Epitope Insertions Define Functional and Topological Features of the Escherichia coli Ferric Enterobactin Receptor

doi:10.1074/jbc.270.6.2483

HOME

HELP

FEEDBACK

SUBSCRIPTIONS

Volume 270, Number 6, Issue of February 10, 1995 pp. 2483-2488
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.

Epitope Insertions Define Functional and Topological Features of the Escherichia coli Ferric Enterobactin Receptor (*)

(Received for publication, June 9, 1994; and in revised form, October 18, 1994)

Sandra K. Armstrong (1), (§), Mark A. McIntosh (2)(¶)

From the (1)Department of Microbiology and Immunology, East Carolina University School of Medicine, Greenville, North Carolina 27858-4354 and the (2)Department of Molecular Microbiology and Immunology, University of Missouri School of Medicine, Columbia, Missouri 65212

ABSTRACT

INTRODUCTION

EXPERIMENTAL PROCEDURES

RESULTS

DISCUSSION

FOOTNOTES

ACKNOWLEDGEMENTS

REFERENCES

ABSTRACT

The outer membrane protein FepA of Escherichia coli is the receptor for the ferric enterobactin siderophore complex and colicins B and D. A foreign antigenic determinant inserted into selected FepA sites allowed mutational analysis of receptor function and in situ immunological tracking of specific protein domains with respect to the bacterial cell compartment. Immunoblot analysis of bacterial proteins using an epitope-specific antibody detected the peptide determinant in the receptor fusions. The impact of the insertions on FepA function was examined by ferric enterobactin-mediated iron uptake experiments and colicin sensitivity tests. In all cases, FepA retained biological activity despite introduction of the foreign sequence. To further develop the topological model of FepA, the peptide-specific antibody was used to localize epitope-carrying FepA domains in intact bacterial cells and their isolated membranes. One epitope resided in a region on the exterior of the cell, at the surface of the FepA protein, while other epitopes appeared to be localized to the periplasm or within the outer membrane.

INTRODUCTION

The outer membrane of a Gram-negative bacterium is a permeability barrier controlling passage of solutes to the periplasmic space surrounding the cytoplasmic membrane. Many substances traverse the outer membrane through nonspecific porin channels. Other porins such as LamB of Escherichia coli form substrate-specific outer membrane channels(1) . Vitamin B and iron-sequestering microbial siderophores are presumed too large or sterically unsuitable for passage through E. coli porins, thus necessitating expression of ligand-specific outer and inner membrane transport proteins which act in conjunction with accessory proteins TonB and ExbB (2, 3, 4, 5) .

The E. coli outer membrane protein FepA is the high affinity receptor for the siderophore ferric enterobactin and the antibacterial colicins B and D(2) . Although the best characterized outer membrane transporters function as porin diffusion channels, the specific mode of FepA function is unknown. Nutrients pass through porins by simple or facilitated diffusion, yet FepA has the capacity to concentrate ferric enterobactin in the periplasm against a gradient(6) . Because the outer membrane supplies no membrane potential as an energy source for such transport, it is postulated that TonB, with ExbB, transduces energy by an unknown mechanism from the cytoplasmic membrane(7, 8) . In support of this concept, one region of FepA and other TonB-dependent outer membrane receptors, the ``TonB box'', has been implicated through genetic studies as a contact point for TonB(9, 10, 11) . Furthermore, chemical cross-linking experiments provided evidence for physical interaction of TonB and FepA(12) . FepA has been proposed to function as a ligand-specific gated porin channel(13) . The model invokes the existence of ligand-specific external domains that occlude the FepA channel; upon binding of the ligand to FepA, TonB is stimulated to allow ligand passage through the channel and to the periplasm.

As the topology of FepA in the outer membrane is key to understanding its function, efforts are directed toward resolving its native structure. Because outer membrane proteins generally lack the characteristic alpha -helical regions predicted to span a lipid bilayer (14) , the topology of these proteins must be deduced from a variety of experimental approaches. Monoclonal antibodies (mAbs) ()to FepA linear epitopes have been used to map seven cell surface-exposed regions of the receptor, two of which appeared to be involved in ligand binding(15) . A linker insertion mutational study defined regions of FepA required for activity of all three ligands, as well as two domains required for colicin function but not ferric enterobactin uptake(16) .

To develop the model of FepA as an integral component of a prototypic TonB-dependent nutrient uptake system, we have constructed FepA-epitope fusions to determine the subcellular location of specific domains and to identify epitope insertion sites that affect FepA receptor activity.

EXPERIMENTAL PROCEDURES

Bacterial Strains, Growth Conditions, and General Genetic Techniques

E. coli strain RWB18-60 (F

, thi proC leuB trpE entA fepA Delta

recA) (16) was the host for fepA plasmids. Bacteria were grown at 37 °C in Luria-Bertani medium (LB) (17) or on LB agar, unless otherwise indicated. Ampicillin was used for plasmid selection at a final concentration of 100 µg/ml. Isolation of plasmid DNA and other standard molecular genetic techniques have been described(18) . Synthetic oligonucleotide primers for sequencing and fusion construction were provided by the University of Missouri DNA Core Facility. Nucleotide sequencing was performed as detailed previously (19) ; the published fepA sequence (20) was used for comparison.

Construction of Epitope Insertions in fepA

The parent plasmid, pITS549, contains a 2.5-kilobase pair SspI-StuI fragment encoding the E. coli K-12 fepA gene(16) . As the upstream Fur-binding sequences are absent, iron deprivation is not required for expression of fepA. The fepA plasmid targets for insertion of oligonucleotides specifying the epitope used in this study have been described(16) . Each of the plasmids is a derivative of pITS549 and has a unique XhoI linker at one of six locations in the fepA coding region. Epitope fusions were created by digestion with XhoI and ligation with a double-stranded 30-mer oligonucleotide (Fig. 1), which specifies an epitope termed M2 (also referred to as Flag(TM)), designed by the Immunex Corp. (Seattle, WA; now licensed to International Biotechnologies, Inc., New Haven, CT) (21) , flanked by XhoI half-sites for insertion at preexisting XhoI linker sites(16) . Transformants carrying the M2 insertions were identified by colony hybridization(18) . The location, sequence junctions, and orientation of each insertion were confirmed by nucleotide sequencing.

Figure 1: Map of FepA epitope insertions. The mature FepA protein is depicted as the open bar; mutant alleles are denoted above the numbers indicating the amino acid residue after which the M2 epitope was inserted into preexisting XhoI linker sites. The oligonucleotide specifying the M2 epitope is shown.

Colicin Sensitivity Tests

Preparation of colicins B and D has been described(16) . Colicin sensitivity was determined by spotting dilutions of colicin onto lawns of test bacteria on LB agar. The reciprocal of the last dilution resulting in inhibition of growth was the titer defining colicin sensitivity. Values are reported as the percent of the titers of cells carrying wild-type fepA on pITS549.

Iron Transport Experiments

Ferric enterobactin uptake was evaluated by monitoring

Fe accumulation using the method (16) modified from Langman et al.(22) . The rate of uptake was determined by linear regression analysis using the least squares method. The reported values are representative of at least three experiments. E. coli RWB18-60 containing either the vector pGEM3Z or pITS549 served as the negative and positive controls, respectively.

SDS-Polyacrylamide Gel Electrophoresis and Immunoblotting

Cellular proteins (

organisms) were denatured in SDS and treated at 100 °C for 7 min prior to fractionation by SDS-polyacrylamide gel electrophoresis on 7.5-20% acrylamide gradient gels and immunoblotting as described (16) . The mAb recognizing the M2 epitope (21) was from Immunex Corp. (Seattle, WA) and was supplied at 1.9 mg/ml and used at a 1:1000 dilution. The FepA-specific polyclonal mouse antiserum has been described(16) .

Intact Cell Immunodots

Logarithmic phase cells cultured in LB plus ampicillin were harvested, washed with 10 mM NaCl, and resuspended to an OD

of 0.3. Dilutions of cells were applied to nitrocellulose and air-dried. The nitrocellulose was rewetted in 0.9% NaCl, 10 mM Tris-HCl (pH 7.4), blocked with 3% BSA (in the same buffer) for 1 h at 37 °C, and reacted with a 1:1000 dilution of anti-M2 mAb. After incubation for 2 h at 37 °C, the nitrocellulose was washed in the Tris-saline buffer before incubation with a secondary antibody horseradish peroxidase conjugate and development by conventional techniques(16) .

Cell Membrane Immunodots

Logarithmic phase bacteria grown in LB plus ampicillin were washed in 0.01 M HEPES (pH 7.4) and resuspended in the same buffer prior to lysis using a French pressure cell. The lysate was centrifuged at 3,000 times

g and the supernatant retained. The insoluble fraction containing both inner and outer membranes was obtained by centrifugation of the supernatant at 100,000 times

g and resuspension in 0.01 M HEPES (pH 7.4). Equivalent quantities of protein samples were diluted and applied to nitrocellulose. In some experiments, membrane samples were either pretreated at 100 °C for 5 min, then cooled on ice or diluted in 10 mM Tris-HCl (pH 8.0), 1 mM EDTA before testing. The nitrocellulose was air-dried, blocked with BSA, reacted with anti-M2 mAb, and treated as the intact cell immunodots described above.

Flow Cytometry

A modification of the technique of Murphy et al.(15) was used to monitor cell surface-exposure of the M2 epitope. Mid-logarithmic phase bacteria were washed in PBS (pH 7.4), and 1 times

cells incubated at room temperature for 45 min with the anti-M2 mAb (a 1:250 dilution in 1% BSA in PBS (PBS-BSA)). The cells were washed in PBS-BSA and incubated in fluorescein isothiocyanate-conjugated goat antibody to murine IgG for 45 min. After a final wash in PBS-BSA, the bacteria were resuspended in the same buffered solution. The cells were analyzed by the East Carolina University Research Flow Cytometry Core Facility on a FACStar Plus flow cytometer (Becton Dickinson, San Jose, CA). For each sample, from 6,500 to 10,000 events were measured and the reported results are the averages of three separate experiments.

Computer Analyses

Sequences were analyzed using programs included in the program package MacPROT (EMBNet Bioinformation Resources Network, European Molecular Biology Laboratory, Heidelberg, Germany). The program PLOT.A/GGR was used to predict protein secondary structure and is based on the algorithm of Gibrat et al. (23) Formation of transmembrane helices was predicted by the PLOT.A/TMH program(24) .

RESULTS

Construction of Epitope Fusions

A previous study localized FepA functional domains by linker insertion mutagenesis(16) . Each linker mutation consisted of a 6-base pair XhoI site inserted at a specific position in fepA, resulting in the addition of amino acids Leu-Glu to the FepA polypeptide. The linker-bearing fepA plasmids HX1, HX2, VX1, RX3, VX2, and HX4 were used in the present study to create epitope fusions at each unique XhoI site (Fig. 1). An oligonucleotide of 30 base pairs encoding the M2 epitope (21) with XhoI-compatible ends allowed ligation of the epitope cassette to the sites supplied by the original linker mutations, resulting in the addition of 10 novel amino acids to the site containing Leu-Glu contributed previously by the linker insertion. The M2 residues Tyr-Lys-Asp-Asp-X-Asp comprise the epitope recognized by the mAb used in this protein tracking system. Nucleotide sequencing of each of the fusion plasmids confirmed the in-frame insertion of a single epitope cassette at the appropriate site.

Expression of FepA Fusion Proteins

Expression of FepA

To determine if the foreign insertion prevented FepA expression, immunoblot analysis was performed on bacteria carrying the wild-type fepA in multicopy and mutated M2 epitope fusion derivatives. With a FepA-specific antiserum, the receptor protein was detected in all M2 mutants (Fig. 2A). As anticipated, the insertion of twelve foreign residues decreased the electrophoretic mobilities of all FepA-epitope hybrids. Cells carrying the HX4F allele appeared to produce less total immunoreactive FepA than the other mutants when analyzed with various FepA-specific polyclonal and monoclonal antisera (Fig. 2A and data not shown). Insertion of the epitope near the amino terminus appeared to destabilize FepA, as HX1F, HX2F, and to some extent, VX1F exhibited faster migrating immunoreactive bands, possibly representing degradation products.

Figure 2: Immunoblot analysis of RWB18-60 carrying fepA fusion alleles. Panel A, detection of the FepA protein. Cells were solubilized and immunoblotted using an anti-FepA polyclonal antiserum(16) . PanelB, expression of the M2 epitope on FepA hybrids. Cells were immunoblotted using the anti-M2 mAb. Lane C, cells carrying wild-type fepA on pITS549; lane 1, HX1F; lane 2, HX2F; lane 3, VX1F; lane 4, RX3F; lane 5, VX2F; lane 6, HX4F. Mature wild-type FepA (panelA, lane C) migrates with a molecular weight of 80,000. Molecular size standards are shown in kilodaltons: phosphorylase b, 97; BSA, 66; ovalbumin, 45.

Expression of the M2 Epitope in FepA

Immunoblot analysis using the M2-specific mAb demonstrated the expression of the FepA-borne epitope in all of the mutants (Fig. 2B). The pattern of anti-epitope reactivity of the mutant proteins was virtually identical to that obtained with the FepA-specific antibody. Neither the original Leu-Glu insertion mutants (data not shown), wild-type FepA, nor any other E. coli protein reacted with the M2 epitope-specific antibody.

Biological Function of FepA Hybrids

It was possible that some FepA-M2 hybrids were altered in structure such that they were no longer biologically active as receptors for ferric enterobactin and colicins B and D. Functional characterization experiments indicated that the hybrid FepA molecules of all six fusion types were localized to the outer membrane and retained a conformation that permitted interaction with the ligands to allow their uptake.

The original XhoI linker insertion mutants were characterized previously with respect to FepA receptor function (16) (Table 1). Depending on the location of the Leu-Glu linker insertion, sensitivity to colicins B and D was either dramatically reduced (mutants HX1, HX2, VX1, and RX3) or remained at wild-type or near wild-type levels (mutants VX2 and HX4). Introduction of the DNA encoding the M2 epitope into each of the linker mutation sites caused a decrease in colicin sensitivity in some cases (HX1F, VX1F, VX2F, and HX4F), yet exerted no additional effect on the colicin receptor function of others (HX2F and RX3F) (Table 1). Although the colicin D sensitivity of HX1F was very weak, this mutant was nevertheless sensitive when undiluted colicin D preparations were tested. fepA null mutants were insensitive when exposed to the same concentrations of colicins B and D. VX2F and HX4F, whose linker mutant parents retained 38-100% of colicin function, lost some susceptibility to the colicins but remained the most sensitive of the epitope fusion mutants.

Introduction of Leu-Glu into FepA after residues 55 (HX1), 142 (HX2), or 324 (RX3) significantly diminished the ability of E. coli to transport ferric enterobactin(16) . Insertion of the M2 epitope into those linker sites exerted little additional effect on enterobactin-mediated iron uptake of HX1F, whereas the insertion in HX2F critically disrupted iron transport and appeared to cause a modest increase in transport for RX3F. The M2 insertion in mutant VX1F reduced the ferric enterobactin transport activity of the original VX1 mutant to 46% of the wild-type value, a decrease from the previously observed 81% level. VX2F and HX4F demonstrated moderate decreases in ferric enterobactin receptor function.

In Situ Localization of the Epitope on FepA

To exploit the M2 epitope as a probe for native FepA protein structure, experiments were performed to localize the epitope with respect to the cell compartments.

Cell Surface Exposure

Exposure of the M2 epitope on the cell surface was determined by subjecting viable intact mid-logarithmic growth phase bacteria to immune recognition by the M2-specific mAb. In one series of experiments, intact cells were exposed to the epitope-specific antibody in dot immunoblots (Fig. 3). Bacteria expressing fusion RX3F consistently exhibited the strongest in situ reactivity with the antibody, indicating the epitope was located on the cell surface and at the surface of the FepA molecule. This surface exposed RX3F M2 epitope was positioned at FepA residue 324, yet at nearby residues 339 (VX2F) and 359 (HX4F), the same epitope was virtually undetected at the cell surface. Hybrids HX1F, HX2F, and VX1F displayed low but variable levels of cell surface reactivity with the mAb, which suggested a poorly accessible epitope or background levels of cell lysis. To obtain quantitative data, flow cytometry was performed on intact mid-logarithmic growth phase bacteria exposed to the epitope-specific antibody (Table 1, Fig. 4). Again, cells expressing RX3F displayed the strongest surface reactivity, and fusions HX1F, HX2F, VX1F, and VX2F showed very low levels of reactivity with the intact bacteria. Cells carrying the HX4F fusion consistently demonstrated virtually no reactivity, a result that may relate to the poor expression noted previously. Intact cells carrying wild-type fepA did not react with the M2-specific antibody in any experiments.

Figure 3: Cell surface expression of the M2 epitope on intact cells. RWB18-60 cells carrying different fepA fusions were washed and dilutions representing equivalent cell numbers were applied to nitrocellulose as detailed under ``Experimental Procedures.'' The bound cells were reacted with the M2-specific mAb and processed as immunoblots. RowC, cells containing wild-type fepA on pITS549; rowsHX1F-HX4F denote the appropriate fepA::M2 fusion-bearing cells.

Figure 4: Flow cytometry profiles of intact RWB18-60 cells expressing FepA-epitope fusions. Bacteria were stained by indirect immunofluorescence with the M2 epitope-specific mAb as described under ``Experimental Procedures.'' x axis, log fluorescence; y axis, counts per channel. fepA alleles are indicated. RWB18-60 carrying pITS549 (549), lacking the epitope insertion, was the negative control. Three typical epitope fusion profiles are shown; cells carrying HX2F, VX1F, and HX4F fusions demonstrated cytometric profiles similar to those of HX1F and VX2F. Quantitative results of these experiments are shown in Table 1.

Exposure at the Periplasmic Face of the Outer Membrane

Lack of surface recognition by anti-M2 for epitope fusions other than RX3F suggested that the epitopes in the other fusion locations were: 1) on the cell surface but antibody-inaccessible, possibly by virtue of masking FepA domains or other outer membrane molecules, 2) located in the periplasm, or 3) embedded in the outer membrane.

Incubation of isolated native outer and inner membranes with the anti-M2 antibody in dot immunoblots demonstrated strong reactivity for HX1F as well as RX3F (Fig. 5A). Comparatively less reactivity was observed for HX2F, VX1F, and VX2F, while minimal signal was detected for HX4F. Since membranes of HX1F reacted strongly, relative to those of RX3F, and intact cells reacted weakly compared with RX3F cells, the epitope may be displayed at the surface of the FepA protein but exposed to the periplasm. Likewise, the increase in immunoreactivity for isolated membranes over whole cells implies that the HX2F, VX1F, and VX2F epitopes also may be located on FepA periplasmic domains. Densitometric comparisons of several intact cell dot blots with membrane dot blots (data not shown) using the reactivity of the RX3F samples as a reference point confirmed the increased antibody reactivities of membranes bearing the HX1F, HX2F, VX1F, and VX2F hybrids.

Figure 5: Detection of the M2 epitope on FepA in isolated bacterial membranes. RWB18-60 cellular fractions containing both cytoplasmic and outer membranes were spotted onto nitrocellulose and immunoblotted using the M2-specific mAb. For all samples, equal quantities of protein were used for each of two dilutions. Rows indicate the fepA::M2 fusion membrane sample. PanelA, membranes were applied without pretreatment; panelB, membranes were pretreated at 100 °C; panelC, membranes were suspended in 10 mM Tris-HCl, 1 mM EDTA (pH 8.0) prior to application to nitrocellulose.

Denaturing treatment of the same isolated membranes at 100 °C either slightly decreased reactivity with the mAb or elicited no change (Fig. 5B). One exception, the epitope of HX4F, appeared to be unmasked by heat treatment, resulting in a small but visible increase in antibody reactivity. These results suggest that for all fusions except HX4F, the epitope location is sensitive to heat denaturation, while such treatment unmasks or enhances antibody access to the epitope positioned at amino acid 359. The M2 epitope itself is heat-stable, as boiled and denatured samples were readily detectable in Western blots (Fig. 2B). Incubation of the membranes in buffer containing 1 mM EDTA enhanced antibody reactivity with all of the FepA fusions except HX4F (Fig. 5C). As EDTA would chelate divalent cations stabilizing lipopolysaccharide and possibly other membrane proteins or FepA itself, such treatment may reveal the M2 epitope by altering the conformational states of such membrane components.

DISCUSSION

Reporter enzyme fusions to study outer membrane protein topology and function may not yield accurate information relating to the native molecule, as the hybrid protein might not exist in a conformation for proper function or translocation to the outer membrane(25) . To circumvent these difficulties, small reporter epitopes have been employed to study the structure and function of a bacterial porin protein(25) . LamB was found to tolerate variously located epitope fusions without significant loss of function, defining permissive sites of insertion. Our analysis of FepA used similar reporter epitope technology for domain localization and identification of permissive insertion sites.

A previous linker mutation study examined the effect of Leu-Glu insertions on FepA function(16) . Insertions after amino acids 55 (mutant HX1), 142 (HX2), and 324 (RX3) dramatically decreased receptor activity for colicins B and D and ferric enterobactin, whereas the same Leu-Glu insertions after residues 339 (VX2) and 359 (HX4) had little or no effect. Linker mutation VX1 (after residue 204) was unique in that it decreased colicin B and D function but had minimal impact on enterobactin-mediated iron uptake, providing the first evidence that the receptor functions were separable. In this report, insertion of the M2 epitope cassette into these linker mutation sites generally caused modest or no obvious additional inactivation of receptor function. The decrease in FepA function for cells carrying the HX4F fusion may simply result from the observed diminished receptor levels. Although it might be expected that the charged amino acids contributed by the M2 epitope might promote the formation of a surface-seeking domain, they did not further alter FepA conformation so as to significantly impair function beyond the effect of the original linker insertion.

A current FepA topological model (13, 15) is schematically depicted in Fig. 6A, while a revamped model based on results from the present study is shown in Fig. 6B. Computer analysis of protein secondary structure (24, 25, 26) of wild-type and mutant FepA protein sequences predicted that a surface-seeking region encompassing residue 55 is extended in the M2 fusion HX1F. The existing FepA topological model positions this region at the external cell surface(13, 15) . As the HX1F epitope was strongly reactive in isolated membranes but not intact cells, it is likely that this region is on the surface of the FepA protein but exposed to the periplasm. It is possible, however, that the membrane isolation procedure, although designed to be minimally disruptive, caused the HX1F epitope hidden at the cell surface to become unmasked and reactive with the M2-specific antibody. Although it is clear that the HX2F epitope (after residue 142) is not exposed at the surface of FepA on the cell exterior, the M2 localization experiments provided suggestive evidence for periplasmic location. The original FepA model predicted this region to be within the outer membrane bilayer. Computer analysis suggested that insertion of the epitope at this position causes loss of a surface-seeking domain and reduces the size of a transmembrane segment. The domain encompassing amino acid 204 containing the epitope VX1F is postulated to reside in a cell surface-exposed region of FepA, which is not involved in ligand binding(13, 15) . A mAb that recognizes this domain reacted only with intact bacteria of a rough lipopolysaccharide chemotype(13) . Because the M2 epitope was undetected at the surface of cells carrying the VX1F allele and the membrane dot immunoblots were inconclusive, this region of FepA may indeed be at the cell surface but masked by lipopolysaccharide O side chains. Computer predictions suggested no changes in local FepA globular conformation as a result of M2 insertion at VX1F residue 204. Analysis of RX3F (after residue 324) and VX2F (after residue 339) predicts no significant FepA structural changes, as the regions remain in the wild-type globular conformation. Probing bacteria expressing the fusion proteins with the M2-specific antibody demonstrated cell surface-exposure of the RX3F epitope on the exterior face of the FepA molecule. This result agrees with the existing FepA topological model, which shows the region at residue 324 in a cell surface-exposed loop predicted to be involved in ligand binding(13, 15) . The epitope insertion of VX2F, which is only 15 amino acids from that of RX3F, was not antibody-accessible at the cell surface to any significant degree. Because it was also not strongly reactive in isolated membranes where the antibody could interact with FepA regions on both sides of the outer membrane, it is likely that the VX2F epitope is masked. The previous FepA model places amino acid 339 within the outer membrane bilayer(13, 15) . Insertion of the M2 epitope after residue 359 (HX4F) is predicted to create a new transmembrane helical domain in a globular region previously positioned within the wild-type outer membrane. Consistent with both predictions, the M2 localization experiments indicated the epitope was embedded in the outer membrane or masked by domains that appeared to be denatured by high temperatures.

Figure 6: Topological models of FepA. ModelA was proposed previously(13, 15) . ModelB has been modified to include new information from this study. The positions of the M2 epitope insertions in mature FepA (723 amino acids in length) are indicated: 1, HX1F; 2, HX2F; 3, VX1F; 4, RX3F; 5, VX2F; 6, HX4F.

The protein region containing the epitope of fusion RX3F is in a cell surface conformation freely accessible to the M2-specific antibody. The hybrid receptor was expressed at wild-type levels, retained some receptor function, and was well tolerated by the cells. These characteristics make the XhoI linker site of RX3 an ideal candidate for heterologous epitope display. Epitope display systems using the E. coli LamB (26) and PhoE (27) outer membrane proteins and P fimbriae (28) as host molecules have proven effective at inducing epitope-specific immune responses. Such technology would be useful for oral/mucosal immunization with live attenuated bacteria. The FepA protein is immunogenic and is expressed by several species of enteric bacteria(29) . An especially appealing trait of FepA is its strong expression under low iron growth conditions. As the receptor is pivotal to iron uptake, the natural iron stress of the in vivo environment would ensure expression of an epitope-bearing hybrid FepA.

FOOTNOTES

*: This research was supported in part by Grant DHHS 1 R01 GM40565 from NIGMS, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§: Recipient of National Research Service Award DHHS 1 F32 AI 07610 from NIAID, National Institutes of Health.
¶: To whom correspondence and reprint requests should be addressed: Dept. of Molecular Microbiology and Immunology, M613 School of Medicine, University of Missouri, Columbia, MO 65212. Tel.: 314-882-4133.
(): The abbreviations used are: mAb, monoclonal antibody; BSA, bovine serum albumin; PBS, phosphate-buffered saline.

ACKNOWLEDGEMENTS

We thank Robert Kadner and Phillip Klebba for sharing ideas. We are grateful to Timothy Brickman for bringing the M2 epitope to our attention, for discussions, and computer expertise. We acknowledge John Warren for technical contributions.

REFERENCES

Luckey, M., and Nikaido, H. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 167-171 [Abstract/Free Full Text]
Pugsley, A. P., and Reeves, P. (1976) J. Bacteriol. 127, 218-228 [Abstract/Free Full Text]
Pugsley, A. P., and Reeves, P. (1976) J. Bacteriol. 126, 1052-1062 [Abstract/Free Full Text]
Pierce, J. R., and Earhart, C. F. (1986) J. Bacteriol. 166, 930-936 [Abstract/Free Full Text]
Shea, C. M., and McIntosh, M. A. (1991) Mol. Microbiol. 5, 1415-1428 [Medline] [Order article via Infotrieve]
Pugsley, A. P., and Reeves, P. (1977) J. Bacteriol. 130, 26-36 [Abstract/Free Full Text]
Kadner, R. J. (1990) Mol. Microbiol. 4, 2027-2033 [Medline] [Order article via Infotrieve]
Postle, K. (1990) Mol. Microbiol. 4, 2019-2025 [CrossRef][Medline] [Order article via Infotrieve]
Heller, K., and Kadner, R. J. (1985) J. Bacteriol. 161, 904-908 [Abstract/Free Full Text]
Schoffler, H., and Braun, V. (1989) Mol. & Gen. Genet. 217, 378-383
Sauer, M., Hantke, K., and Braun, V. (1990) Mol. Microbiol. 4, 427-437 [CrossRef][Medline] [Order article via Infotrieve]
Skare, J. T., Ahmer, B. M. M., Seachord, C. L., Darveau, R. P., and Postle, K. (1993) J. Biol. Chem. 268, 16302-16308 [Abstract/Free Full Text]
Rutz, J. M., Liu, J., Lyons, J. A., Goranson, J., Armstrong, S. K., McIntosh, M. A., Feix, J. B., and Klebba, P. E. (1992) Science 258, 471-475 [Abstract/Free Full Text]
Nikaido, H., and Saier, M. H., Jr. (1992) Science 258, 936-942 [Abstract/Free Full Text]
Murphy, C. K., Kalve, V. I., and Klebba, P. E. (1990) J. Bacteriol. 172, 2736-2746 [Abstract/Free Full Text]
Armstrong, S. K., Francis, C. L., and McIntosh, M. A. (1990) J. Biol. Chem. 265, 14536-14543 [Abstract/Free Full Text]
Miller, J. H. (1972) Experiments in Molecular Genetics , p. 433, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual , Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
Armstrong, S. K., Pettis, G. S., Forrester, L. J., and McIntosh, M. A. (1989) Mol. Microbiol. 3, 757-766 [CrossRef][Medline] [Order article via Infotrieve]
Lundrigan, M. D., and Kadner, R. J. (1986) J. Biol. Chem. 261, 10797-10801 [Abstract/Free Full Text]
Hopp, T. P., Prickett, K. S., Price, V. L., Libby, R. T., March, C. J., Cerretti, D. P., Urdal, D. L., and Conlon, P. J. (1988) Bio/Technology 6, 1204-1210 [CrossRef]
Langman, L., Young, I. G., Frost, G. E., Rosenberg, H., and Gibson, F. (1972) J. Bacteriol. 112, 1142-1149 [Abstract/Free Full Text]
Gibrat, J., Garnier, J., and Robson, B. (1987) J. Mol. Biol. 198, 425-443 [CrossRef][Medline] [Order article via Infotrieve]
Eisenberg, D., Schwarz, E., Komaromy, M., and Wall, R. (1984) J. Mol. Biol. 179, 125-142 [CrossRef][Medline] [Order article via Infotrieve]
Charbit, A., Ronco, J., Michel, V., Werts, C., and Hofnung, M. (1991) J. Bacteriol. 173, 262-275 [Abstract/Free Full Text]
Charbit, A., Sobczak, E., Michel, M., Molla, A., Tiollais, P., and Hofnung, M. (1987) J. Immunol. 139, 1658-1664 [Abstract]
Agterberg, M., Adriaanse, H., Lankhof, H., Meloen, R., and Tommassen, J. (1990) Vaccine 8, 85-91 [CrossRef][Medline] [Order article via Infotrieve]
van Die, I., van Oosterhout, J., van Megen, I., Bergmans, H., Hoekstra, W., Enger-Valk, B., Barteling, S., and Mooi, F. (1990) Mol. & Gen. Genet. 222, 297-303
Rutz, J. M., Abdullah, T., Singh, S. P., Kalve, V. I., and Klebba, P. E. (1991) J. Bacteriol. 173, 5964-5974 [Abstract/Free Full Text]