ABSTRACT
The time-dependent fluorescence intensity of an intravesicular potential-sensitive dye was used to probe the real-time kinetics of potential difference (PD)-dependent amino acid/Na+ symport at pH 9 into brush-border membrane vesicles obtained from larval Manduca sexta midgut. Neutral amino acids (alanine, proline) are symported at higher rates as the vesicles are hyperpolarized. The symport rates of acidic (glutamate) and basic (arginine) amino acids are almost PD-independent. The half-saturation constant of alanine is PD-independent between −108 and −78 mV, although the maximal symport velocity increases by half as the voltage is increased. Amino acid throughput is evidently enhanced as the relatively high transmembrane PDs (>150 mV, lumen positive) measured in vivo are approached. The half-saturation concentrations of Na+ were in the range 15–40 mmol l−1 for most of the amino acids examined and increased with voltage for alanine. The Vmax observed as a function of cation or amino acid concentration increased as the vesicle was hyperpolarized in the case of leucine and alanine. The data support the hypothesis that carrier and substrates are at equilibrium inasmuch as substrate translocation seems to be the rate-determining step of symport.
INTRODUCTION
The transport of amino acids across the lepidopteran larval midgut epithelium from the lumen to the haemolymph is K+-dependent (Nedergaard, 1972). In the larval tobacco hornworm (Manduca sexta) midgut the luminal pH is 10–11 whereas that of the haemolymph is 7 (Dow, 1984). The pH gradient is more or less at equilibrium with a transapical potential difference (PD) of greater than 150 mV, lumen positive. The PD, in turn, depends upon an electrogenic ‘K+ pump’ in goblet cell apical membrane (GCAM) (Harvey et al. 1983; Dow et al. 1984) that is electrically coupled (Dow and Peacock, 1989) to the columnar cell apical membrane (brush-border membrane, BBM). There is essentially no K+ activity gradient across the apical membrane in vivo (Dow et al. 1984) and the PD alone suffices to drive amino acid uptake. Finally, the ‘K+ pump’ consists of an H+-translocating V-ATPase in parallel with a K+/nH+ antiporter in GCAM (Wieczorek et al. 1991). This M. sexta ‘midgut model’ and its relevance to cation-dependent, PD-driven, pH-sensitive amino acid uptake by lepidopteran midgut has been reviewed recently in several articles (see Harvey and Nelson, 1992).
The K+-selective, cation-dependent uptake of amino acids (symport) across midgut BBM vesicles (BBMVs) has been well-characterized in the larval lepidopteran Philosamia cynthia (Hanozet et al. 1980; Giordana et al. 1989). In BBMVs from larval M. sexta, amino acid/cation symport is similarly K+-selective; Na+ is a good substitute in vitro (Hennigan et al. 1993a,b) but is a minor cation in vivo. The amino acid and cation specificity of a broad-spectrum ‘B-type’ neutral amino acid symporter present in M. sexta midgut has been characterized at pH 10, near the physiological value (Hennigan et al. 1993a,b). In the present report, we examine the effects at high pH of electrical PD on amino acid/Na+ symport rates of neutral, cationic and anionic amino acids into larval M. sexta BBMVs using a potential-sensitive fluorescent dye.
Fluorescent dyes have been used to probe the kinetics of symport processes involving electrophoretic fluxes (Beck and Sacktor, 1978; Cassano et al. 1988; Maffia et al. 1990). Insofar as electrophoretic fluxes modulate transmembrane voltage, the time-dependence of the voltage is a measure of the rate of cation flux and, thus, of symport. Spectroscopic studies (which monitor cation flow) of symport kinetics are therefore complementary to the more traditional tracer techniques with the advantage of economically providing a complete kinetic record in real time.
Earlier spectroscopic work (Beck and Sacktor, 1978; Wright et al. 1981) empirically related fluorescence intensity (and thus PD) to the amount of substrate symported. The comparatively recent work of Cassano et al. (1988) and Maffia et al. (1990), in contrast, explores the real-time modulation of a pre-set voltage by incoming alkali ions. In view of this development, we felt that a spectroscopic exploration of amino acid symport in M. sexta at high pH values (close to those in vivo) would be opportune.
Using tracer experiments, at least six types of symporters with overlapping substrate specificities have been identified in the BBMVs of larval lepidopteran midguts (Giordana et al. 1989). A measurement of symport rates in BBMVs will therefore inevitably measure a rate that is averaged over the participating symporters. This complication is method-independent and will influence this spectroscopic study to the same degree that it has influenced tracer studies in the past.
In this paper, we describe experiments using the dye, 3,3’-diethylthiadicarbocyanine iodide [DiSC2(5)], belonging to the well-understood carbocyanine family (Waggoner, 1976; Cabrini and Verkman, 1986), to characterize (a) the PD-dependence of fluxes of alanine, proline, glutamate and arginine into BBMVs; (b) the half-saturation concentrations of a representative group of amino acids, alanine, proline, arginine, leucine, 2,4-diaminobutyric acid (DABA) and glutamate; and (c) the half-saturation concentrations of sodium in the presence of alanine, leucine, arginine and glutamate. The symport of the L-stereoisomer was monitored in all cases with Na+ as the cosymported cation instead of K+ (which predominates in vivo), since it was necessary to use the latter to establish a valinomycin-mediated diffusion potential.
MATERIALS AND METHODS
Brush-border membranes were isolated from midguts of fifth-instar M. sexta larvae using Mg2+ precipitation followed by differential centrifugation (Biber et al. 1981; Eisen et al. 1989). The resulting pellet was homogenized into buffer of the desired composition and the suspension was incubated for 1 h on ice. After centrifugation at 31 000 g for 1 h, the resulting pellet was resuspended by passage through a 26 gauge needle (30 times) to give a protein concentration of 7 mg ml−1. The vesicles were then left overnight on ice and used the next day. Protein assays (Bradford, 1976) used the Bio-Rad Kit (Bio-Rad, Culver City, CA) with bovine serum albumin as standard. All experiments were conducted at pH 9, unless otherwise stated, since the sensitivity of the dye to PD was diminished at higher pH values. The absence of a pH gradient across the vesicle boundary is at variance with the in vivo situation. However, Z. Liu (unpublished data) has found lysine uptake to be independent of intravesicular pH. No difference was observed in fluorescence properties between 2-(N-cyclohexylamino)-ethanesulphonic acid (CHES), 3-(cyclohexylamino)-1-propane sulphonic acid (CAPS) or Bis–Tris propane (BTP) buffers.
Fluorescence measurements were conducted using a Perkin-Elmer LS-50 luminescence spectrophotometer. DiSC2(5) was obtained from Molecular Probes (Eugene, OR) and used without further purification. A 0.6 mmol l−1 stock solution of the dye was prepared in ethanol. The excitation and emission wavelengths were 645 and 665 nm, respectively; 7 nm slitwidths were used in both paths. Data were collected every second with an integration time of 1 s. The sources for other reagents were as follows: mannitol, KCl and NaCl (Fisher Scientific, Pittsburgh, PA); all of the natural amino acids, valinomycin, as well as the buffer components, CHES, CAPS and BTP (Sigma Chemical Company, St Louis, MO); DABA (Aldrich Chemical Company, Milwaukee, WI) and choline chloride (ICN Pharmaceuticals, Costa Mesa, CA).
In a typical experiment, appropriate volumes of buffer, mannitol, amino acid (where appropriate) and salt solutions were added to an acrylic cuvette to yield a total volume of 2 ml. 9 μl each of valinomycin and dye in ethanol were added to this volume to yield final concentrations of 4.5 μmol l−1 and 3 μmol l−1, respectively. The mixture was allowed to equilibrate for 5 min at 28°C under stirring in the cuvette holder of the spectrometer. 20 μl of the BBMVs was added to the contents of the cuvette after a baseline had been established. The immediate quench of fluorescence intensity was monitored for approximately 90 s after vesicle addition. Data were analysed using the graphics program Slide-Rite Plus (Advanced Graphics Software Inc., Sunnyvale, CA). Measurements were usually performed in duplicate, or in triplicate if the differential slope was small, and repeated on two preparations. Unless otherwise indicated, the data shown are averages over both preparations with the error bars representing ± s.e.m.
The PD was calculated using the Nernst equation since the permeability of potassium is considerably higher than those of the other ions in the presence of the K+-specific ionophore valinomycin. The potassium concentration within the vesicles was assumed to be equal to that in the incubation buffer since the vesicles attain equilibrium within an hour at room temperature (Hennigan et al. 1993b).
RESULTS
Fluorescence intensities obtained in the absence of amino acid at [K+]in/[K+]out ratios of 1, 10 and 100 and with alanine at a potassium concentration ratio of 63 (Fig. 1) reveal that the drop in fluorescence upon vesicle addition is related to the ratio of potassium concentrations and thus to the equilibrium PD developed. The rate of recovery is linear with time between 20 and 90 s after the addition of vesicles. The upper limit of this interval decreased to about 60 s when the extravesicular concentration of amino acid was greater than 2 mmol l−1. Recovery rates were measured only over intervals where linearity could be ensured. Even in the absence of amino acid, the rate of recovery increased with PD as a result of PD-dependent ionic movements (Fig. 1). The differential rate of fluorescence recovery, rateamino acid −rateno amino acid, isolates an amino-acid-dependent component of Na+ movement, i.e. symport; accordingly, we use the differential rate of fluorescence recovery as a measure of the symport rate. Since rate measurements are initiated as close to the time of vesicle addition as is feasible, we identify the differential rate as the initial symport rate. Cation-independent modes of amino acid entry cannot be resolved using this approach.
The initial fluorescence quench is proportional to the PD between 0 and approximately −100 mV (Fig. 1, inset) and saturates at more negative voltages. Our symport studies are limited therefore to voltages less negative than −108 mV. The relationship between fluorescence intensity and PD that we observe is not as linear as that observed by Cassano et al. (1988), perhaps because of differences in vesicle preparation or pH.
PD-dependence of amino acid uptake
Using the differential rate of fluorescence recovery as a measure of symport rate, we observed that alanine and proline were symported more rapidly than arginine or glutamate (whose uptake rates were similar) at voltages more negative than −60 mV (Fig. 2). The uptake rates of alanine and proline scale linearly with PD whereas glutamate is symported at a PD-independent rate. The uptake of arginine may be saturable at voltages more negative than −80 mV, although the dye response is weaker in this domain. The sodium and amino acid concentrations used in this experiment (90 mmol l−1 and 4 mmol l−1, respectively) are greater than the half-saturation concentrations (see below). The pH of the cuvette buffer changed by less than 0.1 unit after amino acid addition. The flux rate of sodium through the BBM via non-symport mechanisms (the slope of the control run in the absence of the amino acid) was linear for PDs less than −108 mV (data not shown).
Binding parameters of the amino acid
Plots of uptake velocity versus amino acid concentration for alanine and proline (Fig. 3) at −108 mV and −76 mV reveal that the Vmax for each amino acid increased strongly as the PD was increased, in parallel to the trend seen in Fig. 2. Leucine (Fig. 4) displays a weak dependence of Vmax on PD, whereas seems to be PD-independent (Fig. 5); the measurements with arginine had to be carried out at −108 mV and −96 mV since the data proved to be unusably noisy at lower potentials. The extravesicular concentration of NaCl in these experiments was 80 mmol l−1. Using a Lineweaver–Burke transformation of the data in Figs 3–5, we arrived at values of the binding parameters which were used in computing the curves shown in the figures. The half-saturation concentrations of amino acid, are substantially PD-independent, although more scatter in the binding data was encountered between runs for proline than for alanine. The and values at the indicated PDs are listed in Table 1. The data for glutamate have a large error since repeated attempts to obtain the K0.5 for this amino acid yielded divergent results owing to the small differential rates (Fig. 2). The uptake rates were too small at lower PDs to permit measurements in the case of glutamate. To assess the effect that substrate charge may have upon , we measured symport rate as a function of amino acid concentration with DABA at pH 9, where the amino acid is zwitterionic (Fig. 6) and pH 7, where it is cationic. The trace in Fig. 6 reveals the participation of more than one pathway. The data at pH 7 proved not to be reproducible and therefore are not shown.
Amino acid-dependent sodium binding parameters
Measurements of the half-saturation concentration of sodium, , with extravesicular amino acid concentrations greater than the respective values, were undertaken with alanine (Fig. 7), leucine (Fig. 8), arginine (Fig. 9) and glutamate. was 19±4 mmol l−1 at −78 mV and 39±13 mmol l−1 at −108 mV. Vmax increased by half as the voltage increased from −78 to −108 mV. decreased from approximately 35 mmol l−1 at −78 mV to around 16 mmol l−1 at −108 mV, whereas Vmax increased from around 14 to around 21 arbitrary units min−1 mg−1 protein. However, the plot at the higher voltage is suggestive of amino acid:cation stoichiometry that is not 1:1. Since the calibre of data obtained in the present experiments did not justify Hill index determinations, we are unable to address the question of transport stoichiometry. Similar experiments as a function of [NaCl] at −108 mV with glutamate (data not shown) suggested that the uptake rate of the amino acid was independent of [Na+] at concentrations greater than 5 mmol l−1, implying that was smaller than 5 mmol l−1 (our data yield a Na+ half-saturation concentration of approximately 0.7 mmol l−1). The extravesicular concentration of glutamate was 2 mmol l−1 in these experiments.
The saturable part of the arginine uptake versus [NaCl] curve could not reliably be identified; an Eadie–Hofstee plot, which fitted the data poorly, yielded an estimate of 34±5 mmol l−1 for (Fig. 9). A 10-fold change in amino acid concentration did not affect the symport rate versus [NaCl] profile. Because of the low symport rates, experiments were not attempted at lower PDs.
DISCUSSION
In view of the overlap between symporters in mediating the cotransport of a given amino acid in BBMVs, kinetic studies do not reflect the properties of any single symporter and can only be interpreted in the context of the overall symport process. The rate of symport measured experimentally is that of the slowest step and symport will appear to be PD-dependent only if the slowest step is PD-dependent.
The half-saturation concentrations of all of the amino acids investigated were invariant over the range −50 to −110 mV. Evidently, PD differences greater than −50 mV suffice to bias the carrier distribution in favour of the loaded state. With Philosamia cynthia, Sacchi et al. (1990) used rapid filtration experiments to determine that the and Vmax of leucine were lower and higher, respectively, at negative PDs than in the absence of a PD. With M. sexta, the variation of with PD was also confined to potentials less negative than approximately −70 mV.
In current theoretical models (Lauger and Jauch, 1986; Sanders et al. 1984; Turner, 1981), reaction rate theory is applied to a symport cycle wherein the different states are assumed to be separated by activation barriers. In terms of these models, the variation of Vmax with PD in Fig. 2 suggests that some phase of substrate translocation is likely to be the rate-determining step. Assuming that the rate of return of the unloaded carrier across the membrane is the same for all symporters, we discuss field-dependent uptake in terms of the transit rate of the fully loaded carrier.
The uptake rates of the different amino acids at the lowest PDs are not substantially different (Fig. 2). At pH 9, alanine, proline and arginine carry net charges of 0, 0 and 0.5, respectively. The uptake of zwitterionic forms is accelerated by an applied negative PD whereas that of charged species is less affected, suggesting that the transit activation barrier is higher for neutral amino acids than for their charged counterparts. (The following remarks are phrased in terms of a single hypothetical symporter instead of a symporter ensemble only for brevity.) An activation barrier can be represented by zF×m× Δψ, where Δψ is the charge on the translocated species, F is the Faraday, × Δ is the PD across the membrane and m (0 ⩽ m ⩽;1) accommodates the (assumed) linearity of the transmembrane voltage drop. If the field were the same at the energy maximum (m constant) for each of these symporters, the magnitude of the barrier would depend only on the charge on the fully loaded carrier. Assuming a 1:1 amino acid:cation stoichiometry, the combined charges at pH 9 due to substrate and cation are 1 (alanine, proline) and 1.5 (arginine). If zala,pro symporters > zarg symporter in the fully loaded case, it follows that the charge on the empty symporter must be significantly more negative for the arginine symporter than for the other two. (This implies that the mobilities of these entities will not be equal in an electric field.) If the unloaded arginine symporter is indeed anionic, a high binding affinity for arginine is to be expected and has been found in rapid filtration studies (Z. Liu, unpublished data). Although amino acid binding may be PD-dependent and possibly rate-determining at PDs more positive than −50 mV, at the more negative potentials used in this study and those in feeding larvae, potential-dependent translocation appears to be rate-determining. Among the amino acids studied, only the symport of arginine displays a reduction in the rate of increase at the highest potentials examined. The uptake of glutamate has been found to be electroneutral in M. sexta (T. Xie, unpublished data), as it is in P. cynthia (Giordana et al. 1989).
It is tempting to ascribe the observed differences in values to amino acid charge. Although the experiment with DABA was inconclusive, Cassano et al. (1988) have shown that the range of amino acid half-saturation concentrations spans an order of magnitude even among zwitterionic amino acids. Because amino acids probably use diverse ensembles of symporters, charge differences among unloaded symporters may be as important as those between their substrates. Much information is now available on the overlap between different symporters in terms of amino acid specificity. Thus, work in our laboratory has revealed that lysine may be symported into BBMVs of M. sexta by the neutral amino acid symporter (Hennigan et al. 1993a) as well as by a ‘lysine/arginine symporter’ (Z. Liu, unpublished results). We have found that DABA shares uptake pathways with lysine (R. Parthasarathy, unpublished results). For this reason, differences in the half-saturation concentrations for the dissimilarly charged forms of DABA probably reflect interactions with different symporter ensembles. Giordana et al. (1989) have found an exclusive symporter for proline in P. cynthia. Recent work using rapid filtration methods suggests that glutamate may be transported only marginally by the symporters of the other amino acids (T. Xie, unpublished results). The zwitterionic form of arginine seems not to be symported by the neutral amino acid symporter (Z. Liu, unpublished results). The small symport rates, measured using fluorescence spectroscopy, for these two amino acids are in accord with the results of rapid filtration experiments. Such congruity between fluorescence spectroscopy and rapid filtration heightens confidence in the applicability of the former technique to transport problems.
Under the present no-slip conditions, the affinity of the symporter for sodium could also influence the amino acid symport rate. However, symport rates of the amino acids tested at the lowest PDs of this study are broadly similar to one another even when the affinity for Na+ in the presence of these amino acids is quite different. Evidently the symport rate at approximately −50 mV is most strongly influenced by the kinetics of charge translocation.
In conclusion, we find that amino acid symporter half-saturation concentrations are PD-independent between −50 and −108 mV regardless of the charge borne by the substrate. The translocation phase of symport seems to be rate-determining so that symporter, substrate and cation are inferred to be at equilibrium in the extravesicular medium, as has been assumed in some theoretical models (Lauger and Jauch, 1986).
ACKNOWLEDGEMENTS
This work has been supported in part by grant number AI30464 from the US Public Health Service. We are grateful to Ms Tao Xie for assistance with some of the spectroscopic experiments and to Dr Michael G. Wolfersberger for his valuable comments.