Determinants for the Activity of the Neutral Amino Acid/K⁺ Symport in Lepidopteran Larval Midgut

The identification of the primary active ion movements in an epithelium is very important, since a large number of secondary active mechanisms are expected to depend upon it. The physiology of the midgut epithelium of lepidopteran larvae is dominated by the activity of a proton pump, the H⁺-translocating V-ATPase (Wieczorek et al. 1989) located in the luminal membrane of specialized cells, the goblet cells. The V-ATPase is associated in the goblet membrane with a K⁺/nH⁺ antiporter (Wieczorek et al. 1991; Wieczorek, 1992), thus being responsible for the ultimate active extrusion into the midgut lumen of potassium ions (Harvey and Nedergaard, 1964). The functional identification in the same epithelium of a K⁺ symport, responsible for the uptake of amino acids across the brush-border membrane of midgut columnar cells, dates back to 1980 (Hanozet et al. 1980). Its occurrence was integrated into a model for the K⁺-dependent transepithelial absorption of amino acids across the gut (Giordana et al. 1982). The model emphasized that it is primarily the voltage built up across the luminal membrane of columnar cells by the activity of the electrogenic proton pump, together with the ability of the symporter to bind K⁺, that provides the driving force for secondary amino acid absorption. Fig. 1 gives a schematic presentation of K⁺ activities, pH values and the electrical potential differences (Δ Ψ) in the midgut of B. mori larvae. The estimated K⁺ electrochemical potential across the luminal membrane of columnar cells in vivo, which drives amino acid uptake, is -15.9 kJ mol⁻¹.

In the following years, with larvae of Philosamia cynthia as the experimental animal, the presence of at least five different K⁺-dependent transport systems was established (Hanozet et al. 1984; Giordana et al. 1985, 1989). A detailed analysis of the kinetics of the K⁺/symporter for neutral amino acids (Parenti et al. 1992) revealed the presence of mixed-type kinetics, instead of an affinity-type mechanism that appears to be the model shared by the Na⁺-dependent class of amino acid symporters. This feature might be related to the ability of the K⁺ binding site to accept alkali metal cations other than K⁺ (Hanozet et al. 1980; Giordana et al. 1989; Hennigan et al. 1993b; Sacchi et al. 1994). Moreover, both the fully loaded complex, symport/leucine/cation, and the partially loaded complex, symport/leucine, are able to perform amino acid translocation across the brush-border membrane (Giordana et al. 1989; Sacchi et al. 1990; Parenti et al. 1992).

The K⁺/symport for neutral amino acids shares with the B° Na⁺-dependent transport system of mammalian small intestine and kidney brush-border membranes (Stevens et al. 1984; Lynch and McGivan, 1987; Doyle and McGivan, 1992) a broad specificity towards small, branched-chain and aromatic amino acids and it is unable to accept methylaminoisobutyric acid and cationic amino acids. The K⁺-dependent broad specificity system appears to be common to most lepidopteran larvae ( Hennigan et al. 1993a; Giordana et al. 1994).

The midgut of lepidopteran larvae can be divided into three different regions, identifiable by their morphological and functional differentiations (Cioffi and Harvey, 1981; Chamberlin, 1990; Azuma et al. 1991; Dow, 1992). The time courses of leucine uptake into brush-border membrane vesicles (BBMV) prepared by the Ca²⁺-precipitation method (Giordana et al. 1982) from the anterior–middle (AM) and the posterior (P) regions of B. mori midgut were measured in simulated physiological conditions: i.e. an intravesicular pH of 7.2 similar to the cytoplasmic value (Chao et al. 1991; Dow, 1992) and an extravesicular pH of 8.7. A transmembrane electrical potential of approximately – 90 mV was obtained by adding the protonophore FCCP. Fig. 2 shows that a K⁺ gradient was able to drive the intravesicular accumulation of leucine in both preparations, but that in P-BBMV much higher uptake rates and accumulation were observed. Leucine uptakes into BBMV from the two separate anterior and middle tracts were almost identical (Giordana et al. 1994). The kinetics of leucine uptake into AM- and P-BBMV as a function of external amino acid or K⁺ concentrations were then determined. The kinetic constants reported in Table 1 indicate (a) that whereas the affinity of the symporter for leucine is similar in the two midgut regions, V_max is markedly higher in P-BBMV, confirming that this midgut region is specifically specialized for amino acid absorption; (b) that the cation-dependent component, when both substrates are saturating, is much higher in P-BBMV than in AM-BBMV (apparently most of the amino acid taken up in the latter region is translocated under the binary form carrier/leucine); and (c) that leucine kinetics as a function of external K⁺ concentration is hyperbolic in P-BBMV, but sigmoidal in AM-BBMV with a very high Hill coefficient. The Iso-Random Bi Bi system proposed for P. cynthia (Parenti et al. 1992) can account for nonhyperbolic velocity curves, assuming a steady-state model in which one route to the formation of the ternary complex is significantly favoured over the other. Alternatively, at saturating leucine concentrations, the translocation as a binary complex, kinetically less favoured, prevails when the K⁺ concentration is low, whereas the more rapid translocation under the ternary form takes over as K⁺ concentration increases further.

The apical membrane of the columnar cells of the larval midgut in vivo separates a very alkaline environment (up to a pH of 11 or more) from a nearly neutral cytoplasmic side (reviewed by Dow, 1992). When the extravesicular pH was lowered to neutrality, a drastic reduction of leucine uptake was observed only in P-BBMV. Minor effects were recorded in AM-BBMV, whereas in P-BBMV of the BC 20 B. mori strain the K_m value increased from 0.09±0.01 to 0.87±0.21 mmol l⁻¹ (mean ± S.E.M. of three different determinations), with no modification of V_max. The dependence of the maximal rate of uptake of different neutral amino acids upon an extravesicular alkaline pH has been tested in AM-BBMV and P-BBMV from six different strains of B. mori larvae; in all cases, the effect was far more marked in P-BBMV, and it was always exerted on the affinity of the symporter for the amino acid. However, not all strains were equally sensitive, the decrease in affinity varying from five-to tenfold. Since the posterior midgut is the region where most of the amino acids are absorbed, a failure by the anterior–middle region of the midgut to alkalize the luminal contents (Dow, 1992) would severely hamper amino acid availability for the larva.

It had been clearly established that amino acid/K⁺ symports are very sensitive to the presence of an electrical potential difference across the luminal membrane in the isolated gut in vitro and in vivo (Nedergaard, 1973; Giordana et al. 1982) and of a transmembrane voltage in BBMV (Hanozet et al. 1980; Giordana et al. 1985; Sacchi et al. 1990; Hennigan et al. 1993b). Table 2 shows that, in B. mori, it is in P-BBMV that the Δ Ψ-dependence is more marked, and that it is the translocation step that is affected, because V_max decreases threefold when the transmembrane potential is lowered. These results suggest that, in the absence of V-ATPase activity, the cation-dependent amino acid translocation involves the movement of a net positive charge(s), and that leucine uptake should therefore induce a depolarization of the membrane potential in BBMV. We tested this hypothesis with the fluorescent potential-sensitive dye 3,3 ′ -diethylthiacarbocyanine iodide [DiS-C₂(5)] (Stieger et al. 1983; reviewed by Wright, 1984), by measuring the depolarization of an imposed inside-negative potential across the vesicle membrane, according to Cassano et al. (1988). The quenching of fluorescence was calibrated by generating membrane potentials of different magnitudes (Fig. 3A) and plotting the differences in fluorescence quenching versus the logarithmic ratio of the external and internal [K⁺] (Fig. 3A, inset). The mean value of the Δ F (%) per mV, where Δ F is fluorescence quenching, was 0.41±0.01 (±S.E.M., five experiments). The ability of the amino acid symport to accept Na⁺ instead of K⁺ (Sacchi et al. 1994) was employed to evaluate the dissipation of an imposed ([K⁺]_i=100 mmol l⁻¹, [K⁺]_o=1 mmol l⁻¹ plus valinomycin; see legend to Fig. 3) inside-negative potential by two concentrations of leucine, in the presence of a Na⁺ gradient ([Na⁺]_o>[Na⁺]_i). Fig. 3B shows that the fluorescence quenching dissipated more rapidly with the sodium gradient and even faster when the amino acid was also present, according to its concentration.

Thus, the functional characterization of leucine transport along the larval midgut of B. mori suggests the presence of two different symporters for neutral amino acids in AM and P midgut regions.

A series of questions regarding those features of the amino acid structure that determine its affinity for the neutral amino acid/K⁺ symporters of the two midgut regions was addressed. The topics investigated were: (a) the relevance of the free amino and carboxyl groups; (b) the position of the amino group; (c) the importance of the configuration and structure of the α -carbon; (d) the structure of the sidechain. Leucine uptake was measured in AM-BBMV and P-BBMV in the presence of a 20-fold excess of specific amino acids and analogues. The results (P. Parenti, M. Casartelli, M. Castagna, G. Leonardi and B. Giordana, in preparation), some of which are reported in Table 3, can be summarized as follows. Modifications to the COOH group drastically reduced the affinity for the AM cotransporter, whereas these were better tolerated by the P symporter. As expected, the free NH₂ group was also crucial for substrate recognition. However, for the K⁺/symporter of the AM region, the presence in the sidechain of an imidazolic ring facilitated the interaction of the amino acid or analogue with the binding site. However, both symporters have in common a poor stereoselectivity and a relatively low selectivity for α -methyl amino acids, α -ketoacids and δ -amino acids.

These results give further support to the hypothesis that two structurally different carrier proteins for neutral amino acids are expressed along the midgut of lepidopteran larvae.

Bacillus thuringiensis produces during sporulation a parasporal crystalline inclusion highly specific against several orders of insects (Höfte and Whiteley, 1989). In lepidopteran larvae, the crystal is processed by gut juice proteases to yield the entomocidal polypeptides called δ -endotoxins or Cry proteins. The molecular mechanism of action of these toxins consists of the binding to a brush-border-specific receptor, which determines the host specificity (Hofmann et al. 1988). Binding appears to be followed by a lytic phase with an alteration in the permeability of the apical membrane to solutes, until membrane disruption occurs (reviewed by Wolfersberger, 1992; Knowles and Dow, 1993). The activity and specificity of a given Cry protein against a lepidopteran species can be rapidly assessed by testing its ability to inhibit K⁺-dependent amino acid transport into midgut BBMV (Sacchi et al. 1986; Wolfersberger, 1991; Parenti et al. 1993).

Recently, evidence has been presented that B. thuringiensis subsp. aizawai δ -endotoxin acts as a non-competitive inhibitor of amino acid/K⁺ symport in B. mori larval midgut even in the absence of a K⁺ gradient (Giordana et al. 1993).

The activity of three crystal proteins, CryIA(a), CryIA(b) and CryIA(c), obtained from Escherichia coli clones, was tested on leucine uptake into AM-BBMV and P-BBMV at a pHe of 8.7 (P. Parenti et al. 1994). Fig. 4 shows that the activated CryIA(a) toxin produced a dose-dependent inhibition of amino acid uptake in both midgut regions. The effect is not due to leakage of the substrate from the vesicles and it is specific, since the two related toxins CryIA(b) and CryIA(c) did not cause significant inhibition in the same assay.

The ability of the active toxin to inhibit amino acid uptake can be observed (a) in the presence of a 100 mmol l⁻¹ K⁺ gradient (not shown); (b) with 100 mmol l⁻¹ K⁺ inside and outside the vesicle (Fig. 4: IC₅₀ 6.9±1.3 μg mg⁻¹ BBMV protein for AM-BBMVand 7.0±0.3 μg mg⁻¹ BBMV protein for P-BBMV); (c) in the complete absence of K⁺ (Fig. 4: IC₅₀ 3.5±0.3 μg mg⁻¹ BBMV protein for AM-BBMV and 2.5±0.6 μg mg⁻¹ BBMV protein for P-BBMV). Very similar IC₅₀ values, though with different Hill coefficients (discussed in P. Parenti et al. 1994), were also obtained in the absence of a pH gradient (pHi=pHe=7.2). Therefore, inhibition of amino acid transport in larval midgut BBMV is not secondary to the well-documented pore formation, since the toxin inhibited leucine uptake irrespective of the presence of K⁺ and in the absence of any other ion gradient.

To explain the inhibition of the amino acid/K⁺ symporters by B. thuringiensis subsp. aizawai toxin, we suggested (Giordana et al. 1993) that the complex mechanism of action of B. thuringiensis toxins could be mediated by the binding to a membrane protein represented by the K⁺/amino acid symporters or a strictly associated protein. The receptor for CryIA(c) toxin active on Manduca sexta larvae has now been identified with the midgut brush-border membrane ectoenzyme aminopeptidase N (Knight et al. 1994; Sangadala et al. 1994). McGivan and coauthors have shown that the integrity of this enzyme is essential for an efficient amino acid transport by the Na⁺-dependent B° system in bovine renal brush-border membranes (Plakidou-Dymock et al. 1993).

Our present working hypothesis is that one action – but not the only one (Knowles and Dow, 1993) – of lepidopteran entomocidal B. thuringiensis o-endotoxins is to impair the activity of the amino acid/K⁺ symporters either directly or through their interaction with the functionally associated aminopeptidase N.