Lack of Wall Teichoic Acids in Staphylococcus aureus Leads to Reduced Interactions with Endothelial Cells and to Attenuated Virulence in a Rabbit Model of Endocarditis

Journal of Infectious Diseases, May 2005

Wall teichoic acids (WTAs) are major surface components of gram-positive bacteria that have recently been shown to play a key role in nasal colonization by Staphylococcus aureus. In the present study, we assessed the impact that WTAs have on endovascular infections by using a WTA-deficient S. aureus mutant (ΔtagO). There were no significant differences detected between the isogenic parental strain (SA113) and the ΔtagO mutant in polymorphonuclear leukocyte–mediated opsonophagocytosis; killing by a prototypic platelet microbicidal protein; or binding to platelets, fibronectin, or fibrinogen. However, compared with the parental strain, the ΔtagO mutant adhered considerably less well to human endothelial cells, especially under flow conditions (70.3% reduction; P<.05). Beads coated with WTA bound to endothelium in a dose-dependent manner, suggesting that WTA contributes specifically to this interaction. These in vitro data closely paralleled those from a rabbit model of infective endocarditis in which the ΔtagO mutant was compared with the parental strain. Clearances of staphylococcus from the bloodstream were equivalent, but the ΔtagO mutant showed a significantly reduced capacity to both colonize sterile cardiac vegetations (P<.05) and proliferate within these vegetations, the kidneys, and the spleen (P<.001). We conclude that WTA is an important factor in the induction and progression of endovascular S. aureus infection, likely through a specific interaction with endothelial cells

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

http://jid.oxfordjournals.org/content/191/10/1771.full.pdf

Lack of Wall Teichoic Acids in Staphylococcus aureus Leads to Reduced Interactions with Endothelial Cells and to Attenuated Virulence in a Rabbit Model of Endocarditis

Christopher Weidenmaier 2 3 Andreas Peschel () 2 3 Yan-Qiong Xiong 0 2 4 Sascha A. Kristian 2 5 Klaus Dietz 1 2 Michael R. Yeaman 0 2 4 Arnold S. Bayer 0 2 4 0 Geffen School of Medicine at UCLA, University of California at Los Angeles , Los Angeles 1 Department of Medical Biometry, University of Tu bingen , Tu bingen , Germany 2 Received 15 June 2004; accepted 17 December 2004; electronically published 11 April 2005. Presented in part: Status seminar of Priority Program SPP1130 of the German Research Council , Dresden, Germany, 9-11 September 2003. Financial support: German Research Council (grants SPP1130 and FOR 449/T2 to A.P.); National Institutes of Health (grant AI-39108 to A.S.B. and grants AI- 48031 and RR-13004 to M.R.Y.); American Heart Association (Western Affiliate) (grant 0265054Y to Y.-Q.X.). crobiology Div., Medical Microbiology and Hygiene Dept., University of Tu bingen , 72076 Tu bingen , Germany 3 Cellular and Molecular Microbiology Division, Medical Microbiology and Hygiene Department 4 Department of Medicine, Division of Infectious Diseases, LA Biomedical Research Institute, St. John's Cardiovascular Research Center, Harbor-UCLA Medical Center , Torrance 5 Department of Pediatrics, Division of Infectious Diseases, University of California at San Diego , La Jolla Wall teichoic acids (WTAs) are major surface components of gram-positive bacteria that have recently been shown to play a key role in nasal colonization by Staphylococcus aureus. In the present study, we assessed the impact that WTAs have on endovascular infections by using a WTA-deficient S. aureus mutant (DtagO). There were no significant differences detected between the isogenic parental strain (SA113) and the DtagO mutant in polymorphonuclear leukocyte-mediated opsonophagocytosis; killing by a prototypic platelet microbicidal protein; or binding to platelets, fibronectin, or fibrinogen. However, compared with the parental strain, the DtagO mutant adhered considerably less well to human endothelial cells, especially under flow conditions (70.3% reduction; P ! .05). Beads coated with WTA bound to endothelium in a dose-dependent manner, suggesting that WTA contributes specifically to this interaction. These in vitro data closely paralleled those from a rabbit model of infective endocarditis in which the DtagO mutant was compared with the parental strain. Clearances of staphylococcus from the bloodstream were equivalent, but the DtagO mutant showed a significantly reduced capacity to both colonize sterile cardiac vegetations (P ! .05 ) and proliferate within these vegetations, the kidneys, and the spleen (P ! .001). We conclude that WTA is an important factor in the induction and progression of endovascular S. aureus infection, likely through a specific interaction with endothelial cells. - An important feature of Staphylococcus aureus bacteremia is the high frequency with which the organism spreads from the bloodstream to other targets, such as bone and joints, the lungs, the central nervous system, and heart valves [1]. Of note, S. aureus is the most prevalent bacterial pathogen isolated from patients with endovascular infections (e.g., vascular catheter sepsis and infective endocarditis [IE]) [2, 3]. The process of endovascular infection is believed to involve, at least in part, a complex interactionincluding binding, induction of phagocytosis, microbial phenotypic switching (e.g., to small-colony variants [4]), and modulation of cellular functions (e.g., apoptosis)between circulating bacteria and vascular endothelial cells [5]. S. aureus is relatively unique among common endovascular pathogens in its ability to bind to endothelial cells by multiple, distinct mechanisms [1, 2]. For instance, this organism can adhere to damaged endothelium by exploiting platelets and matrix ligands deposited at such sites (e.g., fibrin, fibronectin, and fibrinogen). In contrast, platelets may also play a role in the host defense by limiting the progression of endovascular infections through its local secretion of endogenous cationic antimicrobial peptides, termed platelet microbicidal proteins (PMPs) [6]. Moreover, S. aureus is capable of adhering to normal endothelium in platelet-independent ways [7], although the underlying mechanisms are only partly understood and only some of the molecules involved are known. Although fibronectin appears to be an important factor in the adhesion of S. aureus to endothelial cells [7], which factors other than fibronectin may be necessary or contribute to this adhesion remain unclear. S. aureus can induce its own uptake into human endothelial cells and then persist in this niche, where it is inaccessible to host defenses and antibiotics [8]. This latter event is fibronectin dependent [7]. Studies from the 1980s implicated wall teichoic acid (WTA), a staphylococcal cell wall polymer [9], in binding to a variety of human cells [10]. Of note, a WTA-deficient S. aureus mutant (DtagO) adhered less efficiently to human nasal epithelial cells than did its parental strain, and it was unable to colonize the nares of cotton rats in an experimental infection model [11]. Thus, although WTA is not essential for cell viability [11], it appears to function as a biologically relevant adhesin. The present study was designed to broaden the scope of our prior investigations and assess the role that WTA plays in interactions between S. aureus and endothelial cells, both in vitro and in vivo. We demonstrate that a deficiency in WTA leads to a profound reduction in the attachment of S. aureus to endothelial cells and to the attenuation of virulence in a rabbit model of bacteremia and endocarditis. MATERIALS AND METHODS Bacterial strains. The WTA-deficient DtagO mutant [11] (figure 1) was generated in S. aureus SA113 (ATCC 35556), a frequently used laboratory strain [1215]. Plasmid pRBtagO used in selected in vitro studies contains a wild-type copy of tagO and its promoter, and it reconstitutes WTA biosynthesis at the level in the parental strain and all relevant phenotypic properties [11]. This plasmid was recently shown to be very unstable in the absence of antibiotic selection [11], which precluded its use in the in vivo experiments. An isogenic S. aureus DdltA mutant lacking the d-alanine residues in WTAs [12] was used in some control experiments. Opsonophagocytosis by human polymorphonuclear leukocytes (PMNLs). Bacteria were grown to midlogarithmic phase in Iscoves modified Dulbeccos medium (IMDM; Gibco), were washed twice in Hanks balanced salt solution, and were resuspended in RPMI 1640 with 0.05% human serum albumin (HSA; CLB). Human PMNLs were isolated from healthy volunteers, as described elsewhere [16]. Fifty-microliter bacterial suspensions (2.5 108 cfu/mL) were mixed with 10 mL of normal human serum (AB serum pooled from 10 healthy donors) and were incubated in 96-well microtiter plates for 10 min at 37 C, as described elsewhere [17]. Fifty microliters of the PMNL mixture containing 5 106 cells/ mL in RPMI with or without 0.05% HSA was added to the plates (yielding a final PMNL:bacteria ratio of 1:50), and the phagocytosis assay was performed under rotation at 37 C. To remove extracellular bacteria, 10 mL of lysostaphin (final concentration, 5 mg/mL; Sigma) was added after 20 min and again after 50 min, and incubation was continued for another 10 min at 37 C until all extracellular bacteria were destroyed. Samples were run in duplicate for each time point. As a control, bacteria and PMNLs were incubated in the absence of HSA. After staining cytospin slides with Wrights stain, the number of bacteria per PMNL was counted. All experiments were performed in triplicate, and at least 14 PMNLs/slide were examined in randomly chosen high-power magnification fields. Susceptibility phenotypes to PMPs. Mammalian platelets secrete PMPs on stimulation with agonists relevant to endovascular infection, such as thrombin [1822]. Susceptibility to thrombininduced PMP (tPMP) was assayed in vitro, as described elsewhere [2022]; the results were expressed as the percentage of the initial inoculum (1 103 cfu) that survived a 2-h exposure to tPMP at a specific activity of 2 mg/mL. For each isolate, tPMP susceptibility was expressed as the average of the results from 2 independent experimental assays performed on separate days. Preparation of tPMP, bioactivity assays, and the quantitative in vitro assay methods for tPMP susceptibility phenotypes have been described in detail elsewhere [21]. Interactions of S. aureus with platelets. S. aureus was mixed with rabbit platelets, and binding studies were performed at 20 C. Flow cytometry, by use of fluorometrically distinct bacterial and platelet fluorophores and a FACSCalibur cytometer (Becton Dickinson), was performed to detect and quantify the extent of bacterial binding to platelets, as described elsewhere [23]. The percentage of bacteria that bound to platelets was calculated by dividing the number of dually labeled particles by the total number of bacteria and multiplying by 100. Data were expressed as the means SD for at least 3 independent experiments performed on separate days. Pilot studies indicated that a 1:1 bacterial:platelet ratio (1 108 cells) yielded optimal binding results and that maximal binding occurred within 5 min of coincubation. These parameters were used for the formal analyses. Adherence to fibronectin or fibrinogen. Bacterial adherence was studied as described elsewhere [11, 24]. Briefly, 96well microtiter plates (Costar) were coated with 50 mg of either fibronectin or fibrinogen (Sigma) in PBS for 15 h at 4 C. Subsequently, wells were blocked with 2% bovine serum albumin (BSA) in PBS for 1 h and were washed twice with PBS. Bacteria were grown in IMDM to midlogarithmic phase, were washed twice with PBS, and were adjusted to a concentration of 1 109 cells/mL using a Neubauer chamber. One hundred microliters of bacterial suspension was added to each well. After 1 h of incubation at 37 C, the wells were washed 3 times with PBS and were stained with safranin for 1 min, and absorption as measured at an optical density of 600 nm was determined in a FluostarOptima microtiter plate reader (BMG Labtek). Interactions of S. aureus with human umbilical vein endothelial cells (HUVECs). HUVECs that had been cultured in endothelial cell growth medium (Promocell) were used up to passage number six. HUVECs were seeded to 24-well culture plates at a concentration of 1 105 HUVECs/well and were grown at 37 C and in 5% CO2 until confluence. Bacteria were prepared and labeled with fluorescein isothiocyanate (FITC), as described elsewhere [11], and were resuspended in IMDM. The concentration of bacteria was adjusted using a Neubauer chamber. Confluent HUVEC monolayers grown in 24-well plates were washed twice with IMDM and were inoculated with FITClabeled bacteria. In pilot studies, the dependence of the level of bacterial adherence on the concentration of the inoculum was confirmed by using bacteria:HUVEC ratios ranging from 5:1 to 100:1. A ratio of 30:1 was found to be optimal and was used in the analyses. After incubation for 1 h at 37 C and in 5% CO2, the wells were washed 3 times with IMDM and were fixed with 3.5% paraformaldehyde in PBS. No morphological changes were observed in HUVECs after this procedure. The number of adherent bacteria/0.1 mm2 was counted using a fluorescence microscope (Leica). Experiments were performed in duplicate; 10 random fields were counted in each well. In control experiments with unlabeled bacteria, HUVECs were cultured in collagen-precoated chamber slides (Lab-Tek II chambers, 2.4 cm2; Nunc) (seeding density, 2 105 HUVECs/well) until confluence. Bacteria were prepared as described above except that the labeling procedure was omitted, and then they were incubated with HUVEC monolayers as described above. After being fixed with 3.5% paraformaldehyde in PBS, the wells were blocked with 1% BSA in PBS. Bacteria were stained with a monoclonal anti-Staphylococcus antibody (USBiological) for 1 h. The wells were washed with IMDM and were incubated with a tetramethylrhodamine-isothiocyanate (TRITC)conjugated antimouse antibody (Sigma) for 1 h. The wells were washed with IMDM, and the number of adherent bacteria/0.1 mm2 was counted using a fluorescence microscope. Bacterial adhesion to HUVECs was also monitored under flow conditions. HUVECs were cultured as described above in collagen-precoated chamber slides, which were manipulated by insertion of sterile tubing into the chamber walls. The chambers were inoculated with FITC-labeled bacteria by use of 2 peristaltic pumps, under flow conditions of 2.5 mL/min and 2 108 bacteria/mL in IMDM, for 20 min. After this period, the flow was stopped, the cells were fixed, and the number of adherent bacteria/0.1 mm2 was counted as described above. Adherence of WTA-coated beads. Amine-modified fluorescent beads (1.0-mm-diameter FluoSpheres; Molecular Probes) were washed with 10 mmol/L potassium phosphate buffer (PPB; pH, 7.0) and were incubated with 500 mg/mL purified WTA for 30 min. Purification of WTA has been described in detail elsewhere [11]. The coated beads were washed twice with PPB and were resuspended in PPB containing 1% BSA, to block hydrophobic areas. The amount of adsorbed WTA was determined by measuring the amount of n-acetylglucosamine that was released when 100 mL of WTA-coated beads without BSA was boiled at 100 C for 10 min, as described elsewhere [11]. The amount of adsorbed WTA was found to be the same for the parental strain and the DdltA mutant. WTA-coated beads were diluted in RPMI and were used in adhesion assays on confluently grown endothelial cells with bead:HUVEC ratios of 12.5:1, 25:1, and 50:1, and the procedures described above for FITC-labeled bacteria were followed. The relative fluorescence at 505 nm/well (excitation) and that at 515 nm/well (emission) were quantified using a fluororeader (BMG Labtek). In some experiments, the number of WTA-coated beads/0.1 mm2 was also counted microscopically, as an internal control for the fluorescence data. Rabbit model of IE. All animal experimentation was performed in accordance with the guidelines for animal health and welfare required by the Research and Education Institute at Harbor-UCLA Medical Center. Female outbred New Zealand White rabbits (Irish Farms) underwent carotid arterytoleft ventricle catheterization as described elsewhere [25]. For the comparative virulence assessments, 910 rabbits each received an inoculum of either 5 105 cfu or 5 106 cfu of either the parental strain or the DtagO mutant (which encompasses the ID95 for these strains for inducing IE, as established by pilot studies) intravenously into the marginal ear vein at 24 h after catheterization. At 48 h after challenge, rabbits were euthanized, and cardiac vegetations, the kidneys, and the spleen were aseptically removed. Vegetations, renal tissues, and splenic tissues were then quantitatively cultured as described elsewhere [25]. Quantitative culture data were expressed as log10 colony-forming units per gram of tissue. Only rabbits that had proper placement of the transaortic valve catheter and macroscopic culture-positive vegetations were considered to have active IE and were included in the final analyses. To examine the impact of the tagO mutation on early valvular colonization, an additional set of rabbits underwent catheterization as described above, to induce sterile vegetations on the aortic valve. Six rabbits each were challenged at 24 h after catheterization with an intravenous inoculum of 5 107 cfu of either the parental strain or the DtagO mutant. Rabbits were then euthanized at 60 min after challenge, their cardiac vegetations were removed and quantitatively cultured, and the number of bacteria adhering at this site was determined as described elsewhere [26]. Maximal valvular adhesion generally occurs within this initial 60-min postchallenge period [26]. To assure that these early valvular adhesion data were not influenced by potential differences in clearance of S. aureus from the bloodstream during the 60-min postchallenge period, quantitative blood cultures were performed for all rabbits at 10 min, 30 min, and 60 min after challenge. Statistics. In vitro and in vivo data were analyzed by either Students t test or Kruskal-Wallis analysis of variance. P ! .05 was considered to be statistically significant. In Vitro Studies PMNL-mediated opsonophagocytosis. Phagocytosis of the parental strain of S. aureus, the DtagO mutant, and the DtagO mutant complemented with plasmid pRBtagO was strongly increased on opsonization with pooled normal human serum (figure 2). No significant differences in the opsonophagocytosis of the WTA-bearing parental strain and the WTA-deficient DtagO mutant were detectable, indicating that WTA is not a dominant target for opsonins under the conditions used. Profiles of susceptibility to tPMP. Both the parental strain and the DtagO mutant were resistant in vitro to tPMP at 2 mg/ mL (mean SD survival at 2 h of incubation, 71% 10% and 85% 6%, respectively). The resistance of these strains to tPMP is consistent with that to other endogenous cationic antimicrobial peptides, including those of PMNL origin (e.g., defensins 13, cathelicidin LL-37, and lactoferrin), as reported elsewhere [11]. Adherence to platelets, fibronectin, and fibrinogen. The parental strain and the DtagO mutant adhered equally well to rabbit platelets, with a maximal mean SD binding of 45% 25% and 46% 21%, respectively. Moreover, the time to maximal binding was equivalent for the strains (15 min). There was also no difference in the ability of the 2 strains to interact with fibronectin or fibrinogen. Adherence of the parental strain to microtiter plates coated with these matrix proteins yielded absorptions, after staining with safranin, of 0.267 0.027 alone and of 0.217 0.073 with fibronectin; adherence of the DtagO mutant yielded absorptions of 0.215 0.039 alone and of 0.210 0.039 with fibrinogen (all data are means SD). Adherence to HUVECs. Under static conditions, adherence of the DtagO mutant was considerably reduced (by 44.7%), compared with that of the parental strain (P ! .001) (figure 3A). Adherence of the DtagO mutant was reduced to a similar extent in the presence and in the absence of human serum. Omitting human serum led to a moderate reduction of adherence of all 3 strains, indicating that serum factors such as fibronectin may contribute to efficient binding. In a control experiment, a significant reduction in adherence of the DtagO mutant (mean SD, 44.6% 6.6%) was also observed when unlabeled bacteria were used (data not shown), confirming that labeling of bacteria with FITC had no independent influence on adhesion. Figure 3. Reduced adherence of Staphylococcus aureus to human umbilical vein endothelial cells (HUVECs). A, Attachment of the S. aureus parental strain (PS), the DtagO mutant, and the DtagO mutant complemented with pRBtagO to confluently grown HUVECs under static conditions, in the absence or the presence of 1% human serum. B, Attachment of S. aureus strains under flow conditions of 2.5 mL/min and 2 108 bacteria/mL in IMDM for 20 min. The means SD of 45 independent experiments are shown. *P ! .05, vs. PS. **P ! .01, vs. PS. ***P ! .001, vs. PS. Under flow conditions, the absolute number of bacteria that adhered to HUVECs was reduced, compared with that under static conditions, whereas the significant difference in adherence between the parental strain and the DtagO mutant was even more pronounced (70.3% reduced binding of the DtagO mutant, compared with that of the parental strain) (P ! .05). Complementation with plasmid pRBtagO, which bears a wildtype copy of tagO and its promoter region, restored adherence profiles to parental strain levels, confirming that the absence of WTA played a role in this reduced adherence. Adherence of WTA-coated beads to HUVECs. Only beads coated with purified WTA adhered efficiently to HUVECs (figure 4), suggesting a specific interaction between WTA and cognate structures on HUVECs. Beads coated with WTA from the DdltA mutant (lacking d-alanylation of WTA) or with the DtagO mutant interacted poorly with HUVECs, indicating that an intact WTA structure is important for this binding interaction. In Vivo Studies Attenuated virulence of the WTA-deficient DtagO mutant. At 60 min after challenge, the DtagO mutant bound to sterile cardiac vegetations to a substantially lower extent than did the parental strain (mean SD, 4.29 0.37 vs. 5.07 0.77 log10 cfu/g tissue, respectively) (P ! .05). There were no significant differences between clearance of the parental strain and that of the DtagO mutant from the bloodstream during the initial 60min postchallenge period (paralleling our in vitro opsonophagocytosis data). The means SD for the parental strain at 10 min, 30 min, and 60 min after challenge were 2.42 0.21, 2.04 0.49, and 2.10 0.59 log10 cfu/mL blood, respectively, and for the DtagO mutant were 2.29 0.36, 2.02 0.25, and 2.00 0.44 log10 cfu/mL blood, respectively. Table 1 shows the comparative bacterial proliferation within cardiac vegetations and the extent of hematogenous dissemination to target tissues, for the parental strain and the DtagO mutant. Of note, the tagO mutation resulted in attenuated virulence in this model at both intravenous challenge inocula (5 105 and 5 106 cfu). Thus, bacterial densities of the DtagO mutant in cardiac vegetations were significantly reduced (by 2 logs), compared with those of the parental strain. The difference in bacterial proliferation within the spleen and the kidneys was even more pronounced between the DtagO mutant and the parental strain (a reduction of 23 logs for the DtagO mutant). These data indicate the significant impact that the tagO mutation has on endovascular pathogenesis in this model of multisystem endovascular infection. The initial site to which endovascular pathogens adhere are sterile cardiac vegetations. These matrices are a composite of endothelial cells, activated platelets, and matrix adhesinssuch as fibrinogen, fibronectin, and fibrinand are relatively devoid of inflammatory cells [27]. Thus, successful pathogens such as S. aureus possess a broad array of surface adhesins that enable the organism to bind to 1 of these relevant tissues and matrix surfaces [3, 28]. In response to pathogenic strategies, the host counterrespondsfor example, by secreting tPMPs at the site of endovascular bacterial colonization or by recruiting PMNLs and subsequent PMNL-mediated opsonophagocytic clearance of the organism [17]. Among the surface-exposed S. aureus molecules, teichoic acid (TA) polymers have recently attracted increasing interest because of their role in nasal colonization [11], resistance to antimicrobial host peptides [11, 12], and biofilm formation [29]. TAs may be linked either to glycolipids (lipoteiochoic acids; LTAs) or to peptidoglycan (a WTA). These 2 TA polymers have different structures, and their individual roles in bacterial physTable 1. Bacterial counts in tissues of rabbits challenged with Staphylococcus aureus SA113 (parental strain) or the DtagO mutant lacking wall teichoic acid. Bacterial count, log10 cfu/g tissue In cardiac vegetations In the kidneys In the spleen iological processes have remained elusive [30]. Although LTA is considered to be essential for bacterial viability, when WTA is deleted, there is surprisingly little impact on the growth and fitness of S. aureus in vitro [11]. Using a well-characterized and discriminative model of endovascular infections in rabbitsexperimental aortic valve IE and the WTA-deficient DtagO mutant, we studied the impact that WTA has on in vitro phenotypes relevant to the induction and propagation of endovascular infections and on in vivo virulence. Because plasmid pRBtagO used to complement the DtagO mutation is relatively unstable in the absence of antibiotics in vivo ([11] and unpublished data), we were not able to reliably employ this additional strain in our in vivo studies. However, selected in vitro experiments presented in the present study and in a previous study [11] demonstrate that complementation restores all relevant parental strain phenotypes. A number of interesting observations emerged from the present study. In vitro, the parental strain and the DtagO mutant were statistically equivalent in their capacities to bind to fibrinogen, fibronectin, and platelets; in their susceptibility to tPMP; and in their susceptibility to be opsonophagocytosed by PMNLs. However, the DtagO mutant was significantly impaired in its capacity to bind to HUVECs, compared with that of the parental strain. Our data confirmed that this defective endothelial cellbinding phenotype was specifically related to the altered WTA profile of the DtagO mutant. Remarkably, these in vitro findings were paralleled by our in vivo findings in the experimental IE model. Thus, there were no differences between the parental strain and the DtagO mutant in terms of their ability to be removed from the bloodstream after intravenous challenge (an in vivo parallel to in vitro opsonophagocytosis). However, the DtagO mutant was significantly less able than the parental strain to bind to sterile cardiac vegetations to initiate IE. In addition, progression during the later stages of IE pathogenesisincluding proliferation within the cardiac vegetations and hematogenous dissemination to and proliferation within distant target tissues such as the kidneys and the spleenwas significantly reduced in the DtagO mutant, compared with that in the parental strain. Because the composite outcomes of the overall IE process reflect the initial colonization and proliferation in the cardiac vegetations, the spread to distant organs, and the hematogenous reseeding of the vegetations, it is likely that the tagO mutation impacts the infection process at more than a single step [20, 23]. Moreover, it is conceivable that the lack of WTA has further consequences for the fitness of S. aureus that might also have affected the outcome of our in vivo experiments. As noted above, interaction with the endothelium, a key step in the induction of S. aureus endovascular infections, was impaired in the WTA-deficient DtagO mutant, indicating that WTA may play an important role in staphylococcal adhesion to endothelium. S. aureus produces a cadre of proteins that mediate binding to endothelial cells [31]. For example, the intercellular adhesion molecule 1binding extracellular adherence protein [32] and the fibronectin-binding proteins FnbA and FnbB, which mediate contact to endothelial a5b1 integrins via fibronectin [33], have been studied in detail. However, in the present study, the tagO mutation did not significantly alter the in vitro capacity of S. aureus to bind to fibronectin. Moreover, the direct and dose-dependent adherence of beads coated with purified WTA to endothelial cells in vitro suggested that WTA has a specific interaction with unidentified structures on these host cells. Furthermore, the negative impact that the tagO mutation had on binding to endothelial cells in vitro was magnified under flow conditions, paralleling the in vivo circumstance likely encountered by circulating staphylococci at damaged endothelial sites. Ongoing investigations into the molecular basis of these putative WTA-mediated interactions with both endothelial and epithelial cells are in progress in our laboratories. Acknowledgments We thank Volkhard Kempf, for providing the human umbilical vein endothelial cells, and Yin Li, for excellent technical assistance with the rabbit model studies and in vitro binding assays. References ing of 45% 25% and 46% 21% respectively. Moreover, the time to maximal binding was equivalent for the strains (1 5 min). There was also no difference in the ability of the 2 strains to interact with fibronectin or fibrinogen. After staining with safranin, adherence of the parental strain to microtiter plates coated with these matrix proteins yielded absorptions of 0.267 0.027 with fibronectin and 0.215 0.039 with fibrinogen, whereas adherence of the DtagO mutant yielded absorptions of 0.217 0.073 with fibronectin and 0.210 0.039 with fibrinogen (all data are means SD). The publisher regrets this error.


This is a preview of a remote PDF: http://jid.oxfordjournals.org/content/191/10/1771.full.pdf

Christopher Weidenmaier, Andreas Peschel, Yan-Qiong Xiong, Sascha A. Kristian, Klaus Dietz, Michael R. Yeaman, Arnold S. Bayer. Lack of Wall Teichoic Acids in Staphylococcus aureus Leads to Reduced Interactions with Endothelial Cells and to Attenuated Virulence in a Rabbit Model of Endocarditis, Journal of Infectious Diseases, 2005, 1771-1777, DOI: 10.1086/429692