The Transport of Alanine, Serine, and Cysteine in Cultured Human Fibroblasts*

The transport of L-alanine, L-serine, and L-cysteine has been studied in skin-derived diploid human fibroblasts in culture. Competition analysis, mathematical discrimination by nonlinear regression, and conditions varying the relative contribution of the various mediations have been used to characterize the systems en- gaged in the inward transport of these amino acids. All the adopted criteria yielded results showing that L- alanine, L-serine, and L-cysteine enter the cell by two Na’-dependent systems, System A and System ASC, and by a Na+-independent route, whose major component has been identified as System L. The apparent affinity of L-alanine, L-serine, and L-cysteine for the putative carrier was higher for System ASC than for System A. The transport V,, for System A increased in response to cell starvation; after 12 h, its values were similar or higher than those exhibited by System ASC. At amino acid concentrations approaching those pres- ent in human plasma, System ASC appeared to be the primary mediation for the inward transport of L-ala- nine, L-serine, and L-cysteine in human fibroblasts. The contribution of System A was negligible in nonstarved cells and became appreciable under conditions of cell starvation. The Na+-independent System L made no substantial contribution to the uptake of L-alanine and L-serine and accounted for approximately one-fourth of the total uptake of L-cysteine. ~-[~~S]cysteine (33 mCi/mmol), [meth~l-~Hlthymidine (5 Ci/mmol), and [%!urea (59 mCi/mmol) were obtained from Amer- sham; [methoxy-'Hlinulin (102.5 mCi/g) was purchased from New England Nuclear. 2-(Methy1amino)isobutyric acid was from Aldrich. 2-Aminobicyclo(2,2,1)heptane-2-carboxylic acid (BCH, isomeric form b(&)) and 4-amino-1-guanylpiperidine-4-carboxylic acid were gener- ous gifts from Professor Halvor N. Christensen, University of Michi-gan, Ann Arbor. Sigma was the source of all unlabeled amino acids and other chemicals.

In order to characterize System ASC in cultured human fibroblasts, we have further extended the criteria of identification to inchde competition analysis, mathematical discrimination by nonlinear regression, and conditions in which the activity of System A was minimized by repression; we have also extended the study to three putative substrates of System ASC, namely L-alanine, L-serine, and L-cysteine. L-Serine is known to be a preferential substrate of System ASC in a number of tissues and cells (5-9); L-cysteine has been identified as the transport-specific substrate of System ASC in rat hepatocytes (10, 11).
In the present paper, some indication of the physiological role of System ASC has been derived from the study of the transport of L-alanine, L-serine, and L-cysteine at external concentrations miming those present in human plasma. The mutual interactions between System ASC and System A (as the preponderant mediations for the Na+-dependent uptake of these amino acids by human fibroblasts) have been investigated by varying the nutritional state of the cells (starvation versus amino acid refeeding) which is known to alter the activity of System A (12, 13).
We report here a formal characterization of System ASC in cultured human fibroblasts, its relevance to the uphill transport of neutral amino acids, and its participation in determining the physiological level at which amino acids are held intracellularly.

EXPERIMENTAL PROCEDURES
Cell Culture-Human fibroblasts, obtained from skin biopsy explants as described previously (I), were routinely grown in 10-cm diameter dishes (Costar, Cambridge, MA) in Medium 199 containing 10% fetal calf serum. Conditions of culturing were: pH 7.4; atmosphere, 5% COe in air; temperature, 37 "C. All measurements of amino acid transport were made on fibroblast subcultures resulting from 4 X IO4 cells seeded onto 2-cm2 wells of disposable multiwell trays (Costar) and incubated for 96 h in 1 ml of growth medium (always renewed 24 h before the experiment). At this time, the cells were confluent, as estimated visually with an inverted phase microscope, and slowly growing, as assessed by [methyl-'H]thymidine incorporation and conventional autoradiography of human fibroblasts exposed to ["Hlthymidine pulses (14, 15).
Human fibroblasts were used between the 10th and the 20th passages; in this interval, they retained a normal complement of chromosomes, as assessed by a trypsin-Leishman banding technique (16).
Incubations and Uptake Assay-In all experiments described in in Earle's balanced salt solution always containing 10% dialyzed fetal this paper (depletion, derepression), cell monolayers were incubated calf serum (1,2). Depletion of intracellular amino acid pools lasted 90 Alanine, Serine, and Cysteine Transport 9583 min before uptake assay (cf. Ref. 1). Amino acid uptake was assessed under conditions approaching initial entry rates in 10-s incubations at 37 "C using the cluster-tray method for rapid measurement of solute fluxes in adherent cells described by Gazzola et al. (17). During this period, serum was omitted. In designated experiments, the medium contained nonmetabolizable amino acid analogues (MeAIB or BCH) at the desired concentrations and/or choline replaced Na' in the sodium salts of the Earle's mixture; in the latter case, cell monolayers were incubated for 10 min in the Na+-free medium before uptake assay. When L-cysteine was the tracer substrate for transport, the uptake medium contained 10 mM dithiothreitol to ensure complete reduction of the amino acid. Assays were terminated by rapidly rinsing the cell monolayers with ice-cold 0.9% NaC1. Acid-soluble pools were extracted with 10% trichloroacetic acid and counted in a liquid scintillation spectrometer. The cells were dissolved in 1 N NaOH and assayed for protein directly in the wells using a modified Lowry prccadure (18) as described previously (17). All samples out of the range 20 to 30 pg of protein/cm' were discarded. Cell Wuter and Calculations-The intracellular fluid volume was estimated from the difference between total water (["Clurea space) and extracellular fluid volume ([3H]inulin space) as described previously (1). The rates of amino acid uptake, corrected for unsaturable components (19), were expressed as micromoles/ml of cell water/min. Kinetic parameters were determined by a computer (Hewlett-Packard 98455) using the Marquardt's algorithm, a general procedure for least squares estimation of nonlinear parameters (20).

RESULTS
Nu' Dependence of t-Alanine, L-Serine, and L-Cysteine Uptake by Human Fibroblasts- Fig. 1 shows that L-alanine and L-serine are primarily taken up by a Na'-dependent route in cultured human fibroblasts. This observation confirms previous results for L-alanine (1, 4). L-Cysteine also entered the cell largely by a Na'-dependent mediation, but a Na+-independent component contributed substantially to its total uptake. As shown in Fig. 2, the latter component was strongly inhibited by BCH, as expected for a System L-mediated uptake, and much less by 4-amino-1-guanylpiperidine-4-carboxylic acid, a transport-specific substrate of the Na'-independent System y+ (for cationic amino acids) in human fibroblasts (21). The inhibition of the Na'4ndependent entry of L-cysteine by BCH proved to be competitive in a Dixon plot (results not shown).
Discrimination of the Nu+-dependent Transport Systems-The study of the reactivity of L-alanine, L-serine, and Lcysteine for the Na'-dependent systems required proper conditions to PULI the activity of these sysems toward levels susceptible to unambiguous discrimination and to avoid the interference of trans-effects (trans-inhibition, trans-stimulation) by internal amino acid pools (1). These requirements were met by measuring amino acid uptake before and after a 12-h starvation of the cell monolayers in Earle's balanced salt solution (2) and by depleting nonstarved fibroblasts of their cellular amino acid pools by a 90-min preincubation before uptake assay (1). Discrimination among components mediating the inward transport of the three amino acids has been approached by measuring their uptake (under intervals short enough to approximate initial entry rates) in the absence and in the presence of excess MeAIB, a transport-specific substrate of System A in human fibroblasts (MeAIB/natural amino acid ratio, 1O:l) and by programming a digital computer to read out the total transport velocity from the summation of independent Michaelis-Menten expressions.
The analysis of the results presented in Figs. 3, 4, and 5 indicates that: (i) when the transport of L-alanine, L-serine, and L-cysteine was determined in nonstarved cells (parts a), FIG . 1 (left). Na+ dependence of L-alanine, L-serine, and L-individual amino acid. cysteine uptake by cultured human fibroblasts. Human fibro-FIG. 2 (right). Na+-independent uptake of L-cysteine by culblast cultures, grown to confluence in Medium 199 (96 h), were tured human fibroblasts.
Human fibroblast cultures, grown to depleted of intracellular amino acids (90-min incubation at 37 "C in confluence in Medium 199 (96 h), were depleted of intracellular amino Earle's balanced salt solution containing 10% dialyzed fetal calf se-acids as described in Fig. 1 and incubated for 10 rnin in a Na+-free rum). Cell monolayers were washed and the transport assay was made medium. Cell monolayers were washed in the same medium and the during 10 S at 37 " c over the indicated range of values of amino acid transport assay was made during 10 s at 37 "C Over the indicated ("-labeled L-alanine and L-serine; 35S-labeled L-cysteine) concentra-range of values of 35S-labeled L-cysteine concentration in a Na+-free tion either in a Na+-containing medium in the form of Earle's balanced medium (in which choline replaced the cation in the sodium salts of salt solution (total uptake, 0 ) or in a Na+-free medium in which the Earle's mixture) containing 10 m~ dithiothreitol, in the absence choline replaced the cation in the sodium salts of the Earle's mixture (m) and in the presence of 20 r n~ BCH (A) or 4-amino-l.guanylpi-(Na+-independent uptake, m). When L-cysteine was the tracer amino peridine-4-carboxylic acid (0) . The transport assay was made during 10 s at 37 "C over a range of 0.02-5 nm labeled amino acid either in a Na+-containing or in a Na'-free medium with both starved and nonstarved cells. In experiments with starved cells, the measurements in Na+-containing medium were performed in the absence and in the presence of a IO-fold excess of MeAIB. In a, the Na'-dependent fractions of L-alanine uptake by starved (0) and by nonstarved cells (B) were calculated by subtracting Na+-independent from total uptake (as measured in the presence of Na+). Data from experiments with nonstarved cells were fitted by linear regression analysis (r = -0.99). In b, the Na+-dependent fraction of starved cells (0) was further divided into a MeAIB-noninhibitable ( 0 ) and a MeAIB-inhibitable component (A), the latter being calculated by difference. The data relevant to each component were fitted by linear regression analysis (r = -0.99 for both MeAIB noninhibitable and MeAIB inhibitable components). From the parameters of the 2 regressions, a combined curve was calculated that was fitted to the experimental points representing the total Na+-dependent uptake of the amino acid (coeficient of determination, r2 = 0.998). In c, the experimental data of the Na'-dependent fraction of starved cells (0) were best fitted by computer (r2 = 0.999) on the basis of the sum of 2 independent rectangular hyperbolas, whose parameters defined the 2 components presented as straight lines. suggested that a single mediation was primarily involved in their uptake; (ii) starvation enhanced the uptake of each amino acid, the increase being more pronounced at high than at low external amino acid concentrations (parts a). As a consequence, the v versus v / S plots became curvilinear, indicating that more than one system of mediation was active under these conditions. The well known increase in activity of transport System A upon starvation (2) supports the involvement of this agency in causing the observed heterogeneity; (iii) when the uptakes of L-alanine, L-serine, and L-cysteine were measured in starved cells in the presence of excess MeAIB to obstruct transport by System A (parts b), the relevant to each component were fitted by linear regression analysis (r = -0.99 for MeAIB-noninhibitable component; r = -0.98 for MeAIB-inhibitable component). The combined curve was obtained by the parameters of the 2 regressions, fitted to total Na'-dependent uptake (r2 = 0.996). c, Na+-dependent fraction of starved cells (0) best fitted (r2 = 0,999) and subdivided into components (straight lines) by computer.
linearity of the plots (noninhibitable components) was restored and the contribution of System A to the total uptakes could be estimated by difference. The latter contribution was appreciable for all the amino acids studied, particularly at high external concentrations; under the latter conditions, it became predominant when L-cysteine was the tracer substrate; (iv) the largest fraction of the Na+-dependent uptake of L-alanine, L-serine, and L-cysteine, not inhibited by excess MeAIB, is attributable to System ASC. This component encompassed almost completely the Na+-dependent uptake of all these amino acids in nonstarved cells and its absolute value was comparable in starved and nonstarved fibroblasts (parts a and b); (v) when the total Na'-dependent uptake of Lalanine, L-serine, and L-cysteine was resolved into components Their kinetic properties were strikingly similar to those of the two fractions identified above in starved cells (iii and iv), where discrimination was attained by competition analysis using MeAIB as transport-specific inhibitor. Table I were calculated from the data of the kinetic experiments shown in Figs.

3, 4, and 5 by linear regression analysis of competition experiments (parts b) and of experiments with unstarved cells (parts a), and by nonlinear regression analysis using a computerized procedure based on the Marquardt's algorithm (parts c).
The values of each kinetic parameter were remarkably similar irrespective of the method adopted for their determination. The only exception was K,,, for L-cysteine uptake by System A, whose value calculated by computer analysis was almost double that derived from competition analysis. The average K , values for the transport of L-alanine and L-serine by System A were comparable (0.6-0.7 mM) and a little lower than the corresponding value for L-cysteine as obtained by inhibition analysis (0.96 m); those for the transport by System ASC were approximately 5 to 6 times lower for Lalanine and L-serine and 18 times lower for L-cysteine (when compared to the K,,, value for System A as obtained by inhibition analysis). The average Vmax values for the transport of L-alanine and L-serine by System ASC were similar, in a range of 3-4 pmol/min/ml of cell water; that for L-cysteine was 2 pmol/min/ml of cell water. The corresponding values for the transport of L-alanine and L-serine by System A in starved cells approximated those reported for System ASC and the V,,, value for the transport of L-cysteine by System A was definitely higher than that obtained for System ASC.

Uptake by Cultured Human Fibroblasts of L-Alanine, L -Serine, and L-Cysteine at Human Blood Plasma Levels-
The relative role of the transport systems involved in the uptake of L-alanine, L-serine, and L-cysteine by cultured human fibroblasts was assessed by transport measurements (under conditions allowing discrimination) at external amino acid concentrations approaching those present in human plasma (22) (Fig. 6). The Na'-independent fraction of transport has been determined in a Na+-free medium and the relative contribution of System L to it was evaluated by the addition of (A) components. Data relevant to each component were fitted by linear regression analysis (r = -0.97 for MeAIB-noninhibitable component; r = -0.96 for MeAIB-inhibitable component). The combined curve was obtained by the parameters of the 2 regressions, fitted to total Na+-dependent uptake (r2 = 0.992). c, Na'-dependent fraction of starved cells (0) best fitted (r2 = 0.999) and subdivided into components (straight lines) by computer.

Kinetic parameters for Nu'-dependent uptake of L-alanine, L-serine, and L-cysteine by cultured human fibroblasts
Conditions of cell culturing, transport assay, and calculations were as described in Figs. 3,4,  increasing concentrations of BCH to the assay medium. The Na+-dependent fraction of transport was calculated by subtracting the value obtained in the Na+-free medium from total uptake as measured in the presence of Na+. The Na+-dependent fraction has been resolved into components (Systems A and ASC) by the addition of MeAIB (at increasing concentrations) to the incubation medium of the uptake assay. The inhibition analysis by increasing concentrations of each transport-specific substrate (BCH or MeAIB) allowed the fit of experimental data to a rectangular hyperbola, whose extrapolation to infinite concentration of the inhibitor yielded the maximal inhibition of the relevant transport component. All the transport measurements were performed with nonstarved and with starved cells. L-Alanine, at an external concentration of 0.33 a, entered the cell primarily by a Na'-dependent route. System ASC contributed by about 96% to this component in nonstarved cells. Its absolute activity remained the same in starved cells, as expected for a system insensitive to adaptive regulation ( latter conditions, the total Na+-dependent uptake increased by about 25% over that observed in nonstarved cells and this increment was fully contributed to by System A. The small Na'-independent fraction (approximately ?ha of the Na'-dependent uptake) was completely inhibitable by BCH, as ex-pected for a System L-mediated uptake. The Na+-dependent uptake of L-serine, at an external concentration of 0.11 m, was shared by Systems ASC and A, with a preponderant contribution of the former system in nonstarved cells. Again, fibroblast starvation increased the total Na'-dependent uptake through the enhanced activity of System A. The fraction of L-serine uptake by Na'-independent mediations was negligible (3% of total uptake) and vanished upon addition of BCH. L-Cysteine, at the concentration found in human plasma (0.03 mM), was taken up by nonstarved cells through the Na+dependent System ASC (71%) and the Na'-independent BCH-inhibitable System L (28%). The contribution of System A, scarcely existent in nonstarved fibroblasts, became appreciable in starved cells, where it accounted for 13% of the Na+dependent fraction.

DISCUSSION
The results presented in this paper provide evidence that L-alanine, L-serine, and L-cysteine share the same transport pathways to gain entrance into cultured human fibroblasts. At low external amino acid concentrations (comparable to those present in human plasma), the major transport component corresponds to a Na+-dependent mediation endowed with a low K , value (Table I). This mediation accounts for almost the entire Na'-dependent amino acid transport in nonstarved cells. Its activity does not change upon either cell starvation or addition of excess MeAIB. Insensitivity to starvation and lack of inhibition by a typical Site A-reactive substrate indicate that this mediation can be identified with transport System ASC (cf. Refs. 7 and 13) known to operate in cultured human fibroblasts (1). A rigorous assessment of the relative contribution of System ASC to amino acid uptake using high concentrations of transport-specific substrates to obstruct other electrogenic mediations (ie. MeAIB for System A) may be hampered by changes in membrane potential caused by the presence of the added inhibitor (23). The consequent alteration in the overall rate of Na+-dependent transport, however, must be of minor importance under our experimental conditions. In fact, at least three different criteria (competition analysis, mathematical discrimination by computer, and transport measurements in nonstarved, repressed cells) yielded comparable values for the kinetic parameters defining the activity of System ASC.
A second Na'-dependent mediation for L-alanine, L-serine, and L-cysteine became significant in cultured human fibroblasts upon starvation. Its activity increased steadily with the progress of starvation up to 20 h (results not shown) and was fully inhibited by the addition of MeAIB. These features allow us to ascribe this transport component to System A, the sole Na+-dependent mediation for neutral amino acids which exhibits adaptive regulation in cultured human fibroblasts (Ref. 2 and results not shown).
As judged from the relevant kinetic parameters of transport (Table I), the apparent affinity of L-alanine, L-serine, and Lcysteine toward the putative carrier is higher for System ASC than for System A. Among these amino acids, the lowest Km value for System ASC is that of L-cysteine. On the other hand, the apparent Km value for the uptake of these amino acids by System A is by no means physiologically insignificant, being of the same order of magnitude as that observed for such natural or model amino acid substrates of System A as Lproline and MeAIB in human fibroblasts (24). On this ground, we cannot confiim for our biological model the results reported by 11) in cultured rat hepatocytes, indicating that L-cysteine could be considered a transport-specific substrate for System ASC. only in un-starved human fibroblasts, when System A is fully repressed (2), is the Na'dependent uptake of L-cysteine attributable to System ASC, but even under these conditions, a Na'-independent fraction contributes substantially to the total uptake of this amino acid. Under the same circumstances, L-serine is perhaps a better substrate than L-cysteine and L-alanine, its entry into the cells by Systems A and L being almost negligible. L-Serine has been found the best substrate of System ASC also in Chinese hamster ovary cell line CHO-K1 (25). Fig. 6 shows that, at amino acid concentrations close to those present in human plasma, L-alanine, L-serine, and Lcysteine enter the fibroblast mainly by System ASC. On the other hand, the K, values for System ASC reported in Table   I are of the same order of magnitude as the above mentioned plasma concentrations. Therefore, under conditions approaching the physiological plasma levels, the uptake of L-alanine, L-serine, and L-cysteine by the fibroblast via System ASC must be remarkably dependent on the fluctuations in substrate concentrations. These considerations emphasize the physiological importance of System ASC, possibly a constitutive agency. In addition, one must consider that L-alanine, L-serine, and L-cysteine are substrates for both System ASC and System A. The latter is typically an adaptive system. Its activity is subject to a fast modulation through trans-effects (trans-inhibition and its release) by the intracellular concentration of Site A-reactive amino acids (1) and to a long lasting modulation through adaptive regulation. This control mechanism correlates inversely the activity of System A with the extracellular concentration of Site A-reactive amino acids (2). Provided that the behavior of Systems A and ASC in fibroblasts of human tissues in viuo parallels that reported here for cultured cells, one may speculate that these two mediations assist each other in regulating the uptake of common amino acid substrates: when the external substrate concentration decreases, the rate of their entry via System ASC should slow down and the increasing activity of System A (release from trans-inhibition, derepression) would tend to counteract this slackening; conversely, when the external substrate concentration increases, the enhanced rate of their entry via System ASC could be buffered by the progressive decrease of the activity of System A (trans-inhibition, repression).