The location of a disulfonic stilbene binding site in band 3, the anion transport protein of the red blood cell membrane

The binding site for 4,4′-diisothiocyano-2,2′-stilbenedisulfonic acid, a specific, potent, irreversible inhibitor of anion transport in red blood cells is located in a 15 000 dalton transmembrane segment of band 3, produced by chymotrypsin treatment of ghosts stripped of extrinsic proteins. The segment was cleaved into three fragments of 7000, 4000 and 4000 daltons by CNBr. The C-terminus of the segment is located in the 7000 dalton fragment; the N-terminus in one of the 4000 dalton fragments; and the binding site for 4,4′-diisothiocyano-2,2′-stilbenedisulfonic acid in the middle 4000 dalton fragment. The latter was cleaved by $\text{N-}\text{bromosuccinimide}$ into two fragments of 2000 daltons. The binding site for 4,4′-diisothiocyano-2,2′-stilbenedisulfonic acid was located on the fragment containing the newly formed N-terminus. It is concluded that the binding site is located about 9000 daltons from the C-terminus (at the outside face of the membrane) and 6000 daltons from the N-terminus (at the cytoplasmic face). In view of the existing evidence that the binding site may be located near the outside face of the membrane, it is suggested that the 15 000 dalton segment is folded, so that it crosses the bilayer three times.     

Inhibition of anion transport associated with chymotryptic cleavages of red blood cell band 3 protein

Right-side-out vesicles derived from red blood cells treated with chymotrypsin retain specific anion transport function (defined as transport sensitive to the specific inhibitor, 4,4′-diisothiocyano-2,2′-stilbenedisulfonic acid (DIDS)), even though the transport protein, band 3, is cleaved into two segments of 60 and 35 kdaltons. In contrast, vesicles derived from alkali-stripped ghosts treated with relatively high concentrations of chymotrypsin retain almost no specific anion function. The loss of function appears to be related to additional cleavages of band 3 protein that occur in treated ghosts, the 60-kdalton segment being reduced first to a 17- and then to a 15-kdalton segment and the 35-kdalton segment being reduced to a 9-kdalton segment plus a carbohydrate containing fragment. The chymotryptic cleavages of band 3 protein of ghosts are preferentially inhibited by high ionic strength, the production of the 9-kdalton segment being somewhat slower than that of the 15-kdalton segment. Vesicles derived from ghosts treated with chymotrypsin at different ionic strengths show a graded reduction in specific anion transport activity, but it was not possible to determine, definitively, which of the additional cleavages was inhibitory. In the light of these data and other information, the functional role of the segments of band 3 is discussed.     

Anion transport across the red blood cell membrane and the protein in band 3.

The paper reviews existing evidence for the participation of the protein in band 3 (nomenclature of Steck, [1]) in anion transport across the red cell membrane and discusses the possible role of common binding sites on band 3 for 1-fluoro-2,4-dinitrobenzene, 2-(4'-aminophenyl)-6-methylbenzenethiazol-3',7-disulfonic acid and dihydro 4,4'-diisothiocyanato stilbene-2,2'-disulfonic acid in the transport process.     

The red cell band 3 protein: its role in anion transport

Studies of anion transport across the red blood cell membrane fall generally into two categories: (1) those concerned with the operational characterization of the transport system, largely by kinetic analysis and inhibitor studies; and (2) those concerned with the structure of band 3, a transmembrane peptide identified as the transport protein. The kinetics are consistent with a ping-pong model in which positively charged anion-binding sites can alternate between exposure to the inside and outside compartments but can only shift one position to the other when occupied by an anion. The structural studies on band 3 indicate that only 60% of the peptide is essential for transport. That particular portion is in the form of a dimer consisting of an assembly of membrane-crossing strands (each monomer appears to cross at least five times). The assembly presents its hydrophobic residues toward the interior of the bilayer, but its hydrophilic residues provide an aqueous core. The transport involves a small conformational change in which an anion-binding site (involving positively charged residues) can alternate between positions that are topologically in and topologically out.     

Interaction of phloretin with the anion transport protein of the red blood cell membrane

Phloretin is an inhibitor of anion exchange and glucose and urea transport in human red cells. Equilibrium binding and kinetic studies indicate that phloretin binds to band 3, a major integral protein of the red cell membrane. Equilibrium phloretin binding has been found to be competitive with the binding of the anion transport inhibitor, 4,4′-dibenzamido-2,2′-disulfonic stilbene (DBDS), which binds specifically to band 3. The apparent binding (dissociation) constant of phloretin to red cell ghost band 3 in 28.5 mM citrate buffer, pH 7.4, 25°C, determined from equilibrium binding competition, is $\text{1.8 ± 0.1}\text{μM}$. Stopped-flow kinetic studies show that phloretin decreases the rate of DBDS binding to band 3 in a purely competitive manner, with an apparent phloretin inhibition constant of $\text{1.6 ± 0.4}\text{μM}$. The pH dependence of equilibrium binding studies show that it is the charged, anionic form of phloretin that competes with DBDS binding, with an apparent phloretin inhibition constant of 1.4 μM. The phloretin binding and inhibition constants determined by equilibrium binding, kinetic and pH studies are all similar to the inhibition constant of phloretin for anion exchange. These studies suggest that phloretin inhibits anion exchange in red cells by a specific interaction between phloretin and band 3.     

Inhibition of inorganic anion transport across the human red blood cell membrane by chloride-dependent association of dipyridamole with a stilbene disulfonate binding site on the band 3 protein

The inhibition of inorganic anion transport by dipyridamole (2,6-bis(diethanolamino)-4,8-dipiperidinopyrimido[5,4-d] pyrimidine) takes place only in the presence of Cl-, other halides, nitrate or bicarbonate. At any given dipyridamole concentration, the anion flux relative to the flux in the absence of dipyridamole follows the equation: Jrel = (1 + alpha 2[Cl-])/(1 + alpha 4[Cl-]) where alpha 2 and alpha 4 are independent of [Cl-] but dependent on dipyridamole concentration. At high [Cl-] the flux approaches alpha 2/alpha 4, which decreases with increasing dipyridamole concentration. Even when both [Cl-] and dipyridamole concentration assume large values, a small residual flux remains. The equation can be deduced on the assumption that Cl- binding allosterically increases the affinity for dipyridamole binding to band 3 and that the bound dipyridamole produces a non-competitive inhibition of sulfate transport. The mass-law constants for the binding of Cl- and dipyridamole to their respective-binding sites are about 24 mM and 1.5 microM, respectively (pH 6.9, 26 degrees C). Dipyridamole binding leads to a displacement of 4,4'-dibenzoylstilbene-2,2'-disulfonate (DBDS) from the stilbenedisulfonate binding site of band 3. The effect can be predicted quantitatively on the assumption that the Cl- -promoted dipyridamole binding leads to a competitive replacement of the stilbenedisulfonates. For the calculations, the same mass-law constants for binding of Cl- and dipyridamole can be used that were derived from the kinetic studies on Cl- -promoted anion transport inhibition. The newly described Cl- binding site is highly selective with respect to Cl- and other monovalent anion species. There is little competition with SO4(2-), indicating that Cl- binding involves other than purely electrostative forces. The affinity of the binding site to Cl- does not change over the pH range 6.0-7.5. Dipyridamole binds only in its deprotonated state. Binding of the deprotonated dipyridamole is pH-independent over the same range as Cl- binding.     

Two-dimensional structure of the membrane domain of human band 3, the anion transport protein of the erythrocyte membrane.

The membrane domain of human erythrocyte Band 3 protein (M(r) 52,000) was reconstituted with lipids into two-dimensional crystals in the form of sheets or tubes. Crystalline sheets were monolayers with six-fold symmetry (layer group p6, a = b = 170 A, gamma = 60 degrees), whereas the symmetry of the tubular crystals was p2 (a = 104 A, b = 63 A, gamma = 104 degrees). Electron image analysis of negatively stained specimens yielded projection maps of the protein at 20 A resolution. Maps derived from both crystal forms show that the membrane domain is a dimer of two monomers related by two-fold symmetry, with each monomer consisting of three subdomains. In the dimer, two subdomains of each monomer form a roughly rectangular core (40 x 50 A in projection), surrounding a central depression. The third subdomain of the monomer measures approximately 15 x 25 A in projection and appears to be connected to the other two by a flexible link. We propose that the central depression may represent the channel for anion transport while the third subdomain appears not to be directly involved in channel formation.     

Temperature dependence of anion transport in the human red blood cell

Arrhenius plots of chloride and bromide transport yield two regions with different activation energies (Ea). Below 15 or 25 degrees C (for Cl- and Br-, respectively), Ea is about 32.5 kcal/mol; above these temperatures, about 22.5 kcal/mol (Brahm, J. (1977) J. Gen. Physiol. 70, 283-306). For the temperature dependence of SO4(2-) transport up to 37 degrees C, no such break could be observed. We were able to show that the temperature coefficient for the rate of SO4(2-) transport is higher than that for the rate of denaturation of the band 3 protein (as measured by NMR) or the destruction of the permeability barrier in the red cell membrane. It was possible, therefore, to extend the range of flux measurements up to 60 degrees C and to show that, even for the slowly permeating SO4(2-) in the Arrhenius plot, there appears a break, which is located somewhere between 30 and 37 degrees C and where Ea changes from 32.5 to 24.1 kcal/mol. At the break, the turnover number is approx. 6.9 ions/band 3 per s. Using 35Cl- -NMR (Falke, Pace and Chan (1984) J. Biol. Chem. 259, 6472-6480), we also determined the temperature dependence of Cl- -binding. We found no significant change over the entire range from 0 to 57 degrees C, regardless of whether the measurements were performed in the absence or presence of competing SO4(2-). We conclude that the enthalpy changes associated with Cl- - or SO4(2-)-binding are negligible as compared to the Ea values observed. It was possible, therefore, to calculate the thermodynamic parameters defined by transition-state theory for the transition of the anion-loaded transport protein to the activated state for Cl-, Br- and SO4(2-) below and above the temperatures at which the breaks in the Arrhenius plots are seen. We found in both regions a high positive activation entropy, resulting in a low free enthalpy of activation. Thus the internal energy required for carrying the complex between anion and transport protein over the rate-limiting energy barrier is largely compensated for by an increase of randomness in the protein and/or its aqueous environment.     

The band 3 protein of the human red cell membrane: A review

Band 3 is the predominant polypetide and the purported mediator of anion transport in the human erythrocyte membrane. Against a background of minor and apparently unrelated polypeptides of similar electrophoretic mobility, and despite apparent heterogeneity in its glycosylation, the bulk of band 3 exhibits uniform and characteristic behavior. This integral glycoprotein appears to exist as a noncovalent dimer of two ～ 93,000-dalton chains which span the membrane asymmetrically. The protein is hydrophobic in its composition and in its behaviour in aqueous solution and is best solubilized and purified in detergent. It can be cleaved while membrane-bound into large, topographically defined segments. An integral, outer-surface, 38,000-dalton fragment bears most of the band 3 carbohydrate. A 17,000-dalton, hydrophobic glycopeptide fragment spans the membrane. A ～ 40,000-dalton hydrophilic segment represents the cytoplasmic domain. In vitro, glyceraldehyde 3-P dehydrogenase and aldolase bind reversibly, in a metabolite-sensitive fashion, to this cytoplasmic segment. The cytoplasmic domain also bears the amino terminus of this polypetide, in contrast to other integral membrane proteins. Recent electron microscopic analysis suggests that the poles of the band 3 molecule can be seen by freezeetching at the two original membrane surfaces, while freeze-fracture reveals the transmembrane disposition of band 3 dimer particles. There is strong evidence that band 3 mediates 1:1 anion exchange across the membrane through a conformational cycle while remaining fixed and asymmetrical. Its cytoplasmic pole can be variously perturbed and even excised without a significant alteration of transport function. However, digestion of the outer-surface region leads to inhibition of transport, so that both this segment and the membrane-spanning piece (which is slectively labeled by covalent inhibitors of transport) may be presumed to be involved in transport. Genetic polymorphism has been observed in the structure and immunogenicity of the band 3 polypeptide but this feature has not been related to variation in anion transport or other band 3 activities.     

Temperature dependence of anion transport in the human red blood cell

Arrhenius plots of chloride and bromide transport yield two regions with different activation energies (E_a). Below 15 or 25°C (for Cl⁻ and Br⁻, respectively), E_a is about 32.5 kcal/mol; above these temperatures, about 22.5 kcal/mol (Brahm, J. (1977) J. Gen. Physiol. 70, 283–306). For the temperature dependence of SO₄²⁻ transport up to 37°C, no such break could be observed. We were able to show that the temperature coefficient for the rate of SO₄²⁻ transport is higher than that for the rate of denaturation of the band 3 protein (as measured by NMR) or the destruction of the permeability barrier in the red cell membrane. It was possible, therefore, to extend the range of flux measurements up to 60°C and to show that, even for the slowly permeating SO₄²⁻ in the Arrhenius plot, there appears a break, which is located somewhere between 30 and 37°C and where E_a changes from 32.5 to 24.1 kcal/mol. At the break, the turnover number is approx. 6.9 ions/band 3 per s. Using ³⁵Cl⁻-NMR (Falke, Pace and Chan (1984) J. Biol. Chem. 259, 6472–6480), we also determined the temperature dependence of Cl⁻-binding. We found no significant change over the entire range from 0 to 57°C, regardless of whether the measurements were performed in the absence or presence of competing SO₄²⁻. We conclude that the enthalpy changes associated with Cl⁻-or SO₄²⁻-binding are negligible as compared to the E_a values observed. It was possible, therefore, to calculate the thermodynamic parameters defined by transition-state theory for the transition of the anion-loaded transport protein to the activated state for Cl⁻, Br⁻ and SO₄²⁻ below and above the temperatures at which the breaks in the Arrhenius plots are seen. We found in both regions a high positive activation entropy, resulting in a low free enthalpy of activation. Thus the internal energy required for carrying the complex between anion and transport protein over the rate-limiting energy barrier is largely compensated for by an increase of randomness in the protein and/or its aqueous environment.     

Intrinsic segments of band 3 that are associated with anion transport across red blood cell membranes

Summary After treatment of red cell ghosts with chymotrypsin, the predominant intrinsic peptides remaining in the membrane fraction are 15,000 and 9,000 daltons mol wt. After partial extraction with Triton X-100, the residual membrane vesicles have almost no other stained peptides and such vesicles are reported to carry out anion transport activities sensitive to specific inhibitors. In vesicles derived from cells treated with DIDS(4,4-diisothiocyano-2,2-stilbene disulfonic acid), an irreversible inhibitor of anion transport that is highly localized in an abundant intrinsic protein known as band 3, the probe is largely recovered in the 15,000 dalton peptide. The part of band 3 from which it is derived is a previously reported 17,000 transmembrane segment (Steck, T.L., Ramos, R., Strapazon, E., 1976,Biochemistry 15:1154). The 9,000-dalton peptide is present in the vesicles in a one-to-one mole ratio with the 15,000-dalton peptide, suggesting that both are derived from the same protein. This conclusion is supported by the finding that the 35,000-dalton C-terminal end of band 3, derived by chymotrypsin treatment of cells, is further proteolysed if the cells are converted to ghosts and its disappearance coincides with the appearance of the 9,000-dalton fragment. Evidence is presented that the 9,000-dalton fragment crosses the bilayer and that it is closely associated with the 15,000-dalton peptide.This paper is dedicated to the memory of Walther Wilbrandt.     

The preparation of hydrophilic derivatives of band 3 protein by acylation of the human red blood cell membrane

In situ reaction of erythrocyte membranes with dicarboxylic anhydrides leads to solubilization of hydrophobic integral proteins. Removal of peripheral proteins and bulk lipid by appropriate sedimentation and dialysis steps yields hydrophilic band 3 protein derivatives. These acyl compounds display size heterogeneity upon gel filtration. A chromatographically homogeneous acyl band 3 protein is obtained if the acylation is conducted in the presence of detergent and the detergent subsequently removed. Hydrophilic acyl derivatives of band 3 protein can be subjected to conventional analytical techniques without the use of detergents.     

Carboxypeptidase Y digestion of band 3, the anion transport protein of human erythrocyte membranes

The exposure of the carboxyl-terminal of the Band 3 protein of human erythrocyte membranes in intact cells and membrane preparations to proteolytic digestion was determined. Carboxypeptidase Y digestion of purified Band 3 in the presence of non-ionic detergent released amino acids from the carboxyl-terminal of Band 3. The release of amino acids was very pH dependent, digestion being most extensive at pH 3, with limited digestion at pH 6 or above. The 55,000 dalton carboxyl-terminal fragment of Band 3, generated by mild trypsin digestion of ghost membranes, had the same carboxyl-terminal sequence as intact Band 3, based on carboxypeptidase Y digestion. Treatment of intact cells with trypsin or carboxypeptidase Y did not release any amino acids from the carboxyl-terminal of Band 3. In contrast, carboxypeptidase Y readily digested the carboxyl-terminal of Band 3 in ghosts that were stripped of extrinsic membrane proteins by alkali or high salt. This was shown by a decrease in the molecular weight of a carboxyl-terminal fragment of Band 3 after carboxypeptidase Y digestion of stripped ghost membranes. No such decrease was observed after carboxypeptidase Y treatment of intact cells. In addition, Band 3 purified from carboxypeptidase Y-treated stripped ghost membranes had a different carboxyl-terminal sequence from intact Band 3. Cleavage of the carboxyl-terminal of Band 3 was also observed when non-stripped ghosts or inside-out vesicles were treated with carboxypeptidase Y. However, the digestion was less extensive. These results suggest that the carboxyl-terminal of Band 3 may be protected from digestion by its association with extrinsic membrane proteins. We conclude, therefore, that the carboxyl-terminal of Band 3 is located on the cytoplasmic side of the red cell membrane. Since the amino-terminal of Band 3 is also located on the cytoplasmic side of the erythrocyte membrane, the Band 3 polypeptide crosses the membrane an even number of times. A model for the folding of Band 3 in the erythrocyte membrane is presented.     

Oligomeric structure and the anion transport function of human erythrocyte band 3 protein

Conclusions Evidence from many laboratories using several different techniques strongly suggests that, in the intact red cell, band 3 exists as dimers which can associate with other dimers to form tetramers. The kinetics of anion transport inhibition by stilbenedisulfonates indicate that irreversible inhibition of one subunit does not detectably affect anion transport by the other subunit. This does not imply that monomeric band 3 could necessarily transport anions; the native conformation of each subunit may require stabilizing interactions with another subunit, as indicated by the recent work of Boodhoo and Reithmeier [10]. A more detailed understanding of the structure of the band 3 dimer/tetramer will require information on which specific segments of the primary structure are involved in subunit-subunit contact. The combination of chemical cross-linking with proteolysis [136] is a promising approach to this problem.     

Peroxynitrite oxidizes erythrocyte membrane band 3 protein and diminishes its anion transport capacity

We describe an altered membrane band 3 protein-mediated anion transport in erythrocytes exposed to peroxynitrite, and relate the loss of anion transport to cell damage and to band 3 oxidative modifications. We found that peroxynitrite down-regulate anion transport in a dose dependent relation (100–300 μmoles/l). Hemoglobin oxidation was found at all peroxynitrite concentrations studied. A dose-dependent band 3 protein crosslinking and tyrosine nitration were also observed. Band 3 protein modifications were concomitant with a decrease in transport activity. ( ? )-Epicatechin avoids band 3 protein nitration but barely affects its transport capacity, suggesting that both processes are unrelated. N-acetyl cysteine partially reverted the loss of band 3 transport capacity. It is concluded that peroxynitrite promotes a decrease in anion transport that is partially due to the reversible oxidation of band 3 cysteine residues. Additionally, band 3 tyrosine nitration seems not to be relevant for the loss of its anion transport capacity.     

Peroxynitrite oxidizes erythrocyte membrane band 3 protein and diminishes its anion transport capacity

We describe an altered membrane band 3 protein-mediated anion transport in erythrocytes exposed to peroxynitrite, and relate the loss of anion transport to cell damage and to band 3 oxidative modifications. We found that peroxynitrite down-regulate anion transport in a dose dependent relation (100-300 μmoles/l). Hemoglobin oxidation was found at all peroxynitrite concentrations studied. A dose-dependent band 3 protein crosslinking and tyrosine nitration were also observed. Band 3 protein modifications were concomitant with a decrease in transport activity. ( - )-Epicatechin avoids band 3 protein nitration but barely affects its transport capacity, suggesting that both processes are unrelated. N-acetyl cysteine partially reverted the loss of band 3 transport capacity. It is concluded that peroxynitrite promotes a decrease in anion transport that is partially due to the reversible oxidation of band 3 cysteine residues. Additionally, band 3 tyrosine nitration seems not to be relevant for the loss of its anion transport capacity.     

The relationship between anion exchange and net anion flow across the human red blood cell membrane

The conductive (net) anion permeability of human red blood cells was determined from net KCl or K2SO4 effluxes into low K+ media at high valinomycin concentrations, conditions under which the salt efflux is limited primarily by the net anion permeability. Disulfonic stilbenes, inhibitors of anion exchange, also inhibited KCl or K2SO4 efflux under these conditions, but were less effective at lower valinomycin concentrations where K+ permeability is the primary limiting factor. Various concentrations of 4,4'-diisothiocyanostilbene-2,2'-disulfonate (DIDS) had similar inhibitory effects on net and exchange sulfate fluxes, both of which were almost completely DIDS sensitive. In the case of Cl-, a high correlation was also found between inhibition of net and exchange fluxes, but in this case about 35% of the net flux was insensitive to DIDS. The net and exchange transport processes differed strikingly in their anion selectivity. Net chloride permeability was only four times as high as net sulfate permeability, whereas chloride exchange is over 10,000 times faster than sulfate exchange. Net OH-permeability, determined by an analogous method, was over four orders of magnitude larger than that of Cl-, but was also sensitive to DIDS. These data and others are discussed in terms of the possibility that a common element may be involved in both net and exchange anion transport.     

Binding of DTNB to band 3 in the human red cell membrane

Inhibition of red cell water transport by the sulfhydryl reagent 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) has been reported by Naccache and Sha'afi ((1974) J. Cell Physiol. 84, 449-456) but other investigators have not been able to confirm this observation. Brown et al. ((1975) Nature 254, 523-525) have shown that, under appropriate conditions, DTNB binds only to band 3 in the red cell membrane. We have made a detailed investigation of DTNB binding to red cell membranes that had been treated with the sulfhydryl reagent N-ethylmaleimide (NEM), and our results confirm the observation of Brown et al. Since this covalent binding site does not react with either N-ethylmaleimide or the sulfhydryl reagent pCMBS (p-chloromercuribenzenesulfonate), its presence has not previously been reported. This covalent site does not inhibit water transport nor does it affect any transport process we have studied. There is an additional low-affinity (non-covalent) DTNB site that Reithmeier ((1983) Biochim. Biophys. Acta 732, 122-125) has shown to inhibit anion transport. In N-ethylmaleimide-treated red cells, we have found that this binding site inhibits water transport and that the inhibition can be partially reversed by the specific stilbene anion exchange transport inhibitor 4,4'-diisothiocyanostilbene-2,2'-disulfonate (DIDS), thus linking water transport to anion exchange. DTNB binding to this low-affinity site also inhibits ethylene glycol and methyl urea transport with the same KI as that for water inhibition, thus linking these transport systems to that for water and anions. These results support the view that band 3 is a principal constituent of the red cell aqueous channel, through which urea and ethylene glycol also enter the cell.     

Proteolytic cleavage of band 3 protein in relation to anion transport in fish (Oncorhynchus mykiss) red blood cells

• 1.1. The effects of trypsin and chymotrypsin on HCO₃⁻/Cl⁻ exchange through red blood cell membranes of humans and trout were studied.
• 2.2. To measure the anion exchange we used a right-angle light-scattering technique by applying the Jacobs-Stewart cycle in ammonium solution and the osmotiration method at constant cell volume.
• 3.3. The Cl⁻ flux in human red blood cells remained unaltered after treatment with external trypsin and chymotrypsin while in trout red blood cells the flux decreased.
• 4.4. This partial inhibition of anion transport in fish, ranging from 30 to 40%,suggest that one or several of the cleavage sites in band 3 protein, essential for anion transport function, are exposed in fish red blood cells.
• 5.5. In human red blood cells the fragments of band 3 which are affected by proteolytic digestion, retain their tertiary structure because there is no influence on anion transport.

     

Inhibition of anion transport in the red blood cell by anionic amphiphilic compounds II. Chemical properties of the flufenamate-binding site on the band 3 protein

Flufenamate is a powerful inhibitor of anion exchange in red blood cells. It binds to the band 3 protein involved in the transport as discussed in the preceding paper (Cousin, J.-L. and Motais, R. (1982) Biochim. Biophys. Acta 687, 147–155). The present study is concerned with the chemical properties of the inhibitory binding site. Structure-activity studies were performed with two sets of compounds derivated from anthranilate (considered as the basic structure of flufenamate). The molar concentrations required to produce 50% inhibition ($\text{I}{}_{50}$) varied over more than a 10⁴ range. The inhibitory activity was quantitatively correlated with the hydrophobic character of the molecules and the electron-withdrawing capacity of the substituents. Comparison between the inhibitory potency of flufenamate analogs made a definition of the contribution of each part of the molecule in the binding to the receptor possible. The results suggest that anionic inhibitors bind to a site which presents a positively charged groups at the water-protein interface whereas the hydrophobic part of the molecule is inserted into an hydrophobic and electron-donor region of the protein. The specificity of amphiphilic compounds towards anion transport is discussed.