Distinctive cationic proteins of the human eosinophil granule: major basic protein, eosinophil cationic protein, and eosinophil-derived neurotoxin

The human eosinophil granule contains a number of cationic proteins that have been identified and purified to homogeneity, including the major basic protein (MBP), the eosinophil cationic protein (ECP), and the eosinophil-derived neurotoxin (EDN). Because of confusion in the literature regarding the distinctiveness of MBP and ECP, we investigated the immunochemical and physicochemical properties of these purified proteins by electrophoresis on sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE), by specific double antibody radioimmunoassays (RIA) for MBP and ECP, and by fractionation of acid-solubilized eosinophil granules on Sephadex G-50 columns. Analysis of a mixture of the three purified proteins by SDS-PAGE showed that they migrated as three distinct bands with differing m.w. Comparison by specific RIA for MBP and ECP did not demonstrate any appreciable immunochemical cross-reactivities among the three proteins. Sephadex G-50 column fractions of acid-solubilized eosinophil granules were analyzed by RIA and by SDS-PAGE analysis of individual column fractions. MBP, ECP, and EDN eluted at different volumes from Sephadex G-50 columns as determined by RIA and SDS-PAGE. Soluble extracts of eosinophil granules from patients with the hypereosinophilic syndrome contained between six and 64 times more MBP than ECP on a weight basis. These observations demonstrate that MBP, ECP, and EDN are distinctive cationic proteins of the human eosinophil granule and that eosinophil granules from patients with eosinophilia contain considerably greater quantities of MBP than ECP.     

Localization of eosinophil cationic protein, major basic protein, and eosinophil peroxidase in human eosinophils by immunoelectron microscopic technique

An immunoelectron microscopic technique using protein A-gold as a specific marker was used for precise intracellular localization of eosinophil granule proteins. Eosinophils from healthy individuals were isolated in metrizamide gradients. Eosinophil cationic protein (ECP) and eosinophil peroxidase (EPO) were clearly located in the matrix of the large crystalloid-containing granules. In addition, ECP was probably present in the small granules of eosinophils. Major basic protein (MBP) was present in the crystalloid structure of specific granules. This method can be applied in studies of eosinophil degranulation to trace the release of biological effector molecules.     

Eosinophil cationic protein high-affinity binding to bacteria-wall lipopolysaccharides and peptidoglycans

The eosinophil cationic protein (ECP) is an eosinophil-secreted RNase involved in the immune host defense, with a cytotoxic activity against a wide range of pathogens. The protein displays antimicrobial activity against both Gram-negative and Gram-positive strains. The protein can destabilize lipid bilayers, although the action at the membrane level can only partially account for its bactericidal activity. We have now shown that ECP can bind with high affinity to the bacteria-wall components. We have analyzed its specific association to lipopolysaccharides (LPSs), its lipid A component, and peptidoglycans (PGNs). ECP high-affinity binding capacity to LPSs and lipid A has been analyzed by a fluorescent displacement assay, and the corresponding dissociation constants were calculated using the protein labeled with a fluorophor. The protein also binds in vivo to bacteria cells. Ultrastructural analysis of cell bacteria wall and morphology have been visualized by scanning and transmission electron microscopy in both Escherichia coli and Staphylococcus aureus strains. The protein damages the bacteria surface and induces the cell population aggregation on E. coli cultures. Although both bacteria strain cells retain their shape and no cell lysis is patent, the protein can induce in E. coli the outer membrane detachment. ECP also activates the cytoplasmic membrane depolarization in both strains. Moreover, the depolarization activity on E. coli does not require any pretreatment to overcome the outer membrane barrier. The protein binding to the bacteria-wall surface would represent a first encounter step key in its antimicrobial mechanism of action.     

Synergism between eosinophil cationic protein and oxygen metabolites in killing of schistosomula of Schistosoma mansoni

To study the cytotoxic reactions responsible for mediating eosinophil damage to schistosomula of Schistosoma mansoni, we have used cytoplasts (eosinophil or neutrophil vesicles devoid of granules and nuclei, with an intact oxidase in their plasma membrane) in combination with purified eosinophil cationic protein (ECP) or major basic protein (MBP) in a cytotoxicity test toward schistosomula. Suboptimal concentrations of ECP (10(-6) M) or MBP (10(-6) M) resulting in less than 10% killing were used in combination with cytoplasts. Cytoplasts alone in the presence of immune serum tested over a wide range of cytoplast:schistosomula ratios generated superoxide and hydrogen peroxide, but were unable to damage schistosomula. However, when a suboptimal ECP concentration (10(-6) M) was combined with neutroplasts or eosinoplasts, 43.9% +/- 8.5 (n = 7) and 24.7% +/- 9.8 (n = 3), respectively, of the schistosomula were killed. Oxygen metabolites were responsible for the synergism, because cytoplasts from a patient with chronic granulomatous disease were unable to act in synergy with ECP. In contrast to ECP, no synergism was found between cytoplasts and MBP (10(-6) to 2 X 10(-5)M). These results show that oxygen metabolites are important for the killing of schistosomula by lowering the concentration of ECP needed to inflict damage.     

Peptides of major basic protein and eosinophil cationic protein activate human mast cells

The eosinophil granule proteins, major basic protein (MBP) and eosinophil cationic protein (ECP), activate mast cells during inflammation; however the mechanism responsible for this activity is poorly understood. We found that some theoretical tryptase-digested fragments of MBP and ECP induced degranulation of human cord blood-derived mast cells (HCMCs). The spectrum of activities of these peptides in HCMCs coincided with intracellular Ca²⁺ mobilization activities in Mas-related G-protein coupled receptor family member X2 (MRGPRX2)-expressing HEK293 cells. Two peptides corresponding to MBP residues 99–110 (MBP (99–110)) and ECP residues 29–45 (ECP (29–45)), respectively, induced degranulation of HCMCs and intracellular Ca²⁺ mobilization in MRGPRX2-expressing HEK293 cells in a concentration-dependent manner. Stimulation with MBP (99–110) or ECP (29–45) induced the production of prostaglandin D2 by HCMCs. The activities of MBP (99–110) and ECP (29–45) in both HCMCs and MRGPRX2-expressing HEK293 cells were inhibited by MRGPRX2-specific antagonists. In conclusion, these results indicated that MBP and ECP fragments activate HCMCs, and it may occur via MRGPRX2. Our findings suggest that tryptase-digested fragments of eosinophil cationic proteins acting via the MRGPRX2 pathway may further our understanding of mast cell/eosinophil communication.     

In vitro killing of microfilariae of Brugia pahangi and Brugia malayi by eosinophil granule proteins

Eosinophil infiltration and degranulation around the tissue-invasive stages of several species of helminths have been observed. Release of eosinophil granule contents upon the worms is supported by localization of two of the major granule proteins, major basic protein (MBP) and eosinophil peroxidase (EPO), on and around species of trematodes, nematodes, and cestodes. In the case of filarial worms, MBP is deposited on degenerating microfilariae (mf) of Onchocerca volvulus. Here, we performed in vitro assays of the toxicity of four purified eosinophil granule proteins, namely, MBP, EPO, eosinophil cationic protein (ECP), and eosinophil-derived neurotoxin (EDN), for the mf of Brugia pahangi and Brugia malayi. MBP, ECP, and EDN killed these worms in a dose-related manner although relatively high concentrations of EDN were necessary. EPO, in the presence of a H2O2-generating system and a halide, was the most potent toxin on a molar basis; here, the most potent halide was I- followed by Br- and Cl-. Surprisingly, EPO in the absence of H2O2 killed mf at concentrations comparable to those required for MBP and ECP. The toxicity of EPO + H2O2 + halide was inhibited by heparin, catalase, or 1% BSA, whereas the toxicity of EPO alone was inhibited only by heparin. Heparin also inhibited killing by both MBP and ECP. Despite the homology of ECP with certain RNases, placental RNasin, an RNase inhibitor, was unable to inhibit ECP-mediated toxicity. These results indicate that all of the eosinophil granule proteins are toxic to mf and they support the hypothesis that eosinophil degranulation causes death of mf in vivo.     

Comparative toxicity of purified human eosinophil granule proteins for newborn larvae of Trichinella spiralis

Eosinophils have been implicated in both in vivo and in vitro destruction of helminths. One approach toward elucidating the role of the eosinophil in parasite killing has been to test the toxicity of purified eosinophil granule proteins for parasites in vitro. Previously, major basic protein (MBP) and eosinophil cationic protein (ECP) were shown to be toxic for schistosomules of Schistosoma mansoni, while eosinophil-derived neurotoxin (EDN) was only marginally so. We tested the toxicity of MBP, ECP, and EDN over a range of concentrations (0.006-5 X 10(-4) M) for newborn larvae of Trichinella spiralis. Our observations confirm previous reports of toxicity of mildly reduced and alkylated (R & A) MBP. At concentrations of 5 X 10(-5) M and above, R & A MBP killed 75% or more of the larvae within the first hour of culture. ECP was an effective toxin for these larvae after 3 hr of culture, and by 12 hr, dose-related toxicity was evident. After 24 hr, 100% of the larvae were killed at 5 X 10(-5) M ECP. EDN was much less toxic; after 12 hr, 90% of the larvae survived at concentrations of 1 X 10(4) M, while 5 X 10(-4) M EDN killed all the larvae. At the optimal toxic concentrations of 5 X 10(-5) M ECP and 5 X 10(-4) M EDN, kinetics of killing by these 2 proteins were essentially the same. Thus, on a molecular basis, both MBP and ECP appear to be potent helminthotoxins whereas EDN is much less so.     

Stimulation of basophil and rat mast cell histamine release by eosinophil granule-derived cationic proteins

Major basic protein (MBP), an arginine-rich basic polypeptide that constitutes the crystalloid core of the large specific eosinophil granule, has previously been shown to stimulate noncytolytic histamine release from human basophils and rat mast cells by an IgE-independent mechanism. Two additional basic polypeptides present in eosinophil granules, eosinophil cationic protein (ECP) and eosinophil-derived neurotoxin (EDN), were examined for similar activity in the present study. Acid-solubilized eosinophil granules were fractionated by chromatography on a Sephadex G-50 column. Incubation of basophil-containing human mononuclear cells with the individual column fractions demonstrated that histamine release occurred only with the fractions that contained MBP. The selectivity of the basophil response for MBP was confirmed by using equimolar concentrations of purified MBP, ECP, and EDN. In contrast, both MBP and ECP, but not EDN, stimulated histamine release from purified rat peritoneal mast cells. Reduction and alkylation of the MBP molecule diminished the response of human basophils to MBP but enhanced the potency of the molecule with rat mast cells. The distinct potency of MBP as a stimulus for histamine secretion from human basophils suggests that eosinophil release of MBP may be a specific event in the augmentation of immediate hypersensitivity reactions and other disorders characterized by eosinophilia.     

Both aromatic and cationic residues contribute to the membrane-lytic and bactericidal activity of eosinophil cationic protein

Eosinophil cationic protein (ECP) and eosinophil derived neurotoxin (EDN) are proteins of the ribonuclease A (RNase A) superfamily that have developed biological properties related to the function of eosinophils. ECP is a potent cytotoxic molecule, and although the mechanism is still unknown this cytotoxic activity has been associated with its highly cationic character. Using liposome vesicles as a model, we have demonstrated that ECP tends to disrupt preferentially acidic membranes. On the basis of structure analysis, ECP variants modified at basic and hydrophobic residues have been constructed. Changes in the leakage of liposome vesicles by these ECP variants have indicated the role of both aromatic and basic specific amino acids in cellular membrane disruption. This is the case with the two tryptophans at positions 10 and 35, but not phenylalanine 76, and the two arginines 101 and 104. The bactericidal activity of both native ECP and point-mutated variants, tested against Escherichia coli and Staphylococcus aureus, suggests that basic amino acids play, in addition to the effect on the disruption of the cellular membrane, other roles such as specific binding on the surface of the bacteria cell.     

The cytotoxic eosinophil cationic protein (ECP) has ribonuclease activity

The eosinophil cationic protein (ECP) is a specific cytotoxic constituent of granules. In this work we demonstrated that ECP has a ribonuclease activity. Purified ECP was resolved by ion exchange chromatography into subfractions, which all showed ribonuclease activity. Another eosinophil granule protein, EPX, identical with eosinophil-derived neurotoxin (EDN) had a 125-fold higher RNase activity than ECP. ECP may exert its cytotoxic effects on parasites and cells because of its extreme basicity alone or it may be internalized and act by degrading mRNA.     

Eosinophil cationic protein cDNA. Comparison with other toxic cationic proteins and ribonucleases

Human eosinophil granules contain several basic proteins including eosinophil cationic protein (ECP), eosinophil-derived neurotoxin (EDN) and major basic protein (MBP). ECP and MBP are potent helminthotoxins while EDN is less so. Both ECP and EDN possess neurotoxic and ribonuclease activities. A clone representing ECP mRNA was isolated from an eosinophil lambda ZAP cDNA library. The cDNA sequence codes for a preprotein of 160 amino acids and a protein of 133 amino acids, the amino terminus of which is identical to the known partial amino acid sequence of ECP. The ECP nucleotide sequence shows similarity to EDN, rat pancreatic ribonuclease, and human angiogenin; all are members of the ribonuclease gene superfamily. Although the deduced amino acid sequence of ECP shares identical active site and substrate binding site residues with EDN, angiogenin, and human pancreatic ribonuclease, the ribonuclease activity of ECP is 50 to 100 times less than that of EDN possibly because of the lack of a positively charged residue at human pancreatic ribonuclease position 122. The calculated isoelectric point (10.8), electronic charge (14.5), and cationic charge distribution of ECP are different from those of EDN but similar to those of MBP, which may account in part for the greater helminthotoxic activity of ECP when compared to EDN. These data suggest that ECP and EDN are derived from a common ancestral ribonuclease gene and that ECP has evolved into a potent helminthotoxin similar in some respects to MBP, while losing much of its ribonuclease activity.     

Effect of lysozyme or modified lysozyme fragments on DNA and RNA synthesis and membrane permeability of Escherichia coli

Previously we have shown that chicken egg white lysozyme, an efficient bactericidal agent, affects both gram-positive and gram-negative bacteria independently of its muramidase activity. More recently we reported that the digestion of lysozyme by clostripain yielded a pentadecapeptide, IVSDGNGMNAWVAWR (amino acid 98-112 of chicken egg white lysozyme), with moderate bactericidal activity but without muramidase activity. On the basis of this amino acid sequence three polypeptides, in which asparagine 106 was replaced by arginine (IVSDGNGMRAWVAWR, RAWVAWR, RWVAWR), were synthesized which showed to be strongly bactericidal. To elucidate the mechanisms of action of lysozyme and of the modified antimicrobial polypeptides Escherichia coli strain ML-35p was used. It is an ideal organism to study the outer and the inner membrane permeabilization since it is cryptic for periplasmic beta-lactamase and cytoplasmic beta-galactosidase unless the outer or inner membrane becomes damaged. For the first time we present evidence that lysozyme inhibits DNA and RNA synthesis and in contrast to the present view is able to damage the outer membrane of Escherichia coli. Blockage of macromolecular synthesis, outer membrane damage and inner membrane permeabilization bring about bacterial death. Ultrastructural studies indicate that lysozyme does not affect bacterial morphology but impairs stability of the organism. The bactericidal polypeptides derived from lysozyme block at first the synthesis of DNA and RNA which is followed by an increase of the outer membrane permeabilization causing the bacterial death. Inner membrane permeabilization, caused by RAWVAWR and RWVAWR, follows after the blockage of macromolecular synthesis and outer membrane damage, indicating that inner membrane permeabilization is not the deadly event. Escherichia coli bacteria killed by the substituted bactericidal polypeptides appeared, by electron microscopy, with a condensed cytoplasm and undulated bacterial membrane. So the action of lysozyme and its derived peptides is not identical.     

Localization of the guinea pig eosinophil major basic protein to the core of the granule

The localization of the guinea pig eosinophil major basic protein (MBP) within the cell was investigated by the use of immunoelectron microscopy and by isolation of the granule crystalloids. First, by immunoperoxidase electron microscopy, we found that the MBP of eosinophil granules is contained within the crystalloid core of the granule. Specific staining of cores was present when rabbit antiserum to MBP was used as the first stage antibody in a double antibody staining procedure, whereas staining was not seen when normal rabbit serum was used as the first stage antibody. Second, crystalloids were isolated from eosinophil granules by disruption in 0.1% Triton X-100 and centrifugation through a cushion of 50% sucrose. Highly purified core preparations yielded essentially a single band when analyzed by electrophoresis on polyacrylamide gels containing 1% sodium dodecyl sulfate (SDS). The E1%1cm of the core protein was 26.8 +/- 1.0 (X +/- SEM); the E1%1cm for the MBP was 26.3. The core protein could not be distinguished from the MBP by radioimmunoassay (RIA) and essentially all of the protein in the core preparations could be accounted for as MBP. The results indicate that the MBP is contained in the core of the guinea pig eosinophil granule and that it is probably the only protein present in the core.     

Mapping the eosinophil cationic protein antimicrobial activity by chemical and enzymatic cleavage

The eosinophil cationic protein (ECP) is a human antimicrobial protein involved in the host immune defense that belongs to the pancreatic RNase A family. ECP displays a wide range of antipathogen activities. The protein is highly cationic and its bactericidal activity is dependant on both cationic and hydrophobic surface exposed residues. Previous studies on ECP by site-directed mutagenesis indicated that the RNase activity is not essential for its bactericidal activity. To further understand the ECP bactericidal mechanism, we have applied enzymatic and chemical limited cleavage to search for active sequence determinants.Following a search for potential peptidases we selected the Lys-endoproteinase, which cleaves the ECP polypeptide at the carboxyl side of its unique Lys residue, releasing the N-terminal fragment (0-38).Chemical digestion using cyanogen bromide released several complementary peptides at the protein N-terminus. Interestingly, ECP treatment with cyanogen bromide represents a new example of selective chemical cleavage at the carboxyl side of not only Met but also Trp residues. Recombinant ECP was denatured and carboxyamidomethylated prior to enzymatic and chemical cleavage. Irreversible denaturation abolishes the protein bactericidal activity.The characterization of the digestion products by both enzymatic and chemical approaches identifies a region at the protein N-terminus, from residues 11 to 35, that retains the bactericidal activity. The most active fragment, ECP(0-38), is further compared to ECP derived synthetic peptides. The region includes previously identified stretches related to lipopolysaccharide binding and bacteria agglutination. The results contribute to define the shortest ECP minimized version that would retain its antimicrobial properties. The data suggest that the antimicrobial RNase can provide a scaffold for the selective release of cytotoxic peptides.     

Structural determinants of the eosinophil cationic protein antimicrobial activity

Abstract Antimicrobial RNases are small cationic proteins belonging to the vertebrate RNase A superfamily and endowed with a wide range of antipathogen activities. Vertebrate RNases, while sharing the active site architecture, are found to display a variety of noncatalytical biological properties, providing an excellent example of multitask proteins. The antibacterial activity of distant related RNases suggested that the family evolved from an ancestral host-defence function. The review provides a structural insight into antimicrobial RNases, taking as a reference the human RNase 3, also named eosinophil cationic protein (ECP). A particular high binding affinity against bacterial wall structures mediates the protein action. In particular, the interaction with the lipopolysaccharides at the Gram-negative outer membrane correlates with the protein antimicrobial and specific cell agglutinating activity. Although a direct mechanical action at the bacteria wall seems to be sufficient to trigger bacterial death, a potential intracellular target cannot be discarded. Indeed, the cationic clusters at the protein surface may serve both to interact with nucleic acids and cell surface heterosaccharides. Sequence determinants for ECP activity were screened by prediction tools, proteolysis and peptide synthesis. Docking results are complementing the structural analysis to delineate the protein anchoring sites for anionic targets of biological significance.     

Escherichia coli signal peptides direct inefficient secretion of an outer membrane protein (OmpA) and periplasmic proteins (maltose-binding protein, ribose-binding protein, and alkaline phosphatase) in Bacillus subtilis.

Signal peptides of gram-positive exoproteins generally carry a higher net positive charge at their amino termini (N regions) and have longer hydrophobic cores (h regions) and carboxy termini (C regions) than do signal peptides of Escherichia coli envelope proteins. To determine if these differences are functionally significant, the ability of Bacillus subtilis to secrete four different E. coli envelope proteins was tested. A pulse-chase analysis demonstrated that the periplasmic maltose-binding protein (MBP), ribose-binding protein (RBP), alkaline phosphatase (PhoA), and outer membrane protein OmpA were only inefficiently secreted. Inefficient secretion could be ascribed largely to properties of the homologous signal peptides, since replacing them with the B. amyloliquefaciens alkaline protease signal peptide resulted in significant increases in both the rate and extent of export. The relative efficiency with which the native precursors were secreted (OmpA >> RBP > MBP > PhoA) was most closely correlated with the overall hydrophobicity of their h regions. This correlation was strengthened by the observation that the B. amyloliquefaciens levansucrase signal peptide, whose h region has an overall hydrophobicity similar to that of E. coli signal peptides, was able to direct secretion of only modest levels of MBP and OmpA. These results imply that there are differences between the secretion machineries of B. subtilis and E. coli and demonstrate that the outer membrane protein OmpA can be translocated across the cytoplasmic membrane of B. subtilis.     

Ribonuclease activity associated with human eosinophil-derived neurotoxin and eosinophil cationic protein

The eosinophil granule contains a series of basic proteins, including major basic protein, eosinophil peroxidase, eosinophil-derived neurotoxin (EDN), and eosinophil cationic protein (ECP). Both EDN and ECP are neurotoxins and helminthotoxins. Comparison of the partial N-terminal amino acid sequences of EDN and ECP showed 67% identity; surprisingly, they also showed structural homology to pancreatic ribonuclease (RNase). Therefore, we determined whether EDN and ECP possess RNase enzymatic activity. By spectrophotometric assay of acid soluble nucleotides formed from yeast RNA, purified EDN showed RNase activity similar to bovine pancreatic RNase, whereas ECP was 50 to 100 times less active. The RNase activity associated with ECP was not significantly inhibited after exposure of ECP to polyclonal or monoclonal antibody to EDN. These results indicate that EDN and ECP both possess RNase activity, the RNase activity of EDN and ECP is specific, and EDN and ECP have maintained not only structural but also functional homology to pancreatic RNase.     

Analysing the eosinophil cationic protein - a clue to the function of the eosinophil granulocyte

Eosinophil granulocytes reside in respiratory mucosa including lungs, in the gastro-intestinal tract, and in lymphocyte associated organs, the thymus, lymph nodes and the spleen. In parasitic infections, atopic diseases such as atopic dermatitis and asthma, the numbers of the circulating eosinophils are frequently elevated. In conditions such as Hypereosinophilic Syndrome (HES) circulating eosinophil levels are even further raised. Although, eosinophils were identified more than hundred years ago, their roles in homeostasis and in disease still remain unclear. The most prominent feature of the eosinophils are their large secondary granules, each containing four basic proteins, the best known being the eosinophil cationic protein (ECP). This protein has been developed as a marker for eosinophilic disease and quantified in biological fluids including serum, bronchoalveolar lavage and nasal secretions. Elevated ECP levels are found in T helper lymphocyte type 2 (atopic) diseases such as allergic asthma and allergic rhinitis but also occasionally in other diseases such as bacterial sinusitis. ECP is a ribonuclease which has been attributed with cytotoxic, neurotoxic, fibrosis promoting and immune-regulatory functions. ECP regulates mucosal and immune cells and may directly act against helminth, bacterial and viral infections. The levels of ECP measured in disease in combination with the catalogue of known functions of the protein and its polymorphisms presented here will build a foundation for further speculations of the role of ECP, and ultimately the role of the eosinophil.     

The biosynthesis of neutrophil and eosinophil granule proteins

Composition of azurophil and specific granules from human polymorphonuclear neutrophils and granules from eosinophils is presented. Biosynthesis of the granule proteins is discussed in detail with particular emphasis on neutrophil myeloperoxidase (MPO) and eosinophil cationic protein (ECP).     

Allele-specific malE mutations that restore interactions between maltose-binding protein and the inner-membrane components of the maltose transport system

Active accumulation of maltose and maltodextrins by Escherichia coli depends on an outer-membrane protein. LamB, a periplasmic maltose-binding protein (MalE, MBP) and three inner-membrane proteins, MalF, MalG and MalK. MalF and MalG are integral transmembrane proteins, while MalK is associated with the inner aspect of the cytoplasmic membrane via an interaction with MalG. Previously we have shown that MBP is essential for movement of maltose across the inner membrane. We have taken advantage of malF and malG mutants in which MBP interacts improperly with the membrane proteins. We describe the properties of malE mutations in which a proper interaction between MBP and defective MalF and MalG proteins has been restored. We found that these malE suppressor mutations are able to restore transport activity in an allele-specific manner. That is, a given malE mutation restores transport activity to different extents in different malF and malG mutants. Since both malF and malG mutations could be suppressed by allele-specific malE suppressors, we propose that, in wild-type bacteria, MBP interacts with sites on both MalF and MalG during active transport. The locations of different malE suppressor mutations indicate specific regions on MBP that are important for interacting with MalF and MalG.