Role of rsbU and staphyloxanthin in phagocytosis and intracellular growth of Staphylococcus aureus in human macrophages and endothelial cells

Abstract

In Staphylococcus aureus, rsbU dowregulates agr and stimulates production of staphyloxanthin (STX), an anti-oxidant that may contribute to intracellular survival after phagocytosis. Using isogenic rsbU^- and rsbU⁺ strains, we show that rsbU causes increased internalization and intracellular growth in both THP-1 macrophages and HUVEC cells (more so for the latter) without change in sub-cellular localization. Inhibition of STX biosynthesis markedly reduces intracellular growth of the rsbU⁺ strain (and of clinical isolates, including USA300; A5 tested with macrophages only), without affecting internalization. rsbU is important for uptake, and for STX biosynthesis, critical for intracellular multiplication of S. aureus.

INTRODUCTION

Staphylococcus aureus, a frequent cause of severe nosocomial and community-acquired bacterial infections in humans, produces a large collection of virulence factors that induce immediate local and general damage during infection. In this context, we have examined the role of rsbU, a phosphatase that positively controls σ^B, which itself down-regulates the expression of agr [1]. agr, indeed, is a regulator of S. aureus virulence and the transcriptional factor that plays a central role in stress response [2] and persistence of infection, in vivo [3]. The persistent and recurrent nature of staphylococcal infections has also been related in several studies to the existence of an intracellular pool of bacteria in both professional and non-professional phagocytes. The expression of virulence factors that damage the host is reduced in intracellular S. aureus (in order to maintain itself in the intracellular milieu [4]), but expression of other genes appears essential for resistance to host defenses. The lack of expression of the corresponding gene (rsbU) is also associated with reduced H₂O₂ tolerance related to impairment of the biosynthesis of staphyloxanthin (STX) [5], a carotenoid pigment that acts as an antioxidant, blocking attack by host reactive oxygen species [6].

Here, we take advantage of the fact that rsbU is absent in the common laboratory S. aureus strain 8325-4, but is present in the isogenic variant, SH1000. This enables an investigation of the internalization and intracellular growth of both strains in professional phagocytes (THP-1 macrophages) as well as in endothelial cells (HUVEC). Since in addition to enhanced pigmentation by STX, SH1000 also shows reduced secretion of exoproteins and dowregulation of agr [7, 8], we complemented our investigation by using a newly described inhibitor of STX biosynthesis, BPH-652 [9], in order to determine the role of STX in intracellular growth in 8325-4, SH1000, as well as in several clinical strains, including US300, one of the more virulent community-associated MRSA strains.

MATERIALS & METHODS

Bacteria

Experiments were performed with S. aureus 8325-4, a weak producer of STX with a natural deletion in rsbU, and in SH1000, a highly pigmented rsbU+ construct obtained from S.J. Foster [7]. We also tested three MRSA clinical isolates showing marked pigmentation, obtained from the Statens Serum Institute, Copenhagen, Denmark, and with the community-associated MRSA strain, US300 (NRS384; obtained from the Network on Antimicrobial Resistance in Staphylococcus aureus, NARSA; Focus Technologies, Inc., Herndon, VA). Bacteria were grown in Mueller-Hinton broth and CFU counting was carried out by plating on Tryptic Soy Agar.

Cells and intracellular infection

Human THP-1 cells (ATCC TIB-202; LGC Promochem Ltd, Teddington, UK) were cultivated using RMPI 1640 medium supplemented with 10% fetal calf serum (Invitrogen Ltd, Paisley, UK) and infected as described [10]. Human umbilical vein endothelial cells (HUVEC; Lonza Inc., Walkersville, MD) with less than 8 passages were seeded in gelatin-coated plates in DMEM-glutamax medium (Invitrogen) supplemented with 10% FCS, and infected following our protocol for adherent cells [11]. Phagocytosis was continued for 1 h at an initial inoculum of 4 pre-opsonized bacteria per cell, after which extracellular bacteria were removed by washing, and intracellular growth allowed to proceed for 24 h in the presence of gentamicin at 1x MIC [10, 11], to minimize growth of extracellular bacteria (< 10 % of total bacterial counts at 24 h). Cell viability was tested at the end of the experiments by Trypan blue staining (< 15 % stained cells in all conditions; a low initial inoculum was critical in these studies to avoid rapid cell killing by apoptosis, commonly seen with larger inocula).

Resistance of bacteria to hydrogen peroxide

We used the method described [6] but with a larger concentration of catalase (100 U/mL), in order to fully destroy any H₂O₂ remaining at the end of incubation period.

Microscopy

For in situ confocal fluorescence microscopy (MRC 1024 confocal microscope; Bio-Rad laboratories, Richmond, CA) of infected cells, bacteria were stained overnight with 0.25 mg/mL fluorescein-5-isothiocyanate (FITC) (Invitrogen) prior to phagocytosis, and 50 μM LysoTracker® Red DND 99 (Invitrogen) was added 1 h before the end of the incubation period, to stain lysosomes. Images were processed with the Confocal Assistant software version 4.02 (http://www.nephrology.iupui.edu/imaging/software.htm). For electron microscopy, infected cells were harvested, washed with PBS, fixed with 2 % glutaraldehyde and 1 % osmium tetroxide, then stained “en bloc” with uranyl acetate [11].

Dehydrosqualene synthase inhibitor and other reagents

The dehydroqualene inhibitor BPH-652, which blocks STX synthesis, was prepared as described previously [9, 12]. Gentamicin was obtained from Glaxo-SmithKline s.a., Genval, Belgium as the commercial product registered for human administration. Gelatin and H₂O₂ were from Sigma-Aldrich (St Louis, MO).

Statistical analyses

Statistical analyses (ANOVA) were performed with GraphPad Instat 3.06.

RESULTS

In a first series of experiments, we compared the susceptibility to phagocytosis and intracellular growth of 8325-4 (rsbU^-) and SH1000 (rsbU⁺) strains. A3 To measure phagocytosis, cells were exposed to opsonized bacteria for 1 h and were then collected after extensive washing (to remove non internalized S. aureus). As shown in Figure 1A, phagocytosis by HUVEC cells was significantly more important with SH1000 (about a 1.8-fold increase) than with 8325-4. With THP-1 cells, phagocytosis was globally more efficient, but, again, with a significant difference between the two strains (about 1.5-fold). To measure intracellular growth, cells exposed to bacteria as indicated above were returned to bacteria-free medium and incubated for 24 h before being collected (also after extensive washing). Intracellular infection could be obtained for both strains in both cell types, but there was significantly more growth of SH1000 compared to 8325-4 (Figure 1B). A2-1 In view of this result, 4 additional studies were made. First, we examined whether SH1000 and 8325-4 would not grow at different rates in broth under conditions mimicking what could take place in cells (i.e. logarithmic growth and up to densities of about 10⁹ CFU/mL [see 10 for details]), but no difference was seen. Second, confocal and electron microscopy was used to detect potential differences in the subcellular localization of both strains. These were found confined to phagolysosomes in THP-1 cells at 3 and 24 h, as reported previously for other strains of S. aureus (11). In HUVEC, most bacteria were also found in phagolysosomes, although a small number were also observed in the cytosol, but with no difference between strains. Third, and because STX had been reported to protect phagocytized S. aureus against reactive oxygen species-dependent cell defense mechanisms [6], we tested whether SH1000 in broth was more resistant to H₂O₂ than was 8325-4. Figure 1C shows that this is indeed the case, with concentrations of H₂O₂ required to kill 50 and 90 % of bacteria being about 42 and 100 mM for SH1000, versus 16 and 22 mM for 8325-4.

Figure 1.

Phagocytosis, intracellular growth, and susceptibility to H₂O₂ of the isogenic S. aureus 8325-4 rsbU^- and SH1000 rsbU⁺ strains. A, enumeration of cell-associated CFU in HUVEC (open symbols) and THP-1 macrophages (closed symbols) after 1 h phagocytosis of 8325-4 (circles) or SH 1000 strains (triangles). Each data point corresponds to the actual counts from independent samples (n = 30 for 8325-4 in HUVEC and 24 for all other conditions). The horizontal bar represents the corresponding mean values. Statistical analysis (ANOVA, Tukey post-hoc test): all conditions A3 were significantly different from one another (p < 0.001 for all comparisons except THP-1 8325-4 vs. THP-1 SH1000 for which p is < 0.01). B, intracellular growth of 8325-4 or SH1000 in HUVEC (white bars) or in THP-1 macrophages over 24 h post-phagocytosis. Values are expressed as the change in CFU per mg protein over the incubation period ± SD of 3 independent samples. Statistical analysis (ANOVA, Tukey post-hoc test): bars with different letters are significantly different from one another (p < 0.001). C, susceptibilities of 8325-4 (circles) or SH1000 (triangles) strains to of H₂O₂ in broth. Bacteria were incubated for 45 min with increasing concentrations of H₂O₂ (0 - 100 mM). Values are expressed as the percentage of CFU compared to controls (no H₂O₂) and are the means ± SD of 3 independent samples.

In a second series of experiments, we examined the influence of the STX biosynthesis inhibitor BPH-652 on bacterial phagocytosis and growth of SH1000 in HUVEC and THP-1 macrophages, as well as of other pigmented strains in THP-1 macrophages. A2 We first verified that BPH-652 (100 μM) added to broth (i) impaired the biosynthesis of staphyloxanthin (illustrated for SH1000 in Figure 2A); (ii) did not affect bacterial growth (checked by change in OD_620nm [turbidimetry] and bacterial counts over 24 h). Bacteria pre-exposed to BPH-652 were then used to infect HUVEC or THP-1 macrophages, and intracellular growth measured as described above. BPH 652 was maintained at 100 μM throughout the experiment, which neither significantly modified the extent of phagocytosis A1 (less than 1 % difference in 2 independent experiments), nor affected cell viability (≥ 85 %, based on the trypan blue exclusion test; no significant difference in total protein content of samples prepared from treated vs. control cells). A4 However, it significantly reduced the intracellular growth SH-1000 (about 1 log₁₀ CFU in both cell types (Figure 2B). BPH 652 also reduced the growth of S8325-4 but to a considerably lesser extent, in relation to the lower production of STX (as evidenced by pigment loss). We also compared in THP-1 cells the growth of a series of clinical isolates of various level of pigmentation (assessed by visual inspection). As shown in Figure 2, their growth (i) was proportional to pigmentation and (ii), except for SA63062, was markedly reduced in the presence of BPH 652, to a level that was similar for all strains including 8325-4.

Figure 2.

A: influence of the dehydrosqualene synthase inhibitor BPH-652 on the pigmentation of the SH1000 strain grown in broth. Bacteria were grown for 2 days in the absence (CONTR) or in the presence of the inhibitor (100 μM) before being pelleted for photography (8325-4 strain subjected to the same treatments is also shown, to demonstrate the near absence of pigmentation under all conditions). B: Intracellular growth of SH1000 and 8325-4 in HUVEC cells, and of SH1000, SA63062, SA63049, SA63063, US300 (4 clinical isolates) and 8325-4 in THP-1 macrophages in the absence (CONTROL; strains are ranked according to the intensity of their pigmentation as assessed by visual inspection of colonies) or in the presence of BPH 652. Bacteria, treated as in the upper panel, were opsonized and phagocytized by the cells using the same protocol as in the experiments shown in Figure 1 with inhibitors maintained at 100 μM in the culture medium for the whole incubation period. Data are presented as in Figure 1B. Statistical analysis: (i) paired comparisons for each strain between control and BPH 652 (unpaired t-test two-tailed): all differences are significant at p < 0.002; (ii) multiple comparisons between strains in control conditions or in the presence of BPH 652 (one way ANOVA with Tukey-Kramer Multiple Comparison test applied independently for HUVEC and THP-1 cells): bars with different letters (a to e for controls; x and y for BPH 652) are significantly different from each other at p < 0.01.

DISCUSSION

The main conclusions that can be drawn from the present study are: 1) that restoration of rsbU in the naturally deficient S. aureus laboratory strain 8325-4 enhances its internalization and intracellular growth in both HUVEC and THP-1 cells, without change in sub-cellular localization and 2), that inhibition of STX biosynthesis impairs the intracellular growth of SH1000 A5 in both HUVEC and macrophages. A similar effect was seen A5 in macrophages for several other pigmented strains of clinical interest. Thus, rsbU appears to control two key steps in the establishment of an intracellular infection: phagocytosis and intracellular growth, but by two distinct downstream mechanisms.

Phagocytosis of S. aureus involves binding of the bacterium to the cell surface, which, in endothelial cells, takes place via fibronectin-binding proteins that in agr- mutants are upregulated and show improved binding [13]. This may explain the larger internalization seen with the SH1000 strain, since rsbU negatively controls agr expression through σ^B. A similar influence of agr on S. aureus internalization is probably also important in THP-1 macrophages, although the basal phagocytic activity of these cells is globally higher, since internalization in professional phagocytes operates through a trigger mechanism in which cells produces pseudopodes, making the overall process more efficient [14].

We showed previously with Listeria-infected macrophages and epithelial cells that intracellular growth (of an intracellular bacterium) is the result of a dynamic balance between bacterial multiplication capabilities and host defense destructive mechanisms, mainly related to the production of reactive oxygen and reactive nitrogen species [15]. The present experiments suggest that the same phenomenon is taking place in S. aureus, with STX providing a primary defense against oxidative damage. This STX-related mechanism is not expected to affect phagocytosis, as was indeed observed. These results are, therefore, consistent with in vivo observations which show that STX accumulation plays a critical role in triggering sustained tissue infections with abscess formation, but do not affect mucosal colonization (9). And while intracellular survival of S. aureus is determined by multiple virulence factors (1), our results point to a critical role for rsbU in the control of virulence in clinical isolates from persistent infections collected from intracellular foci. A5 While generalization to more clinical isolates and more cell types is desirable, the present data already suggest ways for new therapeutic approaches based on the control of STX biosynthesis in S. aureus, as a complement or alternative to direct intervention at the level of this gene product and the corresponding downstream products.