Legionella pneumophila LbtU Acts as a Novel, TonB-Independent Receptor for the Legiobactin Siderophore

Abstract

Gram-negative Legionella pneumophila produces a siderophore (legiobactin) that promotes lung infection. We previously determined that lbtA and lbtB are required for the synthesis and secretion of legiobactin. DNA sequence and reverse transcription-PCR (RT-PCR) analyses now reveal the presence of an iron-repressed gene (lbtU) directly upstream of the lbtAB-containing operon. In silico analysis predicted that LbtU is an outer membrane protein consisting of a 16-stranded transmembrane β-barrel, multiple extracellular domains, and short periplasmic tails. Immunoblot analysis of cell fractions confirmed an outer membrane location for LbtU. Although replicating normally in standard media, lbtU mutants, like lbtA mutants, were impaired for growth on iron-depleted agar media. While producing typical levels of legiobactin, lbtU mutants were unable to use supplied legiobactin to stimulate growth on iron-depleted media and displayed an inability to take up iron. Complemented lbtU mutants behaved as the wild type did. The lbtU mutants were also impaired for infection in a legiobactin-dependent manner. Together, these data indicate that LbtU is involved in the uptake of legiobactin and, based upon its location, is most likely the Legionella siderophore receptor. The sequence and predicted two-dimensional (2D) and 3D structures of LbtU were distinct from those of all known siderophore receptors, which generally contain a 22-stranded β-barrel and an extended N terminus that binds TonB in order to transduce energy from the inner membrane. This observation coupled with the fact that L. pneumophila does not encode TonB suggests that LbtU is a new type of receptor that participates in a form of iron uptake that is mechanistically distinct from the existing paradigm.

Legionella pneumophila is the primary agent of Legionnaires' disease, a potentially fatal form of pneumonia (45, 46). This bacterium is particularly pathogenic for the immunocompromised, the elderly, and smokers, and recent studies have highlighted the growing significance of travel-associated disease and cases among middle-aged adults (96, 110). In aquatic environments, Gram-negative L. pneumophila survives planktonically, within multiorganism biofilms, and as an intracellular parasite of protozoa (43, 80, 129). Infection occurs after the inhalation of contaminated droplets that originate from a variety of aerosol-generating devices (37, 102). In the lung, L. pneumophila grows in alveolar macrophages, although persistence can also involve growth in epithelia and extracellular survival (2, 33, 98). The organism undergoes various forms of differentiation as it moves between intracellular and extracellular habitats (40, 95). Much of the pathogenesis and ecology of L. pneumophila is mediated by a vast array of secreted proteins that are released into the extracellular milieu or host cells via type II and type IV secretion systems (32, 49, 67, 98, 120, 131).

Iron acquisition is another critical aspect of L. pneumophila physiology, intracellular infection, and virulence (30, 31, 34). The mediators of Fe²⁺ assimilation include an inner membrane (IM) Fe²⁺ transport (FeoB) system and a secreted pyomelanin pigment that has Fe³⁺ reductase activity (26, 111). The principal means of L. pneumophila Fe³⁺ assimilation is legiobactin. When L. pneumophila is grown in a low-iron, chemically defined medium (CDM), it secretes this low-molecular-weight siderophore, which is most readily detected by the chrome azurol S (CAS) assay (3, 84). Legiobactin is also able to stimulate the growth of iron-starved legionellae, including the wild type and a feoB mutant (2, 3). Legiobactin contains 13 aliphatic carbons (3 carbonyls) and no aromatic carbons, and spectra further indicate that it is a polycarboxylate (2). Many but not all other Legionella species appear to express legiobactin (3, 123). Two linked genes, lbtA and lbtB, are required for the expression of legiobactin; i.e., supernatants from mutants with lbtA or lbtB inactivated lack CAS reactivity and show a complete inability to stimulate the growth of iron-starved legionellae (2, 3). LbtA has homology to siderophore synthetases, and LbtB is akin to IM siderophore exporters. Thus, cytoplasmic LbtA is likely involved in the synthesis of legiobactin, whereas LbtB promotes the transit of the siderophore across the IM prior to its final export. Importantly, lbtA mutants, but not their complemented derivatives, are defective for infection of the murine lung, demonstrating a role for legiobactin in L. pneumophila virulence (2). In the current study, we report the identification and characterization of an iron-regulated outer membrane (OM) protein (LbtU) that promotes the utilization of legiobactin and therefore is likely to be the receptor for the Legionella siderophore.

MATERIALS AND METHODS

Bacterial strains.

L. pneumophila 130b (ATCC BAA-74, also known as AA100 or Wadsworth) served as our wild type (2). This serogroup 1 strain is a virulent clinical isolate. NU302 is a mutant derivative of 130b that contains a deletion in lbtA, rendering the strain incapable of making legiobactin (3). NU302(plbtA) is the complemented version of NU302, carrying lbtA on low-copy-number plasmid pMMB2002 (3). NU269 is a 130b feoB mutant (111). The Legionella longbeachae strain (ATCC 33462) used in this study was also previously described (90). Escherichia coli DH5α was the host for recombinant plasmids (Invitrogen, Carlsbad, CA).

Bacteriological media and extracellular growth experiments.

L. pneumophila strains were routinely cultured at 37°C on buffered charcoal yeast extract (BCYE) agar, which has an iron supplement consisting of 0.25 g of ferric pyrophosphate per liter (3). When appropriate, the agar was supplemented with chloramphenicol at 6 μg/ml, kanamycin at 25 μg/ml, or gentamicin at 2.5 μg/ml. To monitor the basic extracellular growth capacity of L. pneumophila, bacteria grown on BCYE agar were inoculated into buffered yeast extract (BYE) or CDM broth, and then the optical densities at 660 nm (OD₆₆₀) of the cultures were determined (62, 84, 108, 132). BYE and CDM broths also typically contain the iron supplement. E. coli cells were grown in Luria-Bertani media containing kanamycin (50 μg/ml), gentamicin (2.5 μg/ml), chloramphenicol (30 μg/ml), or ampicillin (100 μg/ml). Chemicals were from Sigma-Aldrich (St. Louis, MO).

To assess the extracellular growth of L. pneumophila under iron-limiting conditions, strains were inoculated in deferrated CDM and growth was monitored spectrophotometrically (3). To judge growth on iron-limited solid media, legionellae were tested for their ability to form colonies on BCYE agar that lacked its iron supplement (3, 111, 133). Also, strains were tested on non-iron-supplemented BCYE agar that had been made even more iron limited by inclusion of 400 μM 2,2-dipyridyl (DIP) or 10 μM deferoxamine mesylate (DFX) (3). In all cases, bacteria were precultured for 3 days on standard BCYE agar and suspended in phosphate-buffered saline (PBS) at 1 × 10⁹ CFU per ml, and then 10-μl aliquots taken from 10-fold serial dilutions in PBS were spotted on the assay media. Growth was recorded after 5 days of incubation at 37°C.

Siderophore assays.

In order to assess legiobactin production, secretion, and utilization, L. pneumophila strains were grown in deferrated CDM at 37°C, and then at 24 and 48 h postinoculation, cell-free culture supernatants were obtained as described previously (2, 3). Siderophore activity within the supernatants was quantified using the CAS assay as previously done, with DFX serving as the standard (2, 3, 84, 123). Supernatants were tested for siderophore biological activity by examining their ability to promote the growth of the NU269 feoB mutant on non-iron-supplemented BCYE agar (2, 3). NU269 lacks an IM Fe²⁺ permease and thus is defective for uptake of Fe²⁺ but not Fe³⁺ (111). This mutant's growth deficit can be reversed by the addition of Fe³⁺ salts or supernatants containing legiobactin (3). To compare wild-type and mutant L. pneumophila strains for their ability to use legiobactin, bacteria were precultured for 3 days on standard BCYE agar and suspended in PBS, and then 1 × 10⁴ CFU was spread onto non-iron-supplemented BCYE agar containing 400 μM DIP (3). Small wells cut in the center of the agar were filled with 75 μl of supernatants obtained from deferrated CDM cultures. Control wells contained equal volumes of deferrated CDM and 5 μM Fe³⁺ pyrophosphate or 20 μM Fe²⁺ ammonium sulfate. Growth around the wells was assessed after incubation at 25°C for 8 to 10 days.

In order to judge the ability of heterologous bacteria containing a cloned lbtU gene to bind legiobactin, L. longbeachae carrying the plasmid pLbtU (see below) was grown for 18 h in non-iron-supplemented BYE broth, which contained, during the final 3 to 4 h of incubation, 0.1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for the purpose of inducing LbtU expression from the vector Ptac promoter. The bacteria were then collected by centrifugation, and following a wash step in 50 mM MOPS (morpholinepropanesulfonic acid) buffer (pH 5.8), the bacterial cells were mixed with legiobactin and incubated for various periods of times at 37°C. Next, following the removal of the bacteria (and any bound legiobactin) by centrifugation and subsequent filter sterilization, the amount of legiobactin remaining in the supernatant (i.e., the residual unbound siderophore) was measured using the CAS assay as described above. The legiobactin used in these experiments was obtained by loading acidified (pH < 1) supernatants obtained from deferrated CDM cultures of wild-type 130b cultures (2, 3) onto Oasis WAX SPE cartridges (Waters Corp., Milford, MA) that had been prepared according to the manufacturer's instructions. After the cartridges were washed with pH 3.0 and pH 8.5 MOPS buffer (50 mM), legiobactin was eluted from the columns with pH 12.2 MOPS buffer.

Bacterial uptake of radiolabeled iron.

L. pneumophila strains were cultured for 18 h in deferrated CDM as described above. ⁵⁵FeCl₃ (Perkin Elmer, Downers Grove, IL) was then added to a final concentration of 0.04 mCi/ml (i.e., 320 nM ⁵⁵Fe). After various times of incubation at 25°C, 1 ml of the cell suspension (n = 3) was filtered through a 0.45-μm nitrocellulose MF filter (Millipore, Bedford, MA), and then the recovered bacteria were washed with 5 ml of 0.5% thioglycolate (prepared fresh) to remove external iron salts. Counts per minute (cpm) of radioactivity associated with the bacterial cells were then measured using a Beckman LS6500 scintillation counter and are reported as the averages of the numbers of counts per minute recorded over a 5-min period. As a negative control, the nonspecific binding of ⁵⁵FeCl₃ to the filters in the absence of bacteria was determined and found to be at or below 600 cpm. In order to judge the contribution of proton motive force (PMF) to uptake, the PMF inhibitor KCN was added to the bacterial culture just prior to the addition of the radiolabeled iron at a final concentration of 10 mM (38, 69, 125).

DNA and protein sequence analysis.

DNA was isolated from L. pneumophila as described before (35). DNA sequencing was done at the Northwestern Biotech Laboratory, with primers obtained from Integrated DNA Technologies (Coralville, IA). DNA sequences were analyzed using Lasergene (DNASTAR, Madison, WI). Protein alignments were done using the Clustal method of Lasergene MegAlign. BLAST homology searches were done through GenBank at the National Center for Biotechnology Information (NCBI) and L. pneumophila databases at http://genolist.pasteur.fr/LegioList/, http://www.ncbi.nlm.nih.gov/genomeprj/48801, and http://www.ebi.ac.uk/ena/data/view/FR687201. Signal sequences were detected by SignalP (99), and predictions for the cellular location of proteins were done by PSORTb (53). Predictions involving topology and β-barrel OM proteins were obtained using PRED-TMBB (http://biophysics.biol.uoa.gr/PRED-TMBB/) and ConBBPRED (http://biophysics.biol.uoa.gr/ConBBPRED/index.jsp), with the latter being a combination of PRED-TMBB, TMBETA, B2TMPRED, and PSIPRED (7, 8, 57, 68, 70). Three-dimensional (3D) structural modeling was performed using the I-TASSER (http://zhanglab.ccmb.med.umich.edu/I-TASSER/) and Phyre (http://www.sbg.bio.ic.ac.uk/∼phyre/) modeling servers (11, 21, 54, 83, 138). Only models within the published threshold were considered; i.e., the LbtU model generated by the Phyre server exhibited 95% confidence, and the model produced by the I-TASSER server gave a C score of −0.86 and a total membrane (TM) score of 0.61 ± 0.14. The iMol 3D structural visualization program (http://www.pirx.com/iMol/) was used to make figures for presentation.

RT-PCR analysis of L. pneumophila gene expression.

Reverse transcription (RT)-PCR was performed as previously described (3, 85). RNA was isolated from 18-h CDM cultures of L. pneumophila using the RNA STAT-60 reagent according to the manufacturer's instructions (Tel-Test B, Inc., Friendswood, TX). Total cDNA was amplified with random hexamers (Invitrogen) and then detected using standard PCR. Primer pairs used for amplifying lbtA and mip were as described before (3). Primers CC01 (5′-AAAATTGACCTACGATGGGC-3′) and CC03 (5′-TTCGTTGGTTTGTCCCAA-3′) were used to assess lbtU transcription. To assess potential cotranscription of lbtU and lbtA, primers CC02 (5′-ATCTAAACATTGCTGGTGCG-3′) and CC04 (5′-GCAACCCCTTTTCCTTATCTT-3′) were employed. Control experiments in which reverse transcriptase was omitted from the reaction mixture were done to rule out contributions from contaminating DNA in the DNase-treated samples. Relative, endpoint PCR mixtures were separated by agarose-gel electrophoresis and proteins detected with ethidium bromide staining (3, 85).

Cloning, mutant construction, and genetic complementation.

As the first step toward obtaining an lbtU mutant, lbtU was amplified by PCR from the genomic DNA of strain 130b using CC01 and CC02, and the resulting 1.2-kb PCR fragment was cloned into pGEM-T Easy (Promega, Madison, WI), to give plasmid pGlbtU. Next, the lbtU-containing EcoRI fragment of pGlbtU was cloned into pUC18. The resulting plasmid, pUClbtU, was digested with StyI, treated with the Klenow fragment, and ligated to a HincII fragment of pMB2190 which carries a kanamycin resistance (Km^r) gene (114). The result was a plasmid (pUClbtUK) carrying lbtU with a Km^r insertion circa halfway through the gene's coding region. pUClbtUK was introduced into 130b by transformation (3, 114, 124), and transformants were selected on antibiotic-containing BCYE agar. Insertion of Km^r into the chromosomal lbtU gene was confirmed by PCR using CC01 and CC02. Two independently derived lbtU mutants were designated NU383 and NU384. Double mutants (NU385 and NU386) with both lbtU and lbtA inactivated were obtained by transforming pUClbtUK into lbtA mutant NU302 and selecting for acquisition of Km^r.

For trans-complementation analysis of the lbtU mutant, a 1.6-kb PCR fragment containing the lbtU coding region plus 200 bp of upstream DNA was amplified from 130b DNA using Vent polymerase (Promega) and primer CC05 (5′-GCTGATTTTCTAGAGTATCCATAAAGCATA-3′) and CC02. The resulting fragment was ligated into the SmaI site of pMMB2002 (114), and the new plasmid pLbtU was electroporated (3) into mutant NU383. Chloramphenicol resistant (Cm^r) clones were confirmed as carrying pLbtU by PCR using the vector-specific primer OR77 (5′-TCGGCTCGTATAATGTGTGG-3′) with CC02. Complementation analysis was also done by making a derivative of NU383 that had lbtU inserted into a neutral site in the chromosome (86). To that end, the 1.4-kb HincII fragment of pX1918 containing a gentamicin resistance (Gm^r) cassette (3) was inserted into the BamHI site of pLbtU, yielding plasmid pLbtUGnt. Next, the SacI/SalI fragment of pLbtUGnt containing lbtU and Gm^r was cloned into SacI/XhoI-digested pZL1153 (86), resulting in pZLLbtUGnt, in which lbtU and the Gm^r gene were flanked by ca. 1,000 bp of Legionella DNA encoding open reading frames (ORFs) lpg2528 and lpg2529. After transformation of pZLLbtUGnt into mutant NU383, a strain (designated NU387) in which intact lbtU was inserted between lpg2528 and lpg2529 was obtained. Insertions at this site have been shown to not affect the ability of L. pneumophila to grow in standard media and infection assays (86).

Plasmid pLbtU and its vector control (pMMB2002) were introduced into L. longbeachae by electroporation (3), and the sequences of Cm^r clones were confirmed as detailed above.

Construction of His-tagged LbtU.

For localizing LbtU in cell fractions, a six-His-tagged version of LbtU (LbtU-His₆) was constructed. To that end, lbtU was amplified by PCR using 130b DNA and primers LbtUBamHI (5′-AAAAAAGATCCAAAACCAAATGCTGTATATAAGAA-3′) and LbtUHis (5′-TTAATGATGATGATGATGATGAAAATTCAGTATGGTTTGTAATGC-3′). Primer LbtUHis had been designed to incorporate a hexahistidine tag directly before the stop codon of LbtU. After the PCR product was cloned into pGEM-T Easy, yielding pGEMLbtUHis, an EcoRI fragment containing lbtU::His₆ was excised and cloned into EcoRI-digested pMMB2002. The resultant plasmid, pLbtUHis₆, had lbtU::His₆ cloned under the control of an IPTG-inducible Ptac promoter. Plasmid pLbtUHis₆ was electroporated into 130b and mutant NU383, and Cm^r transformants were isolated on antibiotic-containing BCYE agar.

Bacterial fractionation and protein localization.

Fractionation of L. pneumophila was done using a modification of protocols published by Roy et al. and Vincent et al. (115, 131). Briefly, bacteria carrying pLbtUHis₆ were grown overnight at 37°C in chloramphenicol-containing BYE broth with 1 mM IPTG to induce the expression of LbtU-His₆. After centrifugation of the culture, the bacterial pellet was resuspended in 0.5 ml of 1 M sucrose, 50 mM Tris-HCl, pH 8. Spheroplasts were then produced by the addition of 10 μl of 0.5 M EDTA (pH 8), 10 μl of 10 mg/ml hen egg white lysozyme, and 125 μl of bacterial protease inhibitors (Sigma catalog no. P8465). After the sample was brought to 2 ml by the addition of double-distilled water (ddH₂O) and centrifuged for 10 min at 5,000 × g, the pellet containing the spheroplasts was resuspended in 2 ml of 50 mM Tris-HCl (pH 8) and then lysed by three rounds of 15-s sonication (15% output on a model 450 Branson Sonifier). Residual, unlysed bacteria were removed from the suspension by a 10-min centrifugation at 5,000 × g. Total membranes were isolated from the lysates by centrifugation of the sample at 100,000 × g for 1 h. The membrane-containing pellet was resuspended in 1 ml of 50 mM Tris-HCl (pH 8), and Triton X-100 was added to a final concentration of 1%. Triton X-100-insoluble (i.e., OM) proteins were then separated from the Triton-soluble (i.e., IM) proteins by a second round of 100,000 × g centrifugation for 1 h. The OM protein preparation was then added to 200 μl of 50 mM Tris-HCl (pH 8) plus 200 μl of SDS-PAGE sample buffer.

In order to detect LbtU-His₆ in cellular fractions, a volume of each fraction that was equivalent to 200 μl of the initial BYE culture was boiled in SDS-PAGE sample buffer and separated by 12% SDS-PAGE. Proteins were then transferred to a polyvinylidene difluoride membrane using a semidry transfer system (Bio-Rad, Hercules, CA). After being incubated for 1 h at room temperature in TBS-T (Tris-buffered saline plus 0.1% Tween 20) with 5% skim milk and 1% bovine serum albumin (BSA), the membrane was incubated with murine anti-His antibodies (Genscript, Piscataway, NJ) at a 1:1,000 dilution in blocking buffer at 4°C for 16 h. After being washed with TBS-T, the filter was incubated with horseradish peroxidase-labeled, anti-mouse IgG (Rockland, Gilbertsville, PA) at a 1:5,000 dilution for 1 h at room temperature. After being washed, the blots were developed with Pierce enhanced-chemiluminescence (ECL) reagents (Thermo, Rockford, IL), and the signal identifying the LbtU-His₆-antibody reactions was detected using X-ray film. To confirm the nature of the fractions, immunoblotting was also done using antibodies against the major OM protein (MOMP) LepB, isocitrate dehydrogenase (ICDH), and known OM, IM, and soluble proteins, as previously performed (131).

Infection assays.

To examine the ability of L. pneumophila strains to grow intracellularly, Hartmannella vermiformis and Acanthamoeba castellanii amoebae, human U937 macrophages, A549 lung epithelial cells, and A/J mouse bone marrow-derived macrophages were infected as previously described (2, 3, 104, 113). To assess the virulence of L. pneumophila, strains were inoculated into the tracheas of 6- to 8-week-old A/J mice (Jackson Laboratory, Bar Harbor, ME) (19). As done previously (2, 42, 114), to determine the relative abilities of strains to replicate and survive in mouse lungs, groups of mice (n = 5) were infected separately with 10⁶ CFU of wild-type and mutant bacteria, and at 0, 24, 48, and 72 h postinoculation, the numbers of CFU in the lungs were determined by plating cells on BCYE agar. In vivo competition assays were also done as we have described before (2, 42, 111-114); mice were inoculated with 10⁵ CFU of a ca. 1:1 ratio of wild-type and mutant bacteria, and then 1 and 3 days later, the ratios of wild-type to mutant cells in lung homogenates were determined by plating. Animal experiments were approved by the Animal Care and Use Committee of Northwestern University.

Nucleotide sequence accession number.

The NCBI and GenBank accession number for the lbtU gene of L. pneumophila strain 130b is BankIt1386894 LbtU HQ207565.

RESULTS

Identification of the iron-regulated lbtU gene.

Previous mutational analysis determined that lbtA and lbtB are required for the synthesis and secretion of legiobactin by L. pneumophila strain 130b (3). Previous work also identified the gene lbtC as being transcriptionally linked to lbtA and lbtB; however, mutational analysis failed to identify a role for lbtC in legiobactin production (3). Further sequencing of the 130b chromosome directly upstream of the operon containing lbtA, lbtB, and lbtC revealed the presence of an ORF (lbtU) that appeared to be iron regulated based upon the presence of a putative Fur box in its promoter region (Fig. 1A). The Fur box had three mismatches compared to the consensus Fur box of E. coli, like the promoter regions of the L. pneumophila iron-regulated genes lbtA (Fig. 1A), frgA, feoB, and fur (3, 61, 62, 111). RT-PCR analysis confirmed that the lbtU ORF was monocistronic and not cotranscribed with the lbtABC operon (data not shown). Further RT-PCR analysis indicated that the levels of mRNA from this ORF, like those from lbtA, were greater in L. pneumophila 130b grown in deferrated CDM than in bacteria cultured in iron-supplemented CDM (Fig. 1B), confirming that the gene is sensitive to iron repression. Based on this information and the proximity of the ORF to lbtABC, we hypothesized that the gene might encode a function that is related to legiobactin. For reasons that will become apparent, we designated the ORF lbtU for legiobactin utilization gene and its product LbtU. In the recently completed genome sequence of strain 130b, lbtU appears as lpw_13361 (118). Examination of the other L. pneumophila databases indicated that lbtU is conserved among the five other sequenced strains of L. pneumophila, being denoted lpp1281, lpg1326, lpl1279, lpc0741, and lpa01956 in strains Paris, Philadelphia, Lens, Corby, and Alcoy, respectively (25, 27, 41, 55). The arrangement of the lbt genes was also conserved, with the lengths of the region between lbtU and lbtABC being 132 to 133 bp, depending on the strain. These data are compatible with the fact that legiobactin expression is broadly conserved among strains of L. pneumophila (3). LbtU was predicted to encode a 321-aa protein that is 96 to 98% identical among the six strains examined (data not shown). In all cases, the LbtU sequence contained an N-terminal signal sequence, indicating that the protein is secreted across the Legionella IM via the Sec system (34, 77) and therefore might be involved in some aspect of legiobactin transport.

FIG. 1.

Location and iron regulation of lbtU. (A) Depiction of the region of the L. pneumophila chromosome containing lbtU and lbtABC. The sequences of the two regions predicted to mediate Fur binding are shown below. (B) Expression of lbtU transcripts. Wild-type 130b was grown in deferrated CDM (CDM − Fe) or deferrated CDM that was supplemented with 5 μM ferric pyrophosphate (CDM + Fe), and then RNA was extracted and analyzed by RT-PCR utilizing primers specific to either lbtU (top row), lbtA, a gene that is known to be iron repressed (middle row), or mip, a gene that is not subject to iron regulation (bottom row) (3). That the PCR products obtained resulted from mRNA templates was confirmed by the lack of product obtained when the PCR did not incorporate RT (−RT lanes). Additional PCR products obtained from genomic DNA appear in the left-most lane, indicating that the mRNAs observed are full length. The results presented are representative of at least three independent experiments.

The outer membrane localization of LbtU.

Further bioinformatics analysis predicted that LbtU is an OM protein, consisting of a 16-stranded transmembrane β-barrel, eight extracellular loops, and short N-terminal and C-terminal tails in the periplasmic space (Fig. 2A). As has been done to investigate various types of OM proteins, including other iron transporters (22, 23, 50, 89, 106, 122, 127, 130), we introduced a His₆-tagged version of LbtU into 130b and then used anti-His antibodies and immunoblot analysis to track the location of LbtU in cellular fractions. LbtU-His₆ did not diminish the growth of the wild type and complemented an iron-related growth defect of an lbtU mutant (see below, Fig. 9), indicating that this form of LbtU functions normally in L. pneumophila. The ∼37-kDa LbtU was exclusively present within total membrane fractions (Fig. 3), being absent from the soluble fraction obtained from lysates and culture supernatants (data not shown). As predicted, LbtU was predominately in the OM fraction and minimally present in the IM portion (Fig. 3). As has been documented in past immunoblot analyses (131), a similar partitioning was detected for MOMP (Fig. 3), a known OM protein of L. pneumophila (51, 63). As further validation, the localization of LbtU and MOMP was in marked contrast to that of LepB, a signal peptidase from the IM, and ICDH, an enzyme from the cytoplasm (78, 131) (Fig. 3). Together, these data indicate that LbtU is an integral OM protein. Based on both the chromosomal location and the iron regulation of lbtU and the OM location of LbtU, we hypothesized that LbtU is involved in the transport of legiobactin across the OM.

FIG. 2.

Predicted OM location and transmembrane β-barrel of LbtU. Predictions were based upon the PRED-TMBB (A) and the ConBBPRED (B) servers. Typically for OM proteins, LbtU was analyzed without its signal sequence.

FIG. 3.

Outer membrane localization of LbtU. Cellular fractions were obtained from wild-type 130b expressing LbtU-His₆, separated by PAGE, and subjected to immunoblotting. The preparations tested included total cell lysates (Lyst.), total membrane (TM), inner membrane (IM), and outer membrane (OM), and the immunoblots utilized antibodies directed against His₆ (first panel), OM protein MOMP (second panel), IM protein LepB (third panel), and cytoplasmic protein ICDH (fourth panel). The anti-His immunoblot also shows the migration of molecular mass markers (in kDa). The OM localization of LbtU-His₆ was observed in three independent experiments.

Isolation of lbtU mutants.

In order to characterize the function of LbtU, we mutated lbtU in strain 130b, using allelic exchange, as we have done in the past (3). Two independent mutants (NU383 and NU384) were obtained and further examined. RT-PCR analysis confirmed that the mutation of lbtU eliminated the gene's transcription but did not abolish the expression of lbtABC (data not shown). Both of the lbtU mutants grew normally on BCYE agar and in BYE and CDM broth (data not shown), indicating that LbtU is not required for L. pneumophila growth in the standard, iron-replete media, as one would expect for a protein dedicated to siderophore function.

LbtU promotes the utilization of legiobactin by L. pneumophila.

To determine if LbtU is required for the synthesis or secretion of legiobactin, we grew the LbtU mutants in deferrated CDM and then assayed culture supernatants for the presence of both CAS reactivity and activity in the legiobactin bioassay. The supernatants obtained from mutants NU383 and NU384 had 90 to 100% of the CAS activity of the wild type and were fully capable of rescuing the growth of the L. pneumophila feoB mutant on non-iron-supplemented BCYE agar (Fig. 4). As was previously observed (2, 3), the lbtA mutant, which is unable to synthesize legiobactin, had a ca. 70% loss of CAS reactivity and a complete loss of bioactivity (Fig. 4). These data indicate that LbtU is not required for the production or export of legiobactin, leading to the hypothesis that this OM protein is involved in the import of a siderophore. Thus, we compared the wild type and lbtU mutant NU383 for their abilities to use legiobactin to stimulate growth on low-iron media. As before (3), wild-type 130b was stimulated to grow on non-iron-supplemented BCYE agar containing the iron chelator DIP when supplied with supernatants containing legiobactin (Fig. 5). In contrast, the lbtU mutant was unable to use legiobactin for growth enhancement (Fig. 5). When an intact copy of lbtU was reintroduced into the mutant, either at a neutral site within the chromosome or on a multicopy plasmid (pLbtU), legiobactin utilization was restored (Fig. 5). These data confirm that LbtU is required for the ability of L. pneumophila to utilize legiobactin for growth under low-iron conditions. Because the lbtU mutant, like the wild type and complemented strains, grew on iron-limited BCYE agar when it was supplied with Fe³⁺ or Fe²⁺ salts (Fig. 5), LbtU is not needed for the assimilation of all iron sources and may be specific for legiobactin usage.

FIG. 4.

Legiobactin production and secretion by wild-type L. pneumophila and lbtU mutants. (A) Wild-type (WT) 130b, the lbtA mutant NU302, and the lbtU mutant NU383 were grown in deferrated CDM, and then at 24 (gray bars) and 48 (white bars) hours postinoculation, the CAS reactivities of supernatants were examined. The CAS values (expressed as DFX equivalents) are the means and standard deviations of results from duplicate cultures. Unlike with the lbtA mutant, the CAS reactivity of the lbtU mutant was not significantly different from that of the wild type (P > 0.05; Student's t test). The result presented for the lbtU mutant is representative of data from eight experiments. (B) We plated feoB mutant bacteria onto the surface of non-iron-supplemented BCYE agar, and a center well was filled with a supernatant sample obtained from deferrated CDM cultures of the wild type (left), lbtA mutant NU302 (center), or lbtU mutant NU383 (right). The growth of the bacteria was then recorded. The results shown are representative of three experiments. In both assays, the lbtU mutant NU384 gave results similar to those of NU383 (data not shown).

FIG. 5.

Legiobactin utilization by the L. pneumophila wild type and lbtU mutants. We plated wild-type 130b (first column), the lbtU mutant NU383 (second column), the complemented mutant NU383 (pLbtU) (third column), or NU387 (a derivative of NU383 containing a complementing copy of lbtU in a neutral site in the chromosome) (fourth column) onto non-iron-supplemented BCYE agar containing DIP, and wells cut in the agar were filled with uninoculated CDM (top row), a CAS-positive, legiobactin-containing supernatant (Sup't.) obtained from deferrated CDM cultures of the wild type (second row), ferric pyrophosphate (third row), or ferrous ammonium sulfate (bottom row). Growth results presented are representative of at least three independent experiments.

To provide additional evidence for the role of LbtU in iron-siderophore uptake, we grew the wild type and the lbtU mutant L. pneumophila strain in deferrated CDM, leading to the production of legiobactin, and then determined the relative abilities of the different cultures to take up added radiolabeled Fe³⁺. Wild-type 130b displayed substantial iron uptake after 60 or 90 min of incubation in the presence of the radiolabel (Fig. 6). This uptake was significantly inhibited (P < 0.05, Student's t test) when the culture was treated with KCN, indicating that legiobactin-mediated iron uptake is dependent on the bacterial PMF. In marked contrast, mutant NU383 incorporated label at a level that was significantly below that of the wild type and was not significantly altered by KCN (Fig. 6). A complemented lbtU mutant incorporated radiolabeled Fe³⁺ at and above wild-type levels (Fig. 6), confirming that the lack of uptake shown by the lbtU mutant was due specifically to the loss of LbtU.

FIG. 6.

Iron uptake by the L. pneumophila wild type and lbtU mutants. Wild-type 130b (black bars), lbtU mutant NU383 (gray bars), and complemented mutant NU383 (pLbtU) (white bars) were grown in deferrated CDM, incubated with ⁵⁵FeCl₃ (for 60 or 90 min, as indicated), and then assayed for their incorporation of radiolabeled iron. A portion of each bacterial sample was also exposed to the PMF inhibitor KCN (various hatched bars) and then assayed for label uptake. Data are the means and standard deviations obtained from the results of three replicate cultures. In the samples lacking KCN, the amount of label incorporated into mutant samples was always significantly less than that of the wild type and complemented mutant (P < 0.05, Student's t test). The results presented for the wild type and mutant are representative of three independent experiments, whereas the complemented mutant was assayed on two occasions, with comparable results being obtained.

L. longbeachae is another species of Legionella whose genome has recently been sequenced (24, 75). Examination of the database revealed that L. longbeachae does not contain lbtU or the lbtABC operon, compatible with our previous inability to grow this species in deferrated CDM for the purpose of assessing legiobactin production (123). Therefore, as a means of demonstrating a direct role for LbtU in legiobactin uptake, we introduced the lbtU-containing plasmid pLbtU into L. longbeachae and examined transformants for their relative abilities to bind legiobactin. On six different occasions, L. longbeachae(pLbtU) bound significantly more legiobactin than did L. longbeachae containing only the pMMB2002 vector (Fig. 7). Considering all of the results obtained from the various phenotypic assays as well as the fact that LbtU is an OM protein, the simplest hypothesis to explain our data is that LbtU is a legiobactin receptor which binds and then mediates the transport of ferrisiderophore across the OM.

FIG. 7.

Legiobactin binding by L. longbeachae containing plasmid-borne lbtU. On three separate occasions, L. longbeachae containing the pMMB2002 vector (black bars) or the lbtU-containing plasmid pLbtU (white bars) were mixed with buffer containing legiobactin for either 60, 90, or 120 min of incubation, and then the amount of legiobactin bound by the bacteria was determined based upon the amount of CAS reactivity that was absorbed out of the solution. Data are the means and standard deviations obtained from the results of three replicate cultures. The amount of CAS reactivity absorbed by the lbtU-containing bacteria was significantly above that absorbed by the vector control (P < 0.05; Student's t test). The results obtained are representative of three additional trials that measured binding at 60, 90, or 120 min.

LbtU and legiobactin are required for L. pneumophila extracellular growth on low-iron media.

Previously, we determined that lbtA mutants are not impaired for growth in deferrated CDM broth or on BCYE agar that lacks their iron supplements or on unsupplemented BCYE agar that contains up to a 400 μM concentration of the iron chelator DIP (3). Compatible with these data, inactivation of lbtU did not compromise L. pneumophila growth in deferrated CDM broth or on non-iron-supplemented BCYE agar with or without 400 μM DIP (Fig. 8 and data not shown). However, when the more powerful iron chelator DFX was incorporated into the non-iron-supplemented BCYE agar, both the lbtA mutant NU302 and the lbtU mutant NU383 exhibited a >100-fold reduction in growth compared to that of the wild type (Fig. 8). Because the addition of ferric pyrophosphate, ferric chloride, or ferrous ammonium sulfate reversed the DFX-mediated growth inhibition (data not shown), the phenotype displayed by these mutants was related to iron starvation. The introduction of an intact copy of lbtA on a plasmid into NU302 restored the ability of that mutant to grow on the DFX-containing agar, indicating, for the first time, that LbtA and legiobactin are required for L. pneumophila growth under these iron-limiting conditions (Fig. 8). When lbtU was reintroduced into the mutant, either on a multicopy plasmid or at a neutral site in the chromosome, a wild-type level of growth was observed (Fig. 8). Complementation of the lbtU mutant was also achieved when a His₆-tagged copy of LbtU was used (Fig. 9). Taken together, these data indicate that LbtU is required for L. pneumophila growth under conditions of severe iron depletion. Compatible with LbtU being a receptor for legiobactin, the reduced ability of the lbtU mutant to grow on the low-iron agar was evident not only when the strain was tested on its own plate (as depicted in Fig. 8) but also when it was spotted near the wild type (data not shown). In contrast, the lbtA mutant showed impaired growth only when it was spotted apart from the wild type. Strain NU385, which contains mutations in both lbtA and lbtU, was no more impaired for growth on DFX-containing BCYE agar than was the lbtU single mutant (Fig. 8), confirming that LbtA and LbtU operate within the same iron metabolism pathway.

FIG. 8.

Growth of the L. pneumophila wild type and lbtA and lbtU mutants on media containing differing amounts of iron. We spotted dilutions of the wild type, lbtA mutant (lbtA-) NU302, complemented mutant NU302 (pLbtA), lbtU mutant NU383, complemented mutant NU383 (pLbtU), NU387 (the NU383 derivative containing lbtU at a neutral chromosomal site), and lbtA lbtU double mutant NU385 onto standard BCYE agar (A), BCYE lacking its usual iron supplement (B), or non-iron-supplemented BCYE containing 10 μM DFX (C). After 5 days of growth at 37°C, growth was recorded. Each strain (i.e., each column of growth depicted here) was spotted on its own plate to prevent diffusible factors (e.g., siderophore) produced by some strains from stimulating the growth of others nearby and thereby confounding mutant analysis. The reduced growth of NU302, NU383, and NU385 relative to that of the wild type as depicted in the last panel was observed on at least four independent occasions. However, the slight differences between the three mutants at the highest CFU concentrations in panel C were not observed in all replicate experiments.

FIG. 9.

His₆-tagged LbtU complements the growth defect of the lbtU mutant. We spotted dilutions of the wild type, the wild type containing a plasmid encoding LbtU-His₆ (i.e., plbtU-His₆), the lbtU mutant NU383, and NU383 containing plbtU-His₆ onto standard BCYE agar (indicated as containing iron and lacking DFX and IPTG [Fe +, DFX −, IPTG −]), non-iron-supplemented BCYE containing 10 μM DFX (Fe −, DFX +, IPTG −), or non-iron-supplemented BCYE containing 10 μM DFX as well as IPTG to induce higher levels of expression of His₆-tagged LbtU (Fe −, DFX +, IPTG +). After 5 days of growth at 37°C, growth was recorded. Each strain was tested on at least three independent occasions, with similar results. As seen on the left side of the figure, expression of His₆-tagged LbtU did not compromise the growth of the wild type, but as seen on the right side, it did restore the growth of the lbtU mutant, especially in the presence of inducing IPTG.

The loss of LbtU prevents effective intracellular and intrapulmonary replication by L. pneumophila.

Previously, we reported that lbtA mutants are not impaired for infection of H. vermiformis (3), and the lack of a required role for legiobactin in amoebal infection was confirmed when we used the lbtA mutant to infect A. castellanii (data not shown). Past studies determined that lbtA mutants are also not impaired for in vitro infection of macrophages and lung epithelia, indicating that the demonstrated importance of legiobactin in lung infection is due to processes beyond intracellular growth in resident lung cells (2, 3). However, when we examined the lbtU mutants NU383 and NU384 for their relative abilities to infect host cells, the strains were impaired as much as 100-fold in H. vermiformis, A. castellanii, and U937 cell macrophages (Fig. 10). This defect appeared to involve an extended lag phase during intracellular infection. On two additional occasions, when the U937 cells were treated with 10 μM DFX, the recovery of NU383 at 48 h postinoculation was reduced 140- to 340-fold (compared to the recovery of cells with no DFX treatment), whereas the wild type's growth was attenuated only 10-fold (data not shown), implying that the infection defect of the lbtU mutant is related to intracellular iron acquisition. The lbtU mutants were also impaired in murine bone marrow-derived macrophages and human A549 lung epithelial cells (data not shown). Efforts aimed at complementing the intracellular defect of the lbtU mutants were confounded when control experiments showed that complementing copies of lbtU significantly reduced the growth of the wild type within U937 cells (Fig. 10D) and other host cells (data not shown). However, the fact that multiple independent lbtU mutants showed infection defects coupled with the fact that lbtU is monocistronic strongly indicated that the impaired intracellular growth of lbtU mutants was specifically due to the lbtU mutation rather than a spontaneous second-site mutation and/or a polar effect on some downstream gene(s).

FIG. 10.

Intracellular growth of the L. pneumophila wild type and lbtU mutants. (A to C) H. vermiformis (A), A. castellanii (B), and U937 cells (C) were infected with wild-type 130b (black diamonds) and the lbtU mutant NU383 (gray squares), and then at the indicated times, the numbers of CFU in the infected monolayers were determined by plating. Data are the means and standard deviations of results from four infected wells. Asterisks indicate those data points where NU383 was significantly different from the wild type (Student's t test, P < 0.05). Each figure is representative of at least 3 independent experiments. The LbtU mutant NU384 behaved similarly to NU383 (data not shown). (D) In addition to being infected with 130b and NU383, U937 cells were infected with 130b (white triangles) and NU383 (open gray squares) containing a complementing copy of lbtU in a neutral chromosomal site and then assayed as described above. In another experiment, wild-type and mutant strains carrying lbtU on plasmid pLbtU also showed the level of growth impairment depicted here (data not shown).

The different infection results obtained for the lbtU mutants and lbtA mutants might suggest that LbtU has a role in intracellular infection that goes beyond legiobactin utilization. However, in the case of the lbtU mutants, it was also possible that, because of an absent LbtU receptor, there is an accumulation of extracellular legiobactin that effectively sequesters the available intracellular iron that is needed for full bacterial growth. Compatible with this scenario, previous work had demonstrated that legiobactin genes are expressed intracellularly (3). As a step toward distinguishing between these opposing hypotheses, we tested the mutant NU385, which lacks both lbtA (i.e., legiobactin synthesis) and lbtU. This double mutant did not have the growth defect, and in fact it grew like the wild type in U937 cells and in A. castellanii (Fig. 11) as well as in H. vermiformis and A549 cells (data not shown). Taken together, these data support the second hypothesis and indicate that the loss of LbtU is detrimental to the intracellular growth of L. pneumophila, but only in a bacterium expressing lbtA and legiobactin.

FIG. 11.

Intracellular growth of an L. pneumophila lbtA lbtU double mutant. A. castellanii (A) and U937 cells (B) were infected with wild-type 130b (black diamonds), lbtU mutant NU383 (gray squares), and lbtA lbtU double mutant NU385 (white triangles), and then at the indicated times, the numbers of CFU in the infected monolayers were determined. Data are the means and standard deviations of results from four infected wells. Asterisks indicate those data points where NU383 was significantly different from the wild type and the double mutant (Student's t test, P < 0.05). Each value is representative of results of at least 3 independent experiments.

When we tested the lbtU mutant NU383 and the lbtA lbtU double mutant NU385 for their ability to infect the lungs of A/J mice, a similar type of event occurred. Previously, when we compared the wild type to the lbtA mutant following intratracheal inoculation into separate groups of animals, the lbtA mutant exhibited a 3- to 13-fold reduced recovery of lung CFU, documenting the in vivo importance of legiobactin (2). A defect was not observed when the lbtA mutant was coinoculated with the wild type in a competition assay, compatible with the diffusible nature of the siderophore (2). In contrast, the lbtU mutant was highly impaired when tested in the in vivo competition assay, exhibiting average 17- and 551-fold growth disadvantages relative to the growth of the wild type at 24 and 72 h postinoculation, respectively (Fig. 12A). This large defect was also seen when the numbers of lung CFU of the lbtU mutant and wild type were compared following inoculation into separate groups of animals (see Fig. S1 in the supplemental material). However, akin to the observations made in vitro, when the lbtA lbtU double mutant was tested, the large defect was no longer manifest (Fig. 12B). Thus, the loss of LbtU is also detrimental to the intrapulmonary growth of L. pneumophila, but only in a bacterium expressing legiobactin; again, the simplest hypothesis to explain these data is that, because of an absent LbtU, there is an accumulation of extracellular legiobactin that binds the available iron that is needed for optimal bacterial growth. Regardless of the precise molecular basis for these mutant phenotypes, these experiments further confirm that LbtU and legiobactin are linked. Furthermore, they provide evidence that legiobactin and LbtU are expressed during intracellular and intrapulmonary infection; this is compatible with our past work showing that lbtA transcription occurs during L. pneumophila growth in macrophages and that the lbtA mutant is impaired for growth in the lungs of A/J mice (2, 3).

FIG. 12.

Growth and survival of the L. pneumophila wild type and the lbtU and lbtA lbtU mutants in the lungs of infected mice. Equal numbers of wild-type and lbtU mutant NU383 (A) or lbtA lbtU mutant NU387 (B) cells were introduced into the lungs of A/J mice by intratracheal inoculation. At 24 and 72 h postinoculation, the ratio of the number of CFU of 130b to the number of CFU of the Km^r mutant in infected lungs was determined. Data are representative of the actual values obtained per mouse (n = 3 to 5), and the solid bars represent the mean values. In the experiment depicted in panel A, two mice did not yield any mutant bacteria at 72 h, indicating that the competitive disadvantage for bacteria lacking LbtU was even greater than indicated by the values presented here. In two additional experiments, the effect of the lbtU mutation was determined to be ca. 250-fold at 24 h and 1,000-fold at 72 h.

DISCUSSION

For multiple reasons, we conclude that LbtU specifically promotes L. pneumophila's utilization of legiobactin. First, lbtU was both iron regulated and located just upstream of the lbtA and lbtB siderophore genes. Second, lbtU mutants, like lbtA mutants, were impaired in their ability to grow on iron-depleted BCYE agar but not on iron-replete media. Third, although fully able to make and secrete legiobactin, multiple lbtU mutants (but not the complemented mutants) could not be stimulated to grow on low-iron BCYE by the addition of legiobactin-containing supernatants. Fourth, lbtU mutants were able to grow on the low-iron medium when it was provided with Fe³⁺ or Fe²⁺ salts. Fifth, an lbtU mutant (but not the complemented mutant) exhibited an impaired ability to take up radiolabeled Fe³⁺. Sixth, a plasmid copy of lbtU was able to confer upon a heterologous bacterium (L. longbeachae) the ability to bind legiobactin. Seventh, lbtU mutants were impaired for infection of A/J mouse lungs in a legiobactin-dependent manner. Given the documented OM location of LbtU, along with its predicted surface and OM-spanning domains, we next conclude that LbtU is involved in the uptake stage of ferrilegiobactin utilization. Thus, the simplest and most logical explanation for our data is that LbtU is the OM receptor for the Legionella siderophore; i.e., ferrilegiobactin binds to LbtU surface domains and then passes through an OM-spanning pore created by the protein. An alternative scenario is that Fe³⁺ is released from ferrilegiobactin (and possibly reduced) while still extracellular and that LbtU then provides transport for the iron into the bacterial interior. A final, formal possibility is that LbtU is indirectly involved in iron uptake by promoting the positioning or functioning of another protein that is the actual receptor/pore. Regardless of its precise role, LbtU is the first Legionella protein to be implicated in the process of siderophore utilization and iron uptake.

Based upon in silico analysis utilizing all relevant programs, OM LbtU was predicted to have eight external loops, a 16-stranded transmembrane β-barrel, and short N- and C-terminal periplasmic tails (Fig. 2). Thus, LbtU did not conform to the typical structure that has been ascribed to bacterial OM siderophore receptors. Based largely on the crystal structures of FecA for citrate, FepA for enterobactin, FhuA for hydroxamates, FpvA for pyoverdine, and FptA for pyochelin, the current view of receptor structure is the presence of a C-terminal 22-stranded β-barrel that forms a gated porin across the membrane and a long (∼150-residue) N-terminal periplasmic domain that acts as a “plug” that opens, with the assistance of TonB-ExbBD, when the receptor is engaged and the siderophore is to be imported (14, 28, 48, 88, 92, 100, 103, 117, 136). For comparison to LbtU, the secondary structures of two well-known receptors appear in Fig. S2A and B in the supplemental material. Therefore, we posit that LbtU is a new type of siderophore receptor/OM transporter. The feasibility of this hypothesis has support from several sources. First, since the common definition of the Gram-negative siderophore receptor is based largely on crystals from only two bacteria, i.e., E. coli (for FecA, FepA, and FhuA) and Pseudomonas aeruginosa (for FpvA and FptA), it is reasonable to think that some different bacteria will reveal other kinds of receptors. Second, because L. pneumophila does not encode TonB-ExbBD (29, 30, 74), the bacterium is expected to use a receptor/transporter that is dissimilar from the previously known receptors that are TonB dependent. Third, 14-stranded-β-barrel proteins have been implicated as siderophore receptors in Francisella species, bacteria that also lack TonB and are intracellular parasites (38, 66, 73, 94, 107, 119). A comparison of predicted structures revealed that LbtU and the Francisella proteins, though distinct from the “traditional” receptors, differ from each other, not only in the number of β-strands but also in the presence of an extended periplasmic tail in the Francisella protein (see Fig. S2C in the supplemental material). Finally, based on modeling produced by the I-TASSER server, a 3D structural prediction suggests that the LbtU 16-stranded barrel can provide a channel through the OM and do so in a way that is different from that of the well-known 22-stranded β-barrel receptors (see Fig. S3 in the supplemental material). A similar 3D structure model for LbtU was obtained when the Phyre server was employed (data not shown). Although LbtU lacked primary sequence similarity to any known protein in the database, it exhibited 43 to 45% similarity and 25 to 27% identity (with E values ranging from 8 × 10⁻¹³ to 2 × 10⁻¹¹) to hypothetical proteins that are predicted to be localized to the OM of the obligate intracellular parasites Coxiella burnetii (GenBank accession no. NP_820680.1 and WP_001424501.2) and Rickettsiella grylli (ZP_02062955.1). Therefore, LbtU might be representative of a new type of siderophore receptor/OM transporter. Clearly, the LbtU crystal structure is an important future goal.

For many years, the TonB-ExbBD complex has been cited as the means by which the proton motive force of the IM is coupled to the OM; i.e., the periplasmic face of the receptor associates with TonB, a periplasm-spanning protein, that connects to ExbB and ExbD embedded in the IM (71, 74, 87, 92, 100, 101). Some bacteria have more than one tonB or tonB-like gene, and in all cases, the connection between TonB-ExbBD and iron acquisition, when examined, has been clear (1, 4, 10, 29, 39, 58, 59, 64, 93, 97, 121, 126, 139). But recently, it has become apparent that some bacteria that contain an OM and IM, including L. pneumophila, and species of Chlamydia, Chlamydophila, Coxiella, Ehrlichia, Francisella, and Rickettsia do not have TonB-ExbBD (29, 74). Furthermore, bacteria that do have TonB-ExbBD will sometimes acquire siderophores and iron by TonB-independent means, such as Salmonella enterica serotype Typhimurium, using dihydroxybenzoic acid or ferrioxamines B and E (60, 72). Hence, an intriguing question is what fulfills the role of TonB-ExbBD in L. pneumophila. One possible answer to this question is the TolAQR complex, which, in other bacteria, operates in a fashion akin to TonB-ExbBD, albeit usually with a different end goal, such as the import of colicins (17, 20, 44, 56, 81, 135). In some cases, TolQR have been shown to act, albeit imperfectly, as replacements for ExbBD, even facilitating siderophore and iron uptake (15, 16, 18, 20, 44, 52, 109). On the other hand, it is possible that L. pneumophila utilizes a system that is very different from both TonB-ExbBD and TolAQR. One certainty based upon the literature and confirmed here is that energy generated by proton motive force is required for iron uptake by L. pneumophila and some others that lack TonB-ExbBD but use siderophores (38, 65, 69). Thus, further investigation in this area may well reveal LbtU to be part of a new paradigm for iron uptake by bacteria.

Because LbtU appears to lack an extended periplasmic tail, another interesting question is if and how the protein transitions between plugged/unplugged states so as to provide regulated import of iron and/or ferrilegiobactin. One hypothesis to explain LbtU-modulated import is that the protein itself creates a closed state that can transition to an open configuration when ferrisiderophore is engaged and energy is provided. A second scenario is that another protein, possibly even the “TonB mimic,” serves as the plug, moving away from LbtU when iron is to be imported. Approaches that have identified critical residues and conformational changes for other siderophore receptors (76, 88, 103, 105) may be applicable to LbtU.

Finally, siderophore receptors have often been shown to have more than one function, and it is possible that this is also the case for LbtU. For example, in Salmonella, Bordetella, Bacillus, and Aeromonas, a single receptor can import more than one siderophore, including siderophores (xenosiderophores) produced by other microorganisms (5, 72, 125, 134, 137). L. pneumophila carries a Fur- and iron-regulated gene (frgA) that has a predicted protein product that is highly similar to LbtA (3, 62). Although we have not yet been able to detect a siderophore activity that is dependent upon FrgA (3, 84), frgA mutants of strain 130b are defective for intracellular infection of U937 macrophages (62). Other functions ascribed to siderophore receptors include promoting adherence to or invasion of host cells (47, 82, 116, 128), being a potential receptor for host hormones (6, 91), acting as a signal-transducing molecule (9, 79), and mediating the uptake of metals besides iron (12, 13, 36). Therefore, the role of LbtU, whether in the context of L. pneumophila infection or environmental survival, might be even more significant than documented here.