charges cannot by itself affect NO production. Moreover, published results with this system show that changes in fluorescence intensity are not observed when l-Arg is replaced with d-Arg or l-Arg + l-NAME (27). Altogether, these findings confirm the involvement of NOS activity and endogenously produced NO in the fluorescence changes that follow l-Arg transport. However, these experiments did not distinguish between a negative feedback on l-Arg transport (but see below) and NO inhibition of NOS activity, an effect that has been well documented (1, 29, 32). Negative feedback on substrate transport and NOS inhibition may not be mutually exclusive mechanisms but rather another expression of highly regulated NO synthesis.

Cationic amino acid-activated currents represent the inward transport of these compounds in cardiac ventricular myocytes (27). In this context, results in Fig. 4 show a number of features that support the existence of a negative feedback mechanism. First, l-Lys currents increased monotonically toward the same steady-state level independently of the presence of free Ca²⁺ and/or Ca-CaM, which, by promoting enzyme dimerization, enhances NOS activity. Second, the biphasic time course of l-Arg currents under similar conditions is consistent with l-Arg transport, followed by NOS-mediated NO production and NO inhibition of subsequent l-Arg transport. Third, once the carriers are inhibited, l-Lys also elicited lower steady-state current levels. Fourth, l-Arg currents in myocytes exposed to a NOS blocker resembled l-Lys currents, demonstrating the involvement of NOS in the biphasic behavior of l-Arg transport. Finally, the most straightforward interpretation of data in Fig. 4E is that NO inhibition of l-Arg transport results in NOS reduced NO synthesis, which in turn relieves current inhibition, increasing once again l-Arg transport and NO production, and so on. These experiments strongly suggest that NO can modulate its own biosynthesis through the regulation of l-Arg transport in cardiac myocytes. The lack of effect of sGC and cGMP-dependent protein kinase blockers on the inhibitory transport component allowed us to rule out carrier phosphorylation via NO-sensitive cGMP-dependent protein kinases, a result also supported by uptake experiments in sarcolemmal vesicles.

A NO-mediated negative feedback mechanism on l-Arg transport was confirmed by showing a direct relationship between [l-Arg] and the size and kinetics of the inhibitory current component. The finding that current levels were half-maximally inhibited and the rate constant for current block was half-maximally increased by ∼2 mM l-Arg indicates the occurrence of a low-affinity process and provides further support to the DAF-FM experiments. Whether 2 mM extracellular l-Arg produces ∼800 nM NO (K_i,l) in live cardiac myocytes, although a tempting extrapolation, remains an open question. Interestingly, calculations with Eq. 2 using K_i,h = 275 nM indicate that the high-affinity transporter will be 75% inhibited when [l-Arg] and [NO] are 2 mM and 800 nM, respectively. This result brings about the question of whether this process is physiologically relevant. We have established that, because of its high capacity and higher resistance to NO inhibition, the low-affinity l-Arg transporter must play a role in physiology. Nonetheless, signal-to-noise limitations in electrophysiological and fluorescence techniques prevented us from studying endogenous NO regulation of l-Arg transport at more physiologic concentrations of the amino acid (100–250 μM), where high- and low-affinity carriers contribute more or less equally to total transport in cardiac myocytes. As such, l-Arg K_0.5 values in the millimolar range and inhibition of l-Arg current by NO that lasts ∼3 min after application of 10 mM l-Arg (Fig. 4E) may be interpreted as a nonphysiological negative feedback mechanism. However, the expectation of biphasic fluorescence time courses and NO-mediated current inhibition (with faster recovery) at lower [l-Arg], similar to the biphasic response of l-Lys uptake to exogenous NO (Fig. 2), does not seem unreasonable. Modulation of the low-affinity l-Arg transporter by NO and a negative feedback on NO synthesis may also be pertinent to cardiac function during starvation, trauma, severe burns, cancer, and sepsis; all catabolic states that transiently increase several-fold circulating amino acid levels as a result of protein degradation in peripheral tissues.

With regard to the source of NO, the compound 7-NINA, reported to block nNOS with an IC₅₀ of 0.47 μM (22), and to competitively inhibit l-Arg binding to this NOS isoform with a K_i = 2.8 μM (19), did not prevent NO inhibition of l-Arg current in whole cell voltage-clamped myocytes. Following the description of a caveolar complex between the high-affinity CAT-1 and eNOS in pulmonary artery endothelial cells (20), and the caveolar location of eNOS in cardiac muscle cells (reviewed in Ref. 5), these results suggest a crosstalk between high- and low-affinity cationic amino acid transporters and caveolar eNOS in cardiac myocytes.

The effect of NO is likely to involve S-nitrosylation of cysteine thiol side chains in these cationic amino acid transporters. In this regard, putative Cys residues that react with NEM to block l-Arg currents (27) and l-Lys uptake (17) might be also the targets for NO-mediated S-nitrosylation. Further mutational studies are needed to determine the residues in the transporter that react with NO. Molecular details aside, the simultaneous presence of high-affinity/low-capacity and low-affinity/high-capacity carriers that are selectively modulated by NO, suggests an exquisitely tuned mechanism for l-Arg transport and NO production in cardiac myocytes.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute (NHLBI) Grant R01HL076392 (to R. D. Peluffo).

DISCLAIMER

The content is solely the responsibility of the authors and does not necessarily represent the official views of the NHLBI or the National Institutes of Health.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.