Regulation of High-Affinity Iron Acquisition Homologues in the Tsetse Fly Symbiont Sodalis glossinidius

Abstract

Sodalis glossinidius is a facultative intracellular bacterium that is a secondary symbiont of the tsetse fly (Diptera: Glossinidae). Since studies with other facultative intracellular bacteria have shown that high-affinity iron acquisition genes are upregulated in vivo, we investigated the regulation of several Sodalis genes that encode putative iron acquisition systems. These genes, SG1538 (hemT) and SG1516 (sitA), are homologous to genes encoding periplasmic heme and iron/manganese transporters, respectively. hemT promoter- and sitA promoter-gfp fusions were constructed, and in both Escherichia coli and Sodalis backgrounds, expression levels of these fusions were higher when the bacteria were grown in iron-limiting media than when the bacteria were grown in iron-replete media. The Sodalis promoters were tested for iron regulation in an E. coli strain that lacks the fur gene, which encodes the iron-responsive transcriptional repressor Fur. Expression of the promoter-gfp fusions in the E. coli fur mutant was constitutively high in both iron-replete and iron-deplete media, and addition of either Shigella flexneri fur or Sodalis fur to a plasmid restored normal regulation. A Sodalis fur mutant was constructed by intron mutagenesis, and semiquantitative reverse transcription-PCR (RT-PCR) showed that iron repression of sitA expression was also abolished in this strain. In vivo expression analysis showed that hemT and sitA are expressed when Sodalis is within tsetse fly hosts, suggesting a biological role for these genes when Sodalis is within the tsetse fly.

Many bacteria exist in symbiotic relationships with eukaryotic organisms, where symbiosis is broadly defined as a close physical interaction between two or more individuals of different species. These symbioses exist on a continuum from mutualism to parasitism. Although parasitic symbionts have been intensively studied, especially with respect to bacterial acquisition of iron from the host, there has been less focus on this topic for mutualistic and commensal bacterial symbionts. This is, in part, because many of these symbionts have not been cultured outside their hosts, making molecular analyses difficult. An exception to this is the bacterium Sodalis glossinidius, a member of the family Enterobacteriaceae in the Gammaproteobacteria (1). Sodalis can be cultured outside its host (25, 51), making it an amenable model for molecular analysis of iron acquisition in symbiotic relationships.

Sodalis is a secondary symbiont of the tsetse fly (Diptera: Glossinidae). Tsetse flies are the vector of Trypanosoma brucei subspecies, the parasitic protozoa causing African trypanosomiasis. Sodalis is primarily located, both intracellularly and extracellularly, in the midgut of the tsetse fly and may also be found in the hemolymph, milk and salivary glands, muscle, and fat bodies of some tsetse fly species but not in the eggs or mature sperm (8, 32, 41, 52). Transmission of Sodalis is believed to occur primarily vertically through milk gland secretions from the female tsetse to the progeny developing in utero (2, 17). Although the presence of Sodalis within the tsetse (8) and its population dynamics through host development (35) have been well documented, the effect of this secondary symbiont on fly biology is not as clear. The bacterium is nonpathogenic to the fly, and some studies suggest that the symbiosis may be mutualistic, as tsetse flies cured of Sodalis have a shortened life span (12). Furthermore, Dale and Welburn reported that there is a ∼30% decrease in the susceptibility of Sodalis-cured tsetse flies to infection with the trypanosome T. brucei rhodesiense (12). These studies suggest an important role for Sodalis in tsetse fly biology.

The Sodalis-tsetse symbiosis is likely of a relatively recent origin, in comparison to other bacteria-insect symbioses. There is little significant genetic distance in Sodalis strains isolated from different tsetse species (1, 49), and evolutionarily distant tsetse species can be reconstituted with different Sodalis strains that subsequently resemble natural infections (49). However, the genome sequence of Sodalis does suggest that the symbiosis has been established long enough such that there has been some erosion of the Sodalis genome. There are a high number of pseudogenes (49%) in the Sodalis genome, many carried by defense or carbohydrate transport genes (45).

Although the genome has been sequenced, the molecular mechanisms by which Sodalis survives and multiplies within the tsetse fly have not been extensively examined. Certainly, acquisition of iron, an essential nutrient for almost all bacteria, will be important for Sodalis survival and proliferation within the tsetse fly. In hematophagous (blood-feeding) insects, iron is found in hemoglobin from the blood meal, heme, transferrin, ferritin, and other iron-containing proteins (22, 24, 56).

Many pathogenic, facultative intracellular bacteria use ABC transport systems to mediate the capture of iron from the host environment (31, 33). Each ABC transport system consists of a periplasmic ligand-binding protein, two cytoplasmic membrane permeases, and two subunits of a peripheral cytoplasmic membrane protein with ATP binding motifs. The transported iron ligand binds to the periplasmic binding protein and is transferred through cytoplasmic membrane permeases via energy obtained from ATP hydrolysis by the associated ATPase (9). Although these transport systems have been studied in pathogens, less is known about their role in the biology of beneficial or commensal bacterial symbionts.

The Sodalis chromosome contains homologues of at least two complete ABC transport systems that may mediate high-affinity iron acquisition (45). One of these systems, encoded by Sodalis genes SG1538 to SG1540, is homologous to the well-characterized heme transport systems in Yersinia enterocolitica (hemTUV), Yersinia pestis (hmuTUV), Shigella dysenteriae (shuTUV), and some Escherichia coli strains (chuTUV) (43, 44, 46, 53). Heme transport via these systems has been implicated in virulence of urinary pathogenic E. coli (UPEC) (47) and in growth of Y. pestis within macrophage cells lines (19). Additionally, homologues of the sit and yfe transport system genes, which encode iron and/or manganese transport systems (5, 21, 39, 54), are present in Sodalis (SG1516 to SG1519). The sit and yfe genes have been found in numerous pathogenic bacteria, and among the enteric species, there seems to be a correlation with those species and strains which have an intracellular component to their life cycle (5, 21, 39, 54). Salmonella enterica serovar Typhimurium sit and Y. pestis yfe mutants are attenuated for virulence (4, 21). A Shigella flexneri sitA mutant is defective in intracellular growth in a genetic background in which other iron transport systems are absent (39).

Since the Hem/Hmu and Sit/Yfe systems have been shown to be important for the biology of other intracellular bacteria when they interact with eukaryotic host cells (19, 34, 39), we hypothesized that the homologous Sodalis systems may be important in Sodalis biology and, thus, that they will be expressed when high-affinity iron acquisition is required for proliferation and viability. The purpose of this study was to characterize the expression and regulation of the Sodalis hemT and sitA genes.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The bacterial strains and plasmids used in this work are listed in Table 1. E. coli strains were routinely grown in Luria-Bertani broth (LB) or Luria agar (L agar), unless otherwise indicated (40). E. coli strains containing green fluorescent protein (GFP) were grown in low-salt Luria-Bertani broth (LSLB) or low-salt Luria-Bertani agar (LSL agar), which contain 5 g of NaCl. Sodalis strains were grown in brain heart infusion (BHI) broth or BHI agar plus 10% defibrinated horse blood (BHIB) (Hemostat, Dixon, CA) at 25°C in a 10% CO₂ incubator. To grow strains under reduced iron conditions for regulation studies, 16 μg (per ml) of the iron chelator ethylene diamino-o-dihydroxyphenyl acetic acid (EDDA) was added to the media (36). For the Fur titration experiments, MacConkey II agar (BBL Corp.) was used, and β-galactosidase assays were done as per Miller (26). Antibiotics were used at the following concentrations for E. coli (per ml): 125 μg carbenicillin, 25 μg kanamycin, and 20 μg chloramphenicol. Antibiotics were used at the following concentrations for Sodalis (per ml): 20 μg ampicillin, 25 μg kanamycin, and 3 μg chloramphenicol.

TABLE 1.

Bacterial strains and plasmids

Strain or plasmid	Characteristic(s)	Reference or source
E. coli strains
BW25113	Δ(araD-araB)567, ΔlacZ4787(::rrnB-3) λ−rph-1 Δ(rhaD-rhaB)568 hsdR514	15
JW0669-2	F− Δ(araD-araB)567 ΔlacZ4787(::rrnB-3) Δfur-731::kan λ−rph-1 Δ(rhaD-rhaB)568 hsdR514	3
H1717	aroB fhuF::λplacMu	20
Sodalis strains
SOD	S. glossinidius from Glossina morsitans morsitans	S. Aksoy
URSOD6	fur183::kan	This study
Plasmids
pLR29	Promoterless GFP vector	38
pAB2	hemT-gfp fusion on pLR29	This study
pHT3	sitA-gfp fusion on pLR29	This study
pGEM-T	High-copy-number cloning vector	Promega Corp.
pRJ30	hemT promoter in pGEM-T	This study
pRJ31	hemT promoter in pGEM-T	This study
pIL	Luciferase gene in pGEM-T	50
pACYC184	Cloning vector with p15A origin of replication	40
pAB4	Sodalis fur on pACYC184	This study
pMS1	S. flexneri fur on pACYC184	37
pAR1219	T7 polymerase under the control of lac UV5 promoter for inducing intron mutagenesis	16

Insects.

Tsetse flies, Glossina morsitans morsitans, were maintained at West Virginia University within the Department of Biology insectary at 24 ± 1°C, with 50 to 55% relative humidity on a 12-h light/12-h dark schedule. Tsetse flies were sorted by sex following their emergence from pupae. Tsetse flies received defibrinated bovine blood (Hemostat) every 48 h through an artificial membrane feeding system (28).

DNA isolation and plasmid construction.

Plasmid and chromosomal DNA were isolated using the QIAprep spin miniprep kit or the DNeasy tissue kit (Qiagen, Santa Clarita, CA), respectively. Isolation of DNA fragments from agarose gels was performed using the QIAquick gel extraction kit (Qiagen). All standard PCRs were carried out using either GoTaq (Promega, Madison, WI) or Pfu polymerase (Stratagene Cloning Systems, La Jolla, CA), according to the manufacturers' instructions.

To clone the fur gene, the gene was amplified from Sodalis DNA by PCR with primers UR235 (5′CCCGGGCAGCTGATGTTTTTGACATCGAA) and UR200 (5′GCGCGGATCCTTCAAAATGAAGGAATGACGC) and Pfu polymerase. The fur fragment was digested with PvuII and BamHI and ligated with pACYC184 (40) digested with EcoRV and BamHI to generate pAB4.

To construct the hemT-gfp and sitA-gfp reporter fusions, hemT promoter primers UR181 (5′GCGCGGATCCCGCTCTTTTTCCTCATCAGC) and UR182 (5′CTAGTCTAGACAAGGCTAATATCCCCTCAGC) or sitA promoter primers UR205 (5′GCGCGGATCCAAGAATTTACAGGCCTCACGC) and UR206 (5′CTAGTCTAGACAACATTTCTTGCCATATCCG) were used to amplify the hemT and sitA promoters from Sodalis DNA with Pfu polymerase. The PCR products were digested with BamHI and XbaI and cloned into the promoterless gfp vector pLR29 (38) that was digested with BamHI and XbaI to generate pAB2 (hemT-gfp) and pHT3 (sitA-gfp), respectively.

To construct multicopy plasmids containing the Fur binding sites from the hemT and sitA promoters, hemT promoter primers UR181 and UR182 or sitA promoter primers UR205 and UR206 were used to amplify the hemT and sitA promoters from Sodalis DNA with Taq polymerase. The PCR products were cloned into the high-copy-number vector pGEM-T (Promega Corporation, Madison, WI) by TA cloning to generate pRJ30 (hemT promoter) and pRJ31 (sitA promoter), respectively.

Construction of the Sodalis fur mutant.

The fur mutant was constructed using the TargeTron intron mutagenesis kit (Sigma-Aldrich, St. Louis, MO). Briefly, the group II intron on pACD4K-C was altered according to the manufacturer's instructions to contain a fur-targeting site located 183 bp from and 3′ of the fur start codon. The altered intron plasmid (pRJ18) was electroporated into Sodalis (13). The electroporation mix was transferred to 5 ml of BHI broth and incubated overnight. The recovered electroporation was centrifuged (13,000 − g for 1 min), and the pellet was resuspended in 3 ml BHI broth containing chloramphenicol and 1% glucose and incubated overnight. Intron expression was induced with 500 μM IPTG (isopropyl-β-d-thiogalactopyranoside) for 1 h. The IPTG-treated sample was centrifuged (13,000 − g for 1 min), and the pellet was resuspended in 1 ml BHI broth containing 1% glucose and incubated for 1 h. The sample was centrifuged (13,000 − g for 1 min), and the pellet was resuspended in 400 μl BHI broth and spread on four BHIB plates containing kanamycin. Single colonies were restreaked after 9 days on BHIB plates containing kanamycin. Insertion of the intron into the Sodalis fur gene and elimination of the wild-type gene were confirmed by PCR analysis using Sodalis primer set UR199 (5′GCGCAAGCTTTTGCTGATGTTTTTGACATCG) and UR200 (5′GCGCGGATCCTTCAAAATGAAGGAATGACGC), which flanks the fur gene.

Sodalis hemT and sitA expression studies using gfp reporter fusions.

After growth of E. coli or Sodalis containing the gfp fusions pAB2 or pHT3 under the indicated iron-replete or iron-deplete conditions, samples were fixed in 2% paraformaldehyde, as described in reference 38, and fluorescence was quantified using a FACSCalibur (Becton, Dickinson and Company, Franklin Lakes, NJ) fluorescence-activated cell sorter, with excitation at 488 nm to measure single-cell fluorescence. FACSCalibur settings used were as follows: forward scatter (FSC), E01; side scatter (SSC), 505; and relative fluorescence between 515 and 545 nm (FL1), 798.

Sodalis hemT and sitA expression studies using semiquantitative reverse transcription-PCR (RT-PCR).

Starter cultures of Sodalis strains were subcultured to an optical density at 600 nm (OD₆₀₀) of between 0.02 and 0.04 in BHI broth containing either the iron chelator EDDA (16 μg/ml) or FeSO₄ (40 μM) and incubated for 24 h at 25°C in a 10% CO₂ incubator. Prior to RNA isolation, samples were stabilized by addition of stabilizing buffer (95% ethanol-5% phenol [pH 4.3]) for 5 min. Total RNA was isolated from bacteria using the RNeasy minikit (Qiagen), which included a DNase I treatment step to degrade DNA. Isolated RNA was treated again with DNase I (Qiagen) to remove any residual contaminating DNA. cDNA was then made from 200 ng total RNA using SuperScript III (Invitrogen, Carlsbad, CA). Semiquantitative real-time PCR was performed on the cDNA samples using Taq polymerase. Samples were taken at 15 to 20 cycles, prior to saturation of the PCR, for agarose gel electrophoresis. Primers used for the PCRs were as follows: for sitA, SG1516Forward (5′CGTACAGCGGTTTACCGAAT) and SG1516Reverse (5′ ACTTCACGTGCCGGTTTATC); for hemT, SodhemT1F (GCTGAGGGGATATTAGCCTTG) and SodhemT1R (ATCCCTTCTGAGGACAGTGGT); and for groEL, SggroelForward (5′GTACCTTGCTCCGATTCCAA) and SggroelReverse (5′GTGGCTTTTTCCAGCTCAAG).

In vivo expression analyses of Sodalis hemT and sitA.

Tsetse flies, G. m. morsitans, were sacrificed at the teneral life stage (i.e., newly eclosed adults prior to blood meal consumption) and 48 h following feeding. Whole midgut RNA was isolated from tsetse fly individuals using TRIzol (Invitrogen) and treated with RNase free-DNase I (Invitrogen). The absence of DNA contamination was verified using PCR. First-strand cDNA synthesis was performed with SuperScript II reverse transcriptase (Invitrogen), 200 ng RNA, and a 2-μM primer cocktail consisting of 5′ACTTCACGTGCCGGTTTAT, 5′GCAATTGCCAGACTTTTTCC, and 5′TTCTTTGCCCACTTTCGCCATA for the sitA, hemT, and chaperonin groEL genes, respectively. Second-strand synthesis was performed with the addition of complementary 5′-end gene primers for the sitA, hemT, and chaperonin groEL genes (5′CGTACAGCGGTTTACCGAAT, 5′GGACAGTACCAGCCTGAAGC, and 5′CCAAAGCTATCGCTCAGGTAGG, respectively), with annealing temperatures of 52°C for sitA and hemT and 58°C for groEL. The PCR amplification products were analyzed by agarose gel electrophoresis and visualized with Kodak 1D image analysis software. The expression level of endogenous Sodalis groEL within respective time points was used as a constitutive control.

RESULTS

Regulation of expression of the Sodalis hemT and sitA genes by iron.

Using bioinformatics tools, we generated a list of Sodalis genes that are homologues to high-affinity iron acquisition genes reported in other bacteria (Table 2). For this first examination of iron acquisition genes in Sodalis, we narrowed our focus to the following two sets of genes that are predicted to be inner membrane ABC transporters: a heme ABC transporter system (hemTUV) and an iron/manganese transporter system (sitABCD). Since many genes that are involved in high-affinity iron acquisition have increased expression under conditions of iron limitation, we examined expression of these two systems in response to iron limitation using Sodalis promoter-gfp fusions. Expression of the hemT-gfp fusion in E. coli was 7-fold higher after growth in the EDDA-treated BHI broth (iron limiting) than after growth in BHI broth without EDDA (iron replete) (Fig. 1 A). Likewise, expression of the sitA-gfp fusion was 26-fold higher after growth in the iron-limiting BHI broth-EDDA (Fig. 1A). A similar expression pattern was observed when the promoter fusions were present in Sodalis. Expression of the hemT-gfp fusion and sitA-gfp fusions in Sodalis was 3-fold and 7-fold higher, respectively, after growth in the BHI broth-EDDA than after growth in BHI broth without EDDA (Fig. 1B).

TABLE 2.

Sodalis genes predicted to encode iron utilization or metabolism functions

Gene(s)a	Predicted Fur boxb	Putative iron acquisition/ metabolism function
SG0185 to SG0187	Y	Inner membrane ABC transporter for metal ions
SG0281 to SG0282	Y	TonB accessory proteins ExbB and ExbD
SG0861	Y	Fur transcriptional regulator
SG1151	Y	Cation efflux protein
SG1275	Y	Ferritin-like protein
SG1381	Y	TonB energy transducer for outer membrane
SG1431 to SG1436	Y	Iron-sulfur cluster assembly scaffold proteins
SG1505	Y	HemS cytoplasmic Fe acquisition from heme
SG1516 to SG1519	Y	Sit inner membrane iron/ manganese ABC transporter
SG1526	Y	ViuB homologue, removal of iron from catechol siderophores
SG1538 to SG1540	Y	Inner membrane heme ABC transporter
SG1674	Y	Manganese transporter
SG1998	N	Hemolysin
SG2280	N	Bacterioferritin
SG2322	Y	FeoA
SGP1_0035 to SGP1_0045	Y	Achromobactin siderophore synthesis and transport system

^aGene numbers correspond to the Sodalis genome sequences under NCBI accession numbers NC_007712 and NC_007713.

^bPotential Fur binding sites were identified using Virtual Footprint analyses (27). Y, at least one Fur binding site within 500 bp of the translational start site of the first gene in the putative operon; N, no Fur binding site detected using Virtual Footprint.

FIG. 1.

Expression of the Sodalis hemT and sitA genes is regulated by levels of iron in E. coli and Sodalis. Starter cultures of the indicated strains were subcultured to an OD₆₀₀ of between 0.02 and 0.04 into BHI broth containing ampicillin and either the iron chelator EDDA (16 μg/ml) (black bars) or 40 μM FeSO₄ (gray bars) and incubated for 24 h at 25°C in 10% CO₂. After incubation, the cells were fixed in 2% paraformaldehyde. The fluorescence of each sample was assessed by flow cytometry using the following settings: FSC, E01; SSC, 505; and FL1, 798. Ten thousand bacterial cells were assayed for each sample. The data presented are the means of results from at least three experiments, with standard deviations of the means indicated.

Repression of the Sodalis hemT and sitA genes by iron requires Fur in E. coli.

Fur is an iron-responsive repressor that many bacteria use to repress transcription of certain genes under iron-replete conditions (18). Both the Sodalis hemT and sitA promoters had putative Fur binding sites located upstream of the translational start sites (Fig. 2). These binding sites were identified using Virtual Footprint and by comparison with other published Fur binding site models (7, 30). Thus, we hypothesized that iron regulation of Sodalis hemT and sitA expression would be mediated by Fur. To begin testing this hypothesis, we measured expression of the Sodalis sitA-gfp and hemT-gfp fusions in the E. coli strain JW0669-2, which has a deletion of the fur gene. If Fur is required for repression of gene expression under iron-replete conditions, then elimination of Fur should result in constitutively high levels of gene expression under both iron-replete and iron-limiting conditions. We found that unlike in the Fur⁺ strain, in the fur mutant, JW0669-2 expression of the sitA-gfp fusion was equally high in both iron-replete and iron-limiting media (Fig. 3 A). A similar expression pattern was observed for the hemT-gfp fusion in the fur mutant (data not shown). Addition of either the S. flexneri fur gene (on pMS1) or the Sodalis fur gene (on pAB4) to the E. coli JW0669-2 fur mutant restored normal iron repression of the sitA and hemT promoters (Fig. 3B). Taken together, these data suggest that the hemT and sitA promoters are repressed by Fur under iron-replete conditions.

FIG. 2.

Maps of the Sodalis sitABCD and hemTUV loci. The open reading frames and the nucleotide sequences of the promoter regions are shown. The putative −10/−35 sequences and the transcriptional initiation sites, identified using BPROM (www.softberry.com), are shown as straight lines and asterisks above the DNA sequences, respectively. The putative Fur binding sites, identified using Virtual Footprint (30), are boxed, and the start codons are in boldface.

FIG. 3.

Fur is required for repression of the transcription of Sodalis hemT and sitA in iron-replete media in E. coli. (A, B) Starter cultures of the indicated strains were subcultured to an OD₆₀₀ of between 0.02 and 0.04 into LSLB containing carbenicillin either with (black bars) or without (gray bars) the iron chelator EDDA (16 μg/ml) and incubated for 2 h at 37°C. After incubation, the cells were fixed in 2% paraformaldehyde. The fluorescence of each sample was assessed by flow cytometery using the following settings: FSC, E01; SSC, 505; and FL1, 798. Ten thousand bacterial cells were assayed for each sample. The data presented are the means of results from at least three experiments, and the standard deviations of the means are indicated. (B) The cultures also contained chloramphenicol to maintain plasmids with p15a origins of replication.

The Sodalis hemT and sitA genes have functional Fur binding sites.

To verify that the Fur binding sites in the Sodalis hemT and sitA genes were functional, we utilized the Fur titration assay (FURTA) described by Stojiljkovic et al. (42). The rationale behind this assay is that multicopy plasmids containing Fur binding sites will titrate Fur away from the Fur binding site on the single-copy, chromosomal fhuF-lacZ fusion in E. coli strain H1717. Without the extra plasmid copies of the Fur binding sites, H1717 is white (Lac⁻) on MacConkey plates supplemented with iron, because Fur represses expression of the fhuF-lacZ fusion. However, when multicopy plasmids containing Fur binding sites are introduced into H1717, the strain is red on MacConkey plates supplemented with iron, because Fur no longer represses expression of the fhuF-lacZ fusion. We introduced the hemT and sitA promoters on pRJ30 and pRJ31, respectively, into H1717 and tested the phenotypes of the strains on MacConkey agar. H1717 containing a plasmid without Fur binding sites (pIL1) was white (Lac⁻) on MacConkey plates supplemented with 25 μM FeSO₄; however, H1717 strains carrying the multicopy plasmids containing Fur binding sites from hemT (pRJ30) and sitA (pRJ31) were red on MacConkey plates supplemented with 25 μM FeSO₄. The phenotypes of these strains were also quantitated using a β-galactosidase assay (Table 3). H1717 containing either pRJ30 or pRJ31 was unable to repress fully the fhuF-lacZ fusion under iron-replete conditions (25 μM FeSO₄), compared to H1717 with pIL1. These data provide support for the hypothesis that the Fur binding sites in the hemT and sitA promoters are functional.

TABLE 3.

Fur binding sites in the Sodalis hemT and sitA promoters titrate Fur away from the fhuF-lacZ fusion in E. coli

Plasmid	Characteristic	β-Galactosidase activity (Miller units) in indicated mediuma
		Iron limited	Iron replete
pIL	No Fur binding site	1,210 (±70)	12 (±1)
pRJ30	Fur binding site from hemT promoter	1,182 (±127)	72 (±1)
pRJ31	Fur binding site from sitA promoter	1,384 (±214)	159 (±6)

^aH1717 containing the indicated plasmid was grown for 24 h at 37°C in LB containing carbenicillin, with 16 μg/ml of the iron chelator EDDA (iron limited) or with 25 μM iron sulfate (iron replete). β-Galactosidase assays were done as described by Miller (26). The data presented are the means of results from at least three experiments, and the standard deviations of the means are indicated in parentheses.

Repression of the Sodalis sitA gene by iron requires Fur in Sodalis.

Sodalis has a fur gene with a deduced amino acid sequence that is 84% identical to the S. flexneri and E. coli Fur proteins. To test whether iron regulation of the sitA gene is mediated by Sodalis Fur in Sodalis, we constructed an insertion mutation in the Sodalis fur gene and examined the expression of sitA through semiquantitative RT-PCR. In iron-replete media, expression of sitA was higher in the fur mutant URSOD6 than in parental strain SOD/pAR1219 or in the wild-type strain SOD (Fig. 4). Unlike in the Fur⁺ strains, in the fur mutant, URSOD6 expression of sitA was equally high in both iron-replete and iron-limiting media. These data suggest that the sitA promoter is repressed by Sodalis Fur under iron-replete conditions in Sodalis.

FIG. 4.

Sodalis Fur represses transcription of sitA in iron-replete media. Starter cultures of the indicated strains were subcultured to an OD₆₀₀ of between 0.02 and 0.04 into BHI broth containing either the iron chelator EDDA (16 μg/ml) or 40 μM FeSO₄ and incubated for 24 h at 25°C in 10% CO₂. RNA was isolated from each sample and used to generate cDNAs, which were amplified using semiquantitative PCR with either sitA or groEL primers. The PCRs were stopped during the exponential amplification stage, and 4 μl from each PCR was run on a 2% agarose gel. Sodalis groEL expression served as a constitutive control.

Sodalis hemT and sitA are differentially expressed when Sodalis is in tsetse flies.

To examine whether the hemT or sitA gene is expressed when Sodalis is within its symbiotic partner, the tsetse fly, we performed semiquantitative RT-PCR on total RNA from teneral (newly emerged, unfed) and from 48-h-old flies following blood meal consumption (Fig. 5). sitA expression was higher in Sodalis within teneral females (newly emerged, unfed) than in Sodalis within the females that had taken a blood meal. In contrast, sitA expression appeared constitutive in Sodalis in the males prior to and following blood meal consumption. Expression of Sodalis hemT in Sodalis within both female and male hosts showed a pattern similar to that of sitA expression.

FIG. 5.

Sodalis sitA and hemT are expressed when Sodalis is within the tsetse host. Whole midgut RNA was isolated from either teneral (unfed) tsetse flies or from flies at 48 h following a blood meal. The RNAs were used to generate cDNAs, which were amplified by semiquantitative PCR with sitA, hemT, or groEL primers. TF, teneral female; TM, teneral male; 48h F, 48-h-old female; 48h M, 48-h-old male. Sodalis groEL expression served as a constitutive control.

DISCUSSION

This study was undertaken to understand the regulation of the genes that encode two putative ABC iron transporters (hemT and sitA) in Sodalis. Sodalis represents an interesting case study for examining gene regulation in bacteria that are part of symbiotic relationships because of its relatively recent association with the tsetse host (1, 49). Sodalis has a relatively large genome compared to those of other insect symbionts, with a 4.2-Mb genome, comparable in size to those of related free-living enteric species (45). However, the presence of a high number of pseudogenes and a relatively low coding density in the Sodalis genome suggest the beginning stages of genome erosion as it adapts to the tsetse host. Thus, the Sodalis-tsetse symbiosis may represent an evolutionarily intermediate with respect to host dependence.

The Sodalis chromosome contains homologues of at least two complete ABC transport systems that may mediate high-affinity iron acquisition, in addition to the plasmid-encoded ABC transport system for the achromobactin siderophore (14, 45). Although this is less than what is typically seen in free-living pathogenic bacteria, it is significantly more than what is seen in the genomes of bacteria engaged in more intimate and, often, more coevolved associations. For example S. enterica serovar Typhimurium, S. flexneri, and Y. pestis all have five or more functionally documented high-affinity iron acquisition systems (10, 19, 31). In contrast, the genomes of obligate symbionts (such as “Candidatus Blochmannia floridanus,” Blattella germanica, Buchnera aphidicola, Rickettsia conorii, Rickettsia prowazekii, Wigglesworthia glossinidia, and Xylella fastidiosa) are annotated with few or, in most cases, no complete high-affinity iron transport systems. The presence of several putative iron acquisition systems in Sodalis may reflect the relatively recent association of Sodalis with its tsetse host, such that the iron acquisition genes present in the ancestral species are still intact. It remains to be seen which of these systems is functional in Sodalis biology and in the tsetse-Sodalis symbiosis.

We found that expression of the hemT and sitA genes is repressed by iron, through the transcriptional repressor Fur. This is consistent with expression of homologues of these genes in bacteria that are relatively recently host associated and/or free-living. The sitA homologues have been shown to be iron and Fur repressed in a plethora of bacterial species, including S. flexneri, S. enterica serovar Typhimurium, Sinorhizobium meliloti, Agrobacterium tumefaciens, and Y. pestis (4, 6, 23, 39, 55). Likewise, the Y. pestis hmuT gene is subject to Fur-dependent iron regulation (55). In pathogenic E. coli (enteropathogenic E. coli [EPEC] and UPEC) and S. dysenteriae, although hemT expression has not been specifically examined, expression of the genes for the heme outer membrane receptor is iron repressed (34, 46, 53). The fact that Sodalis, a symbiont know to have a reduced genome with a high number of pseudogenes, has retained Fur-mediated iron repression of these genes via functional Fur binding sites suggests that this regulation may still be important in vivo for a productive Sodalis-tsetse symbiosis. Iron fluxes during tsetse feeding and/or differential iron levels in the various environments in which Sodalis is located within the fly may have selected for Fur regulation of iron acquisition genes, and our in vivo expression data support the idea that periodic blood feeding by the tsetse fly may influence regulation of the sitA and hemT genes. Interestingly, the genomes of other bacteria engaged in more intimate and, often, more coevolved associations (such as those of the bacterial species “Ca. Blochmannia floridanus,” B. germanica, B. aphidicola, R. conorii, R. prowazekii, W. glossinidia, and X. fastidiosa) lack the fur gene, suggesting that at some point in the evolution of these symbioses, Fur became dispensable. Notably, these species all have significantly reduced genomes (reviewed in reference 29).

The fact the both the S. flexneri and Sodalis fur genes were able to repress gene expression of the Sodalis genes in the heterologous E. coli fur mutant system suggests that although the species likely diverged over 100 million years ago (11), both the Fur protein and the Fur binding site have remained relatively conserved. Sodalis and S. flexneri Fur proteins are 84% identical, with the DNA binding motif completely conserved. This is consistent with observations that the E. coli, Bacillus subtilis, and Pseudomonas aeruginosa Fur binding models are highly similar, suggesting maintenance of conservation among even more distantly related bacteria (7). Retention and conservation of the Fur protein, despite ongoing genome erosion in Sodalis, suggests an important function for this protein. In fact, the Sodalis fur mutant grows less well in vitro than the parent strain (L. J. Runyen-Janecky, unpublished observations). Thus, optimal growth of Sodalis may depend on highly regulated expression of Fur-regulated genes at the appropriate times.

Differential expression of sitA and hemT by Sodalis was observed, both between tsetse males and females and pre- and post-blood meals. The tsetse teneral life stage represents the culmination of an approximately 30-day quiescent pupal stage, and consequently, host resources are thought to be severely limited (17). Correspondingly, there was a greater expression of sitA and hemT by Sodalis within teneral females in comparison to gene expression in females at 48 h following blood meal consumption. These results support an iron-replete environment for Sodalis following blood meal acquisition, which would repress sitA and hemT expression. Interestingly, a similar reduction in gene expression was not observed in Sodalis within the male tsetse at 48 h following a blood meal. This suggests that either the free iron levels that Sodalis detects in males are lower than those detected in females, even after feeding, or there is a second environmental signal in males that overrides Fur-mediated iron repression.

The differential regulation of Sodalis iron acquisition genes between males and females may have implications toward tsetse vector competence. Teneral males have been reported to be more susceptible to trypanosome infections than females (27). The higher level of expression of Sodalis iron acquisition systems in the male tsetse may effectively decrease free iron levels within the tsetse host. This, in turn, might ultimately result in less oxidative stress within the male midgut because excess free iron participates in Fenton reactions that generate toxic superoxide (48). Thus, the trypanosomes in the male tsetse may be able to establish infection more readily than within the female, in part due to a decrease in reactive oxygen species encountered.

Most significantly, our data suggest a biological relevance for the sitA and hemT genes, since they are expressed when Sodalis is within the tsetse fly. Work with some, but not all, bacteria systems that are host associated has demonstrated increased expression of sitA homologues when the bacteria are host associated. For instance, sitA expression has been shown to be increased when Shigella is within epithelial cells and when UPEC is within an intracellular bacterial community (34, 38). Less work has been done on the regulation of hemT. However, when urinary pathogenic E. coli is within an intracellular bacterial community, expression of the outer membrane heme receptor chuA is increased (34). Likewise, Sodalis may increase expression of the sitA and hemT genes under certain conditions within the tsetse fly host for high-affinity iron acquisition, such as when iron resources are scarce. Future work will investigate the role of these Fur-regulated genes in Sodalis biology, especially with respect to symbiosis with the tsetse host.