Amino Acid Transport Systems in Brush-Border Membrane Vesicles from Lepidopteran Enterocytes

Amino acid transport in mammalian cells occurs via different well-characterized transport systems (Christensen, 1984; Mircheff et al. 1982; Stevens et al. 1984). The usual strategy for discriminating among these systems is based on kinetic experiments of cross-inhibition or inhibition by amino acid analogues (Christensen, 1985). An unsatisfactory aspect of such an approach is that inhibition, much more than activation, may depend on nonspecific or indirect factors. For instance, in mammalian intestine, glucose and amino acids exert a mutual inhibition of their uptakes, owing to competition for the same source of energy for the transport, i.e. the sodium electrochemical gradient (Murer et al. 1975). Moreover, a molecule can inhibit a carrier without being transported by it. Therefore, it seemed to us that a more promising way to obtain a preliminary identification of different transport systems could be provided by the measurement of uptake activation, such as that obtained by the countertransport phenomenon, which may be influenced by alternative substrates. This approach has been used by many authors (Murer et al. 1975; Hopfer et al. 1973; Bradford & McGivan, 1982), ourselves included (Hanozet et al. 1984), only as further support for evidence obtained from inhibition experiments.

The usefulness of countertransport experiments in identifying different transport systems was tested on brush-border membrane vesicles purified from the larval midgut of Philosamia cynthia (Saturnidae, Lepidoptera), where amino acid absorption takes place via a potassium-dependent cotransport mechanism (Hanozet et al. 1980; Giordana et al. 1982). We preloaded the vesicles with different amino acids and tested their ability to elicit a counterflow accumulation of a labelled amino acid present in the extra vesicular medium. For a representative amino acid, we compared these results with those obtained from inhibition experiments.

Larvae of P. cynthia in the fifth instar were used. The larvae were fed on Ailanthus glandulosa leaves. The midgut was dissected from the larvae, and the peritrophic membrane with intestinal contents was removed. The midguts were either used immediately for preparation of brush-border membrane vesicles or were frozen in liquid nitrogen for 30min and then stored at —80°C. In the latter case, the tissue was rapidly thawed at 37°C prior to use. Brush-border membrane vesicles were prepared by means of Ca²⁺ precipitation as previously described (Giordana et al. 1982). The pellet obtained from the second centrifugation step and the final pellet were resuspended in a medium containing 100 mmol 1⁻¹ mannitol, 10 mmol 1⁻¹ Hepes-Tris, pH 7-4, with salts and other substances as reported in the legends of the tables and figures. The final membrane pellet was resuspended at a protein concentration of 10-15 mg m1⁻¹, as determined according to Bradford (1976) with a Bio-Rad kit, using bovine serum albumin as standard.

Countertransport experiments were performed in quadruplicate by the rapid filtration technique described by Hanozet et al. (1980), with the following modifications: 10 μl of brush-border membrane vesicles, resuspended in the buffer, 50 mmol 1⁻¹ K₂SO₄, and either 40 mmol 1⁻¹ elicitor amino acid or mannitol were preincubated for 10 min with valinomycin (8μgm1⁻¹ brush-border membrane protein), and then mixed with 190μ of a cocktail with the labelled amino acid (10μCi ml⁻¹), whose final composition is reported in the legends of the table and the figures. At selected times, 70 μl samples were withdrawn from the incubation mixture, diluted with 5 ml of ice-cold stop solution (150 mmol 1⁻¹ NaCl, 2 mmol 1⁻¹ Hepes-Tris, pH7-4), filtered through a cellulose nitrate filter (0·65 μm diameter pores) and rapidly rinsed twice with 5 ml of ice-cold stop solution.

The effect of different intravesicular potassium concentrations was tested in the absence of a transmembrane electrical potential difference (Δ ψ). The Δ ψ was short-circuited by equal concentrations of the highly permeant anion NO₃⁻ inside and outside the vesicles. The vesicles, preloaded with suitable concentrations of KNO₃ plus choline nitrate up to 100 mmol 1⁻¹ NO₃⁻, were diluted 1:2 in a medium with a final NO₃⁻¹ concentration of 100 mmol 1⁻ (for details see Table 2).

The inhibition experiments measuring initial histidine uptake rate into brushborder membrane vesicles were performed in the presence of increasing concentrations of inhibitor (up to 30 mmol 1⁻¹) in the incubation medium. Osmolarity was kept constant by adding mannitol. Potassium concentration inside and outside the vesicles was 20 mmol 1⁻¹ (10 mmol 1⁻¹ K₂SO₄). Brush-border membrane vesicles were preincubated for 10min with valinomycin (3 μg mg^{−1 −1} protein). The initial uptake rate was measured after an incubation time of 7 s, histidine influx being linear up to 12s (data not shown), as found also in conditions of faster initial uptake rate (Giordana et al. 1985). The incubation was started by means of an automated device, as previously described (Giordana et al. 1985), mixing 10μl of the vesicle suspension with 10 μl of a cocktail of the proper composition which contained the labelled amino acid.

Valinomycin was always added from ethanol stocks, so that the ethanol concentration in the incubation mixture did not exceed 0·5 %.

L-[U-¹⁴C]Alanine, L-[4,5-³H]leucine, L-phenyl[2,3-³H]alanine, L-[6-³H]gluta-mic acid, L-[4,5-³H]lysine monohydrochloride and L-[2,5-³H]histidine were purchased from Amersham International pic (Little Chalfont, England) and valinomycin from Boehringer (Mannheim, FRG). All other reagents were analytical grade products from Merck (Darmstadt, FRG).

Fig. 1 shows a typical countertransport experiment in which the uptake of a labelled amino acid ([³H]leucine) was measured in the presence of different substances preloaded into the vesicles. Potassium concentration was the same on both sides of the membrane, and any electrical coupling via transmembrane potential was avoided by the addition of the potassium ionophore valinomycin. The concentration of valinomycin used in these experiments should ensure a constant intravesicular K⁺ concentration, since previous experiments (not reported) have shown that the intravesicular accumulation over the equilibrium value of the amino acids driven by a potassium gradient is completely abolished by valinomycin at a concentration of 0·1μg mg⁻¹ membrane protein, which is the lowest dose able to dissipate the potassium gradient. Under these experimental conditions, intravesicular leucine and alanine, but not mannitol, elicited a transient intravesicular accumulation of labelled leucine, the overshoot occurring in both cases after 1 min of incubation. A similar pattern was observed when other amino acids were used (Hanozet et al. 1984; Giordana et al. 1985); in particular, an intravesicular accumulation of labelled histidine has also been observed in vesicles preloaded with unlabelled histidine in similar experimental conditions, with a maximal accumulation of the isotope also after 1 min of incubation (Giordana et al. 1985).

To screen the amino acids that share the transport system with histidine, the vesicles were preloaded with 14 different unlabelled amino acids and the counterflow accumulation of labelled histidine was measured. When P was less than 0-05, the ratio between the uptake value at 1 min and the value measured after 120 min was considered significantly different from 1, and the elicitor was then considered effective in causing a counterflow accumulation of histidine. The pattern shown in Table 1 indicates that at least six of the tested amino acids share with histidine the same transport agency. All these amino acids are neutral. Proline, glycine and serine, as well as the acidic and basic amino acids, failed to cause an intravesicular accumulation of labelled histidine and should, therefore, cross the membrane via other transport systems.

The validity of this experimental approach to discriminate between different amino acid transport systems was verified by testing the ability of the same 14 amino acids to inhibit histidine uptake. The inhibition experiments were carried out in conditions similar to those used for countertransport experiments, i.e. in the absence of a transmembrane electrical potential difference (Δ ψ) and with the same concentration of potassium on both sides of the membrane. In these conditions, any nonspecific inhibition due to the competition for the energy source, i.e. the potassium electrochemical gradient, should be avoided. Since a trans inhibition of histidine uptake occurs in the presence of internal potassium (Table 2), the potassium concentration on both sides of the vesicles was lowered to 20 mmol 1⁻¹, to minimize the inhibitory effect. The initial uptake rate (at 7 s) of 1mmol1⁻¹ histidine was then measured in the presence of increasing concentrations of the test amino acid in the extravesicular medium. Fig. 2A shows the results of a typical inhibition experiment where the effectiveness of serine, alanine and proline (from 1 to 30 mmol 1⁻¹) in inhibiting histidine uptake was tested. The data were then plotted as the reciprocal of the inhibition vs the reciprocal of the inhibitor concentration (Fig. 2B,C,D), to calculate the inhibition value at infinite inhibitor concentration I_max and the apparent inhibition constant K_iapp, i.e. the inhibitor concentration that, at the histidine concentration used, gives half-maximal inhibition (Webb, 1963). The same procedure was followed for all the 14 amino acids tested. Mannitol, up to 30 mmol 1⁻¹, was used to keep the osmolarity of the medium outside the vesicles constant.

Table 3 reports the calculated I_max and K_iapp values. Although all the tested amino acids inhibit histidine uptake to a certain extent, it is apparent that all the amino acids causing a counterflow accumulation of histidine have high I_max and relatively low K_iapp values. Assuming that the inhibition exerted by these amino acids is competitive, their inhibition constants K_i have also been calculated (Table 3) since, for a competitive inhibition, K_iapp is related to K_i in the manner reported in the legend of the table (Webb, 1963). The four amino acids having the highest K_iapp values (from 3·5 to 1·6) show accumulation ratios of less than 1, which means that those amino acids that do not elicit countertransport of histidine inhibit the uptake of this amino acid only at high concentrations.

Only arginine has an accumulation ratio less than 1 and a low K_iapp value: a possible explanation could be that this amino acid is a noncompetitive inhibitor of the histidine transporter. If this were the case, the inhibitor should not be transported and, therefore, would not elicit a counterflow accumulation, the low K_iapp value simply indicating the high affinity for a binding site different from the binding site of histidine. To test this hypothesis, the kinetic parameters of histidine uptake were measured in control conditions and in the presence of arginine. For comparison, the effects on the kinetic parameters of histidine uptake of phenylalanine, which shows an accumulation ratio greater than 1, and of lysine, which has a high K_iapp value and does not cause a counterflow accumulation, were also tested. Figs 3-5 show histidine uptake vs concentration in the absence and in the presence of these three amino acids. The kinetics of histidine uptake was always consistent with the presence of both a saturable component and a nonsaturable one. The kinetics of the nonsaturable components, calculated from the slopes of the linear part of each curve, differed slightly in the different experiments. From the Eadie-Hofstee plots of the data, corrected for the diffusional component, it can be seen that both arginine and lysine exerted a noncompetitive inhibition, whereas a mixed inhibition was exerted by phenylalanine. In fact, arginine and lysine do not affect the affinity of the transport system for histidine, but both influence J_max, the maximal flux rate, with a 36% decrease in the presence of arginine (2 mmol 1⁻¹) and a 43% decrease with lysine (1 mmol 1⁻¹). Phenylalanine (1 mmol 1⁻¹) caused a 2·2-fold increase of K_m and a 32 % decrease of J_max.

The K_m values for histidine obtained in control conditions in Figs 3–5 were, respectively, 0·60 ± 0·13, 0·38 ± 0·03 and 0·55 ± 0·08 mmol 1⁻¹, with a mean value of 0·51 ± 0·07 mmol 1⁻¹, very different from the value of 2·4 mmol 1⁻¹ previously reported in apparently similar conditions, i.e. in the absence of a Δ ψ and a K⁺ gradient but in the presence of K⁺ on both sides of the membrane (Giordana et al. 1985). The main difference between the two sets of experiments could be the K⁺ counterion used: thiocyanate in the former experiment, sulphate in this case. It should be stressed that the vesicles are preloaded with the potassium salt after the second centrifugation step and therefore the K⁺ salt remains in contact with the membranes for rather a long time. We have now realized that long preincubations with thiocyanate strongly affect the transport characteristics of the membranes. Besides, KSCN, though at concentrations 10 times higher than those used by us, does interact with phosphatidylcholine bilayers (Cunningam & Lis, 1986). The effect of thiocyanate could also explain why we previously failed to find an inhibition of histidine uptake by lysine (Giordana et al. 1985).

The inhibition experiments taken together confirm the suitability of the countertransport experiments in identifying transport systems. Therefore, we measured the uptake values after 1 and 120 min of incubation of five labelled amino acids (alanine, leucine, phenylalanine, lysine and glutamic acid) into brush border membrane vesicles of P. cynthia larval midgut preloaded with the same 14 amino acids used in the experiments with histidine (Table 4).

The delineation of the different transport systems involved in the transfer of amino acids into cells has been intensively studied in intact epithelia (Schultz & Curran, 1970), intact cells (Christensen, 1985; Guidotti et al. 1978) and, more recently, in plasma membrane vesicles (Mircheff et al. 1980, 1982; Stevens et al. 1982). Membrane vesicles are especially suitable for studies of trans effects and, in contrast with the intact tissue or cells, countertransport experiments can be easily performed.

Since a counterflow accumulation of a substrate in the presence of an imposed trans gradient of a different substrate is obtained only if they share the same translocator, it can be inferred from Fig. 1 that at the apical border of lepidopteran enterocytes there is a transport system shared by leucine and alanine, since both amino acids elicited an intravesicular accumulation of labelled leucine with a maximum at 1 min of incubation. The same figure shows also that a trans stimulation of leucine uptake occurred at early incubation times (10 s). However, the trans effect on the initial rate by the same or by an alternative substrate could be, in principle, either positive or negative, depending on the kinetic properties of the carrier and the experimental conditions. In contrast, counterflow accumulation over the equilibrium value, which occurs at later incubation times (with a maximum at 1 min in our conditions), is a result of flux coupling, and it takes place only if the internal substrate is transported by the same carrier (Kessler & Semenza, 1983). Therefore, counterflow accumulation can be used in principle to discriminate the number and specificity of amino acid transporters present on the plasma membrane. However, this experimental approach has never been systematically adopted, with most researchers using the cis inhibition of the uptake of a labelled amino acid to identify amino acid carriers. For this reason, we have verified the agreement between counterflow accumulation experiments and inhibition experiments using histidine as the test amino acid. The rationale was that those amino acids that caused an intravesicular accumulation of labelled histidine should also inhibit histidine uptake from the cis side of the membrane, and the inhibition should be, at least in part, competitive. Conversely, the inhibition, if present, of histidine uptake caused by amino acids that are not elicitors should be noncompetitive, and therefore misleading in the identification of a transport system.

Although anionic and cationic amino acids failed to elicit a counterflow accumulation of. histidine, six of the nine neutral amino acids, besides unlabelled histidine, were effective (Table 1). The same amino acids were able to inhibit histidine uptake when present in the external solution (Table 3). Assuming the inhibition to be competitive, the K_i values were calculated (Table 3). It can be observed that the K_i found for histidine is equal to the K_m value found experimentally (see above), and that this is the highest value measured: this should indicate that histidine is not the primary substrate for the carrier, primary substrates possibly being glutamine and phenylalanine, which show the highest affinity.

Since arginine, which is not countertransported, also shows a very low K_iapp, we have tested the type of inhibition of histidine uptake exerted by this amino acid and, for comparison, by phenylalanine, one of the putative primary substrates, and by lysine, an amino acid which does not elicit an accumulation of histidine. As expected, a noncompetitive inhibition was obtained with lysine and arginine, whereas phenylalanine exerted a mixed inhibition. These results indicate that inhibition experiments alone are not sufficient for assessing the number and specificities of amino acid carriers. In addition, they clearly show that nonspecific factors can affect the inhibition pattern observed. In our case, although special care was taken to avoid any electrical and osmotic effects, a nonspecific inhibition was always present, regardless of the nature of the amino acid used as inhibitor (see, for instance, the inhibition values in Table 3 for aspartic acid and glycine). This could explain how a purely competitive inhibition of histidine uptake is not shown even by those amino acids, such as phenylalanine, that are effective in countertransport experiments.

Countertransport can therefore be considered a suitable test for the interactions of amino acids with different transport systems. Of course, only an initial screening can be obtained, since a complete identification and characterization of a transport system requires a detailed experimental strategy (Christensen, 1966, 1985).

The data reported in Tables 1 and 4 suggest that a specific transport system for lysine is present in the luminal membrane of lepidopteran larval enterocytes: this transport agency is different from the y⁺ found in nonepithelial cells (Kilberg, 1982; White, 1985), intestinal epithelial brush-border membranes (Stevens et al. 1984) and lobster hepatopancreatic brush-border membranes (Ahearn & Clay, 1987a), in that it is strongly cation-dependent (Giordana et al. 1985), it does not interact with neutral amino acids and it does not serve arginine.

A specific transport system for glutamic acid is also present, and it seems to be unable to transfer the other anionic amino acid tested, unlike the transport agency operating in mammalian (Schneider & Sacktor, 1980; Corcelli et al. 1982) and crustacean (Ahearn & Clay, 1987b) epithelia.

The lack of a counterflow accumulation of the labelled test amino acids when either glycine or proline are preloaded into the vesicles provides an interesting indication of the possible occurrence in the apical pole of lepidopteran enterocytes of two specific transport systems, possibly similar to system Gly of non-epithelial cells (Winter & Christensen, 1965) and to the IMINO carrier of mammalian epithelial brush-border membranes (Stevens et al. 1984).

Moreover, the transport of neutral amino acids other than glycine and proline seems to occur, in lepidopteran midgut, through a nonspecific transport system - or, if the exclusion of some amino acids is confirmed (see cysteine, serine and histidine), through more transport systems with a broad range of overlapping substrate specificities - apparently similar to the neutral brush-border system (NBB) found in mammalian enterocytes (Stevens et al. 1984). This system can also recognize and handle histidine.

This work was supported by a grant from Ministero della Pubblica Istruzione, Rome, Italy.