Identification of stereoselective transporters for S-nitroso-L-cysteine: role of LAT1 and LAT2 in biological activity of S-nitrosothiols.

Many of the biological effects of nitric oxide are mediated by S-nitrosothiols. However, the mechanisms by which S-nitrosothiols transduce their activity across cell membranes are unclear. We show that the pathway responsible for the cellular effects of S-nitrosothiols is specific for S-nitrosocysteine (CSNO), is stereoselective, and requires direct uptake of intact L-CSNO. Transport is independent of extracellular sodium, competitively inhibited by leucine, and blocked by 2-aminobicyclo[2.2.1]heptane-2-carboxylic acid, a specific inhibitor of the system L amino acid transporter family. Other nitrosothiols such as S-nitrosoglutathione are not substrates for transport and require reaction with L-cysteine for activity. To show that system L family members mediate uptake, we expressed two members, LAT1 and LAT2, in Xenopus oocytes. Both LAT1 and LAT2, when co-expressed with 4F2 heavy chain, were found to efficiently transport L-CSNO. Mammalian cells were shown to express LAT1 and LAT2. A431 cells express both proteins, whereas T24 cells express only LAT1. Overexpression of LAT1 in T24 cells using recombinant adenoviruses led to increased uptake of L-CSNO, whereas knockdown using a specific small interfering RNA led to decreased uptake. These data definitively identify LAT1 and LAT2 as members of system L that mediate transmembrane movement of l-CSNO and suggest that system L family members are involved in the cellular activity of small molecular weight nitrosothiols.

Nitrosothiols including the low molecular weight nitrosothiols, S-nitrosoglutathione (GSNO) 1 and S-nitrosocysteine (CSNO), have been proposed to be important mediators of nitric oxide signaling (1)(2)(3). The biological activity of these compounds resides in their ability to reversibly modify critical proteins in cell signaling pathways through targeted oxidation of cysteine thiols (1)(2)(3)(4). Although their biological activity is well established, the mechanisms by which they transduce their activity from the extracellular space to intracellular targets are largely unknown.
Insight into pathways that may participate in the transfer of NO equivalents across cell membranes comes from experiments that show that nitrosothiols are not equivalent in their ability to alter signaling pathways. For example, CSNO but not GSNO or S-nitroso-N-acetylpenacillamine (SNAP) has been shown to modify protein-tyrosine phosphatase 1B (PTP1B) (5) and glyceraldehyde-3-phosphate dehydrogenase (6) in intact cells. Similarly, CSNO but not GSNO is able to transfer NO equivalents across red cell membranes, leading to the formation of intracellular GSNO and S-nitrosohemoglobin (7,8). The transfer of NO equivalents does not appear to involve NO release through the breakdown of nitrosothiols (1, 9 -11). Rather it seems that transfer requires specific interactions with proteins since the bioactivity of low molecular weight thiols is known to be stereoselective and unrelated to chemical half-life (1, 9 -13). For example, in the circulation, L-CSNO is a potent vasodilator, but D-CSNO is not (14), and in the central nervous system, L-CSNO rather than D-CSNO is important in regulating ventilatory signaling (13). Furthermore, in cells that possess ␥-glutamyl transpeptidase, GSNO is only active following cleavage to S-nitrosocysteinylglycine (15). Together, these data strongly suggest the existence of a stereoselective mechanism that is characterized either by a receptor-mediated process selective for L-CSNO or by selective transport of intact CSNO from outside to inside the cell.
The involvement of a transporter is supported by data from our group and from others demonstrating that amino acids that competitively inhibit members of the system L amino acid transporter family block the biological activity of CSNO in intact cells (5,16,17). These transporters exist as heterodimers between a common 80-kDa heavy chain (4F2HC) and a variable 40-kDa light chain (4F2LC) that confers transport activity (18 -24). System L transporter activity is derived from two characterized light chains, LAT1 and LAT2. LAT1 is relatively specific for large neutral amino acids, whereas LAT2 has a broader specificity and transports both large and small neutral amino acids (18,(22)(23)(24). Both transporters have wide tissue distribution and are critical for amino acid uptake and transfer across endothelial and epithelial barriers, especially in the brain, kidney, and intestine (21)(22)(23)(24)(25). Both are sodium-independent exchangers with symmetrical extracellular and intracellular substrate specificities but greatly different extracellular (high) and intracellular (low) affinities (25). Importantly, both have been reported to transport mercurial derivatives of cysteine (26).
In the present study, we have investigated pathways used by exogenous nitrosothiols to modify cellular proteins. We report * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  the existence of a transport mechanism that stereoselectively transports L-CSNO into cells and show that uptake activity is mediated by proteins that constitute system L amino acid transport. In addition, we provide definitive data that show that L-CSNO is a substrate for LAT1 and LAT2. These studies are important since they show that specific mechanisms exist for cellular responses to nitrosothiols and identify a novel pathway for movement of nitrosothiols into cells, where they can efficiently exert their biological effects.

EXPERIMENTAL PROCEDURES
Materials-GSNO was synthesized as described previously (5). CSNO was prepared by mixing cysteine hydrochloride (pH 2.5) and sodium nitrite at a molar ratio of 2:3 in water. Complete  Determination of PTP1B Oxidation-PTP1B oxidation was determined using MPB as described previously (5). Briefly, after treatment, cells were washed three times with ice-cold HBSH and then scraped into RIPA buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, and 1 mM phenylmethylsulfonyl fluoride) containing 20 mM N-ethyl maleimide. After incubation on ice for 30 min, samples were centrifuged at 14,000 rpm for 10 min at 4°C, and the supernatant was applied to a PD-10 desalting column (Amersham Biosciences) to remove excess N-ethyl maleimide. The eluate that contained cellular protein was treated with 20 mM dithiothreitol for 30 min to reduce all oxidized thiols. Samples were reapplied to a PD-10 column to remove excess dithiothreitol, and the eluates were treated with 50 M MPB. Excess MPB was removed by elution from a PD-10 column, and eluates were treated with 50 M glutathione. Aliquots containing equivalent amounts of protein were immunoprecipitated using 1 g of PTP1B monoclonal antibody (Oncogene) by incubation at 4°C overnight. Immune complexes were precipitated by incubation for 2 h with 20 l of protein A-agarose (Santa Cruz Biotechnology) and collected by centrifugation. Immunoprecipitates were washed five times with RIPA buffer, dissolved in 20 l of electrophoresis sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerin, and 0.01% bromphenol blue), separated by 10% SDS-PAGE, and transferred to polyvinylidene difluoride membranes. After blocking with 10% nonfat dried milk in TBST buffer (20 mM Tris-HCl, pH 7.6, 137 mM sodium chloride, 0.1% Tween 20), MPB-labeled PTP1B was visualized by incubation with 1:4000 dilution of NeutrAvidin-linked horseradish peroxidase (avidin-HRP, Molecular Probes) followed by incubation with the ECL Plus detection system (Amersham Biosciences). Blots were stripped using 100 mM ␤-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl, pH 6.7, at 50°C for 30 min, reprobed using rabbit polyclonal PTP1B antibody (Santa Cruz Biotechnology), and visualized using anti-rabbit IgG coupled to horseradish peroxidase followed by visualization with ECL Plus.
Direct Measurement of Amino Acid Uptake-L-[ 35 S]CSNO was prepared by mixing L-cysteine (30 Ci/mol [ 35 S]cysteine, pH 2.5) and sodium nitrite at a molar ratio of 2:3 in water. The resulting CSNO was neutralized by diluting the reaction mixture in HBSH. Cells were incubated with either L-[ 35 S]CSNO or [ 3 H]leucine (25 Ci/mol) in HBSH at 37°C, washed five times with ice-cold HBSH to remove extracellular radioactivity, and lysed into RIPA buffer with 1% SDS. Extracts collected after trichloroacetic acid precipitation produced identical results. Protein concentration was quantified by BCA assay (Pierce), and radioactivity was determined by liquid scintillation spectrometry. Uptake was linear for at least 15 min.
RT-PCR-Total RNA was isolated from cells using the RNeasy Mini kit (Qiagen) and reverse-transcribed by oligo(dT) using an Omniscript RT kit (Qiagen) at 37°C for 60 min. A fragment of LAT1 (554 bp) was amplified by PCR using primers corresponding to the partial sequence from nucleotides 418 to 971 of human LAT1 (forward, 5Ј-ATC CGG CCT TCA TCG CAG TAC ATC-3Ј, and reverse, 5Ј-GAC ATG ACG CCC AGG TGA TAG TTC-3Ј). A fragment of LAT2 (706 bp) was amplified using primers corresponding to the partial sequence from nucleotides 4 to 710 of human LAT2 (forward, 5Ј-GAA GAA GGA GCC AGG CAC CGA AAC-3Ј, and reverse, 5Ј-CCA GTG CGA CGA GGC CGA TGT CAG-3Ј). PCR products were separated by electrophoresis in a 2% agarose gel and visualized under UV light in the presence of ethidium bromide.
Expression of System L in Xenopus Oocytes and Measurement of L-CSNO Uptake-Xenopus laevis ovarian lobes containing stage V or VI oocytes were purchased from Nasco. Oocytes were treated with collagenase A (2 mg/ml) for 30 -50 min at room temperature in Ca 2ϩ -free medium (82.5 mM NaCl, 2 mM KCl, 1 mM MgCl 2 , and 5 mM HEPES, pH 7.5) to remove the follicular layer and then maintained in buffer containing 100 mM NaCl, 2 mM KCl, 1.8 mM CaCl 2 , 1 mM MgCl 2 , 5 mM HEPES, and 100 g/ml gentamicin, pH 7.6 at 18°C. 4F2HC cDNA was obtained from human aortic smooth muscle cells by RT-PCR using the following primer pair: forward, 5Ј-TAA GGA TCC ATG AGC CAG GAC ACC GAG G-3Ј, and reverse, 5Ј-TAA GAA TTC TCA GGC CGC GTA GGG GAA-3Ј. Human LAT2 cDNA was obtained from A431 cells by RT-PCR using the following primer pair: forward, 5Ј-TA GGA TCC ATG GAA GAA GGA GCC AGG CAC-3Ј, and reverse, 5Ј-TA GAA TTC GCC AGG GAA TGG TGG TCC TCA-3Ј. LAT1 cDNA was obtained as described below. The 4F2HC, LAT1, and LAT2 cDNAs were subcloned into WCV vector (a kind gift of Dr. Antonius M. J. VanDongen) under the control of the T7 promoter. Capped cRNAs were synthesized in vitro using T7 RNA polymerase, purified by RNeasy Mini kit (Qiagen), and injected (73.6 nl of cRNA, 200 ng/l) into the defolliculated oocytes under a dissecting microscope with a micrometer-driven micropipetter. Pure water was injected as control. Two days following injection, oocytes were washed and then incubated in 500 l of uptake buffer (100 mM choline chloride, 2 mM KCl, 1.8 mM CaCl 2 , 1 mM MgCl 2 , 10 mM HEPES, and 5 mM Tris, pH 7.4) containing 25 M L-[ 35 S]CSNO (30 Ci/mol) with 1 mM MTSES for 15 min at room temperature. The oocytes were then washed five times with ice-cold uptake buffer, and radioactivity was determined by liquid scintillation spectrometry.
Construction of Recombinant LAT1-Human LAT1 cDNA was obtained from A431 cells by RT-PCR using the following primer pair: forward, 5Ј-TAG GAT CCA TGG CTG GTG CGG GCC CGA AG-3Ј, and reverse, 5Ј-TAG AAT TCC TAT GTC TCC TGG GGG ACC AC-3Ј. cDNA was subcloned into adenoviral shuttle vector pAdtrack-CMV (ATCC) and then transformed into BJ5183-AD-1 cells (Stratagene) by electroporation. The positive adenoviral recombinants were identified by sequencing and transfected to Ad-293 cells (Stratagene) using PolyFect transfection reagent (Qiagen) following the manufacturer's instructions. Recombinant adenoviruses were generated and titrated as described everywhere (27).
Construction of Vector-based Small Interfering RNA to LAT1-An RNA polymerase III generated hairpin small RNA was used to knock down the expression of LAT1. Oligonucleotides containing sense and antisense 22-nt sequences (underlined) corresponding to nucleotides 1173-1194 of the LAT1 coding region were synthesized as follows: forward, 5Ј-AGC TTC CGG ACA TCT TCT CCG TCA TCA ACA AGC TTG TTG ATG ACG GAG AAG ATG TCC TTT TTA T-3Ј, and reverse, 5Ј-CGA TAA AAA GGA CAT CTT CTC CGT CAT CAA CAA GCT TGT TGA TGA CGG AGA AGA TGT CCG GA-3Ј. The oligonucleotides were annealed and ligated into the linearized pGEM-T-U6 vector with the hairpin sequence under the control of the U6 promoter. The U6 promoter was amplified from Jurkat cells by RT-PCR using the following primer pair: forward, 5Ј-TA GTC GAC CCC CAG TGG AAA GAC GCG CAG GCA-3Ј, and reverse, 5Ј-TTT TTA TCG ATC CAA GCT TGT CCT TTC CAC AAG ATA TAT AA-3Ј. The RT-PCR product was inserted into pGEM-T vector (Promega) by T/A cloning, and the positive recombinants were linearized with HindIII and ClaI for ligation with the annealed oligonucleotides described above. The recombinant plasmid was then digested with SalI and NotI, and the fragment containing the U6 promoter and the hairpin sequence was subcloned into an adenoviral shuttle vector pAdtrack (ATCC). The viruses were prepared as above. All the constructions were confirmed by DNA sequencing.

RESULTS
Biological Activity of CSNO Is Stereoselective-We have previously shown that L-CSNO oxidatively inhibits PTP1B in intact cells, leading to increased phosphorylation of epidermal growth factor receptor. Because others have found that the activity of nitrosothiols is stereospecific, it seemed reasonable to propose that PTP1B oxidation is mediated by L-CSNO but not by D-CSNO. As shown in Fig. 1, when A431 cells were treated with freshly prepared L-CSNO or with an equimolar mixture of SNAP and L-cysteine, intracellular PTP1B was rapidly oxidized as indicated by increased labeling with MPB (upper panel). D-CSNO or the combination of SNAP plus Dcysteine were largely inactive. Consistent with previous studies (5), neither L-cysteine nor SNAP when used alone had any effect. In addition, cystine, NaNO 2 , or spermine NONOate (NO donor) were also inactive (5). It should be pointed out that Land D-CSNO have similar half-lives, leading to the formation of cystine and NO. As L-and D-CSNO have such dramatically different abilities to modify PTP1B, these breakdown products are unlikely important. These findings demonstrate that the mechanism is stereoselective and strongly support our earlier conclusion that a specific transporter or protein receptor that is specific for L-CSNO is involved (5).
CSNO Is Taken Up Intact by Cells-To characterize the mechanism involved in movement of NO equivalents from the extracellular to intracellular space, we examined the effect of MTSES on PTP1B modification. MTSES is well known to rapidly alkylate free thiols but does not enter the cell. Only thiols in the medium or on the extracellular surface are available for modification, leaving intracellular thiols unaffected. The effect of MTSES on PTP1B oxidation was investigated by adding 10 mM MTSES to the medium bathing cells ( Fig. 2A). MTSES had no effect on the ability of L-CSNO to oxidize PTP1B, demonstrating that free cell surface thiols are not involved in the bioactivity of L-CSNO and confirming that MTSES does not enter cells and modify PTP1B. The absolute requirement for L-CSNO was further demonstrated by studying the effect of GSNO in this model. We had previously shown that the incubation of cells with GSNO plus L-cysteine rapidly oxidizes PTP1B (5). However, when we added 0.25 mM L-cysteine to medium containing 10 mM MTSES followed immediately by the addition of 0.25 mM GSNO, we saw no PTP1B oxidation, suggesting that MTSES rapidly alkylates free thiols and prevents the formation of CSNO. This is in contrast with the pronounced PTP1B oxidation seen if L-cysteine and GSNO were mixed for 5 min at room temperature immediately prior to addition. Also, since MTSES would alkylate accessible cell surface thiols, it is unlikely that free thiols participate in a transnitrosation mechanism and suggests that movement of NO equivalents into cells occurs via transport of intact L-CSNO.
In the following experiments, we investigated direct cellular uptake of L-CSNO. When A431 cells were treated with 25 M L-[ 35 S]cysteine, rapid uptake was seen (Fig. 2B). Co-incubation with 1 mM MTSES to alkylate free thiols blocked uptake of cysteine by 95%. The remaining 5% is likely nonspecific binding. These results show that MTSES effectively prevents uptake of cysteine and show that alkylated cysteine is not taken up. In similar studies, we prepared L-[ 35  Finally, the addition of MTSES did not alter [ 3 H]leucine uptake, demonstrating that system L transport is not affected by this reagent. These results suggest that CSNO is taken up intact by cells.
Direct uptake of intact CSNO was further shown in the following experiments in which we investigated the pH dependence of both CSNO formation and uptake. In these experiments, L-cysteine and NaNO 2 were mixed at the pH indicated, and CSNO formation was determined spectrophotometrically at 335 nm. To measure uptake, L-[ 35 S]cysteine and NaNO 2 were mixed under the same conditions, diluted into HBSH (final pH ϭ 7.4), and added to cells incubated in HBSH containing 1 mM MTSES. Because CSNO formation in solution when L-cysteine and NaNO 2 are combined is pH-dependent (28), uptake measured using this approach should exhibit the same pH dependence. As expected, CSNO formation was nearly 100% reaction at pH Ͻ 3 (Fig. 2C). [ 35 S]CSNO uptake showed similar pH dependence (Fig. 2D). In fact, the relationship between CSNO formation and uptake was linear throughout the entire range of pH tested (Fig. 2E), which confirms that CSNO specifically is the molecule responsible for the uptake observed. Cystine uptake is minimal in this model and is unaffected by MTSES (data not shown). Further, cystine is unlikely an important contributor to uptake of label since [ 35 S]CSNO was prepared and immediately added to cells.
Role of System L in Biological Effects of L-CSNO-Amino acid transporters were initially characterized based on their selectivity for a range of amino acids. In the case of system L, transport selectivity resides in two proteins, LAT1 (large neutral amino acid transporter) and LAT2 (both large and small neutral amino acid transporter) (21)(22)(23)(24)(25). The ability of these proteins to transport amino acids depends on their affinity for individual substrates. As such, substrates such as leucine with high affinity are efficiently transported and compete effectively with other amino acid substrates for uptake. Using PTP1B oxidation as an index for uptake and cellular activity of CSNO, we have studied the competition profile of several different amino acids (Fig. 3, A and B). As expected for system L transport, neutral, non-polar amino acids effectively block transport of CSNO and modification of PTP1B (L-cysteine, L-isoleucine, L-leucine, L-methionine, L-phenylalamine, and L-tryptophan). Modification was partially inhibited by L-histidine, L-serine, L-threonine, and L-valine, whereas polar amino acids (L-arginine, L-asparagine, L-glutamic acid, glycine, L-lysine, and Lproline) were not effective. These data are significant since they clearly demonstrate that the biological activity of extracellular nitrosothiols is regulated by manipulation of transport. The pattern seen in Fig. 3B is virtually identical to that reported for inhibition of leucine uptake by system L in other After treatment, cells were washed four times with ice-cold HBSH to remove extracellular label, and intracellular radioactivity was determined by liquid scintillation spectrometry using an aliquot of the cell extract prepared in RIPA buffer. Data are given as means Ϯ S.E., n ϭ 3-12. C, pH dependence of L-CSNO formation was determined by mixing 500 l of 100 mM L-cysteine in HBSH buffered at the pH indicated with 500 l of 150 mM sodium nitrite. After incubation in the dark at room temperature for 10 min, the formation of CSNO was determined by measuring the absorbance at 335 nm of a 1:50 dilution of the reaction mixture. D, the pH dependence of L-CSNO was determined by incubating A431 cells in the presence of 1 mM MTSES for 15 min at 37°C with 25 M L-[ 35 S]CSNO prepared by premixing 100 mM L-[ 35 S]cysteine (30 Ci/mol) with 150 mM sodium nitrite in HBSH at the pH indicated for 10 min prior to dilution and addition to cells. After incubation, cells were washed four times with ice-cold HBSH, and intracellular radioactivity was determined by liquid scintillation spectrometry. E, the relationship between formation and uptake of L-CSNO was linear (r 2 ϭ 0.98) over the range of pH tested. Data are given as means Ϯ S.E., n ϭ 3. cells (18,(22)(23)(24) and strongly suggests the involvement of this amino acid transport system.
We have further characterized system L transport of L-CSNO by investigating its requirement for Na ϩ and inhibi-tion by amino acid analogs. System L is known to mediate Na ϩ -independent transport and to be selectively inhibited by BCH (18,23). In the following experiments, we treated A431 cells with L-CSNO in HBSH buffer or sodium-free HBSH buffer FIG. 3. Role of system L amino acid transporter in L-CSNO biological activity. A, the effect of extracellular amino acids on PTP1B modification by L-CSNO was studied in A431 cells preincubated for 15 min with 10 mM indicated amino acid followed by a 15-min incubation with the amino acid plus 0.25 mM CSNO. Cell extracts were prepared in RIPA buffer, and PTP1B modification was assayed as in Fig. 1. Blots were stripped, and total PTP1B protein (native and oxidized) loaded in each lane was visualized using rabbit anti-PTP1B followed by anti-rabbit IgG-HRP (lower blot) to document equal loading. B, the histogram was calculated based on the density of the modified PTP1B band (A, upper panel) divided by the density of the protein band (A, lower panel) for each amino acid. Data are given as means Ϯ S.E., n ϭ 3. C, extracellular sodium was found to have no effect on L-CSNO uptake by the incubation of A431 cells with L-CSNO for 15 min in HBSH or buffer in which sodium chloride was replaced with choline chloride. PTP1B modification was determined as in Fig. 1. D, the system L inhibitor, BCH, was added to A431 cells incubated with L-CSNO or an equimolar mixture of SNAP plus cysteine for 15 min in HBSH. The effects of ␣-(methylamino)isobutyrate (MeAIB) (specific inhibitor for system A amino acid transport) were studied in parallel experiments. Cell extracts were prepared, and PTP1B modification was determined as in Fig. 1. in which sodium is replaced by choline. In addition, we treated cells with L-CSNO in the presence of BCH or ␣-(methylamino)isobutyrate, a selective inhibitor of system A transport. Again, we used oxidation of PTP1B as an index for transport and biological activity of L-CSNO. As seen in Fig. 3C, PTP1B oxidation is independent of extracellular Na ϩ . Further, as shown in Fig. 3D, oxidation is inhibited by BCH but not by ␣-(methylamino)isobutyrate. BCH also blocked the effect of SNAP plus cysteine, confirming the requirement for CSNO formation and transport for its biological activity.
Role of System L in L-CSNO Uptake-Having shown that system L is required for the biological effects of L-CSNO, we next examined its role in uptake. To accomplish this, we used the approach outlined in Fig. 2B using MTSES to evaluate transport of intact CSNO into cells. As shown in Fig. 4A, increasing concentrations of BCH block uptake of L-[ 35 S]CSNO by A431 cells. In addition, when we compared a range of BCH concentrations on L-[ 35 S]CSNO uptake, we found that inhibition was competitive (Fig. 4B). The effectiveness of BCH as a competitive inhibitor of L-CSNO uptake was similar to that for inhibition of leucine uptake (data not shown) and demonstrates that L-CSNO and BCH act at the same transporter. In addition, inhibition of L-CSNO uptake by BCH shows that cystine uptake is not important in this model because BCH does not inhibit system x Ϫ c , the major transport system for cystine. As a substrate for system L transport, L-CSNO should act as a competitive inhibitor of transport of other amino acids. Because leucine is the prototypical substrate for system L transport with an apparent K m of ϳ3-19 M (18, 19, 29, 30), we chose to investigate the effect of L-CSNO on leucine uptake. When increasing concentrations of L-CSNO or a combination of GSNO and L-cysteine were added to the medium bathing A431 cells, [ 3 H]leucine uptake was effectively blocked (Fig. 5). The inhibitory effect was not likely due to nonspecific thiol nitrosylation since GSNO alone did not alter uptake and since MTSES does not alter leucine uptake (Fig. 2B). On the other hand, L-cysteine, which is a weak substrate for system L transport (transported primarily by other systems), weakly blocked uptake. These findings are important because they demonstrate that leucine and L-CSNO effectively compete with each other for transport (Figs. 3 and 5), confirming that both are substrates for uptake by the same system L transporter, and show that GSNO is not a substrate for transport.
Reconstitution of System L in Oocytes Leads to L-CSNO Uptake-The data presented thus far show that CSNO is taken up in a stereospecific manner by an amino acid transporter with transport characteristics and pharmacologic properties that suggest the involvement of system L proteins. Currently, two proteins that display system L-like activity have been identified (LAT1 and LAT2). To investigate the possibility that these proteins mediate uptake of L-CSNO, we microinjected cRNAs corresponding to the full-length coding sequence of either LAT1 or LAT2 into Xenopus oocytes and studied their ability to transport L-[ 35 S]CSNO. Since both LAT1 and LAT2 require heterodimerization with 4F2 heavy chain for activity, we coinjected cRNA encoding this protein in some oocytes (Fig. 6A). Basal uptake of labeled L-CSNO by oocytes was low. Expression of LAT1, LAT2, or 4F2 heavy chain alone did not appreciably alter uptake. Co-expression of LAT1 plus 4F2 heavy chain or LAT2 plus 4F2 heavy chain led to an ϳ20-fold increase in CSNO uptake for LAT1 and a 4-fold increase for LAT2. In addition, the formation of a functional heterodimer between LAT1 or LAT2 and 4F2 heavy chain appears to be an absolute requirement for uptake and further identifies system L proteins as mediators of L-CSNO uptake. This requirement for co-expression of LAT1 or LAT2 and 4F2 heavy chain for leucine uptake by oocytes has been demonstrated by others and was confirmed by us in experiments not shown (18,(21)(22)(23).
Expression of System L Proteins in Mammalian Cells Mediates L-CSNO Uptake-System L components are widely distributed in mammalian cells and likely account for the L-CSNO uptake we measured in A431 cells (Figs. 2-4). Using RT-PCR, we found that A431 cells express both LAT1 and LAT2 (Fig.  6B). Although there are differences in amino acid substrate competition profiles between LAT1 and LAT2, both proteins transport leucine and both are inhibited by BCH (18,23), making it difficult to determine their contribution to overall transport. Thus to examine the role of LAT proteins in L-CSNO transport, we used T24 cells, a cell line that expresses only LAT1 (Fig. 6B) (31). Our approach was to modulate expression of LAT1 and examine the effect that altered expression had on uptake of L-CSNO. T24 cells were transduced with a recombinant adenovirus containing either full-length human LAT1 to increase LAT1 expression or constructs containing a small interfering RNA that we have shown in previous experiments reduces expression of LAT1 mRNA levels. Cells overexpressing LAT1 or cells in which LAT1 was knocked down were incubated with L-[ 35 S]CSNO in the presence of MTSES and uptake of intact L-CSNO measured as before. Overexpression of transporter led to increased uptake, whereas knockdown led to decreased uptake (Fig. 6C). These data show that manipulation of transporter levels (up or down) produces the corresponding change in L-CSNO transport and definitively shows that LAT1 is capable of transporting intact L-CSNO in mammalian cells. Although the role of LAT2 was not studied in mammalian cells, it is likely that this protein also participates, as shown by expression in oocytes. DISCUSSION Several mechanisms have been proposed for the transfer of NO equivalents from nitrosothiols across biological mem-branes. These include transnitrosation reactions in which NO equivalents are passed between protein thiols across the membrane (6) and breakdown of nitrosothiols and NO release perhaps mediated by protein disulfide isomerase (32,33). However, transnitrosation seems unlikely because nitrosothiols such as GSNO or SNAP, which should be effective donors for these reactions, are not active in models in which CSNO exerts potent biological effects (5,6,8,34). In addition, NO release following the breakdown of nitrosothiols seems unlikely because (a) NO donors such as spermine NONOate do not mimic the effects of nitrosothiols (5); (b) the potency of low molecular weight nitrosothiols as vasodilators is unrelated to their halflife (9); (c) although D-CSNO and L-CSNO have a similar halflives, only L-CSNO is active in several systems (14,35); and (d) Oocytes were collected and washed five times with ice-cold uptake buffer, and radioactivity was determined by liquid scintillation spectrometry. Data represent means Ϯ S.E. (n ϭ 7-13; *, p Ͻ 0.01 when compared with control). B, L-type amino acid transporter expression in A431 and T24 cells. Total RNA was isolated from A431 cells or T24 cells, reverse-transcribed by oligo(dT), and amplified for 30 cycles using LAT1-or LAT2specific primers. The PCR products are separated by electrophoresis in a 1.5% agarose gel and visualized under UV light in the presence of ethidium bromide. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. C, the effect of LAT1 or LAT2 expression on L-CSNO uptake in mammalian cells. T24 cells were transduced with recombinant adenoviruses encoding LAT1 or small interfering RNA to LAT1. After 3 (overexpression) or 6 days (knockdown), cells were incubated for 10 min at 37°C with 0.25 mM L-[ 35 S]CSNO (30 Ci/mole) in HBSH in the presence of 1 mM MTSES. After incubation, cells were washed four times with ice-cold HBSH to remove extracellular label, and intracellular radioactivity was determined by liquid scintillation spectrometry. Data are given as means Ϯ range of three separate experiments. the effect of L-CSNO is blocked by inhibitors of amino acid transport.
In the present study, we have extended this work and demonstrated that cells can take up L-CSNO directly. This is an important finding since it provides an efficient mechanism by which extracellular S-nitrosothiols can enter a cell to exert their effect. To show uptake, we used a novel approach in which the cell-impermeant thiol-alkylating reagent, MTSES, was added to incubation medium to block all free thiols. Under these conditions, uptake of cysteine was completely inhibited, uptake of a mixture of cysteine and nitrite was pH-dependent, and leucine uptake was normal. Thus when cells are incubated with L-[ 35 S]CSNO, the appearance of label in the cell represents transport of the intact molecule. It seems clear, therefore, that transduction of the biological activity of nitrosothiols is mediated by a transport mechanism that offers specificity and selectivity as suggested by us and others (5,8,13,17). In support of this idea, several recent studies have shown that inhibition of system L amino acid transport by BCH reduces cellular effects of CSNO (5,16,17,36). Although suggestive, these pharmacological approaches fall short of defining which transporters are capable of nitrosothiol uptake.
It should also be mentioned that other pathways have been shown to be important for nitrosothiol uptake. Cleavage of GSNO by extracellular ␥-glutamyl transpeptidase (37) appears to be required especially in the brain, where it was shown that ␥-glutamyl transpeptidase knock-out mice do not respond to GSNO or to hypoxia-induced increases in minute ventilation (13). The importance of ␥-glutamyl transpeptidase was confirmed by showing that S-nitrosocysteinylglycine, the product of ␥-glutamyl transpeptidase activity on GSNO, was active. This is interesting since S-nitrosocysteinylglycine is also active in producing S-nitrosohemoglobin when incubated with red cells (8). Whether S-nitrosocysteinylglycine is a substrate for system L transport is not known. However, it has been shown that dipeptidases that cleave cysteinylglycine also cleave Snitrosocysteinylglycine (38) and may function to release Snitrosocysteine that can be taken up subsequently by the system L transport system.
Using molecular approaches, we have shown directly that system L components mediate L-CSNO uptake by cells. First, reconstitution of system L in oocytes demonstrates that both LAT1 and LAT2 can transport L-CSNO. That transport activity was strictly dependent on co-expression of 4F2 heavy chain is similar to previous data from others that show that expression of LAT1 or LAT2 along with heavy chain in oocytes is required for leucine and phenylalanine uptake (18,(21)(22)(23)(24). In addition, we show that overexpression of LAT1 in mammalian cells leads to increased uptake, whereas knockdown leads to decreased uptake of L-[ 35 S]CSNO, demonstrating that manipulation of light chain expression in cells that express only LAT1 alters L-CSNO transport. We have not examined the relative contribution of LAT1 and LAT2 to L-CSNO transport in cells that express both light chains (for example, A431 cells). Clearly both may participate. The activity of individual light chains in this regard will depend on the apparent K m for transport, on levels of protein expression, and on intracellular amino acid concentration. Previous studies have shown that the K m for leucine transport by LAT1 and LAT2 are quite different (21)(22)(23), suggesting important differences in their biological roles.
The ability of LAT1 and LAT2 to transport L-CSNO has several important implications. First, since transport via a transport protein should be stereoselective, these data provide for the first time an explanation for stereoselective activity of nitrosothiols reported by others (13,14,17). It seems likely that the ability of L-CSNO but not D-CSNO to increase minute ventilation at the nucleus tractus solitarius (13) and the ability of L-CSNO but not D-CSNO to relax smooth muscle (14) are due to the selective binding properties of LAT proteins. Furthermore, since these proteins appear to be specific for L-CSNO transport, their involvement in nitrosothiol uptake explains why GSNO and SNAP are inactive unless co-incubated with L-cysteine. Uptake of L-CSNO by a specific transporter also implies rapid access to the intracellular environment. This may be especially important for regulation of proteins in close proximity to the cell membrane or near sites containing high concentrations of transporter and may be illustrated by our earlier work showing that protein-tyrosine phosphatases that are colocalized with receptor tyrosine kinases on the intersurface of the membrane are sensitive targets for reversible inactivation by L-CSNO (5). In addition, rapid, efficient entry of CSNO may also be involved in the ability of CSNO, but not GSNO, to increase intracellular concentrations of GSNO in red blood cells (7). Uptake mediated by a transporter also implies transmembrane movement of intact L-CSNO. We have demonstrated this in the current report using MTSES to show that NO equivalents pass directly through biological membranes.
In conclusion, the ability of LAT1 and LAT2 to transport L-CSNO provides a novel mechanism for the biological activity of nitrosothiols. Although clearly important in situations in which cellular responses show stereoselectivity and specificity to various low molecular weight nitrosothiols, more work is required to establish the physiological role of system L transport in nitric oxide biology.