ABSTRACT
The ability of a pathogen to metabolically adapt to the local environment for optimal expression of virulence determinants is a continued area of research. Orthologs of the Streptococcus iniae LysR family regulator CpsY have been shown to regulate methionine biosynthesis and uptake pathways but appear to influence expression of several virulence genes as well. An S. iniae mutant with an in-frame deletion of cpsY (ΔcpsY mutant) is highly attenuated in a zebrafish infection model. The ΔcpsY mutant displays a methionine-independent growth defect in serum, which differs from the methionine-dependent defect observed for orthologous mutants of Streptococcus mutans and Streptococcus agalactiae. On the contrary, the ΔcpsY mutant can grow in excess of the wild type (WT) when supplemented with proteose peptone, suggesting an inability to properly regulate growth. CpsY is critical for protection of S. iniae from clearance by neutrophils in whole blood but is dispensable for intracellular survival in macrophages. Susceptibility of the ΔcpsY mutant to killing in whole blood is not due to a growth defect, because inhibition of neutrophil phagocytosis rescues the mutant to WT levels. Thus, CpsY appears to have a pleiotropic regulatory role for S. iniae, integrating metabolism and virulence. Furthermore, S. iniae provides a unique model to investigate the paradigm of CpsY-dependent regulation during systemic streptococcal infection.
INTRODUCTION
Our understanding of the diverse repertoire of transcriptional signaling networks that orchestrate optimal expression of virulence genes dependent upon the metabolic status of the cell is quickly expanding (49, 60). Metabolic adaptations to the various environments encountered by a pathogen within a host prove to be tremendously complex (14). Recent work on streptococcal pathogens has revealed components of these regulatory pathways that relate the nutritional status of the cell to the control of growth phase and expression of virulence genes (1, 21, 23, 24, 36, 40, 51).
Streptococcus iniae is a major aquatic pathogen (15, 16) that causes an invasive systemic infection with severe bacteremia culminating in meningoencephalitis (2). Many instances of zoonoses have been reported as a result of handling infected fish, which typically results in a bacteremic cellulitis (17, 28, 56). The severity of an S. iniae infection is due in part to the ability to rapidly disseminate from the site of infection through the bloodstream and invade systemic tissues (34).
The complex pathogenicity of S. iniae, and other systemic streptococcal pathogens, is reflected in the diversity of genes that are critical to a successful infection (3, 22, 55). Several classical streptococcal virulence factors have been shown to be important for the virulence of S. iniae, including the SivS/R-regulated streptolysin S (sagA) and CAMP factor (cfi) (4, 31), as well as the major surface M protein (siM) (30). S. iniae also contains a polysaccharide capsule that functions for protection from phagocytic clearance in whole blood (32, 34, 39). Furthermore, several genes not previously studied for their virulence traits have been shown to have critical roles during S. iniae infection of hybrid striped bass (HSB), including phosphoglucomutase (pgmA) (8) and a novel polysaccharide deacetylase (pdi) (38).
We previously established the importance of a highly conserved LysR family transcriptional regulator, CpsY, in a zebrafish infection model (34). Deletion of cpsY (ΔcpsY) increased zebrafish survival by 40% 4 days postinfection. Moreover, the ΔcpsY mutant displayed a unique inability to disseminate to the brain within the first hour postinfection. CpsY was originally predicted to regulate capsule biosynthesis in Streptococcus agalactiae due to its adjacent location to the capsule operon (25). Further research determined that CpsY has little effect on capsule production in both S. agalactiae (48) and S. iniae (34). Rather, CpsY was renamed MtaR in S. agalactiae (48) and MetR in Streptococcus mutans (50) because of its influence on methionine biosynthesis and uptake. However, the regulatory function of MtaR in S. agalactiae appears to extend beyond methionine metabolism (6).
In the present study, we define a critical role for CpsY during S. iniae systemic infection. We show that CpsY is required for intracellular survival in neutrophils but not macrophages, which is critical for systemic dissemination in the bloodstream. Furthermore, we discuss variations in the CpsY-dependent regulation of methionine supply pathways that exist among streptococcal pathogens.
MATERIALS AND METHODS
Bacterial strains, media, and culture conditions.Streptococcus iniae 9117 is a human clinical blood isolate from a patient with cellulitis (18). A mutant containing an in-frame deletion of the cpsY gene used in this work was constructed previously (34). Streptococcal strains were routinely cultured in Todd-Hewitt broth (BBL) supplemented with 0.2% yeast extract (BBL) (THY) in conical tubes without shaking at 37°C. For bacterial enumeration, serial dilutions were plated on THY agar plates and incubated at 37°C in 5% CO2. In all assays, overnight bacterial cultures were diluted 1:50 in fresh medium and grown to mid-exponential phase (optical density at 600 nm [OD600] of 0.225) unless otherwise stated.
Bacterial growth curves.Overnight bacterial cultures were diluted 1:50 in 96-well plates containing a 200-μl total volume of THY, C medium (35), or a chemically defined medium (CDM) (54) supplemented with either 50 μg ml−1 l-methionine, 400 μg ml−1 l-methionine (Sigma), or 2% proteose peptone 3 (BBL) when required. The plates were incubated at 37°C in 5% CO2, and OD600 values were measured every 30 min using a Versamax microplate reader (Molecular Devices).
Growth assays in human serum were performed as previously described for Streptococcus agalactiae (48). Briefly, nonheparinized blood was obtained from human volunteers by venipuncture, allowed to clot, and centrifuged at 1,000 × g for 5 min. Mid-exponential-phase cultures of S. iniae strains grown in THY were collected by centrifugation, washed 2 times in phosphate-buffered saline (PBS; Invitrogen), and diluted to 1 × 105 CFU ml−1 in PBS. Ten microliters of diluted bacteria was added to 1 ml of Dulbecco's modified Eagle's medium (DMEM) with 50% human serum in a 1.5-ml microcentrifuge tube, supplemented with 400 μg ml−1 l-methionine (Sigma) when required. Tubes were incubated at 37°C with gentle rotation. At the indicated time points, bacterial growth was determined by serial dilution on THY agar.
Zebrafish infections.Groups of 6 zebrafish (Danio rerio) adults were infected by intramuscular (i.m.) injection into the dorsal muscle with either the S. iniae wild type (WT) or the ΔcpsY mutant from mid-exponential-phase cultures as previously described (41). A 30-gauge needle was used to inject 10 μl of a 1 × 107 CFU ml−1 bacterial suspension in PBS for an infectious dose of 1 × 105 CFU. Zebrafish were euthanized with a lethal dose of ethyl 3-aminobenzoate methanesulfonate (Tricaine; Sigma) at 24 h postinfection, and specific organs were harvested and gently homogenized in 300 μl PBS. One hundred microliters of the homogenate was loaded into a cytology funnel (Thermo), centrifuged onto glass slides for 3 min at 700 rpm in a Cytospin 1 centrifuge (Shandon Elliott), and stained using the HEMA 3 stain set (Fisher Scientific). Slides were viewed and documented with an Axioskop 40 microscope (Carl Zeiss, Inc.) fitted with an AxioCam MRc camera.
RNA isolation and quantitative PCR (qPCR).Bacteria were subcultured to logarithmic phase or early stationary phase in either THY or THY supplemented with 2% proteose peptone 3 (TP). When required, 4 mM l-methionine (Sigma) or homocysteine (Sigma) was added for 30 min upon reaching the appropriate growth phase. Five milliliters of culture was centrifuged and resuspended in 1 ml PBS with 25 mg ml−1 lysozyme (Sigma) and 50 U of mutanolysin (Sigma). Following 30 min of incubation at 37°C, the bacteria were centrifuged, resuspended in 1 ml TRIzol (Sigma), and incubated at room temperature for 5 min. Two hundred microliters of chloroform was added to the suspensions, mixed vigorously, and incubated an additional 3 min at room temperature. The preparations were centrifuged at 13,000 × g at 4°C for 15 min. The aqueous phase was transferred to a fresh microcentrifuge tube, mixed with 500 μl isopropanol, and incubated 10 min at room temperature. The samples were centrifuged at 13,000 × g at 4°C for 15 min, washed twice with 70% ethanol, and resuspended in 100 μl of water. Contaminating DNA was digested with 5 μg of DNase I (Qiagen) at 37°C for 2 h, and RNA was purified using the RNeasy MinElute cleanup kit (Qiagen).
For qPCRs, 100 μg of total RNA was combined in 20-μl total reaction volumes with 0.2 μM forward (fwd) and reverse (rev) primers, Express SYBR GreenER (Invitrogen), and Express SuperScript Mix (Invitrogen) per the manufacturer instructions. Primer sequences were as follows: recA fwd, 5′ CTCAGGTGCTGTTGATTTGG 3′; recA rev, 5′-TGCAGAGAGTTTACGCATGG-3′; atmB fwd, 5′-CCCGTTGGGATAAAATTGAA-3′; and atmB rev, 5′-CACCATTTGCAACTGCCTTA-3′. Reactions were performed in triplicate using a Bio-Rad iCycler with the following cycling conditions: 5 min at 50°C, 2 min at 95°C, and 40 cycles of 15 s at 95°C and 1 min at 60°C. Melting curves were performed after completion of each experiment. Relative fold change in gene expression was calculated by 2−ΔΔCT, where ΔΔCT is [CT (target gene) − CT (recA)]mutant − [CT (target gene) − CT (recA)]wild type, with CT being the threshold cycle.
Sequence alignment.S. iniae genomic sequence data were obtained from the Human Genome Sequencing Center at Baylor College of Medicine. Sequences for the CpsY orthologs MtaR of S. agalactiae A909 (accession number YP_329877), MetR of S. mutans UA159 (AAN58910), CpsY of Streptococcus pyogenes (NP_269094), and CpsY of S. iniae (AAY17292), as well as the nucleotide promoter sequences immediately upstream of atmB (metQ1) of S. agalactiae A909 (YP_330257), S. mutans UA159 (AAN59551), and Streptococcus pyogenes (NP_268657), were obtained from GenBank. Sequences were aligned and plotted using ClustalX through the Mobyle portal (http://mobyle.pasteur.fr/cgi-bin/portal.py) (42).
Macrophage adherence, internalization, and survival assays.The RAW 264.7 murine macrophage cell line (45) was maintained in 75-cm2 cell culture flasks (Corning) in DMEM (Gibco) containing 10% heat-inactivated fetal bovine serum (FBS; Gibco) at 37°C with 5% CO2. Macrophage infection assays were performed in 24-well flat-bottom plates (Greiner Bio One). One day prior to each assay, macrophages were seeded at a density of 0.5 × 106 cells per well in 1 ml DMEM with 10% FBS medium. When required, macrophages were activated with 100 U ml−1 of murine gamma interferon (IFN-γ) at 1.5 h postseeding. After overnight incubation, cell culture medium was removed, and macrophages were washed 3 times with PBS. Mid-exponential-phase bacteria were diluted 1:100 into 1 ml DMEM with 10% FBS medium, which was then added to cells at a multiplicity of infection (MOI) of 1 (bacteria-cells). The plates were centrifuged for 5 min at 250 × g to allow for contact between bacteria and the cell monolayer. After 1 h, cell culture medium was removed, and macrophages were washed 5 times with 1 ml of PBS for removal of any nonadherent bacteria. For adherence assays, the cells were lysed in 1 ml sterile double-distilled water (ddH2O), and 100 μl was serially diluted and plated on THY agar for enumeration of bacterial CFU, which represents both adhered and internalized bacteria. For internalization and survival assays, infected cells were incubated with 1 ml fresh DMEM with 10% FBS containing 100 μg ml−1 gentamicin (Gibco) for 1 h to kill all extracellular bacteria. After the 1-h incubation, cells were washed 3 times with PBS. For internalization assays, cells were lysed in 1 ml sterile ddH2O and 100 μl was removed, serially diluted, and plated on TP agar for enumeration of surviving intracellular bacteria. For survival assays, 1 ml of fresh DMEM with 10% FBS was added, which was then considered to be the 0-h time point. At specified times, supernatants were collected, and cells were washed 3 times in PBS and lysed in 1 ml sterile ddH2O. One hundred microliters of supernatants and lysates was serially diluted and plated on THY agar for enumeration of bacterial CFU. Final adherence numbers were calculated as follows: adhered-internalized CFU − internalized CFU. All infections were performed in duplicate, and each experiment was repeated 3 times.
Whole-blood survival.Heparinized human whole blood was collected by venipuncture. When required, cytochalasin D (Sigma) was added to a final concentration of 10 μg ml−1 for 30 min prior to incubation with bacteria. Mid-exponential-phase bacterial cultures were adjusted to 1 × 104 CFU ml−1, inoculated 1:100 in 1 ml whole blood, and incubated for 3 h at 37°C with gentle rotation. One hundred microliters was serially diluted and plated on THY agar for enumeration of bacterial survival. All infections were performed in duplicate, and each experiment was repeated 3 times.
Neutrophil infections.Human polymorphonuclear leukocytes, or neutrophils, were isolated as follows. Heparinized human whole blood was collected by venipuncture and mixed in equal volumes with 3% dextran (MP Biomedicals) in saline using 50-ml conical tubes. Red blood cells were allowed to sediment out, and the supernatant was centrifuged for 10 min at 800 × g at 4°C to pellet the remaining cellular fraction. Pellets were resuspended in 35 ml of 0.9% NaCl, and a 10-ml underlay of Ficoll-Paque (GE Healthcare) was applied. This was centrifuged for 30 min at 410 × g in a hanging bucket rotor at room temperature to separate out the granulocyte fraction. The supernatant was gently aspirated down to the loose pellet of granulocytes and erythrocytes. Remaining erythrocytes were lysed by resuspension of the pellet with 10 ml sterile H2O for 28 s, and isotonicity was quickly restored with 10 ml 1.8% NaCl. Purified neutrophils were pelleted by centrifugation at 500 × g and resuspended in 3 ml DMEM. Neutrophil concentration was determined by hemocytometer count.
For infections, neutrophils were diluted to a final concentration of 1 × 106 cells ml−1 in DMEM with 50% human serum. When required, neutrophils were pretreated 30 min prior to the addition of bacteria with 10 μg ml−1 cytochalasin D (Sigma) at 37°C with gentle rotation. For experiments requiring heat-inactivated serum, human serum samples were incubated at 55°C for 30 min prior to neutrophil addition. Mid-exponential-phase cultures were washed twice in PBS and diluted to 1 × 105 CFU ml−1 in PBS. Ten microliters was inoculated into a 1-ml neutrophil suspension or medium alone using 1.5-ml microcentrifuge tubes. Samples were incubated at 37°C for 3 h with gentle rotation. One hundred microliters of the sample was serial diluted and plated on THY agar for enumeration of bacterial survival. In parallel, 100 μl of the same sample was loaded into a cytology funnel (Thermo), centrifuged onto glass slides for 3 min at 700 rpm in a Cytospin 1 centrifuge (Shandon Elliott), and stained using the HEMA 3 stain set (Fisher Scientific). Slides were viewed with an Axioskop 40 microscope (Carl Zeiss, Inc.) fitted with an AxioCam MRc camera.
Statistical analysis.A statistical analysis for all functional tests was performed by two-tailed Student's t test using StatView analysis software.
Research compliance.All research presented complies with the regulations put forth by the Wayne State University Human Investigation Committee and the Institutional Animal Care and Use Committee.
RESULTS
CpsY influences growth in vitro.Previous work demonstrated that disruption of the CpsY ortholog MtaR in S. agalactiae results in a significant growth defect in human plasma that can be rescued with excess methionine (6, 48). To determine if this same phenotype existed in S. iniae, both the WT and the ΔcpsY mutant were cultured in human serum supplemented with and without 400 μg ml−1 l-methionine. Growth of the ΔcpsY mutant paralleled that of the WT for the first 3 h in serum alone, after which a significant drop in the growth rate of the mutant resulted in approximately 1.5 log fewer CFU ml−1 recovered by 6 h (Fig. 1). This attenuation could not be rescued with excess methionine. The additional methionine had no influence on the growth of WT S. iniae, consistent with previous reports for S. agalactiae (48). To determine if S. iniae was a methionine auxotroph, both the WT and the ΔcpsY mutant were cultured in a chemically defined medium lacking methionine (Fig. 2). In the absence of methionine, neither strain was able to grow; however, the addition of 50 μg ml−1 l-methionine to the medium rescued growth of both strains, indicating that S. iniae is auxotrophic for methionine and that CpsY has no influence on growth in CDM.
Serum growth curves. Mid-logarithmic-phase cultures of S. iniae WT along with the ΔcpsY mutant were diluted into 1 ml DMEM-50% human serum with (+) or without 400 μg ml−1 l-methionine. Samples were incubated at 37°C, and the number of bacterial CFU were determined at the indicated time points by serial dilution and plating on THY agar. Error bars represent ± standard errors. *, P < 0.05.
Effect of methionine on bacterial growth in chemically defined medium. Overnight cultures of S. iniae 9117 WT and the ΔcpsY mutant were diluted 1:50 into 200 μl of chemically defined medium (CDM). When required, medium was supplemented with 50 μg ml−1 methionine (+). Samples were incubated at 37°C, and OD600 measurements were recorded over time. A representative experiment is shown. Error bars represent ± standard errors.
To determine whether the methionine-independent growth attenuation of the ΔcpsY mutant in human serum translated to other media, growth of the ΔcpsY mutant was examined in THY broth. In this medium, the ΔcpsY mutant exhibited a pattern similar to that of growth in serum, demonstrating a significant drop after 3 h, which was independent of methionine (Fig. 3A). However, when cultured in the more nutritionally defined C medium, which contains 0.5% proteose peptone and no glucose, the ΔcpsY mutant grew identically to the WT (Fig. 3B). These identical growth patterns were due to a decrease in cell density at stationary phase of the WT in C medium compared to that in THY and not due to an enhancement in the growth of the ΔcpsY mutant. Again, the addition of excess methionine had no effect on growth of either the WT or the ΔcpsY mutant. To determine if the growth deficiency of the ΔcpsY mutant in THY broth was associated with the lack of proteose peptone, the medium was supplemented with 2% proteose peptone 3. Adding peptone drastically extended the exponential growth of the ΔcpsY mutant, reaching a final cell density over 2.5 times greater than that without peptone (Fig. 4A). This was contrasted by the minimal change in growth observed for WT S. iniae in the presence of peptone (Fig. 4A). A similar trend was observed for growth of the two strains in C medium, although growth phenotypes were not as severe (Fig. 4B). This suggests that WT S. iniae can manage the excess nutrients supplied in peptone with minimal effect on normal growth patterns, whereas the ΔcpsY mutant displays a type of unrestrained exponential growth. Furthermore, the data indicate that CpsY influences the growth patterns of S. iniae in a methionine-independent manner.
Effect of methionine on bacterial growth. Overnight cultures of S. iniae WT and the ΔcpsY mutant were diluted 1:50 into 200 μl of THY (A) or C medium (B). When required, medium was supplemented with 400 μg ml−1 l-methionine (+). Samples were incubated at 37°C, and OD600 measurements were recorded every 30 min. A representative experiment is shown. Error bars represent ± standard errors. *, P < 0.05.
Effect of proteose peptone on bacterial growth. Overnight cultures of S. iniae 9117 WT and the ΔcpsY mutant were diluted 1:50 into 200 μl of THY (A) or C medium (B). When required, medium was supplemented with 2% proteose peptone 3 (+). Samples were incubated at 37°C, and OD600 measurements were recorded every 30 min. A representative experiment is shown. Error bars represent ± standard errors. *, P < 0.05.
CpsY regulation of methionine metabolic pathways.The S. iniae CpsY protein shares 79% amino acid sequence identity to the orthologs MtaR of S. agalactiae and MetR of S. mutans. Both MtaR and MetR have been shown to regulate genes involved in methionine biosynthesis and uptake (48, 50). Thus, we hypothesized that the insensitivity to methionine on growth of the ΔcpsY mutant may be due to differences in these metabolic pathways. To address this hypothesis, the genomes of S. iniae and several other streptococcal species were examined by BLAST analysis for the presence of genes involved in the biosynthesis and uptake of methionine as previously described for S. mutans (Fig. 5) (50). Neither S. iniae nor S. agalactiae possesses the full complement of genes required for de novo methionine biosynthesis from cysteine (Table 1); however, all streptococcal species investigated were found to contain the AtmBDE (MetQ1NP) methionine transport system. Additionally, S. agalactiae contains the genes encoding MetE and MetF for conversion of homocysteine to methionine. S. iniae does not possess metE or metF and thus appears to be purely auxotrophic for methionine.
Schematic of streptococcal methionine biosynthesis and uptake pathways. The pathways represented were previously characterized for S. mutans (50). Genes present in S. iniae as determined by BLASTP analysis are shown in black, while genes absent are in gray. Gene designations are listed in Table 1. SAM, S-adenosylmethionine; SAH, S-adenosylhomocysteine; SRH, S-ribosylhomocysteine; MET, methionine; AI2, autoinducer 2.
Orthologs of the methionine metabolism and transport pathways
The methionine-dependent growth defect of the S. agalactiae mtaR mutant was attributed to the inability to effectively transport exogenous methionine due to reduced expression of the MetQ1NP methionine transport system (6, 48). Because exogenous methionine could not rescue growth of the S. iniae ΔcpsY mutant (Fig. 3A), the role of CpsY in regulation of the AtmBDE locus was investigated. Quantitative PCR (qPCR) was performed to assess CpsY-dependent regulation of the atmB gene under various growth conditions. When cultured in THY broth to either mid-logarithmic or stationary phase, the difference in gene expression for the ΔcpsY mutant relative to that of the WT was less than 2-fold (Table 2). However, the addition of 4 mM exogenous methionine for 30 min at mid-logarithmic or stationary phase resulted in a relative decreased expression of over 2.5- or 3.5-fold, respectively. This suggests that CpsY may function as a transcriptional activator of atmB in the presence of methionine, although whether methionine is a specific cofactor for CpsY is uncertain. The addition of exogenous homocysteine, however, had a minimal effect on atmB gene expression, as previously described for S. mutans (50). In contrast to these observations, the addition of 2% proteose peptone (TP) resulted in a 2.5-fold increased expression of atmB relative to that of the WT during mid-log phase but not stationary phase. Thus, CpsY may function as a transcriptional activator of atmB in the presence of exogenous methionine; however, the presence of proteose peptone disrupts this regulation, suggesting that other regulatory elements may be involved.
Relative atmB expression for the ΔcpsY mutant
Transcriptional control of atmB by CpsY orthologs is proposed to occur through interactions with a specific promoter element termed the Met box (6, 47, 50). Sequence alignment of the atmB promoter region in S. iniae to several other streptococcal species revealed the presence of a Met box in the same position upstream of the predicted −10 and −35 sites (Fig. 6). Overall, these data are consistent with the proposed regulation of the AtmBDE methionine transporter for other streptococcal species. However, this regulation does not explain how CpsY influences S. iniae growth given that exogenous methionine does not rescue the ΔcpsY mutant in vitro. Thus, CpsY may have pleiotropic effects on S. iniae that extend beyond regulation of methionine metabolism, which has also been suggested for S. agalactiae (6).
Promoter alignment for the atmBDE cluster. Sequences obtained from GenBank were aligned and plotted using ClustalX. The previously determined S. mutans transcriptional start site (50) is in bold, predicted −10 and −35 sites are underlined, the translational start codon is italicized, and predicted Met boxes are shaded. Sag, S. agalactiae A909; Spy, S. pyogenes M1; Sin, S. iniae 9117; Smu, S. mutans UA159.
CpsY facilitates S. iniae tissue-specific localization within macrophages.During our previous studies of S. iniae, we observed that CpsY was essential for systemic infection in a zebrafish model (34). Whereas WT S. iniae could disseminate rapidly to the brain following intramuscular injection, the ΔcpsY mutant displayed an inability to efficiently spread beyond the spleen (34). Thus, we sought to determine a more specific function for CpsY during the course of systemic infection. Adult zebrafish were infected with WT S. iniae or the ΔcpsY mutant by intramuscular injection, and systemic organs were harvested 24 h postinfection for histological analysis. Visual examination of HEMA-stained cytospin preparations from spleen homogenates showed that both the WT and the ΔcpsY mutant were contained within macrophage phagosomal compartments (Fig. 7A). For all infected fish (n = 6), the majority of splenic macrophages were filled with cocci, with few extracellular bacteria observed. In contrast, only the WT could be detected in cytospin preparations of brain homogenates (Fig. 7B). Brain macrophage-like cells from ΔcpsY mutant-infected fish showed signs of activation with large empty phagosomal vacuoles but contained no bacteria (Fig. 7B). These results suggest that S. iniae localizes within tissue macrophages during the course of systemic infection and that CpsY affects this localization within specific tissue environments.
Images of tissue homogenates from S. iniae-infected zebrafish. Zebrafish were injected i.m. with 105 CFU of either S. iniae WT or the ΔcpsY mutant. Zebrafish spleens (A) or brains (B) were dissected 24 h postinfection, gently homogenized in PBS, cytocentrifuged onto glass slides, and stained for visualization. Two representative images for each condition are shown. Black arrows depict examples of intracellular bacteria within macrophages. Macrophage nuclei are noted with a white asterisk. Average numbers of disseminated CFU for each organ at 24 h postinfection are indicated. *, P < 0.05.
Intracellular survival in macrophages is independent of CpsY.Because of the tissue-specific macrophage localization observed in the zebrafish infection model, we sought to determine if S. iniae intracellular survival in macrophages was dependent upon CpsY. The RAW 264.7 murine macrophage cell line (45) was used as an in vitro platform to evaluate streptococcus-macrophage interactions. Macrophages were infected with either WT S. iniae or the ΔcpsY mutant for 1 h to allow for bacterial uptake, followed by a 1-h gentamicin treatment to kill all extracellular bacteria. Bacterial growth was then determined at various time points over the following 24 h. No significant difference was observed between S. iniae WT and the ΔcpsY mutant in macrophage cell lysates (Fig. 8A) or culture supernatants (Fig. 8B) at all time points tested. The numbers of bacteria remained relatively constant at the 2- and 4-h time points and by 24 h had increased by over 3 logarithms. Adherence and internalization assays also revealed no significant difference in bacterial uptake by macrophages (Fig. 8C). Furthermore, macrophage activation by pretreatment with IFN-γ resulted in no significant difference from untreated samples (Fig. 8). Because all extracellular bacteria are initially killed during the gentamicin treatment, the observation of bacteria in the culture supernatant suggests that the intracellular bacteria are escaping the macrophage into the culture medium. A possibility is that S. iniae causes lysis of macrophages, allowing the bacteria to escape after lysis; however, S. iniae does not appear to be actively killing macrophages, because viability staining of infected macrophages showed that the majority of macrophages were alive, intact, and filled with bacteria through 48 h (data not shown). This would suggest that S. iniae has the ability to escape from the intracellular environment, but any details of this proposed mechanism have yet to be determined. Alternatively, it is possible that the bacteria isolated in the culture supernatants came from a small population of necrotic macrophages, which resulted in bacterial replication in the culture medium to the observed levels. This, however, is unlikely due to the limited growth of S. iniae in DMEM (data not shown) and the significant number of bacteria observed in the medium over time. Ultimately, these results indicate that CpsY does not play an essential role during interactions with macrophages in vitro and suggest an alternative role in the virulence of S. iniae.
S. iniae macrophage infections. RAW 264.7 macrophages were infected at an MOI of 1 with the WT (gray bars) or the ΔcpsY mutant (white bars) as described in Materials and Methods. At required time points, culture supernatants were collected, and cells were washed and lysed in H2O. Dilutions of the cell lysates (A) and culture supernatants (B) were serially diluted and plated for bacterial enumeration. For certain experiments, macrophages were activated by pretreatment with 100 U ml−1 of murine IFN-γ (γ) for 24 h prior to infection with bacteria. (C) Adherence (A) and internalization (I) assays were performed as described in Materials and Methods. Error bars represent ± standard errors.
CpsY is essential for survival in whole blood.Systemic dissemination from the initial infection site requires a pathogen to effectively enter and survive in the bloodstream. While in the bloodstream, a pathogen must withstand immune clearance by blood leukocytes. Because the ΔcpsY mutant was unable to disseminate to the brain, the requirement of CpsY for S. iniae survival in whole blood was investigated. Both WT S. iniae and the ΔcpsY mutant were incubated in human whole blood for 3 h and serially diluted onto THY agar plates to determine bacterial survival. WT S. iniae displayed an ability to replicate in whole blood, reaching levels approximately 2 logs over that of the inoculum (P < 0.05), while the ΔcpsY mutant was almost completely killed under the same conditions (P < 0.05) (Fig. 9). To determine if this difference in bacterial survival was due to phagocytic killing by blood leukocytes, whole blood was pretreated for 30 min with cytochalasin D (CD), an inhibitor of actin polymerization, to block phagocytosis. The addition of CD completely rescued the ΔcpsY mutant to WT levels (Fig. 9). These results suggest that leukocyte phagocytic killing is the primary bactericidal force against the ΔcpsY mutant and indicate a function for CpsY in the regulation of factors required for protection of S. iniae from phagocytic clearance.
S. iniae survival in human whole blood. Mid-log phase cultures of S. iniae WT (gray bars) or the ΔcpsY mutant (white bars) were diluted into 1 ml heparinized human whole blood and incubated at 37°C with gentle rotation. Samples were serially diluted and plated on THY agar for enumeration of bacterial CFU. For certain conditions, blood was pretreated for 30 min with 10 μg ml−1 CD to inhibit neutrophil phagocytosis. Error bars represent ± standard errors. *, P < 0.05.
CpsY protects S. iniae from neutrophil intracellular killing.To confirm that blood leukocytes were responsible for the extensive killing of the ΔcpsY mutant in whole blood, S. iniae WT and the ΔcpsY mutant were incubated for 3 h in either 50% human serum alone or with the addition of primary peripheral blood neutrophils. Significant growth was observed for both the WT and the ΔcpsY mutant in 50% human serum alone (P < 0.0001) (Fig. 10). The addition of neutrophils resulted in a reduction of WT growth compared to that in serum alone (P = 0.0005) (Fig. 10), which was still significant growth over the inoculum (P < 0.0001). In contrast, the ΔcpsY mutant was highly attenuated with the addition of neutrophils (P = 0.0001, compared to serum alone) (Fig. 10). Pretreatment of the neutrophils with CD rescued both the WT and the ΔcpsY mutant back to the levels observed in serum alone (Fig. 10), confirming the phenotype found in the whole blood.
S. iniae neutrophil infections. Mid-logarithmic-phase cultures of S. iniae WT (gray bars) or the ΔcpsY mutant (white bars) were diluted in 1 ml DMEM-50% human serum (S). When necessary, human primary neutrophils (N) were added to a final MOI of 0.001 (bacteria-neutrophil). If required, neutrophils were pretreated with 10 μg ml−1 CD for 30 min. Samples were incubated at 37°C with gentle rotation, followed by serial dilution and plating on THY for enumeration of bacterial CFU. Error bars represent ± standard errors. *, P < 0.01; **, P = 0.0001.
HEMA-stained cytospin slide preparations of infected neutrophils (in 50% human serum) revealed intracellular localization of both the WT and the ΔcpsY mutant within a vacuole (Fig. 11). Pretreatment with CD completely inhibited phagocytosis, but the bacteria remained closely associated to the neutrophil outer surface (Fig. 11). Similar observations were made when serum was heat inactivated prior to infection (Fig. 11). Thus, cpsY is not involved in preventing phagocytic uptake but rather must function to protect S. iniae within the neutrophil phagosome.
Images of S. iniae neutrophil infections. S. iniae WT or the ΔcpsY mutant were mixed with neutrophils suspended in 1 ml DMEM with 50% serum alone, with CD pretreatment, or with 50% heat-inactivated serum (HIS). After 3 h, 100 μl of the sample was cytocentrifuged onto glass slides and stained for visualization. A representative image for each condition is shown. Black arrows depict bacteria. Cell nuclei are noted with a white asterisk.
DISCUSSION
Bacteria have a variety of mechanisms for maintaining the required supply of nutrients needed for survival in nutrient-limited environments by either de novo synthesis or nutrient scavenging from the external environment. While many of these pathways are functionally conserved in both Gram-positive and Gram-negative bacteria, others differ markedly even within the same genera. Streptococcal species appear to use the same general pathway for managing intracellular methionine supply, although portions of the pathway appear to be missing in some species (Table 1).
The current collection of information on the streptococcal CpsY regulon is focused on its role in methionine metabolism and uptake (48, 50). De novo biosynthesis of methionine begins with the generation of homocysteine through transsulfuration (MetB, MetI, MetC), direct sulfhydrylation (MetB, MetY), or S-adenosylmethionine (SAM) recycling (MetK, Pfs, LuxS) pathways (Fig. 5). Homocysteine is then converted to methionine by the cobalamin-independent methionine synthase (MetE) and 5,10-methylenetetrahydrofolate reductase (MetF) (50). Portions of these pathways are incomplete in S. iniae, S. agalactiae, and S. pyogenes (Table 1) (47), and in these pathogens incomplete de novo biosynthesis is thought to be balanced by environmental methionine scavenging through the conserved AtmBDE (MetQ1NP, also called Abc and YaeEC) ABC transport system (47).
A comprehensive understanding of how streptococci regulate their repertoire of methionine metabolic genes is still emerging. In Gram-negative organisms, the MetJ repressor along with the SAM cofactor controls transcription of the de novo methionine biosynthesis genes and the atmBDE locus when methionine levels are sufficient (27). Repression is relieved under sulfur-limiting conditions when methionine biosynthesis is required (20). In addition, the CpsY ortholog MetR utilizes homocysteine as a cofactor to compete with the MetJ-SAM complex and increases methionine biosynthesis through transcriptional activation of metE and metF when methionine levels are insufficient (9, 11, 33, 43, 52, 53).
MetR is thought to be the primary regulator of methionine biosynthesis and uptake pathways for streptococcal species due to the absence of an MetJ ortholog (26). In S. mutans, MetR-dependent activation of metE and atmBDE occurs under methionine limitation but not excess (50). Expression of metE and atmB can be triggered in the presence of excess methionine only if excess homocysteine is also present in the medium, because homocysteine is the activating cofactor for MetR (50). In S. agalactiae, MtaR was shown to activate metQ1 (atmB) in the presence of excess methionine but had no effect on metE expression (6). The effect on metQ1 or metE in the absence of methionine was not investigated in that study. S. iniae does not carry the metE or metF gene but does contain genes encoding the AtmBDE transporter. We show here that a CpsY-dependent increase in atmB expression occurs in the presence of excess methionine but not homocysteine. Thus, as observed in other species, CpsY appears to increase cellular methionine pools through transcriptional activation of the AtmBDE transport system; however, the environmental condition under which this occurs differs between species. Those species which contain a complete de novo methionine biosynthesis pathway (e.g., S. mutans) limit expression of atmB in the presence of excess methionine and respond to homocysteine, whereas those with incomplete pathways (e.g., S. iniae) increase expression of the transporter and do not respond to homocysteine. This difference is not due to alterations in the promoter structure, because a canonical Met box (TATAGTTTnAAACTATA) was found in the same proximal location to atmB in all species. This may be due to differences in CpsY itself or additional regulatory elements that function either in cis or in trans to influence the regulation of this locus. Furthermore, the cofactor for S. iniae CpsY is unclear. Though the presence of methionine in the medium alters expression of atmB in a CpsY-dependent manner, it has not been determined that this is the actual cofactor.
The physiological explanation for these observed differences in S. iniae has yet to be determined but may be due in part to the absence of both MetE and MetF for conversion of homocysteine to methionine. This conversion is a key step in the recycling of the primary methyl donor SAM (58). An inability to convert homocysteine back to methionine due to the absence of MetEF would necessitate methionine scavenging for generation of SAM. Thus, S. iniae may have evolved alternate regulatory mechanisms to compensate for the lack of MetEF. S. pyogenes, Streptococcus uberis, and Streptococcus equi also do not have the metEF genes (26), but whether their respective CpsY orthologs function similar to that of S. iniae is unknown.
A systemic pathogen must adapt to one or more environments in the host, and as has been shown for many pathogens, these adaptations involve metabolic changes. The work presented here demonstrates a specific role for CpsY during Streptococcus iniae systemic infection. In addition to control of methionine transport through regulation of the atmB gene, CpsY of S. iniae is critical for adaptation and survival in the intracellular environment of neutrophils in whole blood. The enhanced susceptibility of the ΔcpsY mutant to neutrophil-mediated killing correlates with the attenuation initially observed in our zebrafish infection model (34).
While CpsY is absolutely required for S. iniae survival in neutrophils, it appears dispensable for survival in macrophages. This is supported in vivo, where localization of S. iniae within splenic macrophages of infected zebrafish is independent of CpsY. Contradictory was the observation that intracellular localization within zebrafish macrophage-like cells in the brain is dependent upon CpsY. An explanation for this dichotomy in vivo may be that the ability to survive and replicate in the bloodstream is a precursor for the establishment of a central nervous system (CNS) infection (7, 13, 44, 61). Thus, the inability of the ΔcpsY mutant to localize within brain macrophages could be due to limited bacteremia in vivo and not susceptibility to macrophage-mediated killing within the brain. If the ΔcpsY mutant were rapidly cleared in the bloodstream by neutrophils, then a decrease in the bacterial load at distant sites would be expected. This explanation correlates with previous reports where the ΔcpsY mutant displayed a unique inability to acutely disseminate to the brain 30 min after intramuscular infection, while bacterial loads in the spleen were equal to those of the WT (34). Neutrophils are the primary phagocytic cell found in the bloodstream and are vital for protection from invading pathogens (29). The ability to exploit the host inflammatory response and survive within neutrophils is an important virulence trait of several pathogens, including Staphylococcus aureus (19), Streptococcus pyogenes (37), Streptococcus suis (57), and Listeria monocytogenes (12). Further research into how CpsY protects S. iniae from neutrophil-mediated killing is ongoing.
Mutations in orthologs of CpsY have been shown to influence growth in vitro due to the disruption of methionine supply, as described above. This defect could be rescued with the addition of exogenous methionine for S. agalactiae (6, 48) and S. mutans (50) but, as shown here, not for S. iniae. In contrast, the addition of 2% proteose peptone resulted in unrestricted exponential growth for the ΔcpsY mutant, while growth of the WT strain was unaffected. These data suggest that in S. iniae, CpsY functions to maintain proper growth in vitro in a way that differs from that of S. agalactiae or S. mutans. The inability of the ΔcpsY mutant to maintain proper growth in vitro does not explain its virulence attenuation. The ΔcpsY mutant was highly attenuated in whole blood but was able to grow to WT levels upon inhibition of neutrophil phagocytosis by the addition of CD. Thus, the attenuation observed in whole blood is not due to a growth defect but is specifically due to the inability to survive neutrophil phagocytosis. The sensitivity to phagocytosis is specific to neutrophils, because growth in macrophages was unaltered. Both neutrophils and macrophages are important phagocytic cells of the immune system but have very different bactericidal mechanisms. The primary difference pertains to the neutrophil-specific cytoplasmic granules that fuse with the maturing phagosome, delivering an arsenal of bactericidal components that perform various cytotoxic functions (5). These granules are absent from macrophages and will likely provide evidence toward the function of CpsY in future studies. Overall, these data suggest that CpsY provides a unique virulence function that allows S. iniae to adapt for intracellular survival within neutrophils. The growth disparity observed for the ΔcpsY mutant in vitro might indicate an early entry into stationary phase, which could suggest a defect in cell wall biosynthesis, cell stress response pathways, or autolysin regulation rather than an auxotrophic growth phenotype (46).
CpsY orthologs have also been implicated in the regulation of genes unrelated to methionine acquisition. S. agalactiae MtaR affects the expression of a cell surface protease (cspA) and fibrinogen binding protein (fsbB) implicated in streptococcal pathogenesis, as well as numerous genes involved in arginine transport and sugar metabolism (6). In Pseudomonas aeruginosa, MetR has significant effects on swarming motility in semiviscous medium and was shown to control several transcriptional regulators, a type III secretion system, as well as pyoveridine and pyochelin synthesis genes (59). Furthermore, MetR in Vibrio harveii was shown to act directly on the luxCDABE operon to repress luminescence independent of the quorum-sensing AI1 and AI2 signal levels (10).
The data presented here demonstrate that CpsY has a pleiotropic role on S. iniae pathogenesis that extends beyond the traditional methionine metabolic processes. Furthermore, CpsY may have been co-opted to provide regulatory input on virulence gene expression based on the nutritional status of the cell. Moreover, S. iniae provides a unique model to investigate the paradigm of CpsY-dependent regulation during systemic streptococcal infection due to the absence of a methionine-dependent growth deficiency. Future studies on the CpsY regulon will provide an enhanced picture of the complicated regulatory networks that exist among streptococcal pathogens and the critical role they play during systemic infection.
ACKNOWLEDGMENTS
We thank Sarah K. Highlander and the Human Genome Sequencing Center at Baylor College of Medicine for access to the Streptococcus iniae genome sequence.
Genome sequencing was supported by grant 2006-35600-16569 to Sarah Highlander, George Weinstock, and Victor Nizet.
FOOTNOTES
- Received 23 June 2011.
- Returned for modification 26 July 2011.
- Accepted 31 August 2011.
- Accepted manuscript posted online 12 September 2011.
- Copyright © 2011, American Society for Microbiology. All Rights Reserved.
