Cationic amino acid transporters and their modulation by nitric oxide in cardiac muscle cells

Abstract

Cationic amino acid transporters (CATs) play a central role in the supply of the substrate L-arginine to intracellular nitric oxide synthases (NOS), the enzymes responsible for the synthesis of nitric oxide (NO). In heart, NO produced by cardiac myocytes has diverse and even opposite effects on myocardial contractility depending on the subcellular location of its production. Approximately a decade ago, using a combination of biophysical and biochemical approaches, we discovered and characterized high- and low-affinity CATs that function simultaneously in the cardiac myocyte plasma membrane. Later on, we reported a negative feedback regulation of NO on the activity of cardiac CATs. In this way, NO was found to modulate its own biosynthesis by regulating the amount of L-arginine that becomes available as NOS substrate. We have recently solved the molecular determinants for this NO regulation on the low-affinity high-capacity CAT-2A. This review highlights some biophysical and biochemical features of L-arginine transporters and their potential relation to cardiac muscle physiology and pathology.

The system y⁺ family of cationic amino acid transporters (CATs)

System y⁺ includes cationic amino acid transporters CAT-1, CAT-2A, CAT-2B, CAT-3, and CAT-4 (for references, see Nawrath et al. 2000), although the inclusion of CAT-2A in this group remains controversial (see below). CAT-1 is expressed ubiquitously, except in the liver where CAT-2A is the only isoform present. CAT-2A is also expressed in the skeletal muscle, skin, ovary, and stomach, whereas its splice variant CAT-2B is present in the activated macrophages and splenocytes, lung, and testis. CAT-3 is expressed mainly in the brain and CAT-4 in the pancreas, skeletal muscle, heart, and placenta (for references on CAT tissue distribution see Palacín et al. 1998; Kakuda et al. 1993, 1998). In rat’s heart, ventricular myocytes constitutively exhibit CAT-1 transport activity (Devés and Boyd 1998; Lu et al. 2009). Functional expression of CAT-2A has been reported in neonatal rat cardiac myocytes only after extensive cytokine treatment (Simmons et al. 1996). At variance with this, we have functionally identified and characterized a second L-arginine (L-Arg) transporter, with properties consistent with CAT-2A, in acutely-isolated and otherwise untreated cardiac myocytes (Peluffo 2007). In light of this, we will briefly review the main characteristics of system y⁺ transporters, in particular CAT-1 and CAT-2A.

General characteristics

Cationic amino acid transporters are N-glycosylated membrane proteins arranged in 12–14 transmembrane segments (Devés and Boyd 1998), with a molecular weight for the protein portion ranging between 65 and 72 kDa. Accordingly, these transporters show a typical behavior in a western blot upon sample treatment with N-glycosidase F, with a smeared band at ~ 110 kDa that begins to fade as another band becomes apparent at ~ 70 kDa (Zheng et al. 2020). The degree of homology between CAT-1 and CAT-2A is ~ 60% (Closs et al. 1993). Human, rat, and mouse isoforms of these transporters show > 90% homology (Yoshimoto et al. 1991; Puppi and Henning 1995; Closs et al. 1997).

Substrate specificity and kinetic parameters

System-y⁺ transporters are highly selective for the “L” enantiomers of cationic amino acids such as arginine, lysine, and ornithine (Devés and Boyd 1998; Palacín et al. 1998). However, these carriers display differences in apparent affinity (1/K_0.5) and V_max for the transported species. Thus, whereas CAT-1 transports L-Arg with relatively high apparent affinity (K_0.5 = 0.07–0.25 mM) and low capacity, CAT-2A displays higher V_max and lower apparent affinity (K_0.5 = 2–15 mM) for this amino acid (Devés and Boyd 1998; Peluffo 2007; Lu et al. 2009; Zhou et al. 2010). In fact, CAT-2A is the only low-affinity member of the CAT family. We found V_max (CAT-2A)/V_max (CAT-1) ≈ 34 and K_0.5 (CAT-2A)/K_0.5 (CAT-1) ≈ 70 in uptake experiments with giant sarcolemmal vesicles from rat cardiac ventricular myocytes (Fig. 1A, and Lu et al. 2009). Despite K_0.5 values are several-fold larger than circulating levels of cationic amino acids, the high transport capacity of CAT-2A makes this transporter’s contribution to be physiologically relevant (Lu et al. 2009). Within a given isoform, L-Arg, L-lysine (L-Lys), and L-ornithine (L-Orn) activate transport with comparable K_0.5 and maximal rates, and it is assumed that one amino acid molecule is translocated per carrier cycle. The interaction with the protein is stereospecific, i.e., D-isomers fail to activate transport or do so very poorly (Fig. 1B, and White et al. 1982; Peluffo 2007; Zhou and Peluffo 2010). Interestingly, D-arginine appears to bind to CAT-2A at the same binding site as L-Arg and with similar apparent affinity. The lack of transport for the D-enantiomer seems to reside in conformational transitions associated with amino acid inward translocation (Zhou and Peluffo 2010). Members of system y⁺ can also recognize L-Arg analogs that are methylated at the guanidine group such as N^G-nitro-L-arginine methyl ester (L-NAME), an inhibitor of nitric oxide synthase (NOS) activity (Forray et al. 1995). Transport of cationic amino acids mediated by members of system y⁺ is independent of external Na⁺, K⁺, Ca²⁺, or Mg²⁺ (Kim et al. 1991; Kakuda and MacLeod 1994; Nawrath et al. 2000; Peluffo 2007).

Fig. 1.

A L-Lys uptake for the range 0.05–50 mM external L-Lys in giant sarcolemmal vesicles from rat cardiac ventricle. Symbols show the mean ± S.E. of 3–5 experiments performed in triplicate. The curve through the data points represents the sum of two hyperbolic functions fitted to the data with best-fit parameters K_m,h = 222 ± 71 μM, V_max,h = 0.121 ± 0.036 nmol (mg of vesicle protein)⁻¹ min⁻¹, K_m,l = 16.0 ± 3.6 mM, and V_max,l = 4.07 ± 0.36 nmol (mg of vesicle protein)⁻¹ min⁻¹. B L- and D-Arg current–voltage (V_M) relationships in whole-cell voltage-clamped cardiomyocytes, for the range − 100 to + 40 mV. Symbols represent the mean ± S.E. of 5–7 curves performed under each condition. C Time courses of 10 mM L-Lys uptake at 37 °C. Symbols represent the mean ± S.E. of 3 experiments (performed in triplicate) at each incubation time for total L-Lys uptake (⬤), uptake in samples boiled in the presence of 1% SDS (▽), and uptake in samples incubated with 0.2 mM NEM for 10 min at 37 °C (▼). D K_0.5–V_M curve for L-Arg activated currents. K_0.5 values were obtained by fitting Hill equations to current-L-Arg concentration curves (in the range 1–50 mM L-Arg_o) at each V_M

Inhibitors

The sulfhydryl reagent N-ethylmaleimide (NEM) was reported to inhibit members of system y⁺ at concentrations 5- to tenfold lower than those required to block other proteins (Devés et al. 1993; Peluffo 2007; Lu et al. 2009). In addition, only brief NEM exposure periods (7–10 min) are required to fully block these transporters, another distinctive feature with respect to other carriers (Fig. 1C, and Peluffo 2007; Lu et al. 2009). In erythrocytes, system y⁺ was blocked by 0.2 mM NEM with an inactivation rate constant of 0.53 min⁻¹ (25 °C), while cationic amino acid transport system y⁺L, also present in these cells, remained unaffected by similar concentrations of this reagent (Devés and Boyd 1998). Likewise, the reactivity of system y⁺ to NEM is high compared to that of other transporters present in the same cells, such as the choline carrier (Devés and Krupka 1981) and the ASC amino acid transporter (Al-Saleh and Wheeler 1982). Studies in Xenopus oocytes expressing human CAT-2A identified Cys residues 33 (N terminus) and 273 (6th transmembrane domain) as responsible for NEM inhibition of transport (Beyer et al. 2013). Thus, these protein regions must be important for amino acid transport, and the ‒SH moieties in these two Cys residues are likely to be highly exposed and accessible.

Trans-stimulation of uptake

Transport activity mediated by members of system y⁺ is accelerated by the presence of cationic amino acids at the other side of the membrane, a feature called trans-stimulation (Devés and Boyd 1998). High-affinity isoforms such as CAT-1 and CAT-2B show up to tenfold trans-stimulation (Closs et al. 1997; Devés and Boyd 1998; Lu et al. 2009). The low-affinity CAT-2A, on the other hand, has been reported to be much less responsive to the presence of trans-substrate (Closs et al. 1993; Kavanaugh et al. 1994). In fact, CAT-2A expressed in Xenopus oocytes shows only a modest 1.1–1.5-fold increase in activity (Closs et al. 1997). Based on this and other reasons, some authors consider that CAT-2A is not a member of system y⁺ (Beyer et al. 2013). In our hands, native CAT-2A from cardiac sarcolemmal vesicles showed a larger ~ 2.5-fold trans-stimulation of L-Lys uptake (Lu et al. 2009). Furthermore, when parameters from hyperbolic fitting of the respective dose–response curves were converted to initial uptake slopes (υ₀ = V_max/K_0.5), trans-stimulation rates for the low-affinity CAT-2A increased to ~ fivefold. The occurrence of trans-stimulation can be explained through the rates of binding site translocation from one side of the membrane to the other, which are a function of substrate occupancy (Stein 1990; Closs et al. 1993). This unique feature is characteristic of a small number of membrane transport families, and represents a fundamental kinetic difference when compared to the behavior of channel proteins. For the purposes of this review, we will include the low-affinity CAT-2A as a y⁺ or y⁺-like transporter, hopefully, without losing any generality.

Voltage dependence

All system-y⁺ members are electrogenic, i.e., they move charge across the membrane electric field and generate electrical currents. In the nineties, CAT-1 (Kavanaugh 1993) and CAT-2A and -2B (Nawrath et al. 2000) were cloned and expressed in Xenopus oocytes to study the membrane potential (V_M) dependence of cationic amino acid-activated currents with a two-electrode voltage clamp technique. In the case of CAT-1, the concentration of extracellular L-Arg (L-Arg_o) required for half-maximal activation of current, K_0.5, and maximum current levels, I_max, were both found to be dependent on V_M. I_max for L-Arg influx increased with hyperpolarization whereas K_0.5 values increased with depolarization (Kavanaugh 1993). The V_M dependence of K_0.5 implies that less L-Arg_o would be needed to saturate the transporter at hyperpolarizing potentials. With regard to CAT-2A, currents generated by this low-affinity isoform in whole-cell voltage-clamped cardiomyocytes also display a steep V_M dependence, with K_0.5 values for L-Arg activation of current ranging from 5 mM at − 80 mV to 11 mM at 0 mV (Fig. 1D, and Peluffo 2007). I_max values were found to be V_M-dependent as well, suggesting that conformational transitions involving amino-acid-bound translocation steps remain as a function of V_M even at saturating amino acid concentrations. On the other hand, CAT-2A’s high-affinity splice variant CAT-2B, when expressed in oocytes, was reported to show a L-Arg_o-concentration independent conductance and a V_M-independent I_max in the range − 150 to + 60 mV (Nawrath et al. 2000). Whether this differential V_M-dependent behavior has biological implications in excitable tissues remains an open question.

CATs in cardiac myocytes

L-Arg de novo synthesis only occurs in the liver and kidneys (Morris 1992). Nevertheless, vascular endothelial cells (Hecker et al. 1990) and macrophages (Wu and Brosnan 1992), among other NO-producing cell types, are capable of recycling L-Arg from intracellular citrulline pools, making these cells less dependent on the presence of extracellular L-Arg and its transport. Cardiac muscle cells, on the other hand, are among the cell types that fully depend on L-Arg transport to cover intracellular needs for this multifunctional amino acid (Lu et al. 2009 and references therein). Similarly, the essential L-Lys relies on a transport system to cross biological membranes. In this regard, we have functionally identified CAT-2A as responsible for L-Arg currents in whole-cell voltage-clamped cardiomyocytes (Peluffo 2007) as well as CAT-1 and CAT-2A in [¹⁴C]L-Lys uptake experiments with cardiac sarcolemmal vesicles (Lu et al. 2009). Analysis of radiotracer experiments using a sum of two hyperbolic functions yielded high-affinity, low-capacity (CAT-1, K_0.5 ~ 0.2 mM, V_max ~ 0.12 nmol (mg of vesicle protein)⁻¹ min⁻¹) and low-affinity, high-capacity uptake components (CAT-2A, K_0.5 ~ 15 mM, V_max ~ 4.0 nmol (mg of vesicle protein)⁻¹ min⁻¹) (Lu et al. 2009). These two carriers were found to work simultaneously, forming a “homeostatic device” that will likely ensure proper L-Arg supply in cardiac myocytes under a variety of metabolic conditions. Analysis of a two-transporter system with the array of kinetic parameters reported above showed the activity of the low-affinity component to be physiologically relevant, as it is responsible for 50–60% of the total cationic amino acid influx in cardiac muscle cells (Lu et al. 2009). Physiologic plasma levels for L-Arg, L-Lys, and L-Orn in humans and rodents add up to 0.40–0.70 mM (van Haeften and Konings 1989; Noeh et al. 1996), a value range which is far below the K_0.5 for CAT-2A. However, the high transport capacity (V_max) displayed by this carrier accounts for its physiological competence (Lu et al. 2009). In addition, given the V_M dependence of K_0.5 (Fig. 1D), a threefold increase in apparent affinity is anticipated at the cardiac myocyte resting membrane potential (Peluffo 2007).

Nitric oxide modulation of cationic amino acid transport

One important role for L-Arg is as substrate for nitric oxide (NO) biosynthesis via the enzyme nitric oxide synthase (NOS; EC 1.14.13.39), a dioxygenase that uses NADPH and O₂ in the oxidation of L-Arg to yield NO and citrulline (Gross and Wolin 1995). This reaction requires flavin mononucleotide, FAD, tetrahydrobiopterin, and also a Ca-CaM complex in the endothelial (eNOS) and neuronal (nNOS) isoforms (Nathan 1992; Bredt and Snyder 1994). In the heart, NO has been reported to have negative chronotropic and positive lusitropic effects while producing negative or positive inotropy (Massion et al. 2003). Indeed, ventricular myocyte-generated NO affects myocardial contractile performance and relaxation, depending on the intracellular location and the NO-producing NOS isoform. Thus, NO produced by caveolar eNOS activates the soluble guanylyl cyclase (sGC, the main molecular target of NO) producing cGMP which is used by cGMP-dependent protein kinases in the phosphorylation of L-type calcium channels. As a result, phosphorylation decreases β-adrenergic-induced myocardial contractile performance (Balligand et al. 1993; Bloch et al. 1999). nNOS, on the other hand, interacts with ryanodine receptors at the sarcoplasmic reticulum in cardiomyocytes. nNOS-generated NO increases myocardial contractility by stimulating calcium release from the sarcoplasmic reticulum via direct activation of these receptors (Bredt 2003). Clearly, a fine-tuned modulation of NO synthesis is then crucial for correct cardiac function. Then, an appealing hypothesis was whether y⁺ carriers represent a regulatory check-point for cardiomyocyte NO synthesis and metabolism. Equally interesting was to test whether NO synthesis is acutely self-regulated by an inhibitory effect on L-Arg transport.

Effect of NO donors

Membrane currents elicited by exposure of whole-cell voltage-clamped cardiac myocytes to 10 mM extracellular L-Arg were quickly blocked by application of the NO donor sodium nitroprusside (Fig. 2A, and Zhou et al. 2010). Similarly, both L-Lys uptake components described above in cardiac sarcolemmal vesicles were inhibited by S-nitroso-N-acetyl-DL-penicillamine (SNAP)-derived NO (Fig. 2B, and Zhou et al. 2010). Simultaneous non-linear regression of these uptake components with an equation derived from a non-competitive reaction scheme yielded SNAP inhibitory constants (K_i) of 7.3 and 22 μM, for the high- and low-affinity transporters, respectively. Thus, the high-affinity CAT-1 appeared to be ~ 3 times more sensitive to exogenous NO inhibition than the low-affinity CAT-2A. Using a NO selective electrode, SNAP K_i values were converted into 275 and 827 nM NO concentrations, for the high- and low-affinity components, respectively (Zhou et al. 2010). These NO values fall within a broad physiologic range reported for different cell types (Chen et al. 2008). Considering a total concentration of ~ 0.60 mM for circulating levels of L-Arg, L-Lys, and L-Orn, NO K_i values, when replaced in the equation derived from the non-competitive model, will result in ~ 33 and ~ 58% inhibition of total cationic amino acid transport, respectively (Zhou et al. 2010). Thus, this array of high- and low- affinity transporters shows the benefit that total uptake will be just 33% inhibited at a NO concentration of 275 nM, which, otherwise, would inhibit 50% of CAT-1 function. In fact, the low-affinity carrier will be the predominant functional transporter at intermediate NO levels because of its lower sensitivity to inhibition.

Fig. 2.

A Ten millimolar L-Arg-activated inward currents in a voltage-clamped cardiomyocyte at the holding membrane potential of − 40 mV, before or during simultaneous exposure to 2 mM of the NO donor SNP (sodium pentacyanonitrosyl ferrate(III) dihydrate). B L-Lys uptake measured in the absence or presence of 100 µM SNAP (S-nitroso-N-acetyl-DL-penicillamine) at 37 °C for the range 0.05–50 mM external L-Lys in giant sarcolemmal vesicles from rat cardiac ventricle. Symbols are the mean ± S.E. of 0.2 mM NEM-sensitive uptake from 3–5 experiments, each performed in triplicate. Curves represent the best fitting of two hyperbolic functions that include non-competitive inhibitor terms. C Slopes of the time courses of fluorescence changes induced by NO production for the range 1–50 mM L-Arg. Initial fast slopes (⬤) were analyzed by fitting a hyperbola with a best-fit K_0.5 = 2.92 ± 0.75 mM. Secondary slow rates of NO production (◯) were analyzed with an inhibitory hyperbola with K_I = 1.89 ± 0.99 mM. D Effect of endogenously produced NO on cationic amino acid-activated currents in voltage-clamped cardiomyocytes. Cells were held at − 40 mV with an electrode solution containing a calculated [Ca²⁺]_free = 460 nM, and superfused with 10 mM of either L-Lys or L-Arg

Effect of endogenous NO

Aside from the effect of NO donors, it was of central physiological importance to determine whether endogenously produced NO also modulates cationic amino acid transport. To study this aspect, acutely isolated cardiomyocytes preloaded with the dye 4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate (DAF-FM diacetate) were exposed to increasing concentrations of extracellular L-Arg. DAF-FM is highly fluorescent in the presence of NO, so that our experimental design consisted of two steps: CAT-mediated L-Arg import and NOS-mediated NO production. Time courses of spectroscopic fluorescence changes showed a biphasic behavior as a function of external L-Arg concentration, with an initial fast component followed by a slower one (Zhou et al. 2010). The slope (dF/dt) of the fast component increased with the concentration of L-Arg, while the slow component showed the opposite behavior, although both effects occurred at millimolar L-Arg indicating the involvement of CAT-2A (Fig. 2C, and Zhou et al. 2010). Control assays replacing L-Arg with L-Lys (not a NOS substrate) or incubating cardiac myocytes with the NOS inhibitor L-NAME confirmed these results (Zhou et al. 2010). Therefore, millimolar concentrations of L-Arg first accelerated the synthesis of endogenous NO, an effect that was then followed by a decrease in the rate of further NO production. These effects are consistent with a NO negative-feedback modulation of L-Arg transport. However, the well-described “inhibition by product” of NOS activity by NO can also explain these results (Rogers and Ignarro 1992; Assreuy et al. 1993; Vickroy and Malphurs 1995).

Unequivocal demonstration of a NO inhibitory effect on L-Arg transport came from electrophysiological studies. Granted that NO donors also blocked L-Arg currents in whole-cell voltage-clamped cardiac myocytes (Zhou et al. 2010), more challenging experiments were performed under conditions that allowed for NOS activity in these cells. Voltage-clamp experiments in cardiac muscle cells require high electrode concentrations of a calcium chelator to keep these excitable cells under non-contracting, resting conditions. However, as we mentioned above, a Ca-CaM complex is required for constitutive NOS proper function. In fact, calmodulin activation of NOS activity was reported to have a K_0.5 for Ca²⁺ in the range 0.2–0.4 μM (Nathan 1992). Thus, as a compromise between having myocyte resting condition while activating NOS, membrane currents were measured with a nominal concentration of 0.46 μM free intracellular Ca²⁺. L-Arg elicited inward currents that were biphasic in nature, reaching a peak ~ 6 s after exposure and then decaying exponentially in 10–20 s to a steady value that varied between 70 and 20% of the peak level, depending on the concentration of L-Arg_o (Fig. 2D, and Zhou et al. 2010). This biphasic behavior disappeared when L-Arg was replaced by L-Lys (Fig. 2D, inward currents remained at the peak level) or incubating myocytes with the NOS blocker L-NAME before adding extracellular L-Arg. Potential effects of downstream reactions were ruled out by adding sGC and cGMP-dependent protein kinase inhibitors to the electrode solution (Zhou et al. 2010). Finally, myocytes were exposed to 10 mM L-Arg and inward currents were continuously recorded for ~ 5 min at a holding membrane potential of –40 mV, displaying a recurring biphasic behavior that reappeared after ~ 3 min under our experimental conditions. Thus, L-Arg transport was shown to be reversibly blocked by NOS-derived NO in cardiac muscle cells (Zhou et al. 2010). At variance with the traditional enzymatic inhibition by products, NO can acutely self-regulate its biosynthesis by modulating the activity of the transporters that supply the substrate L-Arg to NOS.

L-Arg modulation of NOS activity

Limiting concentrations of L-Arg or tetrahydrobiopterin affect NOS enzymatic activity by uncoupling NADPH oxidation and NO production, with molecular oxygen being the final acceptor of electrons instead of the L-Arg guanidine group. This NOS aberrant activity results in the formation of the free radical superoxide (O₂^●‒) (Xia et al. 1998), which, in a diffusion-limited reaction with NO, produces peroxynitrite (ONOO^‒) (Beckman and Koppenol 1996), an oxidizing molecule implicated in decreased myocardial contractility and congestive heart failure (Ferdinandy et al. 2000). Peroxynitrite is responsible for the oxidative modification of proteins, lipids, carbohydrates, and nucleic acids. One signature effect on target proteins is the nitration in position 3 of tyrosine side-chain rings (Crow and Ischiropoulos 1996; Ducrocq et al. 1999; Reifenberger et al. 2008).

One pertinent question relates to the actual concentration of circulating L-Arg below which NOS will uncouple, triggering peroxynitrite detrimental effects. To address this issue, we studied the extracellular L-Arg dependence of NOS-mediated O₂^●‒/ONOO^‒ production in cardiac myocytes loaded with appropriate fluorescent dyes (Ramachandran and Peluffo 2017). Our results indicate that these reactive oxygen/nitrogen species will be produced by NOS at extracellular L-Arg concentrations below ~ 60 μM or below ~ 100 μM, depending on the relative circulating concentrations of L-Lys and L-Orn, which will efficiently compete with L-Arg for CAT-1 and CAT-2A binding sites (Ramachandran and Peluffo 2017). Furthermore, the endothelial isoform eNOS was identified as responsible for these effects in adult rat cardiac ventricular myocytes. Independent evidence for the L-Arg modulation of NOS activity and the presence of ONOO^‒ came from immunocytochemistry studies using anti-3-NO₂-tyrosine antibodies in cardiomyocytes. We found that the intensity of the fluorescent signal (due to nitration of tyrosine rings) was inversely proportional to the concentration of extracellular L-Arg (Ramachandran and Peluffo 2017). Therefore, extracellular L-Arg levels above 100 μM are important for physiologic intracellular NOS activity and to avoid the detrimental effects of ONOO^‒ on cell function.

The target site(s) for NO modulation in CAT-2A

NO can regulate protein function through S-nitrosation, nitration, or oxidation reactions (Heinrich et al. 2013). In particular, S-nitrosation is the reversible modification of thiol groups to form S-nitroso thiol. Thus, cysteine ‒SH side chains are likely targets for NO inhibition. In this regard, 15 Cys residues are present in CAT-2A, seven or eight of which are located in transmembrane segments (TM), according to hydropathy plots: Cys⁸⁵ (TM2), Cys¹⁷¹ (TM4), Cys²⁶⁴ and Cys²⁷³ (TM6), Cys²⁹⁹ (TM7), Cys³⁴⁷ (TM8), Cys⁴²⁷ (TM10), and Cys⁵³⁶ (TM12) (Fig. 3A, and Zheng et al. 2020). To identify the target residue(s) for NO inhibition of CAT function, the low-affinity CAT-2A was expressed in mammalian cell lines and NO sensitivity was assessed on transporter variants carrying Cys → Ala or Cys → Ser substitutions in various Cys residues. As a result of these studies, Cys³⁴⁷ in TM8 was identified as the target for NO interaction in CAT-2A by using both, NO donors or NOS endogenously generated NO (Fig. 3B, and Zheng et al. 2020). These latter studies, performed by co-expressing CAT-2A and eNOS in mammalian cell lines, highlight the physiological relevance of this NO feedback modulation on L-Arg transport. The ‒SH moiety in Cys³⁴⁷ also proved to be important for transporter function, since the Cys → Ala variant, although NO insensitive, showed a ~ 50% loss of radiolabeled L-Lys uptake levels, compared to wild-type control. Western blot and immunocytochemistry assays showed that this difference was not due to protein expression or location. Generation of the more conservative CAT-2A Cys → Ser variant (also NO insensitive, Fig. 3B) was needed in order to recover control uptake levels (Zheng et al. 2020), indicating that certain polarity in the side chain of this Cys is required for proper cationic amino acid transport.

Fig. 3.

A The 14-transmembrane-domain model for CAT-2A according to hydropathy plots. The yellow circles indicate putative locations for the fifteen Cys residues present in this transporter, seven of which are located in transmembrane segments (although Cys²⁷³ appears to be located at transmembrane domain VI as well). The 4th intracellular loop (in red) shows the 42 amino-acid stretch that distinguishes CAT-2A from its splice variant CAT-2B. There is one N-glycosylation site located at the 2nd extracellular loop (Asn¹⁵⁷) and two in the 3rd extracellular loop (Asn²²⁷, Asn²³⁹). B NO sensitivity of uptake determined with 10 mM L-Lys in control and Cys³⁴⁷Ser variant groups. SNAP concentrations were converted to NO by using the calibration curve shown in Zhou et al. 2010. Bars represent the mean percent ± S.E. relative to that of control CAT-2A (n = 3–8 for each group). Control vs. Cys³⁴⁷Ser ± SNAP treatment, not statistically significantly different. C Scheme summarizing the negative feedback modulation of CATs by NO, highlighting the target Cys³⁴⁷. Also included is the “inhibition by product” of NOS activity by NO (Rogers and Ignarro 1992; Assreuy et al. 1993; Vickroy and Malphurs 1995), a reaction that involves itself the direct modification of ‒SH residues in the synthase by NO (Ravi et al. 2004)

The identification of Cys³⁴⁷ as the target for NO inhibition of transport is also in line with our previous work showing that voltage-clamped cardiomyocytes exposed to NO exhibited a ~ 33% increase in the steepness of the 10 mM L-Arg current–voltage relationship (Zhou et al. 2010). This observation was an early clue that the target residue for NO modulation was likely located within a TM in CAT-2A. We tested three other Cys residues also located in TMs but neither NO sensitivity nor control uptake levels were affected by the Cys → Ala substitutions (Zheng et al. 2020). However, we did not experimentally test other putative target residues such as Cys²⁶⁴ and Cys²⁹⁹. Thus, for the sake of completeness, we generated and optimized a 3D structure for CAT-2A by homology modeling using as a template the high-resolution crystal structure of a bacterial CAT homologue with bound arginine, corresponding to an outward-facing occluded conformation (Kowalczyk et al. 2011; Jungnickel et al. 2018). This structure was then used for computing classical molecular interaction potentials (Gelpí et al. 2001) to predict NO-Cys interactions. At low enough free energy values, NO interacted only with CAT-2A regions surrounding Cys²⁷³ and Cys³⁴⁷ showing specificity for these two residues (Zheng et al. 2020). We were able to rule out Cys²⁷³ as participating in NO inhibition, as follows: (a) all our uptake experiments measured NEM-sensitive uptake so that Cys²⁷³, the target for NEM binding (Beyer et al. 2013), was not available to react with NO; (b) the observed NO block in NEM-sensitive uptake indicates that these two compounds were interacting with different Cys residues in CAT-2A; (c) the size of the NO block that we quantified in voltage-clamp experiments (~ 70%, Zhou et al. 2010) was similar to that measured in NEM-sensitive uptake assays when both approaches used 10 mM L-Arg_o concentration. Zero-current levels in voltage-clamp experiments were obtained just by withdrawing L-Arg from the superfusion solution instead of applying NEM so that Cys²⁷³ was also available, even though the amount of NO block did not increase.

Concluding remarks

High- and low-affinity CATs function together in the cardiac myocyte plasma membrane to, among other roles, ensure the supply of L-Arg for NO biosynthesis by NOS. However, NO cellular levels must be carefully regulated, as too little or too much of this second messenger will be harmful for almost all cell types. Because of its chemical nature, this small diatomic gas cannot be stored in intracellular compartments, unlike Ca²⁺, for example. It has been reported that ‒SH compounds, such as glutathione, cysteine, and cysteine-containing proteins, act as transient intracellular NO reservoirs through the formation of S-nitrosothiols, which are involved in intracellular signaling (Gaston 1999), similar to the described effect of extracellular, circulating S-nitroso-albumin on venodilation (Orie et al. 2005). We discussed here a unique mechanism in which, as the product of a downstream reaction, NO can self-regulate its cellular levels by acting on the membrane proteins that are responsible for delivering the substrate L-Arg to NOS (Fig. 3C).

Funding

This work was supported over the years by the National Heart, Lung, and Blood Institute (Grant R01-HL-076392) and by a Scientist Development Grant from the American Heart Association, National branch.

Declarations

Conflict of interest

The author declares no competing interests.