Response of Haemophilus somnus to iron limitation: expression and identification of a bovine-specific transferrin receptor

Insights from targeting transferrin receptors to develop vaccines for pathogens of humans and food production animals

While developing vaccines targeting surface transferrin receptor proteins in Gram-negative pathogens of humans and food production animals, the common features derived from their evolutionary origins has provided us with insights on how improvements could be implemented in the various stages of research and vaccine development. These pathogens are adapted to live exclusively on the mucosal surfaces of the upper respiratory or genitourinary tract of their host and rely on their receptors to acquire iron from transferrin for survival, indicating that there likely are common mechanisms for delivering transferrin to the mucosal surfaces that should be explored. The modern-day receptors are derived from those present in bacteria that lived over 320 million years ago. The pathogens represent the most host adapted members of their bacterial lineages and may possess factors that enable them to have strong association with the mucosal epithelial cells, thus likely reside in a different niche than the commensal members of the bacterial lineage. The bacterial pathogens normally lead a commensal lifestyle which presents challenges for development of relevant infection models as most infection models either exclude the early stages of colonization or subsequent disease development, and the immune mechanisms at the mucosal surface that would prevent disease are not evident. Development of infection models emulating natural horizontal disease transmission are also lacking. Our aim is to share our insights from the study of pathogens of humans and food production animals with individuals involved in vaccine development, maintaining health or regulation of products in the human and animal health sectors.

Transferrin Binding Protein B and Transferrin Binding Protein A2 Expand the Transferrin Recognition Range of Histophilus somni

The bacterial bipartite transferrin receptor is an iron acquisition system that several important human and animal pathogens require for survival. It consists of the TonB-dependent transporter transferrin binding protein A (TbpA) and the surface lipoprotein transferrin binding protein B (TbpB). Curiously, the Tbps are only found in host-specific pathogens and are themselves host specific, meaning that they will bind to the transferrin of their host species but not to the transferrins of other animal species. While this phenomenon has long been established, neither the steps in the evolutionary process that led to this exquisite adaptation for the host nor the steps that could alter it are known. We sought to gain insight into these processes by studying Tbp specificity in Histophilus somni, an economically important pathogen of cattle. A past study showed that whole cells of H. somni specifically bind bovine transferrin but not transferrin from sheep and goats, two bovids whose transferrins share 93% amino acid sequence identity with bovine transferrin. To our surprise, we found that H. somni can use sheep and goat transferrins as iron sources for growth and that HsTbpB, but not HsTbpA, has detectable affinity for sheep and goat transferrins. Furthermore, a third transferrin binding protein found in H. somni, HsTbpA2, also showed affinity for sheep and goat transferrins. Our results suggest that H. somni TbpB and TbpA2 may contribute to broadening the host transferrin recognition range of H. somni IMPORTANCE Host-restricted pathogens infect a single host species or a narrow range of host species. Histophilus somni, a pathogen that incurs severe economic losses for the cattle industry, infects cattle, sheep, and goats but not other mammals. The transferrin binding proteins, TbpA and TbpB, are thought to be a key iron acquisition system in H. somni; however, despite their importance, H. somni TbpA and TbpB were previously shown to be cattle transferrin specific. In our study, we find that H. somni TbpB and another little-studied Tbp, TbpA2, bind sheep and goat transferrins, as well as bovine transferrin. Our results suggest that TbpB and TbpA2 may allow for host range expansion and provide a mechanism for how host specificity in Tbp-encoding pathogens can be altered.

Are lactoferrin receptors in Gram-negative bacteria viable vaccine targets?

A number of important Gram-negative pathogens that reside exclusively in the upper respiratory or genitourinary tract of their mammalian host rely on surface receptors that specifically bind host transferrin and lactoferrin as a source of iron for growth. The transferrin receptors have been targeted for vaccine development due to their critical role in acquiring iron during invasive infection and for survival on the mucosal surface. In this study, we focus on the lactoferrin receptors, determining their prevalence in pathogenic bacteria and comparing their prevalence in commensal Neisseria to other surface antigens targeted for vaccines; addressing the issue of a reservoir for vaccine escape and impact of vaccination on the microbiome. Since the selective release of the surface lipoprotein lactoferrin binding protein B by the NalP protease in Neisseria meningitidis argues against its utility as a vaccine target, we evaluated the release of outer membrane vesicles, and transferrin and lactoferrin binding in N. meningitidis and Moraxella catarrhalis. The results indicate that the presence of NalP reduces the binding of transferrin and lactoferrin by cells and native outer membrane vesicles, suggesting that NalP may impact all lipoprotein targets, thus this should not exclude lactoferrin binding protein B as a target.

Iron acquisition through the bacterial transferrin receptor

Transferrin is one of the sources of iron that is most readily available to colonizing and invading pathogens. In this review, we look at iron uptake by the bacterial transferrin receptor that is found in the families Neisseriaceae, Pasteurellaceae and Moraxellaceae. This bipartite receptor consists of the TonB-dependent transporter, TbpA, and the surface lipoprotein, TbpB. In the past three decades, major advancements have been made in our understanding of the mechanism through which the Tbps take up iron. We summarize these findings and discuss how they relate to the diversity and specificity of the transferrin receptor. We also outline several of the remaining unanswered questions about iron uptake via the bacterial transferrin receptor and suggest directions for future research.

Histophilus somni host-parasite relationships

Histophilus somni (Haemophilus somnus) is one of the key bacterial pathogens involved in the multifactorial etiology of the Bovine Respiratory Disease Complex. This Gram negative pleomorphic rod also causes bovine septicemia, thrombotic meningencephalitis, myocarditis, arthritis, abortion and infertility, as well as disease in sheep, bison and bighorn sheep. Virulence factors include lipooligosaccharide, immunoglobulin binding proteins (as a surface fibrillar network), a major outer membrane protein (MOMP), other outer membrane proteins (OMPs) and exopolysaccharide. Histamine production, biofilm formation and quorum sensing may also contribute to pathogenesis. Antibodies are very important in protection as shown in passive protection studies. The lack of long-term survival of the organism in macrophages, unlike facultative intracellular bacteria, also suggests that antibodies should be critical in protection. Of the immunoglobulin classes, IgG2 antibodies are most implicated in protection and IgE antibodies in immunopathogenesis. The immunodominant antigen recognized by IgE is the MOMP and by IgG2 is a 40 kDa OMP. Pathogenetic synergy of bovine respiratory syncytial virus (BRSV) and H. somni in calves can be attributed, in part at least, to the higher IgE anti-MOMP antibody responses in dually infected calves. Other antigens are probably involved in stimulating host defense or immunopathology as well.

Complete genome sequence of Haemophilus somnus (Histophilus somni) strain 129Pt and comparison to Haemophilus ducreyi 35000HP and Haemophilus influenzae Rd

Haemophilus somnus can be either a commensal of bovine mucosal surfaces or an opportunistic pathogen. Pathogenic strains of H. somnus are a significant cause of systemic disease in cattle. We report the genome sequence of H. somnus 129Pt, a nonpathogenic commensal preputial isolate, and the results of a genome-wide comparative analysis of H. somnus 129Pt, Haemophilus influenzae Rd, and Haemophilus ducreyi 35000HP. We found unique genes in H. somnus 129Pt involved in lipooligosaccharide biosynthesis, carbohydrate uptake and metabolism, cation transport, amino acid metabolism, ubiquinone and menaquinone biosynthesis, cell surface adhesion, biosynthesis of cofactors, energy metabolism, and electron transport. There were also many genes in common among the three organisms. Our comparative analyses of H. somnus 129Pt, H. influenzae Rd, and H. ducreyi 35000HP revealed similarities and differences in the numbers and compositions of genes involved in metabolism, host colonization, and persistence. These results lay a foundation for research on the host specificities and niche preferences of these organisms. Future comparisons between H. somnus 129Pt and virulent strains will aid in the development of protective strategies and vaccines to protect cattle against H. somnus disease.

Bovine plasma proteins increase virulence of Haemophilus somnus in mice

The role of bovine serum or plasma proteins in Haemophilus somnus virulence was investigated in a mouse model of septicemia. An increase in virulence was detected when the organism was pre-incubated for 5 min and inoculated with fetal calf serum. When purified bovine serum or plasma proteins were pre-incubated with H. somnus before inoculating into mice, transferrin was found to increase virulence. Bovine lactoferrin was also noted to increase virulence, but to a lesser extent and had a delayed time course when compared with transferrin. Using an ELISA assay, an increased amount of H. somnus whole cells and culture supernatant bound to bovine transferrin when the organism was grown in iron-restricted media. Lactoferrin also bound to H. somnus, but binding was not affected by growth in iron-restricted media and it was eliminated with 2M NaCl, which reversed charge mediated binding. Transferrin, but not lactoferrin, supported growth of H. somnus on iron-depleted agar based media using a disk assay. Therefore, lactoferrin increased virulence by an undetermined mechanism whereas transferrin increased virulence of H. somnus by binding to iron-regulated outer-membrane proteins (IROMPs) and providing iron to the pathogen.

Haemophilus somnus possesses two systems for acquisition of transferrin-bound iron

Haemophilus somnus strain 649 was found to acquire iron from ovine, bovine, and goat transferrins (Tfs). Expression of Tf receptors, as evaluated by solid-phase binding assays, required the organisms to be grown under iron-restricted conditions in the presence of Tf. Competition binding assays revealed the presence of two distinct Tf-binding receptor systems, one specific for bovine Tf and the other capable of binding all three ruminant Tfs. Affinity isolation procedures using total membranes yielded three putative bovine Tf-binding polypeptides and one putative ovine and goat Tf-binding polypeptide. PCR amplification followed by DNA sequence analyses revealed that H. somnus strain 649 possesses genes that encode a bipartite TbpA-TbpB receptor along with a homolog of the Histophilus ovis single-component TbpA receptor. Expression of TbpB and the single-component TbpA would appear to be subject to a form of phase variation involving homopolymeric nucleotide tracts within the structural genes.

High resolution structure of an alternate form of the ferric ion binding protein from Haemophilus influenzae

The periplasmic iron binding protein of pathogenic Gram-negative bacteria performs an essential role in iron acquisition from transferrin and other iron sources. Structural analysis of this protein from Haemophilus influenzae identified four amino acids that ligand the bound iron: His(9), Glu(57), Tyr(195), and Tyr(196). A phosphate provides an additional ligand, and the presence of a water molecule is required to complete the octahedral geometry for stable iron binding. We report the 1.14-A resolution crystal structure of the iron-loaded form of the H. influenzae periplasmic ferric ion binding protein (FbpA) mutant H9Q. This protein was produced in the periplasm of Escherichia coli and, after purification and conversion to the apo form, was iron-loaded. H9Q is able to bind ferric iron in an open conformation. A surprising finding in the present high resolution structure is the presence of EDTA located at the previously determined anion ternary binding site, where phosphate is located in the wild type holo and apo structures. EDTA contributes four of the six coordinating ligands for iron, with two Tyr residues, 195 and 196, completing the coordination. This is the first example of a metal binding protein with a bound metal.EDTA complex. The results suggest that FbpA may have the ability to bind and transport iron bound to biological chelators, in addition to bare ferric iron.

Bacterial lactoferrin receptors: insights from characterizing the Moraxella bovis receptors

Moraxella bovis is the causative agent of infectious conjunctivitis in cattle. Moraxella bovis isolates were shown to specifically bind bovine lactoferrin (bLf) and bovine transferrin (bTf) and to use these proteins as a source of iron to support the growth of iron-limited cells. Affinity isolation experiments with immobilized bTf yielded two proteins readily resolved by SDS-PAGE analysis, whereas only a single band of approximately 100 kDa was detected when immobilized bLf was used as the affinity ligand. Using a novel cloning strategy, regions containing the genes encoding the lactoferrin (Lf) and transferrin (Tf) receptor proteins were isolated and sequenced, demonstrating that they both consisted of two genes, with the tbpB or lbpB gene preceding the tbpA or lbpA gene. The cloned lbp genes were used to generate isogenic mutants deficient in lactoferrin binding protein A and (or) B, and the resulting strains were tested in growth and binding assays. The isogenic mutants were deficient in their use of bLf for growth and had substantially diminished bLf binding capability. The predicted amino acid sequence from the segment encoding Lf binding protein B revealed an internal amino acid homology suggesting it is a bi-lobed protein, with a C-lobe enriched in acidic amino acids, but without the evident clustering observed in Lf-binding proteins from other species.

Identification of fur and fldA homologs and a Pasteurella multocida tbpA homolog in Histophilus ovis and effects of iron availability on their transcription

tbpA, fur, and fldA homologs from two strains (9L and 3384Y) of the sheep pathogen Histophilus ovis were sequenced. The predicted TbpA proteins of these strains are homologs of the Pasteurella multocida TbpA protein and collectively represent the second example of a new subfamily of TonB-dependent receptors. tbpA transcripts were readily detected by reverse transcription (RT)-PCR with RNA isolated from strain 9L grown under iron-restricted conditions in the presence or absence of bovine transferrin (Tf). However, with strain 3384Y and depending on the primer pair, tbpA transcripts were detected by RT-PCR predominantly when the RNA was from cells grown under iron-restricted conditions in the presence of bovine Tf. In both strains, the fldA homolog was found to be immediately upstream of fur and, based on RT-PCR, these genes are transcribed as a single unit; the availability of iron and the presence or absence of bovine Tf in the growth medium had no apparent effect on the relative amounts of the fldA-fur transcripts.

Production of transferrin receptors by Histophilus ovis: three of five strains require two signals

Five strains of Histophilus ovis (9L, 642A, 714, 5688T, and 3384Y) were investigated with respect to iron acquisition. All strains used ovine, bovine, and goat transferrins (Tfs), but not porcine or human Tfs, as iron sources for growth. In solid phase binding assays, total membranes from only two (9L and 642A) of the five strains, grown under iron-restricted conditions, were able to bind Tfs (ovine, bovine, and goat, but not porcine or human). However, when the organisms were grown under iron-restricted conditions in the presence of bovine transferrin (Tf), total membranes from all strains exhibited Tf binding (as above); competition experiments demonstrated that all three Tfs (ovine, bovine, and goat) were bound by the same receptor(s). Membranes from organisms grown under iron-replete conditions in the presence or absence of bovine Tf failed to bind any of the test Tfs. An affinity-isolation procedure allowed the isolation of two putative Tf-binding polypeptides (78 and 66 kDa) from total membranes of strains 9L and 642A grown under iron-restricted conditions, and from membranes of all strains if the growth medium also contained Tf. It is concluded that all strains tested acquire Tf-bound iron by means of siderophore-independent mechanisms involving surface receptors analogous to the Tf-binding proteins (TbpA and TbpB) found in comparable organisms; although iron restriction alone is sufficient to promote the expression of these proteins by strains 9L and 642A, their production by strains 714, 5688T, and 3384Y appears to require two signals, iron restriction and the presence of Tf.

Iron loading and disease surveillance

Iron is an oxidant as well as a nutrient for invading microbial and neoplastic cells. Excessive iron in specific tissues and cells (iron loading) promotes development of infection, neoplasia, cardiomyopathy, arthropathy, and various endocrine and possibly neurodegenerative disorders. To contain and detoxify the metal, hosts have evolved an iron withholding defense system, but the system can be compromised by numerous factors. An array of behavioral, medical, and immunologic methods are in place or in development to strengthen iron withholding. Routine screening for iron loading could provide valuable information in epidemiologic, diagnostic, prophylactic, and therapeutic studies of emerging infectious diseases.

Identification of human transferrin-binding sites within meningococcal transferrin-binding protein B

Transferrin-binding protein B (TbpB) from Neisseria meningitidis binds human transferrin (hTf) at the surface of the bacterial cell as part of the iron uptake process. To identify hTf binding sites within the meningococcal TbpB, defined regions of the molecule were produced in Escherichia coli by a translational fusion expression system and the ability of the recombinant proteins (rTbpB) to bind peroxidase-conjugated hTf was characterized by Western blot and dot blot assays. Both the N-terminal domain (amino acids [aa] 2 to 351) and the C-terminal domain (aa 352 to 691) were able to bind hTf, and by a peptide spot synthesis approach, two and five hTf binding sites were identified in the N- and C-terminal domains, respectively. The hTf binding activity of three rTbpB deletion variants constructed within the central region (aa 346 to 543) highlighted the importance of a specific peptide (aa 377 to 394) in the ligand interaction. Taken together, the results indicated that the N- and C-terminal domains bound hTf approximately 10 and 1000 times less, respectively, than the full-length rTbpB (aa 2 to 691), while the central region (aa 346 to 543) had a binding avidity in the same order of magnitude as the C-terminal domain. In contrast with the hTf binding in the N-terminal domain, which was mediated by conformational epitopes, linear determinants seemed to be involved in the hTf binding in the C-terminal domain. The host specificity for transferrin appeared to be mediated by the N-terminal domain of the meningococcal rTbpB rather than the C-terminal domain, since we report that murine Tf binds to the C-terminal domain. Antisera raised to both N- and C-terminal domains were bactericidal for the parent strain, indicating that both domains are accessible at the bacterial surface. We have thus identified hTf binding sites within each domain of the TbpB from N. meningitidis and propose that the N- and C-terminal domains together contribute to the efficient binding of TbpB to hTf with their respective affinities and specificities for determinants of their ligand.

Cloning and characterization of bacteriophage-like DNA from Haemophilus somnus homologous to phages P2 and HP1

In an attempt to identify and characterize components of a heme uptake system of Haemophilus somnus, an Escherichia coli cosmid library of H. somnus genomic DNA was screened for the ability to bind hemin (Hmb+). The Hmb+ phenotype was associated with a 7,814-bp HindIII fragment of H. somnus DNA that was subcloned and sequenced. Thirteen open reading frames (orfs) were identified, all transcribed in one direction, and transposon mutagenesis identified orf7 as the gene associated with the Hmb+ phenotype. Orf7 (178 amino acids) has extensive homology with the lysozymes of bacteriophages P-A2, P21, P22, PZA, phi-29, phi-vML3, T4, or HP1. The orf7 gene complemented the lytic function of the K gene of phage P2 and the R gene of phage lambda. A lysozyme assay using supernatants from whole-cell lysates of E. coli cultures harboring plasmid pRAP501 or pGCH2 (both of which express the orf7 gene product) exhibited significant levels of lysozyme activity. The orf6 gene upstream of orf7 has the dual start motif common to the holins encoded by lambdoid S genes, and the orf6 gene product has significant homology to the holins of phages HP1 and P21. When expressed from a tac promoter, the orf6 gene product caused immediate cell death without lysis, while cultures expressing the orf7 gene product grew at normal rates but lysed immediately after the addition of chloroform. Based on this data, we concluded that the Hmb+ phenotype was an artifact resulting from the expression of cloned lysis genes which were detrimental to the E. coli host. The DNA flanking the cloned lysis genes contains orfs that are similar to structural and DNA packaging genes of phage P2. Polyclonal antiserum against Orf2, which is homologous to the major capsid precursor protein (gpN) of phage P2, detected a 40,000-M(r) protein expressed from pRAP401 but did not detect Orf2 in H. somnus, lysates. The phage-like DNA was detected in the serum-susceptible preputial strains HS-124P and HS-127P but was absent from the serum-resistant preputial strains HS-20P and HS-22P. Elucidation of a potential role for this cryptic prophage in the H. somnus life cycle requires more study.

Utilization of transferrin and salmon serum as sources of iron by typical and atypical strains of Aeromonas salmonicida

The ability of typical and atypical strains of Aeromonas salmonicida to utilize non-haem sources of protein-bound iron was evaluated. (i) In a plate bioassay, the suppression of growth imposed on typical and atypical A. salmonicida by addition of the high-affinity iron chelator ethylenediamine-di(o- hydroxyphenylacetic acid) (EDDA) to the growth medium was reversed by the addition of 30% or 90% iron-saturated bovine or human transferrin (Tf) or lactoferrin (Lf) to the growth medium. (ii) The mechanism of obtaining iron from Tf was investigated by the addition of bovine Tf contained within a dialysis bag. The reversal of iron-restricted growth suppression differed between the strains in that the atypical strains were unable to utilize Tf contained within a dialysis bag while the typical strains were able to do so. This suggested a siderophore-mediated uptake of iron from Tf by the typical strains, which are known to produce siderophores while atypical strains do not. (iii) A solid-phase binding assay using horseradish-peroxidase-conjugated or biotinylated Tf or Lf failed to detect Tf/Lf-binding activity using whole typical or atypical cells. (iv) When atypical extracellular products (ECP) plus bovine Tf or salmon serum were enclosed in a dialysis bag, diffusible products were released which could reverse the EDDA-imposed growth suppression of an atypical strain. This reversal was negated by inhibition of the ECP metalloprotease with EDTA. (v) Purified 70 kDa serine protease of a typical strain was able to digest bovine Tf to low molecular mass fragments as observed in SDS-PAGE. These results indicate that typical and atypical strains of A. salmonicida differ in their mechanism of utilization of non-haem protein- bound sources of iron. Typical strains utilize Tf via a siderophore-mediated mechanism and are also able to digest Tf with the extracellular serine protease. Atypical strains utilize Tf by a siderophore-independent mechanism probably involving the proteolytic degradation of Tf by the extracellular metalloprotease.

Cloning, sequencing and expression of the transferrin-binding protein 1 gene from Actinobacillus pleuropneumoniae

Two outer-membrane proteins are involved in the uptake of iron from transferrin by certain Gram-negative bacteria, transferrin-binding proteins 1 and 2. The gene encoding transferrin-binding protein 1 from a serotype 1 isolate of the Gram-negative pathogen Actinobacillus pleuropneumoniae was cloned, and a fragment encoding 700 amino acids of Tbp1 was expressed in Escherichia coli. We also report here sequencing of the tbpl gene and a comparison of the deduced amino acid sequence with Tbpls from related species. The predicted polypeptide product of tbpl is a 106 kDa protein with a 22-residue signal peptide.

Biochemical analysis of lactoferrin receptors in the Neisseriaceae: identification of a second bacterial lactoferrin receptor protein

Bacterial transferrin receptors that have been described in the families Pasteurellaceae and Neisseriaceae are composed of two receptor proteins, transferrin binding proteins 1 and 2 (Tbp1 and Tbp2). In contrast, bacterial lactoferrin receptors have only been described for human pathogens in the family Neisseriaceae, and were believed to consist of a single protein, Lbp1, which is highly homologous to Tbp1. We describe a modified affinity isolation procedure that facilities isolation of a second lactoferrin receptor protein Lbp2 (a presumptive Tbp2 homologue) from Neisseria meningitidis, Moraxella catarrhalis and Moraxella bovis using immobilized lactoferrin. Antiserum specific for either the M. catarrhalis Tbp1+2 molecules, the M. catarrhalis Lbp1 molecule, or for a commercial preparation of human lactoferrin did not react on western blots with the same organisms' affinity purified Lbp2. In addition, the M. catarrhalis Lbp2 could be isolated in a functional form without contaminating Lbp1 or Tbp1+2. We also demonstrate that the bovine pathogen, M. bovis, produces functional transferrin and lactoferrin receptors specific for the bovine forms of these glycoproteins. A putative lbpB gene, recently speculated to reside immediately upstream of the N. meningitidis Lbp1 structural gene, lbpA, likely encodes the newly isolated Lbp2 protein from this bacterial species.

Comparative study of iron acquisition by biotype 1 and biotype 2 strains of Actinobacillus pleuropneumoniae

Four strains of the swine pathogen, Actinobacillus pleuropneumoniae, namely, the type strain (ATCC 27088; biotype 1), the 'reference' strain of biotype 2 (Bertschinger 2008/76), and two additional biotype 1 strains, strain BC181, which is less virulent than the type strain, and strain K17, which was isolated from a lamb, were investigated with respect to iron acquisition. All strains produced iron-repressible outer membrane proteins. However, only the type and biotype 2 strains could acquire iron from porcine transferrin and no organism could utilize human, bovine or ovine transferrin, or ovine or porcine lactoferrin; haemoglobin supported good growth of all strains except strain K17. In all cases, iron acquisition from transferrin and haemoglobin required direct contact between the organisms and the proteins indicating the existence of specific receptors. An affinity isolation technique, using biotinylated porcine transferrin plus streptavidin-agarose, allowed the isolation of the following polypeptides from total membranes of organisms grown under iron-restricted conditions: 99 kDa and 64 kDa from ATCC 27088; 93 kDa from Bertschinger 2008/76; 95 kDa (trace amounts) and 60 kDa from BC181; none from K17. These results indicate that the 93-99 kDa polypeptides are involved in the acquisition of iron from porcine transferrin and that the inability of strain K17 to use transferrin as an iron source is due, probably, to the lack of, or a defect in, an analogous component.

Identification and characterization of genes encoding the human transferrin-binding proteins from Haemophilus influenzae

Haemophilus influenzae, a strict human pathogen, acquires iron in vivo through the direct binding and removal of iron from human transferrin by an as yet uncharacterized process at the bacterial cell surface. In this study, the tbpA and tbpB genes of H. influenzae, encoding the transferrin-binding proteins Tbp1 and Tbp2, respectively, were cloned and sequenced. Alignments of the H. influenzae Tbp1 and Tbp2 protein sequences with those of related proteins from heterologous species were analyzed. On the basis of similarities between these and previously characterized proteins, Tbp1 appears to be a member of the TonB-dependent family of outer membrane proteins while Tbp2 is lipid modified by signal peptidase II. Isogenic mutants deficient in expression of Tbp1 or Tbp2 or both proteins were prepared by insertion of the Tn903 kanamycin resistance cassette into cloned sequences and reintroduction of the interrupted sequences into the wild-type chromosome. Binding assays with the mutants showed that a significant reduction in transferrin-binding ability resulted from the loss of either of the Tbps and a complete loss of binding was evident when neither protein was expressed. Loss of either Tbp2 or both proteins correlated with an inability to grow on media supplemented with transferrin-bound iron as the sole source of iron, whereas the Tbp1+ Tbp2- mutant was able to grow only at high transferrin concentrations.

Iron piracy: acquisition of transferrin-bound iron by bacterial pathogens

The mechanism of iron utilization from transferrin has been most extensively characterized in the pathogenic Neisseria species and Haemophilus species. Two transferrin-binding proteins, Tbp1 and Tbp2, have been identified in these pathogens and are thought to be components of the transferrin receptor. Tbp1 appears to be an integral, TonB-dependent outer membrane protein while Tbp2, a lipoprotein, may be peripherally associated with the outer membrane. The relative contribution of each of these proteins to transferrin binding and utilization is discussed and a model of iron uptake from transferrin is presented. Sequence comparisons of the genes encoding neisserial transferrin-binding proteins suggest that they are probably under positive selection for variation and may have resulted from inter-species genetic exchange.

Iron uptake mechanisms of pathogenic bacteria

Most of the iron in a mammalian body is complexed with various proteins. Moreover, in response to infection, iron availability is reduced in both extracellular and intracellular compartments. Bacteria need iron for growth and successful bacterial pathogens have therefore evolved to compete successfully for iron in the highly iron-stressed environment of the host's tissues and body fluids. Several strategies have been identified among pathogenic bacteria, including reduction of ferric to ferrous iron, occupation of intracellular niches, utilisation of host iron compounds, and production of siderophores. While direct evidence that high affinity mechanisms for iron acquisition function as bacterial virulence determinants has been provided in only a small number of cases, it is likely that many if not all such systems play a central role in the pathogenesis of infection.

Characterization of an immunoreactive 17.5-kilodalton outer membrane protein of Haemophilus somnus by using a monoclonal antibody

A single outer membrane protein (OMP) of Haemophilus somnus, with an apparent molecular mass of 17.5 kDa, was identified in the sodium dodecyl sulfate (SDS)-insoluble fraction after extraction with 1% SDS-0.5 M NaCl-0.1% beta-mercaptoethanol. A hybridoma derived from mice immunized with H. somnus OMP fractions produced a monoclonal antibody (MAb), designated 20-3-5, that bound to the 17.5-kDa OMP of H. somnus. The MAb 20-3-5 epitope was present on 45 of 45 strains of H. somnus tested. MAb 20-3-5 cross-reacted with Haemophilus agni, Histophilus ovis, and Haemophilus haemoglobinophilus but not with 13 other species and subspecies of gram-negative bacteria. Immunoelectron-microscopic and antibody absorption studies revealed that the MAb 20-3-5 epitope is exposed on the surface of bacteria. In an immunoblot analysis, convalescent-phase sera obtained from calves with experimental H. somnus pneumonia contained antibodies to the 17.5-kDa OMP of H. somnus. Future studies will be directed toward examining the role of the 17.5-kDa OMP in immunity to H. somnus infections.

The region of human transferrin involved in binding to bacterial transferrin receptors is localized in the C-lobe

Iron-saturated human transferrin was digested with either chymotrypsin or trypsin to produce C-lobe and N-lobe protein fragments. Individual protein fragments were purified by a combination of gel filtration and Concanavalin A affinity chromatographic procedures. The C-lobe and N-lobe fragments of human transferrin were then used in binding assays to assess their ability in binding to the bacterial transferrin receptors. Competitive binding assays demonstrated that the C-lobe fragment of human transferrin binds as well as intact human transferrin to bacterial transferrin receptors from Neisseria meningitidis, Neisseria gonorrhoeae and Haemophilus influenzae. Using isogenic mutants of N. meningitidis deficient in either of the transferrin-binding proteins (Tbps), we demonstrated that both transferrin-binding proteins were able to bind to the C-lobe fragment of human transferrin.

Differences in protein expression of Haemophilus somnus grown under conditions of iron-restriction

Outer membrane protein profiles were compared in 14 H. somnus strains isolated from brain and lung lesions as well as from the genital tract of asymptomatic carriers during in vitro growth under iron-restricted conditions. Ethylenediamine-di-O-hydroxyphenyl acetic acid (EDDA) was used to obtain iron-restricted conditions in media used for this study. The outer membrane protein profiles were studied by the discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoretic system (SDS-PAGE), and the proteins were stained with silver or transferred to nitrocellulose sheets and western blots conducted. Growth under iron-restricted conditions resulted in the induction of outer membrane proteins in most H. somnus strains examined. Studies also indicated differences among H. somnus strains in the number of induced proteins and their molecular weights but the results did not indicate a specific relationship between these strain-dependent differences and tissue trophism. Western blot analysis revealed a high degree of immunological relatedness among strains of H. somnus in their iron-regulated proteins. However, hyperimmune serum used in these assays failed to recognize certain iron-regulated proteins expressed by some H. somnus strains, a finding which may have important implications for the induction of protective immunity in cattle against this bovine pathogen.

Characterization of a heat-modifiable outer membrane protein of Haemophilus somnus

In immunoblot analysis, a murine monoclonal antibody (MAb), 27-1, which was produced to an outer membrane protein (OMP) of Haemophilus somnus, showed that a major OMP is heat modifiable, having a molecular mass of 28 kDa when the N-lauroylsarcosine-insoluble OMP preparation was solubilized at 60 degrees C and a mass of 37 kDa when the OMP preparation was solubilized at 100 degrees C. The heat-modifiable OMP reacted intensely with convalescent sera obtained from calves with experimental H. somnus pneumonia in immunoblot analysis. Immunoelectron microscopic and antibody absorption studies revealed that the MAb 27-1 epitope was not surface exposed on the intact bacterium. However, a decrease in antibody reactivity to the heat-modifiable OMP in immunoblot analysis after absorption of convalescent serum with intact bacterial cells of H. somnus suggests that a surface-exposed portion of the heat-modifiable OMP is expressed on the intact bacterium. MAb 27-1 reacted with 45 of 45 strains of H. somnus tested in immunoblot analysis. The apparent molecular mass of the antigen varied among strains, and five reactivity patterns demonstrated by MAb 27-1 were observed. MAb 27-1 also reacted with six species in the family Pasteurellaceae, Escherichia coli, and Salmonella dublin, but not with the other eight species of gram-negative bacteria. The heat-modifiable OMP of H. somnus showed immunological cross-reactivity with the OmpA protein of E. coli K-12 and significant N-terminal amino acid sequence homology with the OmpA proteins of gram-negative bacteria. We conclude that a major, 37-kDa heat-modifiable OMP of H. somnus, which elicits an antibody response in H. somnus-infected animals, is a common antigen among H. somnus strains tested and is structurally related to the OmpA protein of E. coli.

Molecular cloning, nucleotide sequence, and characterization of lppB, encoding an antigenic 40-kilodalton lipoprotein of Haemophilus somnus

Haemophilus somnus is a facultative intracellular pathogen which causes a wide range of diseases in cattle. To identify putative virulence determinants, a genomic library of H. somnus in Escherichia coli was screened for Congo red binding, a property associated with virulence in pathogenic bacteria, and subsequently with bovine hyperimmune sera raised against H. somnus HS25. A Congo red-binding clone carrying a 1.8-kb DNA insert was found to encode a strongly seroreactive LppB protein with an apparent molecular weight of 40,000. The nucleotide sequence of the entire DNA insert was determined. Two open reading frames coding for polypeptides with calculated molecular weights of 21,893 and 30,721 were identified. The larger open reading frame encoded LppB, while the smaller reading frame encoded a nonseroreactive protein with a relative molecular mass of approximately 18 kDa. The 16 amino-terminal amino acids of the deduced LppB polypeptide showed strong sequence homology to the signal peptide of secreted bacterial proteins, and the sequence Leu-Ala-Ala-Cys at the putative cleavage site corresponds to the consensus cleavage sequence of bacterial lipoproteins. Synthesis of the mature LppB lipoprotein in H. somnus was inhibited by globomycin, a specific inhibitor of signal peptidase II. LppB was localized to the outer membrane of H. somnus.

Interaction of ruminant transferrins with transferrin receptors in bovine isolates of Pasteurella haemolytica and Haemophilus somnus

The interactions of ruminant transferrins with receptors on bovine isolates of Pasteurella haemolytica and Haemophilus somnus were compared by growth studies and direct and competitive binding assays. Isolates of P. haemolytica were capable of utilizing and binding transferrin from sheep, goat, or cattle, whereas isolates of H. somnus were capable of utilizing and binding only bovine transferrin.

Effect of porphyrins and host iron transport proteins on outer membrane protein expression in Porphyromonas (Bacteroides) gingivalis: identification of a novel 26 kDa hemin-repressible surface protein

Porphyromonas gingivalis is capable of in vitro growth when iron sources are either complexed to hemin or host iron transport proteins, or exist in an inorganic form. This study examined the effect of these iron sources on outer membrane protein (OMP) expression in P. gingivalis W50. Hemin (iron) starved P. gingivalis was transferred into growth medium containing hemin, hemoglobin, hemin-saturated human serum albumin, hemin-free human serum albumin, transferrin, lactoferrin, or inorganic iron. Surface proteins were identified by 125I-labeling and resolved by SDS-PAGE and autoradiography. When grown under hemin starved conditions, P. gingivalis W50 and related strains expressed a major 26 kDa OMP, as revealed by 125I-autoradiography. Autoradiographic analysis demonstrated the absence of this 26 kDa OMP from the P. gingivalis surface in hemin-containing environments. Growth of P. gingivalis W50 in the presence of host iron transport proteins (hemin-free) or inorganic iron resulted in surface expression of a 26 kDa OMP. The presence of protoporphyrin IX or substitution of hemin-associated iron with zinc, resulted in continued surface expression of the 26 kDa OMP, indicating that repressibility of this OMP required the coordination of iron to the protoporphyrin IX molecule (i.e. hemin). A survey of 125I-labeled OMPs from several hemin starved P. gingivalis and related strains, demonstrated that a hemin-repressible 26 kDa OMP occurred only in P. gingivalis. We report here a newly described 26 kDa hemin-regulated surface protein occurring in several strains of P. gingivalis which is expressed on the cell surface in hemin starved conditions and is lost from the cell surface in response to an environment containing iron coordinated specifically to protoporphyrin IX (i.e. hemin).