Identification and characterization of an iron-regulated hemopexin receptor in Haemophilus influenzae type b.

Heme can serve Haemophilus influenzae as a source of both essential porphyrin and iron. In extracellular mammalian body fluids neither free heme nor free iron is available, since they are tightly bound to hemopexin and transferrin, respectively. Since H. influenzae grows in the presence of iron-transferrin and heme-hemopexin and is known to express a saturable receptor for transferrin, we investigated the process by which this pathogen acquired heme from hemopexin for use as an iron source. The ability of human and rabbit hemopexin to donate heme as a source of iron to H. influenzae type b strains was demonstrated by plate bioassays. With a dot enzyme assay with biotinylated hemopexin as ligand, H. influenzae bound heme-hemopexin and apo-hemopexin following growth in iron-restricted, but not in iron-sufficient, medium. Competitive binding studies with heme-hemopexin and apo-hemopexin demonstrated saturability of binding. Neither heme, protoporphyrin IX, hemoglobin, nor transferrin blocked the binding of hemopexin to whole cells, demonstrating the specificity of binding. Treatment of whole H. influenzae cells with trypsin abolished binding. Taken together, these observations suggest that H. influenzae type b expresses an outer membrane protein(s) which acts as a receptor for hemopexin and which is regulated by the availability of iron in the growth medium. In iron-restricted media, H. influenzae 706705 and DL42 did not express the 100-kDa hemopexin-binding protein previously reported (M.S. Hanson, S.E. Pelzel, J. Latimer, U. Muller-Eberhard, and E.J. Hansen, Proc. Natl. Acad. Sci. USA 89:1973-1977, 1992). The putative iron-regulated hemopexin receptor was solubilized from cell envelopes of H. influenzae 706705, DL42, and Eagan with the detergent CHAPS (3-[(3-cholamidopropyl)-dimethyl-ammonio]-1-propanesulfonate) and isolated by affinity chromatography on heme-hemopexin-Sepharose 4B. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the proteins bound to the affinity resin revealed three proteins of 29, 38, and 57 kDa, of which the 57- and 29-kDa proteins bound hemopexin after Western blotting (immunoblotting). A monoclonal antibody to the 57-kDa hemopexin-binding protein of 706705 recognized a 57-kDa protein on Western blots of the cell envelope proteins of 706705, DL42, and Eagan; no reaction was observed with the 100-kDa hemopexin-binding protein of DL42. These data suggest that some H. influenzae strains possess at least two hemopexin receptors, the expression of which is determined by the prevailing growth environment.     

Protein sources of heme for Haemophilus influenzae.

Although Haemophilus influenzae requires heme for growth, the source of heme during invasive infections is not known. We compared heme, lactoperoxidase, catalase, cytochrome c, myoglobin, and hemoglobin as sources of heme for growth in defined media. The minimum concentration of heme permitting unrestricted growth of strain E1a, an H. influenzae type b isolate from cerebrospinal fluid, was 0.02 micrograms/ml. Using molar equivalents of heme as lactoperoxidase, catalase, cytochrome c, myoglobin, and hemoglobin, we determined that myoglobin and hemoglobin permitted unrestricted growth at this concentration. To determine the ability of host defenses to sequester heme from H. influenzae, we used affinity chromatography to purify human haptoglobin and hemopexin, serum proteins which bind hemoglobin and heme. Plate assays revealed that 12 strains of H. influenzae acquired heme from hemoglobin, hemoglobin-haptoglobin, heme-hemopexin, and heme-albumin. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of outer membrane proteins of strain E1a grown in heme-replete and heme-restricted conditions revealed a heme-repressible outer membrane protein with an apparent molecular mass of 38 kilodaltons. These results demonstrated that, unlike Escherichia coli, H. influenzae may acquire heme from hemoglobin-haptoglobin. H. influenzae also may acquire heme from hemopexin and albumin, which have not been previously investigated. The role of outer membrane proteins in the acquisition of heme is not yet clear.     

Response of Neisseria gonorrhoeae to iron limitation: alterations in expression of membrane proteins without apparent siderophore production.

For acquisition of iron, an essential nutrient, most microorganisms produce siderophores (low-molecular-weight iron-chelating compounds) and membrane proteins to serve as receptors for the iron-siderophore complexes. The gonococcus does not appear to produce a siderophore, since the quantity of siderophore detected by bioassays of culture supernatants from strains F62 and FA19 was never greater than the amount present in the uninoculated medium. Iron limitation of the laboratory strains F62 and FA19 and 12 recent clinical isolates resulted in the expression of several iron-repressible membrane proteins. The expression of proteins in the apparent molecular weight range of 70,000 to 100,000 was strain dependent. All strains expressed 36,000-dalton (36K) and 19.5K proteins. FA19 and F62 were also grown in medium containing iron sources commonly encountered in vivo (i.e., transferrin, lactoferrin, hemoglobin, or hemin). Comparison of growth rates indicates that transferrin and lactoferrin were more readily utilized as iron sources than hemin and hemoglobin were. Expression of the iron-repressible proteins varied depending upon the iron source. Fewer iron-repressible proteins were observed when cells were supplied with transferrin or lactoferrin than when the cultures were grown with either hemin or hemoglobin. The 36K protein was expressed with all four iron sources.     

Iron acquisition by Helicobacter pylori: importance of human lactoferrin.

Helicobacter pylori is known to be an etiologic agent of gastritis and peptic ulcer disease in humans. However, the mechanism by which this organism acquires iron has not been studied. For this investigation, H. pylori was grown in iron-restricted medium. Siderophore production was not detected by chemical assays, and the strains were unable to use enterochelin and pyochelin for growth in low-iron media. Human lactoferrin supported full growth of the bacteria in media lacking other iron sources, but neither human transferrin, bovine lactoferrin, nor hen ovotransferrin served as a source for iron. Since lactoferrin was found in significant amounts in human stomach resections with superficial or atrophic gastritis, the iron acquisition system of H. pylori by the human lactoferrin receptor system may play a major role in the virulence of H. pylori infection.     

Synthesis of aerobactin and a 76,000-dalton iron-regulated outer membrane protein by Escherichia coli K-12-Shigella flexneri hybrids and by enteroinvasive strains of Escherichia coli.

One of the chromosomal segments associated with the virulence of Shigella flexneri and transferred to Escherichia coli K-12 by conjugation has been shown to code for the production of aerobactin and for the synthesis of an iron-regulated 76,000-dalton (76K) outer membrane protein. Analysis of various E. coli K-12-S. flexneri transconjugants showed that the genes involved with the synthesis of aerobactin and with the production of the 76K protein were linked to the mtl region of the S. flexneri chromosome. S. flexneri itself synthesized a 76K protein in its outer membrane under iron restriction as well as traces of 81K and 74K proteins. An examination of four enteroinvasive strains of E. coli showed that each produced aerobactin and a 76K outer membrane protein during iron-restricted growth. The profile of the iron-regulated proteins expressed by the enteroinvasive strains of E. coli was virtually identical to that expressed by the laboratory-constructed E. coli K-12-S. flexneri hybrids under the same growth conditions.     

Comparison of outer membrane protein and biochemical profiles of Haemophilus aegyptius and Haemophilus influenzae biotype III.

Haemophilus aegyptius and Haemophilus influenzae biotype III are morphologically and biochemically similar; however, their outer membrane protein (Sarkosyl insoluble) profiles are distinct. Of 18 strains of H. aegyptius examined, 15 had a type 1 protein profile, and 3 had a type 2 profile, whereas the 5 strains of H. influenzae biotype III examined had three other protein profile types. All Haemophilus strains examined had 31- and 76-kilodalton (kDa) proteins and minor proteins with molecular masses between 20 and 100 kDa. H. aegyptius, with a type 1 protein profile, had major outer membrane proteins with apparent molecular masses of 27, 35.5, and 41.5 kDa, and H. aegyptius, with a type 2 protein profile, had 26-, 29-, 39.5-, and 41-kDa proteins. The type strain of H. influenzae biotype III had three major outer membrane proteins with apparent molecular masses of 29, 38.5 and 40 kDa. Four other strains designated as H. influenzae biotype III had major outer membrane proteins between 27 and 41.5 kDa representing two additional protein profiles.     

Blocking of iron uptake from transferrin by antibodies against the transferrin binding proteins inNeisseria meningitidis

Neisseria meningitidis, when cultured in iron-restricted environments, synthesises new outer-membrane proteins, many of which are necessary for their survival and growth. Some of these proteins e.g. transferrin-binding proteins 1 and 2 (Tbp1 and Tbp2), are required for the acquisition of iron from transferrin and are examples of important iron-regulated meningococcal surface antigens which are not expressed after growth in common laboratory media. The antigenicity and antigenic heterogeneity of these proteins have been extensively studied, and the bactericidal activity of antibodies directed to them have been studied. In this work we analysed the ability of such antibodies to inhibit transferrin binding and to block iron uptake from human transferrin. Antisera from mice immunized with either meningococcal outer membrane vesicles, purified Tbp1/2 complexes, or purified Tbp2, were incorporated in radiolabeled-iron uptake assays. Uptake was blocked by more than 80% in the homologous strains, but blocked much less efficiently in some heterologous strains, correlating well with inhibition of transferrin binding and with an inhibitory effect on bacterial growth. Inhibition of iron uptake from citrate was unaffected which suggests that this effect is due to antibodies against the components of the transferrin binding system, specially Tbp2. Our results support the importance of these proteins and their suitability to be considered in the development of effective vaccines against serogroup B meningococci.     

A set of two monoclonal antibodies specific for the cell surface-exposed 39K major outer membrane protein of Haemophilus influenzae type b defines all strains of this pathogen.

Six murine plasma cell hybridomas producing monoclonal antibodies (mabs) directed against the 39,000-molecular-weight (39K) major outer membrane protein of Haemophilus influenzae type b were employed in the antigenic analysis of the 39K protein. The initial characterization of the mabs by radioimmunoprecipitation analysis showed that four of these mabs reacted with antigenic determinants of the 39K protein that are exposed on the bacterial cell surface and accessible to antibody. The other two mabs reacted with antigenic determinants of the 39K protein that are either not exposed on the H. influenzae type b cell surface or not accessible to antibody (internal determinants). A total of 126 clinical isolates of H. influenzae type b obtained from pediatric research centers throughout the United States were examined for reactivity with the six mabs by using a solid-phase radioimmunoassay in which bacterial colony growth from agar plates was placed on filter paper and used as antigen. The reactivities of these strains with two of the mabs recognizing cell surface-exposed antigenic determinants of the 39K protein were used to divide the 126 strains into four different groups. Group 1 strains reacted with mab 12D9, group 2 strains reacted with mab 4C4, group 3 strains reacted with both mabs 12D9 and 4C4, and group 4 strains (only one was found) did not react with either mab. The reactivities of two other mabs recognizing cell surface-exposed antigenic determinants of the 39K protein were used to further divide the four groups into eight subgroups. A single mab recognizing an internal antigenic determinant of the 39K protein reacted with every H. influenzae type b strain examined in this study. These data indicate that only limited antigenic heterogeneity exists among the cell surface-exposed antigenic determinants of the 39K outer membrane proteins among H. influenzae type b strains and that at least one internal antigenic determinant of the 39K protein is universally present in all H. influenzae type b strains. Radioimmunoprecipitation analysis also demonstrated that H. influenzae type b strains which lacked a 39K major outer membrane protein possessed a 38K major outer membrane protein which reacted with the anti-39K mabs, indicating that the 38K and 39K outer membrane proteins of different H. influenzae type b strains are antigenically related.     

Iron acquisition and hemolysin production by Campylobacter jejuni.

Campylobacter jejuni strains were tested for their ability to acquire iron from various iron sources present in humans. Hemin, hemoglobin, hemin-hemopexin, and hemoglobin-haptoglobin stimulated the growth of C. jejuni strains in low-iron medium. Transferrin, lactoferrin, and ferritin were unable to provide iron to the strains tested. Derivatives of the naturally transformable C. jejuni strain 81-176 were isolated on the basis of their inability to use hemin as an iron source. These mutants were also unable to use hemoglobin, hemin-hemopexin, or hemoglobin-haptoglobin as iron sources. Some mutants lacked a 71,000-Da iron-regulated outer membrane protein, while others appeared to retain all of their outer membrane proteins. Growth curves and a recombination experiment that exploited natural transformation were used to further characterize the mutants. A hemolytic activity was shown to be produced by several C. jejuni strains, but it did not appear to be iron regulated.     

Haemophilus influenzae type b isolates show antigenic variation in a major outer membrane protein.

Antigenic variation of the outer membrane proteins among isolates of Haemophilus influenzae was examined by immunoblotting. Rabbit antisera were raised against six strains of H. influenzae type b and tested against outer membrane preparations of 50 isolates. The principal outer membrane band was not reactive on immunoblotting, so its antigenic heterogeneity could not be examined. Most of the other outer membrane proteins shared common determinants among all strains tested. Absorption of serum with heterologous bacteria removed antibody to nearly all proteins, confirming the extensive cross-reactivity among isolates. The greatest antigenic variation was seen in one major outer membrane band, a heat-modifiable, Zwittergent-soluble protein with a molecular weight of 49,000 to 51,000. One antiserum reacted with the 49,000-to-51,000-molecular-weight protein of the homologous isolate only; the remaining five antisera showed differing patterns of reactivity with heterologous 49,000-to-51,000-molecular-weight proteins. We were able to divide the 50 H. influenzae isolates into 13 antigenic groups based on their reaction patterns. The antigenic groupings may provide an epidemiological tool for studying the prevalence and transmission of strains of H. influenzae type b.     

Common antigenic domains in transferrin-binding protein 2 of Neisseria meningitidis, Neisseria gonorrhoeae, and Haemophilus influenzae type b.

There is now considerable evidence to show that in the Neisseria and Haemophilus species, membrane receptors specific for either transferrin or lactoferrin are involved in the acquisition of iron from these glycoproteins. In Neisseria meningitidis, the transferrin receptor appears to consist of two proteins, one of which (TBP 1) has an M(r) of 95,000 and the other of which (TBP 2) has an M(r) ranging from 68,000 to 85,000, depending on the strain; TBP 2 binds transferrin after sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electroblotting, but TBP 1 does not do so. The relative contributions of these two proteins to the binding reaction observed with intact cells and to iron uptake are presently unknown. However, they are being considered as potential components of a group B meningococcal vaccine. Analogous higher- and lower-molecular-weight proteins associated with transferrin binding have been found in N. gonorrhoeae and Haemophilus influenzae. Previous work with polyclonal antibodies raised in mice with whole cells of iron-restricted N. meningitidis showed that the meningococcal TBP 2 exhibits considerable antigenic heterogeneity. Here, we report that antiserum against purified TBP 2 from one strain of N. meningitidis cross-reacts on immunoblotting with the TBP 2 of all meningococcal isolates examined, as well as with the TBP 2 of N. gonorrhoeae. This antiserum also cross-reacted with the TBP 2 of several strains of H. influenzae type b, thus showing the presence of common antigenic domains among these functionally equivalent proteins in different pathogens; no cross-reaction was detected with a purified sample of the human transferrin receptor.     

Identification of 2,3-dihydroxybenzoic acid as a Brucella abortus siderophore.

Brucella abortus grown in low-iron medium or in the presence of iron chelators [ethylenediamine-di(o-hydroxyphenylacetic acid) and 2,2-dipyridyl] showed reduced cell yields and released a material positive in chemical and biological assays for catechols. This material was purified from culture fluids of B. abortus 2308 by chromatography on agarose-iminodiacetic acid-Fe3+ and identified as 2,3-dihydroxybenzoic acid (2,3-DHBA) by thin-layer chromatography, paper electrophoresis, and UV-visible nuclear magnetic resonance and mass spectroscopy. No other major catechols were observed at different stages of growth, and 2,3-DHBA was also produced upon iron limitation by representative strains of B. abortus biotypes 1, 5, 6, and 9. Both synthetic 2,3-DHBA and the natural catechol relieved the growth inhibition of B. abortus 2308 by ethylenediamine-di(o-hydroxyphenylacetic acid), and 2,3-DHBA promoted 55Fe uptake by B. abortus 2308 by an energy-dependent mechanism. Two other monocatechols tested, 2,3-dihydroxybenzoyl-Ser and 2,3-dihydroxybenzoyl-Gly, also promoted 55Fe uptake. More complex catechol siderophores (agrobactin and enterobactin), hydroxamate siderophores (aerobactin, ferrichrome, and deferriferrioxamine mesylate [Desferal]), and an EDTA-related siderophore (rhizobactin) failed to mediate 55Fe uptake. B. abortus cells grown in low-iron medium or in medium with iron had similar rates of iron uptake when supplied with 55Fe-2,3-DHBA, and the release of 2,3-DHBA under iron starvation was not associated with the expression of new outer membrane proteins. These results suggest an uptake system in which only the synthesis of the siderophore is regulated by the iron available for growth.     

Molecular cloning of DNA coding for outer membrane proteins of Haemophilus influenzae type b.

DNA from Haemophilus influenzae type b was cloned in Escherichia coli with a vector lambda gt11 Amp1. Clones producing antigens reactive with hyperimmune rabbit antisera were identified by colony radioimmune assay. A second screening with hyperimmune serum adsorbed with intact H. influenzae type b bacteria was used to identify those clones producing surface-exposed outer membrane proteins. The proteins expressed in E. coli were coupled to Sepharose and used to affinity purify antibodies which were tested for reactivity with outer membrane vesicles. It was found by Western blotting that the clones were producing antigens corresponding to Mr 49,000, 39,500, and 35,000 major outer membrane protein or antigens of H. influenzae type b. Additional clones could be detected by human serum, but their reactivity was not removed when serum was adsorbed with intact bacteria. One of these studied in more detail was found to produce an antigen present in H. influenzae type b lysate but not in outer membrane vesicle preparations.     

Ferric reduction is a potential iron acquisition mechanism for Histoplasma capsulatum

For the fungus Histoplasma capsulatum, and for other microbial pathogens, iron is an essential nutrient. Iron sequestration in response to infection is a demonstrated host defense mechanism; thus, iron acquisition may be considered an important pathogenic determinant. H. capsulatum is known to secrete Fe(III)-binding hydroxamate siderophores, which is one common microbial process for acquiring iron. Here, we report H. capsulatum ferric reduction activities in whole yeast cells and in both high- and low-molecular-weight fractions of culture supernatants. Each of these activities was induced or derepressed by growth under iron-limiting conditions, a phenomenon often associated with specific iron acquisition mechanisms. The high-molecular-weight culture supernatant activity was enhanced by the addition of reduced glutathione, was proteinase K sensitive and heat labile, and could utilize ferric chloride, ferric citrate, and human holotransferrin as substrates. The low-molecular-weight culture supernatant activity was resistant to proteinase K digestion. These results are consistent with the expression by H. capsulatum of both enzymatic ferric reductase and nonproteinaceous ferric reductant, both of which are regulated by iron availability. Such components could be involved in fungal acquisition of iron from inorganic or organic ferric salts, from H. capsulatum hydroxamate siderophores, or from host Fe(III)-binding proteins, such as transferrin.     

Effect of iron restriction on the outer membrane proteins of Actinobacillus (Haemophilus) pleuropneumoniae.

The outer membrane protein profile of Actinobacillus (Haemophilus) pleuropneumoniae grown under iron-restricted and iron-replete conditions was studied by polyacrylamide gel electrophoresis and immunoblotting. A virulent serotype 1 isolate synthesized a novel protein with an apparent molecular weight of 105,000 (105K) and increased the synthesis of a 76K protein under iron-restricted conditions. Both proteins were synthesized within 15 min of establishment of iron-restricted conditions. Proteins of equivalent molecular weights could also be induced by iron restriction in serotype 2, 3, 4, 5, and 7 isolates of A. pleuropneumoniae. Convalescent-phase sera from serotype 1-infected pigs contained antibodies which recognized both the 105K and 76K proteins from all six serotypes examined, indicating that these proteins were expressed in vivo and were immunologically conserved. Cells expressing the 105K and 76K proteins also displayed an enhanced ability to bind Congo red and hemin, suggesting that one or both of these proteins functioned to acquire complexed iron during in vivo growth.     

Evidence for in vivo expression of transferrin-binding proteins in Haemophilus influenzae type b.

Haemophilus influenzae type b acquires transferrin-bound iron via a siderophore-independent mechanism involving direct contact between the human iron-binding glycoprotein and the bacterial cell surface. Evidence has accumulated to show that the transferrin receptor consists of at least two iron-regulated outer membrane transferrin-binding proteins (TBPs), of which one has a molecular mass of around 100 kDa (TBP1) and the other has a molecular mass of 60 to 90 kDa (TBP2). In H. influenzae type b strain Eagan, proteins of 76, 90, and 107 kDa appear to be involved in transferrin binding. To determine whether these TBPs are expressed during growth in vivo, strain Eagan was recovered without subculture from the intraperitoneal cavities of infected infant rats. By using a dot blot assay, outer membranes prepared from these in vivo-grown bacteria, unlike those grown in iron-sufficient broth, bound human transferrin and produced the 76-, 90-, and 107-kDa TBPs. Immunoblotting experiments using convalescent sera from infected rats also revealed the presence of antibodies to the 76- and 90-kDa strain Eagan TBPs. In addition, convalescent sera from three of four patients recovering from H. influenzae type b meningitis contained antibodies to the 90- and 105-kDa TBPs of the corresponding infecting strain. Furthermore, fresh clinical isolates of H. influenzae type b recovered from blood and cerebrospinal fluid showed constitutive expression of the TBPs, which became iron regulated only after prolonged in vitro subculture on iron-sufficient medium. This contrasted with the laboratory-adapted Eagan strain, in which the TBPs remained iron regulated even after animal passage. These findings indicate that the H. influenzae type b transferrin receptor is expressed during experimental animal and human infections.     

Influence of iron on growth, morphology, outer membrane protein composition, and synthesis of siderophores in Campylobacter jejuni.

Three human isolates of Campylobacter jejuni were grown in a biphasic culture medium with and without the addition of a synthetic chelator to induce iron limitation. Cells grown in low-iron medium exhibited slower growth rates and altered cellular morphology. Increased numbers of longer, more filamentous forms were seen in Gram-stained smears. Three proteins, with apparent Mrs of 82,000, 76,000, and 74,000, were consistently present in the outer membrane of cells grown in low-iron medium. At least one of these proteins (76,000 to 74,000) was exposed on the cell surface. A bioassay was used to look for the production of siderophores by these and other strains of C. jejuni. Seven of 26 strains tested produced detectable amounts of siderophores. Growing strains at 42 degrees C failed to suppress siderophore synthesis or to alter the outer membrane protein profiles of iron-starved cells. The ability of three strains to utilize exogenously supplied siderophores for growth in low-iron medium was also examined. All three strains were able to utilize enterochelin and ferrichrome, but none utilized aerobactin, rhodotorulic acid, or desferrioxamine B. The effect of iron on the virulence of C. jejuni for 11-day-old chicken embryos inoculated via the chorioallantoic membrane was also determined.     

Identification and purification of a conserved heme-regulated hemoglobin-binding outer membrane protein from Haemophilus ducreyi.

A hemoglobin-binding protein (HgbA) from Haemophilus ducreyi was identified and purified. The 100-kDa HgbA was detected in all strains of H. ducreyi tested, and a somewhat larger hemoglobin-binding protein was found in one strain of Haemophilus influenzae. HgbA was purified and the amino acid sequence of the N terminus of HgbA revealed no significant homologies with known proteins. Two different antisera to HgbA from H. ducreyi 35000 recognized HgbA proteins from all tested H. ducreyi strains; they did not recognize proteins from the H. influenzae strain. Expression of HgbA was regulated by the level of heme but not by iron present in the medium. Animal species of hemoglobin competed with iodinated human hemoglobin for binding to whole cells of H. ducreyi and supported the growth of H. ducreyi. The lack of immunological cross-reactivity and the differences in hemoglobin specificities between the H. ducreyi and the H. influenzae hemoglobin-binding proteins suggest that they are unrelated.