|Home | About | Journals | Submit | Contact Us | Français|
Neisseria gonorrhoeae successfully overcomes host strategies to limit essential nutrients, termed nutritional immunity, by production of TonB-dependent transporters (TdTs)—outer membrane proteins that facilitate nutrient transport in an energy-dependent manner. Four gonococcal TdTs facilitate utilization of iron or iron chelates from host-derived proteins, including transferrin (TbpA), lactoferrin (LbpA), and hemoglobin (HpuB), in addition to xenosiderophores from other bacteria (FetA). The roles of the remaining four uncharacterized TdTs (TdfF, TdfG, TdfH, and TdfJ) remain elusive. Regulatory data demonstrating that production of gonococcal TdfH and TdfJ are unresponsive to or upregulated under iron-replete conditions led us to evaluate the role of these TdTs in the acquisition of nutrients other than iron. In this study, we found that production of gonococcal TdfH is both Zn and Zur repressed. We also found that TdfH confers resistance to calprotectin, an immune effector protein highly produced in neutrophils that has antimicrobial activity due to its ability to sequester Zn and Mn. We found that TdfH directly binds calprotectin, which enables gonococcal Zn accumulation in a TdfH-dependent manner and enhances bacterial survival after exposure to neutrophil extracellular traps (NETs). These studies highlight Zn sequestration by calprotectin as a key functional arm of NET-mediated killing of gonococci. We demonstrate for the first time that N. gonorrhoeae exploits this host strategy in a novel defense mechanism, in which TdfH production hijacks and directly utilizes the host protein calprotectin as a zinc source and thereby evades nutritional immunity.
Neisseria gonorrhoeae is an obligate human pathogen and the etiological agent of the sexually transmitted infection (STI) gonorrhea, which was estimated to infect 820,000 people in the United States alone in 2011 (1). Generally, symptomatic infection presents as cervicitis in women and urethritis in men and can be resolved with antibacterial drug treatment. However, infection is often asymptomatic in women, resulting in the lack of treatment and ascension to the upper reproductive tract. This can result in serious downstream sequelae, including pelvic inflammatory disease (PID) and disseminated gonococcal infection (DGI), and can even result in ectopic pregnancy and infertility. In addition to the high morbidity associated with gonococcal infection, the total direct cost of treating N. gonorrhoeae infections in the United States in 2008 was estimated to be between $81.1 and $243.2 million (2), representing a significant economic burden.
The ability to treat and control gonococcal infections has become increasingly challenging due to the development of antibiotic resistance. Resistance to sulfonamides, penicillin, and fluoroquinolones is widespread among gonococcal isolates, rendering extended-spectrum cephalosporins (ESCs) the last class of antibiotics approved for use (3). However, recent reports of treatment failures with cefixime and ceftriaxone (4) have led to revised therapy recommendations, which include combination therapy of ceftriaxone plus azithromycin (5). Characterization of “superbug” gonorrhea, resistant to all approved therapies, has legitimized the threat of untreatable gonorrhea (6, 7). As such, the development of novel strategies to treat gonorrhea has become a top international priority.
In spite of significant efforts, there is no vaccine to protect against gonococcal infection (reviewed in reference 8). Furthermore, natural gonococcal infection confers no protective immunity; as such, previously infected individuals remain susceptible to future infections (8). The ability of the gonococcus to phase and antigenically vary its surface structures is thought to contribute to its evasion of the host adaptive immune response (8, 9). Recent studies have also demonstrated that the gonococcus proactively skews the host response away from humoral immunity and toward innate immunity, which is to the pathogen's advantage (9). New vaccine development strategies will likely include immunomodulatory therapy along with presentation of conserved surface antigens in an effort to develop a protective response (10). Thus, the identification of potential vaccine targets remains imperative. Given their surface exposure, conservation among gonococcal isolates, and their limited sequence variability due to their role in the acquisition of essential nutrients, TonB-dependent transporters (TdTs) have the potential to be ideal targets for novel therapeutic and/or preventative strategies.
TonB-dependent transporters are large outer membrane β-barrel proteins with loop regions that extend into the extracellular space to interact with external molecules. A conserved plug occludes the lumen of the barrel and also extends into the periplasmic space (reviewed in reference 11). TonB, an inner membrane protein in complex with ExbB and ExbD, interacts with the “ton box” region of the plug to energize passage of nutrients across the bacterial outer membrane through the barrel of TdTs (12). N. gonorrhoeae encodes eight putative TdTs, five of which have been at least partially characterized (reviewed in reference 13). The best characterized gonococcal TdTs facilitate the acquisition of iron or iron chelates from host-derived proteins, including transferrin, lactoferrin, and hemoglobin, as well as from siderophores made by other bacteria. Iron acquisition is particularly important to pathogenesis, as genetically engineered gonococcal strains incapable of simultaneously using transferrin and lactoferrin cannot cause experimental infection in human male volunteers (14).
In addition to the ability to hijack host proteins for iron acquisition, Neisseria gonorrhoeae can survive a robust neutrophil response, which it specifically elicits during infection (15, 16). As the first line of defense in the innate immune response, polymorphonuclear leukocytes (PMNs) or neutrophils normally locate and eliminate invading microbes through phagocytosis and subsequent degranulation or the formation of neutrophil extracellular traps (NETs) (17). A growing body of evidence now indicates that N. gonorrhoeae can overcome the antimicrobial functions of PMNs by delaying maturation of phagosomes (18), resisting the oxidative burst (19, 20), and escaping NETs through excretion of a nuclease (21).
The host defense strategy of nutritional immunity (reviewed in reference 22) refers to the sequestration of essential nutrients within the host to restrict the growth of pathogenic bacteria and, therefore, their ability to cause infection. Originally described as a strategy of iron limitation, the limitation of other transition metals such as Zn and Mn in response to microbial infection has also been described (23). Zn, in particular, provides structural or catalytic support to 5 to 6% of bacterial proteins (80% of which are enzymes) (24) that function in bacterial processes such as gene regulation, cellular metabolism, and virulence (23). Calprotectin (CP) is a hetero-oligomeric protein complex that constitutes 45% of the cytosolic protein content of neutrophils (25). This innate immunity protein has been demonstrated to have antimicrobial activity against a wide range of pathogens (26,–33), which has been attributed to its Zn- and Mn-chelating ability. The depletion of Zn within microbial tissue abscesses (26) due to CP and the attenuation of virulence or colonization by microbes with inactivated Zn transport systems (23) indicate the importance of Zn at the host-pathogen interface and underscore the need for better understanding of bacterial mechanisms that overcome nutritional immunity strategies.
Herein, we describe the mutagenesis and phenotypic analysis of two gonococcal TdTs: TdfH and TdfJ. The Neisseria meningitidis homologues of these proteins, CbpA and ZnuD, respectively, have been shown to contribute to growth under Zn-limited conditions (34, 35). CbpA was demonstrated to bind to CP and furthermore to enable growth under Zn limitation. ZnuD was recently shown to bind to Zn in vitro (36). In the present study, we show that gonococcal TdfH and TdfJ are Zn and Zn uptake regulator (Zur) repressed and contribute to gonococcal growth under limited-Zn conditions. We demonstrate that TdfH binds directly to CP in whole-cell binding assays and facilitates gonococcal internalization of Zn from CP. Furthermore, we show that both gonococcal growth in the presence of CP and CP binding to whole cells require TdfH production. Our studies further extend our understanding of the biological significance for production of TdfH in gonococcal pathogenesis, as we demonstrate that CP is accessible to gonococci in NETs and that TdfH enhances gonococcal survival in the presence of NETs. Finally, we show that Zn sequestration is a key component of NET-mediated killing. This is the first report to identify a bacterial surface protein that facilitates CP resistance within a biologically relevant niche for N. gonorrhoeae. Therefore, we present a novel mechanism of gonococcal survival in the presence of neutrophils wherein TdfH mediates Zn utilization to overcome nutritional immunity by its obligate human host.
Escherichia coli was cultured in Luria-Bertani medium with antibiotic selection at 100 μg/ml for ampicillin and 50 μg/ml for kanamycin. N. gonorrhoeae was routinely maintained on GC medium base (Difco) agar with Kellogg's supplement I (37) and 12 μM Fe(NO3)3 at 37°C in 5% CO2. Zn-restricted conditions in liquid culture were achieved in both rich medium and defined medium. For each condition, the Zn chelator concentration added was the minimum required to inhibit the growth of the wild-type strain prior to addition of excess Zn. For Zn-restricted growth in rich medium, individual colonies from GC broth (GCB) agar plates were inoculated into GCB treated with supplement I, 12 μM Fe(NO3)3, and 12.5 μM chelator N,N,N′N′-tetrakis-(2-pyridylmethyl)-ethylenediamine (TPEN), and cultures were grown until they reached log phase. At log phase, cultures were back-diluted with GCB plus supplement I, Fe, and TPEN and further treated with 18.75 μM ZnSO4 for Zn-replete conditions or not further treated for Zn-depleted conditions. For Zn-restricted growth in defined medium, individual colonies from GCB agar plates were inoculated into chemically defined medium (CDM) treated with Chelex-100 (Bio-Rad) (38), and cultures were grown to log phase with or without Fe(NO3)3 before back-dilution and the addition of TPEN and/or ZnSO4 at the stated concentrations. All liquid cultures were grown at 37°C with 5% CO2 and vigorous shaking. For regulation studies, gonococcal strains were grown in GC broth medium treated with supplement I until log-phase growth was achieved. At log phase, ZnSO4 or TPEN was added at a final concentration of 25 μM for the Zn-replete or Zn-depleted conditions, respectively. Cultures were grown for 4 h under the described conditions before whole-cell lysates were harvested for SDS-PAGE analysis. For NET bacterial survival studies, viable exponential-phase gonococci were obtained by sequential dilution in rich medium as previously described (19), except at the final dilution, where cultures were resuspended in phenol red-free RPMI with 1 μM TPEN and grown to an optical density at 550 nm (OD550) of 0.4. Bacterial suspensions were then centrifuged at 10,000 × g for 3 min and washed once with RPMI before exposure to neutrophils.
The strains and plasmids used in this study are listed in Table 1. All plasmids were propagated in E. coli TOP10 cells (Invitrogen). To construct the gonococcal zur mutant, we obtained zur::kan from Alastair G. McEwan, which contains the full-length zur gene in the SmaI site of pUC19 interrupted by a kanamycin resistance cassette. This plasmid was linearized with ScaI and used to transform gonococcal strains FA19 and FA1090. Transformants were selected on GCB agar plates supplemented with kanamycin at 50 μg/ml, and the location of the insertion in the chromosome was confirmed via PCR. The resulting strains were named MCV963 (FA19 zur) and MCV964 (FA1090 zur). Gonococcal strain MCV927 contains an Ω insertion within the coding region of the tdfH gene in an FA19 background and has been described elsewhere (39).
To generate a complemented derivative of MCV927, the full-length tdfH gene, including its native ribosome binding site (RBS) was PCR amplified from FA1090 chromosomal DNA using the following primers: oVCU807 (5′-GGG CAG TAC TGA GGA AAA TAT GAG ATC T-3′ [ScaI site underlined]) and oVCU798 (5′-CAC AGT TTA AAC AAA CGC GGG CTG-3′ [PmeI site underlined]). The amplicon was cloned into pCR2.1 using the TOPO TA Cloning kit (Invitrogen) according to the manufacturer's instructions to generate pVCU944. The tdfH sequence was confirmed by sequencing, and pVCU944 was digested with restriction endonucleases ScaI and PmeI to isolate the tdfH fragment, which was then purified and inserted at the PmeI site of pGCC4 (40). The proper orientation of tdfH in the resulting plasmid, pVCU945, was verified by restriction mapping with ScaI and KpnI. For gonococcal transformation, pVCU945 was digested with NotI, and the fragment containing tdfH, lctP, aspC and the erythromycin resistance cassette was purified and used to transform MCV927. Transformants were selected on GCB agar plates supplemented with erythromycin at 1 μg/ml. The resulting strain, MCV956, contains the chromosomal tdfH mutation and an ectopically inserted copy of the wild-type tdfH gene preceded by its native RBS under the control of a lac promoter.
To generate a double mutant incapable of expressing both TdfH and TdfJ in the FA1090 background, the XbaI- and SacI-flanked tdfJ fragment from pVCU703 was first subcloned into pVCU403, which contains 10-bp gonococcal uptake sequence between the HindIII and PstI sites of pUC18 (41). The resulting plasmid, pVCU937, was subjected to random transposon mutagenesis using the Ez-Tn5<Kan-2> kit (Epicentre). Kanamycin-resistant clones were screened via restriction mapping for Tn5<Kan-2> insertion within the tdfJ gene fragment; the insertionally mutagenized plasmid employed subsequently was named pVCU938. Gonococcal strain MCV661 contains an Ω insertion within the coding region of the tdfH gene in an FA1090 background and has been described elsewhere (42). MCV661 was transformed with pVCU938, and transformants were selected on GCB agar supplemented with 50 μg/ml kanamycin, resulting in strain MCV936.
For NET survival studies, a tdfH mutant in the opacity (Opa) protein-deficient (Opaless) background (43) was constructed. To generate this mutant, the XbaI- and SacI-flanked tdfH fragment from pVCU702 was subcloned into pVCU403. The resulting plasmid, pVCU947, was also subjected to Ez-Tn5<Kan-2> random transposon mutagenesis and restriction mapped for Tn5<Kan-2> insertion within the tdfH gene fragment, and the insertionally mutagenized plasmid was named pVCU948. The Opaless version of strain FA1090 was then transformed with pVCU948, and transformants were selected on GCB agar supplemented with 30 μg/ml kanamycin, resulting in strain MCV955. The complemented Opaless tdfH mutant strain was constructed by transforming MCV955 with chromosomal DNA from strain MCV956, selecting for erythromycin resistance. The transformant was back-crossed twice, again into MCV955. The final, selected transformant (strain MCV956A) was confirmed to contain the tdfH gene disrupted by Ez-Tn5<Kan-2>, in addition to a wild-type copy of tdfH in the ectopic expression locus.
Gonococcal whole-cell lysates were generated by harvesting gonococcal strains at a standardized density based upon the optical density of the culture. One milliliter of culture at a density of 100 Klett units (KU) was centrifuged, and the standardized gonococcal cell pellets were then resuspended in Laemmli solubilizing buffer (44) and stored at −20°C until use. Immediately prior to SDS-PAGE, whole-cell lysates were treated with 5% β-mercaptoethanol and boiled for 3 min. Proteins were separated on a 7.5% polyacrylamide gel and transferred to nitrocellulose. Equivalent loading of whole-cell lysates from well to well was verified by Ponceau S staining of the nitrocellulose membranes. For TdfH detection, membranes were blocked with 5% bovine serum albumin (BSA) in a high-salt Tris-buffered saline (TBS) (20 mM Tris, 500 mM NaCl [pH 7.5], 0.05% Tween 20), probed with polyclonal antiserum against TdfH (45) and washed with high-salt TBS followed by incubation with secondary antibody conjugated to alkaline phosphatase (AP) (Bio-Rad). Blots were developed with the nitroblue tetrazolium–5-bromo-4-chloro-3-indolylphosphate (NBT/BCIP) system (Sigma). For TdfJ detection, we generated polyclonal antipeptide antibodies (New England Peptides) against regions predicted to correspond to loop 2 and loop 5 after alignment of TdfJ to a two-dimensional (2D) topology model of TbpA. Membranes were blocked with 5% skim milk in a low-salt TBS (50 mM Tris, 150 mM NaCl [pH 7.5]), probed with loop 2-specific sera, and washed with low-salt TBS before being incubated with secondary antibody conjugated to horseradish peroxidase (Southern Biotech). Blots were developed with the Pierce-ECL-2 kit (Thermo Scientific). For TbpB detection, rabbit anti-TbpB primary antibody (46) was used after blocking with 5% skim milk. Blots were then washed with low-salt TBS before incubation with secondary antibody conjugated to AP and developed with NBT/BCIP.
Gonococci were grown to log phase in Zn-restricted rich medium (GCB plus supplement I, Fe, and TPEN). At log phase, cultures were back-diluted and were not further treated to represent Zn-depleted conditions or were treated with ZnSO4 (18.75 μM) for Zn-replete conditions. After growth for 6 h, bacterial cells were harvested by centrifugation at 4,000 × g for 15 min at 4°C, washed twice with cold PBS–1 mM EDTA, and resuspended in PBS before being serially diluted and spot plated on GCB agar plates supplemented with V-C-N inhibitor (BD BBL). Plates were incubated at 37°C in 5% CO2 for 16 h before CFU were enumerated.
Gonococci were grown as described above in Zn-restricted chemically defined medium (CDM) without Fe(NO3)3. After the optical density of each culture doubled, independent cultures were diluted to the same optical density (100 KU) and then transferred to a 96-well microtiter plate. Each well contained 7.5 μM 30% saturated human transferrin (hTf) as the sole iron source. hTf (Sigma) was dissolved in a mixture of 40 mM Tris, 150 mM NaCl, and 10 mM NaHCO3 at pH 8.6, to which ferric chloride was added to result in 30% saturation by mass. After equilibration, partially saturated hTf was dialyzed to remove residual unbound iron. In addition to the sole iron source, 2.5 μM apo-bovine transferrin (bTf) was added to sequester residual iron, and 1 μM TPEN was incorporated for Zn chelation. A 2 mM concentration of IPTG (isopropyl-β-d-thiogalactopyranoside) was also present in all wells to induce tdfH expression in MCV956. Some wells were further supplemented with either 10 μM CP plus 5 μM ZnSO4 or 5 μM ZnSO4 alone. Levels of calcium in CDM (0.25 mM) are consistent with excess calcium requirements (0.2 mM) described for optimal Zn binding to CP (47). The microtiter plate was incubated at 37°C in 5% CO2 with vigorous shaking, and OD600 readings were collected every 2 h for 6 to 8 h. CP was a kind gift from Walter Chazin and was purified as previously described (31).
Gonococcal strains were grown as described for Zn restriction in CDM with 24 μM Fe(NO3)3. After one mass doubling (approximately 1.5 h of incubation), 1 μM TPEN and 2 mM IPTG were added, and cultures were incubated under the described conditions for 4 h before being applied to nitrocellulose membranes using a dot blot apparatus (Bio-Rad). Dried membranes were blocked for 1 h with 5% skim milk in low-salt TBS before being incubated with CP plus Zn (0.17 μM CP, 0.01 μM ZnSO4) in blocker. After being washed with low-salt TBS, blots were probed for CP using rabbit anti-s100A9 polyclonal antibody (Thermo-Scientific) and secondary antibody conjugated to AP. Dot blots were then developed with NBT/BCIP. For densitometry, replicate blots were scanned, and intensities of individual dots were quantitated using NIH Image J (48). Quantitated intensities were then normalized to the wild type and expressed as a percentage of CP binding.
Gonococcal strains were grown as described for Zn restriction in rich medium (GCB plus supplement I, Fe, and TPEN) until log-phase growth was achieved, at which point cultures were back-diluted and either treated with 18.75 μM ZnSO4 (Zn replete) or no further addition (Zn depleted). Cultures were grown for 6 h before being harvested by centrifugation at 4°C for 15 min at 4,000 × g and washed twice with cold, Chelexed PBS supplemented with 1 mM EDTA. Bacterial cells were then resuspended in undiluted trace-metal-grade nitric acid, heated at 95°C for 2 h, and cooled at room temperature overnight. For analysis, samples were diluted 12.5- or 22-fold in Chelexed distilled water (dH2O). Metal composition was determined by inductively coupled plasma-optical emission spectrometry (ICP-OES) using an MPX Vista spectrometer (Varian, Inc.). To determine the concentration of metals associated with the cells, a standard curve was generated with a 10 μg/ml multielement standard (CMS-5; Inorganic Ventures) diluted in Chelexed dH2O and then serially diluted 2-fold in 1% HNO3 to generate dilutions ranging from 0.640 μg/ml to 0.020 μg/ml. Cell pellets from parallel cultures were resuspended in PBS, treated with 5% SDS, and vortexed for 30 s for lysis. Protein concentrations were determined using the bicinchoninic acid (BCA) protein assay kit (Pierce) according to the manufacturer's instructions.
Poly-l-lysine-treated wells of chambered cover glasses (Nunc) were seeded with 1 × 106 primary human neutrophils and treated with 20 nM phorbol myristate acetate (PMA) for 30 min to induce NET formation. Opaless parental gonococci were added to wells for 1 h at a multiplicity of infection (MOI) of 1. Samples were fixed in 4% paraformaldehyde (Electron Microscopy Sciences) and then blocked overnight in 1% BSA–PBS (Fisher-Gibco). Primary antibodies against gonococci (Biosource) and calprotectin (Sino Biological, Inc.) and secondary antibodies anti-rabbit Alexa 555 and anti-Mouse Alexa 645 (Life Technologies), respectively, were used for staining. Following washes, samples were stained with Sytox green (Invitrogen) to visualize DNA. Samples were mounted in Fluoromount G (Southern Biotech) with 2.5 mg/ml n-propyl gallate (Acros Organics). Images were acquired on a Zeiss LSM 700 confocal laser scanning microscope (63×/1.40 oil immersion) in the University of Virginia Advanced Microscopy Core. z-stack slices were exported as TIF files from Zen 2012 image processing software (Zeiss).
Survival of gonococcal strains upon exposure to PMA-stimulated neutrophils treated with cytochalasin D or cytochalasin D and DNase I was determined as previously described (21). Neutrophils were purified from venous blood collected from healthy donors according to a protocol approved by the University of Virginia Institutional Review Board for Health Science Research. Briefly, 24-well plates were seeded with neutrophils and treated with 20 nM PMA for 30 min at 37°C in 5% CO2. One set of wells was treated with 10 μg/ml cytochalasin D (Sigma), while another set was treated with cytochalasin D and 1 U of DNase I, each for 15 min. Neutrophils were then exposed to Opaless parent, MCV955 (tdfH), MCV956A (tdfHC) (the complemented strain), or Δnuc gonococci grown as described above at an MOI of 1 for 1 h. After infection, well contents were scraped, serially diluted, and plated for CFU enumeration. Bacterial survival is expressed as a percentage of the initial inocula per well.
Neutrophils were PMA stimulated and cytochalasin D treated to generate NETs, as described above, and infected with Opaless parent or MCV955 (tdfH) gonococci ± 0.5 μM ZnSO4 at an MOI of 1 for 1 h. Well contents were scraped, serially diluted, and plated to enumerate CFU and to calculate the percentage of bacterial survival relative to the initial inoculum per strain.
We first sought to characterize the production of TdfH and TdfJ in response to Zn levels. Gonococcal wild-type strain FA1090 was grown in chemically defined media (CDM) in the presence of the zinc chelator TPEN, ZnSO4, and/or Fe(NO3)3 at the indicated concentrations. We found that TdfH and TdfJ levels were downregulated in the presence of Zn (Fig. 1). Conversely, production of both proteins was upregulated in the presence of increasing concentrations of TPEN. Consistent with previous reports (45), TdfH production levels did not change in response to iron. However, TdfJ production was further elevated when supplemented with iron, even at TPEN concentrations shown to derepress TdfJ production (Fig. 1), suggesting that TdfJ is also iron-induced. To verify that TPEN treatment only altered Zn levels, we assessed production of TbpB (transferrin binding protein B), which is known to be iron regulated. TbpB levels did not change at various TPEN concentrations. TbpB production (except under iron-replete conditions) also serves as a loading control for the TdfH and TdfJ blots shown in Fig. 1A.
To assess whether zinc repression of TdfH and TdfJ was Zur mediated, we generated zur isogenic mutants in gonococcal strains FA19 and FA1090, resulting in strains MCV963 and MCV964, respectively. FA1090, FA19, and their respective zur mutant derivatives were grown in Zn-replete or Zn-depleted medium. We found that production of both TdfH and TdfJ was upregulated under Zn-depleted conditions in the wild-type backgrounds. Production was increased in zur mutant derivatives in both backgrounds. Additionally, TdfH and TdfJ were not differentially produced under Zn-replete versus Zn-depleted conditions in zur mutants (Fig. 1B). These data indicate that Zn repression of TdfH and TdfJ is Zur mediated. We also observed that TdfH production was higher under all conditions in strain FA1090 compared to that in FA19, while levels of TdfJ production were similar in both backgrounds (Fig. 1B). Interestingly, the increased TdfH production in FA1090 versus FA19 was also observed in the respective zur mutants. These data suggest that while Zn repression of TdfH is Zur mediated in strains FA1090 and FA19, Zur does not contribute to the increased production of TdfH in FA1090 versus FA19.
Next, we sought to determine whether production of gonococcal TdfH was altered by the two-component regulatory system MisRS. Microarray studies have identified meningococcal TdfH as part of the MisR regulon (49). Additionally, MisRS has been shown to regulate the inner core structure of gonococcal lipooligosaccharide (LOS) and production of the hemoglobin receptor HmbR in N. meningitidis (50, 51). When gonococci were grown under Zn-replete and Zn-depleted conditions, we found that the misR mutant derivative in the FA1090 background showed no difference in TdfH production compared to its parent (Fig. 1C). In contrast, in FA19 the misR mutant derivative produced less TdfH than its parent under both Zn-replete and Zn-depleted conditions (Fig. 1C). MisR was not found to alter the production of TdfJ in either FA1090 or FA19. These data suggest that TdfH production can be controlled by MisR in addition to Zn and Zur, and this regulatory circuit varies depending on the genetic background.
Given the regulation of the TonB-dependent receptors TdfH and TdfJ by Zn and Zur, we sought to determine whether these TdTs contribute to gonococcal Zn acquisition. We hypothesized that those gonococcal strains lacking Zn receptors would be deficient for growth under Zn-restricted conditions. Likewise, we expected that upon supplementation with Zn, growth of wild-type gonococci would recover, whereas strains lacking the receptors would remain suppressed for growth. To test these hypotheses, wild-type or tdfH, tdfJ, and tonB mutant strains (FA19, MCV927, MCV928, and MCV650, respectively) were grown in GC broth medium treated with TPEN in the presence or absence of additional Zn. After growth for 6 h, viable bacteria were enumerated. As expected, significantly more viable bacteria were recovered from all bacterial strains tested when grown under Zn-replete versus Zn-depleted conditions, except for the tdfH mutant, where addition of Zn did not increase the numbers of recovered bacteria (Fig. 2). Moreover, significantly fewer viable tdfH mutant than wild-type bacteria were recovered after growth in the presence of Zn. These data are consistent with a function for TdfH in Zn acquisition. The tdfJ mutant was the most impaired for growth under Zn-replete or -depleted conditions; however, growth was partially restored by addition of Zn (Fig. 2). This finding suggests that while TdfJ is important for growth under Zn restriction, gonococci have additional mechanisms of Zn acquisition independent of TdfJ. In agreement with the effects of the TdTs TdfH and TdfJ on Zn acquisition, significantly fewer tonB mutant bacteria were also recovered, compared to wild-type bacteria.
Since both TdfH and TdfJ contribute to growth under Zn-restricted conditions and the meningococcal homologue of TdfH binds to the host Zn-binding protein calprotectin (CP) (34, 35), we tested whether these putative transporters are required for the assimilation of Zn from CP. We tested the ability of wild-type or tdfH and tdfJ mutant strains to grow in the presence of CP supplemented with Zn under Zn-restricted conditions. Contrary to the antimicrobial activity reported against other pathogens (26, 27, 30, 31, 52), we found that the growth of wild-type gonococci was enhanced in the presence of Zn plus CP (Fig. 3A). Addition of CP also suppressed growth of the tdfH mutant, which is particularly evident at 8 h (Fig. 3A and andC).C). These data suggest that N. gonorrhoeae is capable of overcoming the antimicrobial effects of CP by production of TdfH. Growth of the tdfJ mutant was similar to that of the wild type (Fig. 3A). To confirm that TdfH was indeed responsible for the observed resistance of gonococci to CP, we generated a complemented derivative of the tdfH mutant strain, MCV956 (tdfHC), which we confirmed produced TdfH in the presence of IPTG (Fig. 3B). Growth in the presence of CP was restored in the tdfHC strain (Fig. 3C), demonstrating that growth in the presence of CP specifically requires TdfH production. Additionally, the tdfH growth defect was specific to CP treatment, as tdfH, tdfHC, and wild-type bacteria all grew similarly when supplemented with Zn alone (Fig. 3C).
Given the ability of N. gonorrhoeae to grow in the presence of CP, we next sought to determine whether the gonococcus was capable of binding to CP. Whole gonococcal cells, grown in CDM with Fe(NO3)3 and TPEN, were applied to nitrocellulose paper, presenting cell surface proteins in their native conformation within an intact bacterial outer membrane. After blocking, membranes were incubated with Zn plus CP, before being washed and probed for CP. CP was specifically detected in association with wild-type FA1090, indicating that N. gonorrhoeae is indeed capable of binding CP (Fig. 4A). CP binding was reduced by 80% in both tdfH single and tdfH tdfJ double mutants in the FA1090 background, but binding was unaffected in the tdfJ mutant. These findings indicate that TdfH mediates the gonococcus-CP interaction. Consistent with the reduced level of TdfH detected in FA19 versus FA1090 whole-cell lysates (Fig. 1B and andC),C), CP binding by FA19 was low and not significantly different between wild-type and tdfH mutant bacteria (Fig. 4B). However, CP binding was enhanced in FA19 tdfHC upon IPTG induction (Fig. 4B). These results demonstrate that TdfH specifically facilitates the binding of CP to gonococci in both FA1090 and FA19 strain backgrounds.
Our Zn regulatory and growth data led us to hypothesize that TdfH, a TonB-dependent transporter, and TonB would contribute to Zn internalization by gonococci. While meningococcal CbpA was shown to contribute to growth under Zn limitation, Zn accumulation inside N. meningitis cells was not directly demonstrated (35). To address whether N. gonorrhoeae accumulated intracellular Zn in a TdfH-dependent manner, we grew the wild-type or tdfH and tonB mutants under Zn-replete and Zn-depleted conditions and measured the amount of internalized Zn using inductively coupled plasma optical emission spectrometry (ICP-OES). All of the gonococcal strains tested accumulated significantly more Zn when grown under Zn-replete versus Zn-depleted conditions (P < 0.05), indicating that the growth conditions effectively modulated Zn availability (Fig. 5A). The tdfH and tonB mutants internalized significantly less Zn when grown under both Zn-replete and Zn-depleted conditions compared to the wild type (P < 0.05). Thus, TdfH and TonB contribute to Zn accumulation.
Given the demonstrated interaction between TdfH and CP (Fig. 3 and and4)4) and the role of TdfH and TonB in Zn accumulation (Fig. 5A), we assessed whether TdfH contributed to Zn accumulation in the presence of CP. Wild-type and tdfH mutant bacteria were grown under Zn-depleted conditions in the presence of Zn plus CP, and the level of internalized Zn was assessed via ICP-OES. We found that the tdfH mutant accumulated less Zn than the wild type in the presence of CP (Fig. 5B). These results show that TdfH enables Zn assimilation from CP.
Gonococcal infection is characterized by a robust recruitment of neutrophils, which migrate to the site of infection and attempt to eliminate the invading microbes via phagocytosis and/or NET formation (17). CP is a major component of the neutrophil cytosol and is released during NET formation. CP remains associated with NETs, where it has been demonstrated to be the active component in NET-mediated killing of Candida albicans (27). Given our findings that TdfH can bind and utilize CP as a Zn source, we hypothesized that gonococci can utilize CP in NETs to enhance its survival in association with NETs. To test this hypothesis, we first sought to determine if CP is accessible to gonococci in the presence of neutrophils and NETs. Gonococcal opacity (Opa) proteins, which are subject to phase variation, have been shown to differentially affect gonococcal survival in association with neutrophils (16). Therefore, all neutrophil experiments were conducted in the FA1090 Opa-deficient (Opaless) background (43). Primary human neutrophils seeded onto chambered cover glasses and induced to form NETs with phorbol myristate acetate (PMA) (53) were stained with antibodies against CP and gonococci. CP demonstrated a punctate staining pattern throughout the NETs and was detected in proximity to gonococci in NETs (Fig. 6). This observation indicates that CP is accessible to N. gonorrhoeae in NETs.
We next sought to determine if gonococcal CP utilization, as mediated by TdfH, conferred a bacterial survival advantage in the presence of NETs. We constructed a tdfH mutant (MCV955) in the FA1090 Opaless background, demonstrated that this strain produced no TdfH (Fig. 7A), and then assessed this mutant's survival in NETs compared to that of the Opaless parent. Primary human neutrophils were treated with PMA and cytochalasin D to activate neutrophils and stimulate NET production while inhibiting phagocytosis, thereby selecting for only extracellular mechanisms of neutrophil killing. DNase I was added under some conditions to assess the role of intact DNA fibers on NET-mediated killing. Neutrophils were then infected with the Opaless parent and isogenic tdfH mutant and tdfHC complement. The Opaless Δnuc strain was used as a positive control for DNase I-reversible sensitivity to NET-mediated killing (21). We found that while the parental strain demonstrated little to no survival defect in the presence of NETs, with or without DNase I, only 40% of tdfH mutant bacteria survived exposure to neutrophils producing NETs (Fig. 7B). The defect in survival exhibited by the tdfH mutant was reversed by complementation (Fig. 7B). Additionally, treatment with DNase I significantly increased survival of the tdfH mutant, indicating that intact DNA fibers contribute to the observed NET sensitivity of the tdfH mutant. Thus, TdfH is critical for optimal gonococcal resistance to NET-mediated killing.
Given our in vitro data that TdfH binds to and can utilize Zn from calprotectin, we postulated that TdfH facilitates gonococcal survival in NETs by overcoming CP-mediated Zn chelation. In support of this hypothesis, addition of Zn successfully restored survival of the tdfH mutant to wild-type levels after exposure to NETs (Fig. 7C). We conclude that NETs can kill N. gonorrhoeae by sequestering the essential metal Zn, and TdfH enables gonococcal survival in the presence of neutrophils and NETs by overcoming CP-mediated Zn chelation.
Nutritional immunity, the host strategy of limiting essential nutrients from invading pathogens, is a prevalent theme in host defense against microbes. While traditionally characterized as the sequestration of iron from pathogens, roles for other transition metals such as Zn and Mn are now being examined in the struggle for nutrients within the context of infection (22, 23). N. gonorrhoeae is particularly adept at overcoming host nutritional immunity strategies. Despite being unable to produce any siderophores to scavenge iron or iron chelates from the iron-limited host environment, N. gonorrhoeae cells express TonB-dependent transporters that enable utilization of siderophores produced by other bacteria as well as the host iron binding proteins transferrin and lactoferrin and the heme-binding protein hemoglobin (13, 54). Given the role of the transferrin and lactoferrin receptors in gonococcal pathogenesis (14), we propose that the remaining TdTs, which have yet to be fully characterized, also have important roles in the ability of N. gonorrhoeae to survive in its human host and cause disease. Moreover, these transporters also have the potential to be manipulated for the development of anti-infective therapies.
In this report, we define for the first time the strategy used by N. gonorrhoeae to overcome nutritional immunity posed by Zn restriction in the host. We demonstrate that N. gonorrhoeae uses the TdT TdfH to overcome growth restriction by the host Zn-chelating immune effector protein CP. TdfH directly facilitates bacterial binding of CP binding and growth in its presence. Due to its sequence similarity to heme receptors, TdfH was initially thought to contribute to heme acquisition; as such it is designated Hup, for heme utilization protein, in neisserial genome annotations. However, a gonococcal mutant unable to express TdfH was not defective for growth under heme-limiting conditions (45), placing in doubt TdfH's role in heme utilization. Additionally, production of TdfH along with the neisserial TonB system could not restore heme utilization in an E. coli hemA mutant (45), as has been shown for other heme receptors (55,–57). Instead, our studies demonstrate that gonococcal TdfH binds to CP and enables Zn acquisition. These findings are consistent with a report that CbpA, the meningococcal TdfH homologue, is required for N. meningitidis growth in the presence of CP and for CP binding (35). The present report extends these findings to N. gonorrhoeae and furthermore demonstrates that TdfH contributes to intracellular Zn accumulation in the presence and absence of CP.
Consistent with its role in nutritional immunity, CP inhibits the growth of numerous microbial pathogens, including Candida albicans (58), Staphylococcus aureus (26), Helicobacter pylori (28), Staphylococcus epidermidis, Staphylococcus lugdunensis, Enterococcus faecalis, Acinetobacter baumannii, Pseudomonas aeruginosa, Escherichia coli, and Shigella flexneri (52) due to its Zn/Mn-chelating activity. To overcome this growth restriction, pathogens, including Aspergillus fumigatus (59), Acinetobacter baumannii (29), Pseudomonas aeruginosa (60), and Salmonella enterica serovar Typhimurium (33), use Zn acquisition systems to promote resistance to CP. For Gram-negative bacteria, our understanding of the bacterial factors contributing to CP resistance have mostly been limited to the high-affinity Zn uptake system ZnuABC, a member of the ATP-binding cassette-type transporters located at the inner membrane (29, 32, 60). However, the mechanism of Zn entry through the outer membrane has yet to be fully resolved, although TonB-dependent transporters such as ZnuD in N. meningitidis have been implicated (34). Our finding that TdfH not only binds CP but is also required for Zn accumulation from CP identifies a bacterial surface protein that can directly interact with CP to overcome CP-mediated Zn sequestration.
We found here that gonococcal TdfH is Zn and Zur repressed. In agreement with previous reports (45, 61), we confirmed that production of TdfH is independent of iron availability, consistent with a function distinct from iron acquisition. Expression of meningococcal tdfH was found to be increased under low-Zn conditions in microarray studies (62), and production of the homologous protein, CbpA, was also increased in a zur mutant derivative (35). Gonococcal tdfH was found to be differentially expressed in microarray studies of PerR, which is described as a Mn-dependent repressor (63). However, gonococcal and meningococcal Zur share 96% protein identity, and the gonococcal PerR and meningococcal Zn-responsive regulons overlap significantly. Moreover, TdfH production was not altered by Mn availability (data not shown). Our data thus support the contention that gonococcal PerR functions more like Zur, responding to Zn as opposed to Mn (64), and agree with our finding that the Zn acquisition protein TdfH is regulated by Zn in a Zur-dependent manner.
TdfH has been reported to be a member of the regulon of the two-component regulatory system MisRS in N. meningitidis (49). A mutant derivative unable to express the response regulator, MisR, had reduced levels of tdfH in microarray studies, confirmed by quantitative reverse transcription (RT)-PCR (49). CbpA production was also reduced in a meningococcal ΔmisRS mutant (35). Our data demonstrate that while TdfH production was unaffected by the presence of MisR in N. gonorrhoeae strain FA1090, it was reduced in the misR mutant in gonococcal strain FA19, indicating that TdfH regulation is distinct in these two gonococcal strains. We also report an increase in basal levels of TdfH production in strain FA1090 versus FA19, which is retained in both zur and misR mutants, indicating that the different regulation patterns between strains are independent of the transcriptional regulators Zur and MisR. This is the first study to report that TdfH production is distinct in different gonococcal strains and may also be controlled by additional factors beyond transcriptional regulators Zur and MisR. However, despite the differences in TdfH production, both the FA1090 (data not shown) and FA19 gonococcal strains demonstrated TdfH-dependent CP resistance.
Our novel observation that TdfH contributes to gonococcal survival within NETs is yet another example of the many ways in which N. gonorrhoeae manipulates host-pathogen interactions to its advantage (9, 16). NETs are comprised of DNA and histones, as well as neutrophil granule and cytosolic proteins. CP has been detected as a major component of NETs (27) (Fig. 6). NET-mediated killing is thought to occur by trapping microbes in chromatin fibers in a microenvironment with high levels of antimicrobial proteins and molecules like CP and cathepsin G (53). DNase treatment degrades NETs and destroys the NET microenvironment and can thus abrogate NET-mediated killing of bacteria, including Neisseria spp. (21, 53, 65). CP, through its Zn-chelating activity, was found to be the major antimicrobial component in NETs against Candida albicans, Klebsiella pneumoniae, and Aspergillus nidulans (27, 30, 66). Antimicrobial metal chelation in NETs can also be mediated by DNA itself (67). As such, a mechanism to outcompete or overcome host nutrient sequestration strategies provides a significant survival advantage in NETs. Indeed, in N. meningitidis, the Zn outer membrane transporter ZnuD was found to contribute to survival within NETs (65). The data reported in this study demonstrate that CP is in close proximity and therefore accessible to gonococci in NETs. Additionally, this study is the first to demonstrate that TdfH is critical for N. gonorrhoeae survival of NET-mediated killing, where CP has been shown to be a key antimicrobial component (27, 30). We also show that Zn supplementation abrogates sensitivity of the tdfH mutant to NETs, demonstrating that Zn sequestration is a critical component of NET-mediated killing and that TdfH contributes to overcoming Zn sequestration in neutrophil NETs. Our findings represent a novel gonococcal neutrophil resistance strategy, wherein TdfH assimilates Zn from CP, neutralizing its antimicrobial activity, to contribute to survival in NETs.
This report describes TdfH as a TonB-dependent transporter that mediates CP binding and Zn internalization. While at this time we do not know whether TdfH preferentially binds to Zn-laden CP as opposed to apo-CP, the data presented in this report clearly demonstrate that TdfH enables Zn transport into gonococci whether CP is present or not, suggesting that both TdfH and TdfJ have the capacity to import Zn that is not chelated to protein. It is also currently unclear why the tonB mutant is not impaired more than either of the single tdfH or tdfJ mutants for growth with or internalization of unchelated Zn. However, TonB-dependent functions in N. gonorrhoeae have to date been exclusive to protein-bound metals such as transferrin and lactoferrin, whereas siderophore-iron internalization can be Ton independent, depending on the strain tested (39). The data presented here suggest that Zn internalization by the gonococcus occurs in both a TonB-dependent manner and a TonB-independent manner. The growth experiments presented here also suggest that growth stimulation with unchelated Zn is more efficient at employing the TdfJ pathway and that TonB-dependent growth with unchelated Zn is primarily accomplished by TdfJ. Explanations for these observations could be related to relative affinities of TdfH and TdfJ for Zn or their respective copy numbers in the gonococcal outer membrane.
We propose that TdfH is an attractive potential vaccine target for several reasons. First, TdfH is surface exposed and not subject to phase variation. Second, it is produced by most pathogenic Neisseria strains. TdfH production has been detected in all meningococcal strains tested (35, 45) and 81% of gonococcal strains (45), but only a single isolate of Neisseria lactamica (45), a commensal that can occasionally cause infection (54). Third, it exhibits a high degree of sequence conservation among gonococcal strains. Fourth, TdfH is produced by gonococci recovered from women with lower genital tract infection (58). Finally, it is likely to be important for gonococcal survival in vivo by mediating acquisition of Zn via CP binding.
In summary, this study exposes a novel gonococcal strategy, centered on production of TdfH, to evade CP-mediated nutritional immunity and thereby contribute to survival in NETs. Given the role we demonstrate for TdfH in gonococcal growth and survival and its potential as a vaccine candidate, we anticipate future studies will fully discern the contributions of TdfH in N. gonorrhoeae pathogenesis and infection of its obligate human host.
We thank Alastair G. McEwan and Karrera Djoko for the zur::kan mutant construct used to generate the FA1090 and FA19 zur mutants. We are also grateful to William M. Shafer for providing the FA1090 misR and FA19 misR strains. We also thank Eric P. Skaar, Walter J. Chazin, and Benjamin Gilston for providing the purified, heterodimeric calprotectin used in this study. We thank Frederick Sparling and James Anderson for providing the TdfH-specific antibody. Finally, we thank Joseph Turner for assistance with ICP-OES and Donna Woodburn who helped with initial regulation studies.
The funders of this study had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.