The iron-repressed, AraC-like regulator MpeR activates expression of fetA in Neisseria gonorrhoeae

Abstract

Neisseria gonorrhoeae is an obligate human pathogen that causes the common sexually transmitted infection gonorrhea. Gonococcal infections cause significant morbidity, particularly among women, as the organism ascends to the upper reproductive tract, resulting in pelvic inflammatory disease, ectopic pregnancy, and infertility. In the last few years, antibiotic resistance rates have risen dramatically, leading to severe restriction of treatment options for gonococcal disease. Gonococcal infections do not elicit protective immunity, nor is there an effective vaccine to prevent the disease. Thus, further understanding of the expression, function, and regulation of surface antigens could lead to better treatment and prevention modalities in the future. In the current study, we determined that an iron-repressed regulator, MpeR, interacted specifically with the DNA sequence upstream of fetA and activated FetA expression. Interestingly, MpeR was previously shown to regulate the expression of gonococcal antimicrobial efflux systems. We confirmed that the outer membrane transporter FetA allows gonococcal strain FA1090 to utilize the xenosiderophore ferric enterobactin as an iron source. However, we further demonstrated that FetA has an extended range of substrates that encompasses other catecholate xenosiderophores, including ferric salmochelin and the dimers and trimers of dihydroxybenzoylserine. We demonstrated that fetA is part of an iron-repressed, MpeR-activated operon which putatively encodes other iron transport proteins. This is the first study to describe a regulatory linkage between antimicrobial efflux and iron transport in N. gonorrhoeae. The regulatory nidus that links these systems, MpeR, is expressed exclusively by pathogenic neisseriae and is therefore expected to be an important virulence factor.

Fig. 1.

FetA expression is regulated by MpeR. (A) SDS-PAGE analysis of protein expression. Total membrane proteins were isolated from wild-type (WT) (FA1090) and mpeR mutant (MCV304) strains grown under iron-depleted (−) and iron-replete (+) conditions for 4 h. Proteins were separated on a 7.5% acrylamide gel. The arrow on the left indicates the band that was excised and identified as FetA by mass spectrometry analysis. The positions of molecular mass markers are indicated on the right. (B) Western blot analysis of FetA expression. The WT (FA1090), mpeR mutant (MCV304), fetA mutant (FA6959), and complemented mpeR^C (MCV305) strains were grown under iron-depleted (−) and iron-replete (+) conditions. Total membrane proteins from each strain were isolated and standardized before being separated by SDS-PAGE and then transferred to nitrocellulose. Blots were probed with an anti-FetA monoclonal antibody.

Fig. 2.

MpeR activates fetA transcription under iron-depleted conditions. The WT (FA1090), fetA mutant (FA6959), mpeR mutant (MCV304), and complemented mpeR^C (MCV305) strains were grown under iron-depleted (−) and iron-replete (+) conditions. RNA samples isolated from each gonococcal strain were analyzed for expression of fetA and mpeR by RT-PCR. The 16S rRNA gene (16s)was used as a positive control because it was constitutively expressed under all growth conditions. Expression of the 16S rRNA gene in the absence of reverse transcriptase [16s(−)] was used as a negative control.

Fig. 3.

MpeR binds upstream of fetA in a specific manner. (A) Sequence of the 500-bp intergenic region immediately upstream of fetA. The sequence highlighted in teal is contained within the fetA1 probe employed for the EMSA shown in panel B. The sequence highlighted in yellow is contained within the fetA2 competitor DNA. The sequence highlighted in blue is contained in both fetA1 and fetA2 amplicons. The Fur binding site (34) is highlighted in gray. The previously mapped (14) promoter elements (underlined) and transcriptional start site (asterisk) are identified. The start codon for FetA is shown in red. (B) Five nanograms of the 250-bp labeled fetA1 probe (lane 1) was incubated with 10 μg of MBP-MpeR in the absence of unlabeled competitor (lane 2) or in the presence of 2×, 10×, or 20× excess unlabeled competitor DNA. Lanes 3 to 5 contain reaction mixtures including increasing concentrations of the specific competitor (fetA1), lanes 6 to 8 contain mixtures including increasing concentrations of fetA2, and lanes 9 to 11 contain mixtures including increasing concentrations of the nonspecific competitor rnpB.

Fig. 4.

FetA is encoded as part of a multigene operon. (A) Genetic locus, including fetA and downstream genes. Genes are depicted by boxes. Hatched regions 5′ of fetA and fetB indicate approximate locations of Fur boxes (34). Below the chromosomal locus are small dark arrows indicating primer locations. Numbered black bars denote the amplicons generated from each primer set. Long, dark gray arrows indicate the lengths and start positions of two proposed transcripts. (B) RT-PCR analysis of the fet operon. RNA was isolated from the indicated gonococcal strains which were grown under iron-depleted (−) and iron-replete (+) conditions. Amplicon numbers correspond to the diagram in panel A. The 16S rRNA gene (16S) was used as a positive control because it is constitutively expressed under all conditions tested. Expression of the 16S rRNA gene in the absence of reverse transcriptase [16S(−)] was used as a negative control.

Fig. 5.

Identification of the fetB transcriptional start site. (A) Sequence of the intergenic region immediately upstream of fetB. The ATG at the end of the sequence represents the FetB start codon. Several overlapping potential −10 promoter elements are underlined. The Fur binding site (34) is highlighted in gray. The transcriptional start site identified in this analysis is identified by the asterisk. (B) Primer extension products generated from RNA samples harvested from wild-type gonococcal strain FA1090 grown under iron-replete (+Fe) and iron-depleted (−Fe) conditions. Equivalency of the amount of RNA template in each sample was confirmed by ethidium bromide staining of RNA separated on an agarose gel. For comparison, the sequencing reaction using the same primer as was used for the primer extension reaction is shown on the left. The T residue highlighted by the asterisk marks the point of transcript initiation on the noncoding strand.

Fig. 6.

Xenosiderophore utilization by gonococcal strain FA1090. CDM plates were supplemented with apo-bovine transferrin, and wells within the plates were inoculated with the following siderophores: ENT, enterobactin; D1, dihydroxybenzoylserine (DHBS); D2, the dimer form of DHBS; D3, the trimer form of DHBS; S2, the linear derivative of salmochelin; and S4, the cyclized form of salmochelin. Ferric citrate (+) was used as the positive control, and apo-bovine transferrin (−) was used as the negative control. Each bar indicates the average growth in millimeters around each siderophore source; the averages and standard deviations were determined from seven independent experiments, each conducted in triplicate. Bars represent average growth zones for the wild-type strain FA1090, fetA mutant strain FA6959, tonB mutant strain MCV656, the mpeR mutant strain, the mpeR^C complemented strain, and the fetA mpeR double mutant strain. The horizontal dashed line indicates the diameter of the well containing each iron source. Pairwise comparisons between the wild-type and mutant strains resulted in the following P values: *, P < 0.001; #, P = 0.0124; ^, P = 0.0197.

Fig. 7.

FetA expression is not induced by the presence of xenosiderophores. (A) WT strain FA1090 was grown in CDM with the indicated ferrated xenosiderophores (final concentration of 10 μM) as the sole iron source: ENT, enterobactin; D1, dihydroxybenzoylserine (DHBS); D2, the dimer form of DHBS; D3, the trimer form of DHBS; S2, the linear derivative of salmochelin; and S4, the cyclized form of salmochelin. As controls, the WT strain was grown in the absence of iron (−) or with ferric nitrate (+) but without the addition of siderophores. Aliquots collected at 2, 4, and 6 h (indicated above the blots) were lysed and subjected to SDS-PAGE. After separation, proteins were transferred to nitrocellulose. Blots were probed with anti-FetA (top) or anti-TbpA (bottom) antibodies. (B) As in panel A, except the WT strain FA1090 was grown in CDM with the indicated xenosiderophores in the iron-free or apo form.