Anomalous Glutamate/Alkali Cation Symport In Larval Manduca Sexta Midgut

Glutamate uptake in lepidopteran midgut (Wolfersberger et al. 1987; Giordana et al. 1989; Hanozet et al. 1989; Wolfersberger, 1993) presents a paradox. This anionic amino acid links the Krebs and urea cycles and hence is produced continuously in all cells. In the midgut of Manduca sexta, a 5000-fold electrical gradient favors the exit of glutamate from the cell to the lumen. For the amino acid to be driven into cells against this powerful force, a symport cycle would have to be electrophoretic and transfer net positive charge. Nevertheless, experimental data suggest that glutamate transport is electroneutral in lepidopteran (Giordana et al. 1989) and mammalian (Schneider et al. 1980; Corcelli et al. 1982) tissues. Glutamate symport in lepidopteran midgut, therefore, must either be extremely unusual or, in the absence of selective pressure for the evolution of a glutamate-specific symporter (‘glutamate symporter’), not exist at any physiologically relevant level.

M. sexta, an intensively studied lepidopteran, provides a tractable experimental model with which to resolve this paradox. The 1000-fold gain in body mass between egg and fifth-instar larva requires a substantial flux of amino acids to be used as an energy source (Parenti et al. 1985) and for growth. Not surprisingly, studies on larval Philosamia cynthia midgut brush-border membrane vesicles (BBMV) have identified six putative symporter families, including one each for acidic, neutral and basic amino acid groups (Giordana et al. 1989).

Earlier studies on amino acid symport in lepidopteran larvae have been conducted at near-neutral pH values (e.g. Giordana et al. 1982; Sacchi et al. 1984, 1990). Recently, however, Parenti et al. (1992) characterized leucine uptake at pH 8.8 in larval Philosamia cynthia midgut BBMV. Hennigan et al. (1993a,b) identified a system-B-type symporter in larval M. sexta midgut BBMV that symports zwitterionic amino acids optimally at pH 10, which is close to the average pH (10.5) found in the midgut contents (Dow, 1984). K⁺ is the preferred cation for the symport of both neutral and basic amino acids, especially at pH 10 (Hennigan et al. 1993b). This finding is not altogether unexpected, since K⁺ is the predominant cation in the lumen (Dow et al. 1984) and is required for transepithelial transport of amino acids (Nedergaard, 1972).

Amino acid uptake by lepidopteran midgut differs from uptake in more familiar crustacean and vertebrate tissues in four ways. (1) Uptake is energized by an H⁺-translocating V-ATPase coupled tightly to a K⁺/2H⁺ antiporter; the resulting potassium electrochemical potential gradient is transformed into a large potential difference (PD) that drives uptake across the apical membrane. (2) K⁺ and not Na⁺ is the predominant cation. (3) CO₃2⁻ is the dominant anion, the concentration of Cl⁻ being small, so that the lumen is quite alkaline. The extracellular side is close to neutral pH. In contrast, crustacean and vertebrate membranes are energized by ion gradients secondary to K⁺/Na⁺-ATPase activity and Na⁺ is the major extracellular cation and Cl⁻ the major anion. In short, amino acid/K⁺ uptake in lepidopteran midgut is driven electrophoretically by an electrical PD, whereas amino acid/Na⁺/Cl⁻ uptake in crustacean and vertebrate tissues is driven by an ionic gradient. Inasmuch as glutamate was found to be an extremely poor substrate for the neutral/basic amino acid symporter (Hennigan et al. 1993a) or any other symporter characterized to date in M. sexta, an evaluation of glutamate uptake under alkaline conditions seemed important.

Second-and third-day, fifth-instar M. sexta larvae, weighing 5.2±0.6 g, were reared from eggs under constant light at 27°C on an artifical diet (Carolina Biological Supply Company, Burlington, NC).

BBMV were isolated from midguts following the methods of Biber et al. (1981), as modified by Wolfersberger et al. (1987). Estimates of BBM purity via marker enzyme assays were reported by Eisen et al. (1989). The vesicles were loaded with resuspension buffer by equilibration for 1 h on ice, followed by centrifugation for 30 min at 31 000 g. The method of Bradford (1976), using a commercial dye reagent (Bio-Rad Laboratories, Richmond, CA) and bovine serum albumin as standard, was used to determine protein concentrations.

Uptake was measured using published protocols (Hanozet et al. 1980), as modified by Wolfersberger et al. (1987) and Hennigan et al. (1993a,b). The vesicles were loaded with resuspension buffer [100 mmol l⁻¹ mannitol, 50 mmol l⁻¹ aminomethylpropanediol (AMPD) acidified with HCl to pH 10] for experiments with a salt gradient. For experiments without a salt gradient, the resuspension buffer also contained 100 mmol l⁻¹ KSCN and the vesicles were treated with 8 μg valinomycin per mg protein to preclude the development of a cation gradient and the associated diffusion potential. 10 μl samples of the vesicle suspension were incubated with an equal volume of transport buffer (compositions indicated in the figure legends) for the desired interval and transport was quenched by the rapid addition of 4 ml of stop buffer. The diluted reaction mixture was filtered through a nitrocellulose filter supported on a sintered glass funnel. In countertransport experiments, the amino acid used as elicitor was added at a concentration of 20 mmol l⁻¹ to the resuspension buffer. The transport buffer contained 50 mmol l⁻¹ AMPD chloride at pH 10, 100 mmol l⁻¹ mannitol and 100 mmol l⁻¹ KSCN, together with 1 mmol l⁻¹ labeled glutamate. The stop buffer had the same composition except that the amino acid was eliminated. In these experiments, 5 μl of the vesicle suspension were diluted into 95 μl of transport buffer and the incubation was allowed to proceed for the desired interval.

Experiments were conducted in triplicate on three separate preparations of vesicles and data were averaged over the three preparations; means ± S.E.M. are reported. Statistical analyses were performed using Student’s t-test. A probability of 97.5% was considered to represent a statistically significant difference.

Labeled glutamate (L-[U-³H]glutamic acid) was purchased from Amersham (Arlington Heights, IL) and unlabeled amino acids were obtained from Sigma (St Louis, MO). The glutamate esters were obtained from Fluka (Ronkonkoma, NY). Other chemicals were obtained from Fisher (Pittsburgh, PA).

Preliminary experiments with a K⁺ gradient showed that maximum L-glutamate uptake always occurred between 1.5 and 5 min of incubation and that equilibrium distribution of glutamate between the intravesicular and the extravesicular volumes was always reached within 60 min. Therefore, we chose sampling intervals (t) of 30 s, 60 s, 90 s, 2 min, 5 min, 10 min and 60 min in all subsequent experiments. The amount of non-specifically bound glutamate, approximated by the (extrapolated) t=0 uptake, was 0.3± 0.1 nmol mg⁻¹ protein (see Fig. 6). Non-specific substrate binding contributes less than 5% of the peak accumulation and 12% of the equilibrium values, and the small correction was ignored. The time-dependence of glutamate uptake was insensitive to osmotic boundary conditions, as shown previously for leucine uptake (Hennigan et al. 1993b). Thus, a similar time course was obtained irrespective of whether iso-osmoticity was present at t=0 or t=∞ (Fig. 1). All of the work reported here was performed with iso-osmotic conditions present at t=∞. A K⁺ gradient promoted the development of a transient maximum in the intravesicular concentration of labeled glutamate (Figs 1 and 2). The maximum accumulation, as well as the initial glutamate uptake rate (Fig. 2, inset), was somewhat higher with SCN⁻ than with Cl⁻ as the counterion (Fig. 2A). The addition of valinomycin, a K⁺-specific ionophore that dissipates the cation gradient, abolished the maximum (Fig. 2B). A choline gradient did not promote glutamate accumulations over those at equilibrium (Fig. 2B).

The effect of the alkali ions on glutamate uptake was examined at pH values of 8.9, 9.5 and 10.5, chosen to bracket the pK‴ of glutamate (9.67). The highest pH investigated approaches the in vivo luminal pH. At pH 8.9 (Fig. 3A) the uptake of glutamate was stimulated by initial gradients of K⁺, Na⁺ and Li⁺. The maximum accumulations as a function of cotransported cation followed the sequence K⁺≈Na⁺⪢Li⁺, with the difference between K⁺/Na⁺ and Li⁺ being statistically significant at a level of 97.5% (Fig. 4).

At pH 9.7 (Fig. 3B) there is little change from the situation at pH 8.9, except for a marked decrease in the accumulation in the presence of Li⁺ that is not significantly greater than 1. The difference between the effects of K⁺ and Na⁺ is again not statistically significant at the 97.5% level (Fig. 4).

At pH 10.5 (Fig. 3C) the relative accumulations of glutamate in the presence of both K⁺ and Na⁺ were lower than at pH 9.7 or 8.9, but the maximum accumulation in the presence of K⁺ was now significantly greater than that in the presence of Na⁺ (97.5% level). Li⁺ does not stimulate glutamate uptake at pH 10.5 (Fig. 4). Neither Rb⁺ nor Cs⁺ stimulated glutamate uptake at any of the pH values tested. The equilibrium (60 min) glutamate concentration in all incubation mixtures shown in Fig. 3 was 2.60± 0.21 nmol mg⁻¹ protein.

The uptake of glutamate, leucine and lysine was investigated at three different membrane potentials. A proton diffusion potential was generated using a transmembrane pH imbalance in the presence of a protonophore, carbonylcyanide p-trifluoromethoxyphenylhydrazone (FCCP). When the permeability of a given ion species (H⁺ in this case) far exceeds that of the other ions, the Goldman–Hodgkin–Katz diffusion equation reduces to the Nernst–Einstein equilibrium equation. A pH difference of 1 unit then corresponds to a limiting potential difference (PD) of 59 mV at 25°C. A K⁺ gradient in the presence of valinomycin could not be used to generate a PD since K⁺, the physiologically relevant anion, was the object of study as the symported cation.

The initial rates for glutamate uptake were found to be constant within experimental error at intravesicular PD values of +59, 0 and −59 mV (relative to the extravesicular medium) (Fig. 5). By contrast, the initial rates of leucine and lysine uptake, measured as positive controls, were dependent upon intravesicular PD (Fig. 5).

Initial rates of glutamate uptake with several inhibiting amino acids were tested in the presence and absence of a K⁺ gradient. Uptake in the absence of inhibitor was used as a control in each case. Several likely inhibitors were used to demonstrate that the initial time courses were linear and that they had a common intercept, with a variation of less than 10% (Fig. 6). The relative inhibitions, 1-(rate with inhibitor/rate without inhibitor), were then calculated, with the uptake at a single point in time substituted for the rate. The inhibition of uptake of radiolabeled glutamate in the presence and in the absence of a KSCN gradient by the 20 commonly occurring amino acids, D-glutamate, L-glutamic acid 5-methyl ester (5-MG) and L-glutamic acid dimethyl ester (DMG) is shown in Table 1. The amino acids are listed in their rank order of ability to inhibit K⁺-gradient-driven uptake of labeled glutamate. Among the 23 amino acids tested, tryptophan, aspartate, glutamate and phenylalanine inhibited glutamate uptake most strongly.

To eliminate the possibility that symport of the ‘inhibiting’ amino acid diminishes the K⁺ gradient, cis-inhibition experiments in the absence of a salt gradient were carried out (Table 1). The uptakes were reduced to as much as one-third of their ‘with-gradient’ values and their range was reduced sharply. In the absence of a K⁺ gradient, phenylalanine inhibited the most (relative inhibition of 50%) whereas leucine, tryptophan, serine, isoleucine, proline, lysine and 5-MG inhibited the least. There was no apparent difference among the remaining amino acids, with initial rates of glutamate uptake ranging from 0.37±0.03 to 0.48±0.08 nmol mg⁻¹ protein 6 s⁻¹.

The relative accumulation of labeled glutamate at time t′, was calculated as uptake_t=t′/uptake_{t=equilibrium}. A study of the relative accumulation of labeled glutamate elicited by intravesicular glutamate (Fig. 7) revealed that maximum accumulation occurred around 90 s. Increasing the intravesicular glutamate concentration by a factor of 2 had no effect on the time course of uptake. t′ was accordingly chosen to be 90 s in the other countertransport experiments. Amino acids that were found to be the strongest and weakest inhibitors were used as elicitors in order to outline glutamate countertransport patterns (Table 2).

In order to estimate the binding affinity of glutamate, initial uptake rates as a function of [glutamate]_out between 0.001 and 10.0 mmol l⁻¹ were measured. These rates were very low and independent of [glutamate]_out to within experimental error.

The initial rate of glutamate uptake was accelerated by a K⁺ gradient (Fig. 2), and glutamate concentration in larval Manduca sexta midgut BBMV reached a maximum of 2–3 times the equilibrium value in the presence of an initial alkali ion gradient (Fig. 2). No accumulation was observed, either in the absence of an alkali cation or in the absence of an alkali ion electrochemical gradient (Fig. 2). We infer from these results that glutamate is symported since its flux is (a) coupled to alkali ion entry and (b) driven by the electrochemical gradient of the cation.

In contrast to its effect on the Na⁺-dependent mechanism in crustacean hepatopancreatic BBMV (Balon and Ahearn, 1991), a Cl⁻ gradient does not seem to increase the glutamate uptake rate (Fig. 2). This may not be surprising, considering that the concentration of Cl⁻ in the anterior midgut fluid is about 25-fold lower than that of K⁺ (Dow et al. 1984). In any event, the maximum accumulation reached with a given alkali cation is pH-dependent. Thus, glutamate uptake is stimulated by both K⁺ and Na⁺ gradients, with increasing specificity for K⁺ at higher pH values. A Li⁺ gradient was able to stimulate glutamate uptake at pH 8.9 but not at pH 9.7 nor at pH 10.5, whereas Rb⁺ and Cs⁺ gradients were ineffective at any of the pH values examined (Figs 3, 4). These results indicate that, at physiological pH values, glutamate symport occurs preferentially with K⁺, which is present in vivo at far higher concentrations than is Na⁺ (179±23 mmol kg⁻¹ wet mass versus <1 mmol kg⁻¹ wet mass; Dow et al. 1984).

Wolfersberger (1993) found no evidence for above-equilibrium accumulations of glutamate in larval Lymantria dispar BBMV with inwardly directed alkali ion gradients. BBMV from the larval midgut of Pieris brassicae (Wolfersberger et al. 1987) were similar to mammalian systems in that modest accumulations of glutamate were found in the presence of Na⁺ but not K⁺. Working at pH 7.4 with larval P. cynthia midgut BBMV, Giordana et al. (1989) demonstrated that glutamate uptake was higher in the presence of Na⁺ than in the presence of either Li⁺, Rb⁺ or Cs⁺ as the symported cation. Accumulations twice that at equilibrium were observed. In these respects, M. sexta seems to resemble P. cynthia rather than the other lepidopterans.

Fig. 5 reveals that the application of a transmembrane PD between −59 and +59 mV accelerates amino acid uptake in the order lysine>leucine ⪢glutamate. Fluorescence spectroscopic experiments with a potential-sensitive dye had earlier shown that glutamate uptake rates were PD-independent between −50 and −110 mV (Parthasarathy and Harvey, 1994). In contrast, neutral amino acid uptake rates increased as the vesicle was depolarized. Recollecting that amino acid uptake across the BBM is driven by a PD in vivo, a feature that can be reproduced in vitro (Fig. 5), the insensitivity of glutamate uptake rates to PD strongly suggests that symport may not be an important means of glutamate transport across the BBM in the living organism. Glutamate symport has been found to be electroneutral in other tissues, such as P. cynthia midgut (Giordana et al. 1989) and mammalian intestine (Corcelli et al. 1982; Schneider et al. 1980). A number of factors, including an electroneutral limiting step or changes in substrate stoichiometry (compensating for changes in amino acid or carrier charge), could be responsible for the invariant glutamate uptake rate.

Amino acid symport systems can often be distinguished from one another by their inhibition and countertransport patterns (Christensen, 1984). Inhibition of amino acid uptake can arise from (a) competitive inhibition, (b) non-competitive inhibition (symport of the inhibitor through a different symporter dissipates the cation electrochemical gradient) or (c) uncompetitive inhibition. Cases a and b can be resolved by inhibition experiments in the absence of a salt gradient. Cases a and c can be resolved by countertransport experiments.

Each of the 20 naturally occurring amino acids depressed glutamate uptakes below that of the control in the presence of a K⁺ gradient (Table 1). Tryptophan, aspartate and phenylalanine inhibited glutamate uptake by 72%, 67% and 62%, respectively. These results could reflect competition either for a glutamate-binding site or for the K⁺ gradient. At the other end of the spectrum, lysine and 5-MG inhibited uptake only by 12% and 6%, respectively. Other amino acids, including D-glutamate and DMG, inhibited uptake by less than 50%. These results suggest that lysine and 5-MG do not compete with glutamate for the K⁺ gradient. There is little correspondence between the salt gradient data reported in Table 1 and those reported in rabbit jejunal BBMV at pH 6.0 by Maenz et al. (1992).

The degree of inhibition of glutamate uptake was lower without, than with, a K⁺ gradient. Inasmuch as the amino acids examined, including glutamate and aspartate, were found to inhibit glutamate uptake to similar extents, it appears that the putative glutamate-specific symporter does not have a well-defined set of inhibitors.

Countertransport experiments were attempted with several amino acids, chosen to represent either strong or weak inhibitors of glutamate uptake. The relative accumulations ranged from 0.77±0.09 for 5-MG to 0.97±0.18 for aspartate. At no time did any of the amino acids examined elicit intravesicular labeled glutamate uptake greater than equilibrium values. Glutamate itself elicited no accumulation of its labeled congener (Fig. 7). Hanozet et al. (1989) found that unlabeled glutamate elicited a relative accumulation of about 1.5 that of labeled glutamate in P. cynthia at neutral pH. We conclude that none of these amino acids (including glutamate) is symported by a putative glutamate-specific symporter.

In M. sexta, glutamate uptake resembles neutral amino acid symport in that it is K⁺-specific above pH 9.7, but differs in that (a) glutamate uptake rates are pH-independent between pH values of 8 and 10, (b) the process is electroneutral, (c) unlabelled glutamate appears to be unable to elicit counterflow accumulations of its labeled counterpart and (e) a well-defined set of inhibitors is absent. The absence of a symporter specific for glutamate would serve to explain the low uptake rate of this amino acid as well as the non-specific inhibition, the high K_m values and the failure of intravesicular glutamate to elicit labeled glutamate uptake. The accumulations observed in the presence of an alkali ion gradient may reflect glutamate symport via symporters of other amino acids. The binding constant of glutamate in P. cynthia was 3–5 times higher than that of the neutral amino acids and the V_max was between 15 and 50% of that of the neutral amino acids (Giordana et al. 1989). However, at pH 7.4, the K_m and V_max of aspartate symport in M. sexta (Reuveni and Dunn, 1993) were similar to those of leucine. The differences between aspartate and glutamate symport at the same pH remain to be investigated. In any event, glutamate symport in lepidopterans, including M. sexta, seems to be marginal.

Glutamate is present at a very low concentration in the haemolymph but is relatively abundant in the midgut lumen and within the midgut epithelial cells of larval P. cynthia (Giordana et al. 1989) and B. mori (Parenti et al. 1985). Considering the pivotal position that glutamate occupies in intermediary metabolism, it is likely that glutamate is mainly synthesized by endogenous mechanisms so that a pathway optimized for glutamate uptake from the diet may not have evolved. Glutamate synthesis, rather than symport, may account for the high concentrations of glutamate in the epithelial cells.

This work was supported in part by research grant number AI 30464 from the US Public Health Service.