Portal de Periódicos da CAPES

Comparison of Lipoteichoic Acid from Different Serotypes of Streptococcus pneumoniae

2006; Elsevier BV; Volume: 281; Issue: 45 Linguagem: Inglês

10.1074/jbc.m602676200

1083-351X

Christian Draing, Markus Pfitzenmaier, Sebastiana Zummo, Giuseppe Mancuso, Armin Geyer, Thomas Härtung, Sonja von Aulock,

Antibiotics Pharmacokinetics and Efficacy

Pneumococcal lipoteichoic acid (LTA) is known to have a completely different chemical structure compared with that of Staphylococcus aureus: the polyglycerophosphate in the backbone is replaced in the pneumococcal LTA by a pentamer repeating unit consisting of one ribitol and a tetrasaccharide carrying the unusual substituents phosphocholine and N-acetyl-d-galactosamine. Neither d-alanine nor N-acetyl-d-glucosamine, which play central roles in the biological activity of the staphylococcal LTA, has been reported. The extraction using butanol is more gentle compared with the previously reported chloroform-methanol extraction and results in a higher yield of LTA. We characterized the LTA of two different strains of Streptococcus pneumoniae:R6 (serotype 2) and Fp23 (serotype 4). NMR analysis confirmed the structure of LTA from R6 but showed that its ribitol carries an N-acetyl-d-galactosamine substituent. The NMR data for the LTA from Fp23 indicate that this LTA additionally contains ribitol-bound d-alanine. Dose-response curves of the two pneumococcal LTAs in human whole blood revealed that LTA from Fp23 was significantly more potent than LTA from R6 with regard to the induction of all cytokines measured (tumor necrosis factor, interleukin-1 (IL-1), IL-8, IL-10, granulocyte colony-stimulating factor, and interferon γ). However, other characteristics, such as lack of inhibition by endotoxin-specific LAL-F, Toll-like receptor 2 and not 4 dependence, and lack of stimulation of neutrophilic granulocytes, were shared by both LTAs. This is the first report of a difference in the structure of LTA between two pneumococcal serotypes resulting in different immunostimulatory potencies. Pneumococcal lipoteichoic acid (LTA) is known to have a completely different chemical structure compared with that of Staphylococcus aureus: the polyglycerophosphate in the backbone is replaced in the pneumococcal LTA by a pentamer repeating unit consisting of one ribitol and a tetrasaccharide carrying the unusual substituents phosphocholine and N-acetyl-d-galactosamine. Neither d-alanine nor N-acetyl-d-glucosamine, which play central roles in the biological activity of the staphylococcal LTA, has been reported. The extraction using butanol is more gentle compared with the previously reported chloroform-methanol extraction and results in a higher yield of LTA. We characterized the LTA of two different strains of Streptococcus pneumoniae:R6 (serotype 2) and Fp23 (serotype 4). NMR analysis confirmed the structure of LTA from R6 but showed that its ribitol carries an N-acetyl-d-galactosamine substituent. The NMR data for the LTA from Fp23 indicate that this LTA additionally contains ribitol-bound d-alanine. Dose-response curves of the two pneumococcal LTAs in human whole blood revealed that LTA from Fp23 was significantly more potent than LTA from R6 with regard to the induction of all cytokines measured (tumor necrosis factor, interleukin-1 (IL-1), IL-8, IL-10, granulocyte colony-stimulating factor, and interferon γ). However, other characteristics, such as lack of inhibition by endotoxin-specific LAL-F, Toll-like receptor 2 and not 4 dependence, and lack of stimulation of neutrophilic granulocytes, were shared by both LTAs. This is the first report of a difference in the structure of LTA between two pneumococcal serotypes resulting in different immunostimulatory potencies. Streptococcus pneumoniae is one of the most common Gram-positive pathogens that colonizes the upper respiratory tract and causes many severe infections like otitis media, sinusitis, and more life-threatening diseases like pneumonia, bacteremia, and meningitis, when it gains access to the lower respiratory tract or the bloodstream (1Organization W.H. Wkly. Epidemiol. Rec. 1998; 73: 187-188PubMed Google Scholar, 2Zangwill K.M. Vadheim C.M. Vannier A.M. Hemenway L.S. Greenberg D.P. Ward J.I. J. Infect. Dis. 1996; 174: 752-759Crossref PubMed Scopus (172) Google Scholar). In the United States alone, there were in the last 20 years approximately 7 million cases of otitis media each year, 500,000 cases of pneumonia, 50,000 cases of bacteremia, and 3,000 cases of meningitis (3Stool S.E. Field M.J. Pediatr. Infect. Dis. J. 1989; 8: S11-S14Crossref PubMed Scopus (166) Google Scholar, 4Jedrzejas M.J. Microbiol. Mol. Biol. Rev. 2001; 65 (first page, table of contents): 187-207Crossref PubMed Scopus (370) Google Scholar). S. pneumoniae also causes a high mortality rate of 40,000 per year in the United States. When the infection has cleared, patients often retain neurological sequelae like hearing impairment or learning disabilities. Most bacteria are surrounded by a capsule, which makes recognition by the immune system more difficult. On the basis of the differences in composition of the capsular polysaccharides, S. pneumoniae can be divided into >90 serotypes (5Henrichsen J. J. Clin. Microbiol. 1995; 33: 2759-2762Crossref PubMed Google Scholar, 6Trollfors B. Burman L. Dannetun E. Llompart J. Norrby R. J. Clin. Microbiol. 1983; 18: 978-980Crossref PubMed Google Scholar). But, only seven serotypes (4, 6B, 9V, 14, 18C, 19F, and 23F) are responsible for 65% of all cases of pneumococcal disease (7Farrell D.J. Jenkins S.G. Reinert R.R. J. Med. Microbiol. 2004; 53: 1109-1117Crossref PubMed Scopus (22) Google Scholar) and 23 serotypes for 90% (8Bridy-Pappas A.E. Margolis M.B. Center K.J. Isaacman D.J. Pharmacotherapy. 2005; 25: 1193-1212Crossref PubMed Scopus (85) Google Scholar). A prevalent problem with S. pneumoniae infections is the emergence of antibiotic-resistant strains in the last years. Previous studies show an increase in penicillin resistance of S. pneumoniae from 4% to 21% over a 3-year period (2Zangwill K.M. Vadheim C.M. Vannier A.M. Hemenway L.S. Greenberg D.P. Ward J.I. J. Infect. Dis. 1996; 174: 752-759Crossref PubMed Scopus (172) Google Scholar), and there are some isolates that are resistant to vancomycin (9Novak R. Henriques B. Charpentier E. Normark S. Tuomanen E. Nature. 1999; 399: 590-593Crossref PubMed Scopus (309) Google Scholar) and levofloxacin (10Canton R. Morosini M. Enright M.C. Morrissey I. J. Antimicrob. Chemother. 2003; 52: 944-952Crossref PubMed Scopus (115) Google Scholar). A thorough understanding of the immune response to S. pneumoniae might open up new treatment opportunities. Like all Gram-positive bacteria, the cell envelope of S. pneumoniae consists of a cell wall containing several layers of peptidoglycan with bound teichoic acids and lipoteichoic acids (LTA), 3The abbreviations used are: LTA, lipoteichoic acid; LPS, lipopolysaccharide; TLR, Toll-like receptor; HIC, hydrophobic interaction chromatography; LAL, Limulus amoebocyte lysate test; LAL-F, Limulus antilipopolysaccharide factor; GalpNAc, N-acetyl-d-galactosamine;β-Glcp, glucose; AATGalp, 2-acetamido-4-amino-2,4,6-trideoxygalactose; G-CSF, granulocyte colony-stimulating factor; IFNγ, interferon γ. 3The abbreviations used are: LTA, lipoteichoic acid; LPS, lipopolysaccharide; TLR, Toll-like receptor; HIC, hydrophobic interaction chromatography; LAL, Limulus amoebocyte lysate test; LAL-F, Limulus antilipopolysaccharide factor; GalpNAc, N-acetyl-d-galactosamine;β-Glcp, glucose; AATGalp, 2-acetamido-4-amino-2,4,6-trideoxygalactose; G-CSF, granulocyte colony-stimulating factor; IFNγ, interferon γ. which are anchored in the cell membrane. According to previous reports (11Fischer W. Microb. Drug Resist. 1997; 3: 309-325Crossref PubMed Scopus (48) Google Scholar, 12Behr T. Fischer W. Peter-Katalinic J. Egge H. Eur. J. Biochem. 1992; 207: 1063-1075Crossref PubMed Scopus (122) Google Scholar) the pneumococcal lipoteichoic acid has a completely different chemical structure to the well characterized LTA from Staphylococcus aureus. The polyglycerophosphate in the staphylococcal LTA backbone is replaced by a pentamer repeating unit composed of ribitol and a tetrasaccharide, and the phosphate content in the pneumococcal LTA is much lower than that of LTA from S. aureus. The LTA backbone of staphylococcal LTA carries N-acetyl-d-glucosamine and d-alanine, which both play a central role in the biological activity of this LTA. Instead of these substituents, phosphocholine and N-acetyl-d-galactosamine are found in the pneumococcal LTA (12Behr T. Fischer W. Peter-Katalinic J. Egge H. Eur. J. Biochem. 1992; 207: 1063-1075Crossref PubMed Scopus (122) Google Scholar). The pneumococcal LTA stimulates the release of pro-inflammatory cytokines but was reported to be less potent in comparison to LTA from S. aureus (13Han S.H. Kim J.H. Martin M. Michalek S.M. Nahm M.H. Infect. Immun. 2003; 71: 5541-5548Crossref PubMed Scopus (148) Google Scholar). Activation of the monocytes occurs via the Toll-like receptor 2 (TLR-2) with CD14 as co-receptor (14Schroder N.W. Morath S. Alexander C. Hamann L. Hartung T. Zahringer U. Gobel U.B. Weber J.R. Schumann R.R. J. Biol. Chem. 2003; 278: 15587-15594Abstract Full Text Full Text PDF PubMed Scopus (496) Google Scholar). The LTA from S. pneumoniae described initially was isolated by a very complex and time-consuming method (15Briles E.B. Tomasz A. J. Biol. Chem. 1973; 248: 6394-6397Abstract Full Text PDF PubMed Google Scholar, 16Briese T. Hakenbeck R. Eur. J. Biochem. 1985; 146: 417-427Crossref PubMed Scopus (111) Google Scholar). After autolysis of the bacteria and several enzyme treatments, a co-fractionation of radioactive choline label and antigenic activity was used for the preparation of the LTA. This complex isolation procedure and the low yield of LTA were the reasons why structural studies were not performed before 1992. Using a chloroform-methanol extraction, Behr et al. (12Behr T. Fischer W. Peter-Katalinic J. Egge H. Eur. J. Biochem. 1992; 207: 1063-1075Crossref PubMed Scopus (122) Google Scholar) developed a simpler and more effective isolation procedure resulting in the establishment of the structure of pneumococcal LTA from strain R6 (12Behr T. Fischer W. Peter-Katalinic J. Egge H. Eur. J. Biochem. 1992; 207: 1063-1075Crossref PubMed Scopus (122) Google Scholar). 5 years ago a new isolation procedure for the LTA from S. aureus was published using butanol for the extraction instead of hot phenol or chloroform-methanol (17Morath S. Geyer A. Hartung T. J. Exp. Med. 2001; 193: 393-397Crossref PubMed Scopus (373) Google Scholar). It was shown that the previously used methods resulted in a decomposition of the LTA characterized by the loss of glycerophosphate units as well as d-alanine and N-acetyl-d-glucosamine substituents (18Morath S. Geyer A. Spreitzer I. Hermann C. Hartung T. Infect. Immun. 2002; 70: 938-944Crossref PubMed Scopus (137) Google Scholar). The components of the LTA backbone seemed to play a very important role in the biological activity of the LTA as indicated by the activity of synthetic derivatives of the LTA from S. aureus, which induced cytokine release in human monocytes that depended on the stereoisomer of d-alanine substituents (19Deininger S. Stadelmaier A. von Aulock S. Morath S. Schmidt R.R. Hartung T. J. Immunol. 2003; 170: 4134-4138Crossref PubMed Scopus (82) Google Scholar). Since then, LTA from further bacteria like Bacillus subtilis, Streptococcus pyogenes, Streptococcus agalactiae, and Lactobacillus plantarum (18Morath S. Geyer A. Spreitzer I. Hermann C. Hartung T. Infect. Immun. 2002; 70: 938-944Crossref PubMed Scopus (137) Google Scholar, 20Henneke P. Morath S. Uematsu S. Weichert S. Pfitzenmaier M. Takeuchi O. Muller A. Poyart C. Akira S. Berner R. Teti G. Geyer A. Hartung T. Trieu-Cuot P. Kasper D.L. Golenbock D.T. J. Immunol. 2005; 174: 6449-6455Crossref PubMed Scopus (101) Google Scholar, 21Grangette C. Nutten S. Palumbo E. Morath S. Hermann C. Dewulf J. Pot B. Hartung T. Hols P. Mercenier A. Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 10321-10326Crossref PubMed Scopus (341) Google Scholar) have been isolated by this method. All of these have a similar molecular structure and immunostimulatory potency. In the present study we asked whether the butanol-extracted LTA from S. pneumoniae shows differences in its chemical structure in comparison to the reported chloroform-methanol-extracted LTA, possibly resulting in differences in the immunostimulatory potency of these pneumococcal LTA. For the biological characterization of the LTA we focused our interest on the cytokine induction by human whole blood, the activation of neutrophilic granulocytes, and the TLR dependence. We chose two different strains of S. pneumoniae, the well known laboratory R6 strain (serotype 2), which has been used in most studies since the 1950s, and the Fp23 strain (serotype 4), which was isolated in 1987 and whose complete genome sequence is published (22Tettelin H. Nelson K.E. Paulsen I.T. Eisen J.A. Read T.D. Peterson S. Heidelberg J. DeBoy R.T. Haft D.H. Dodson R.J. Durkin A.S. Gwinn M. Kolonay J.F. Nelson W.C. Peterson J.D. Umayam L.A. White O. Salzberg S.L. Lewis M.R. Radune D. Holtzapple E. Khouri H. Wolf A.M. Utterback T.R. Hansen C.L. McDonald L.A. Feldblyum T.V. Angiuoli S. Dickinson T. Hickey E.K. Holt I.E. Loftus B.J. Yang F. Smith H.O. Venter J.C. Dougherty B.A. Morrison D.A. Hollingshead S.K. Fraser C.M. Science. 2001; 293: 498-506Crossref PubMed Scopus (1110) Google Scholar). The isolation and characterization of the LTA from these strains might help to elucidate the interaction between the pathogen and the immune system, giving a better insight into the pathogenesis of S. pneumoniae. Bacterial Strains and Cultivation—S. pneumoniae strain R6 (serotype 2), kindly provided by E. Tuomanen, St. Jude's Children's Research Hospital, Memphis, TN, and S. pneumoniae strain Fp23 (serotype 4) provided by M. R. Oggioni, Dipartimento di Biologia Molecolare, Università di Siena, Siena, Italy, were grown to the late logarithmic phase (A580 of 0.9) in Todd-Hewitt broth (Oxoid, Milan, Italy). Bacteria were washed extensively with distilled water and lyophilized. S. aureus (DSM 20233) was cultured aerobically and harvested after 18 h (stirring at 37 °C, 150 rpm) at an A578 of 3 by centrifugation at 4225 × g for 20 min. Integrity of bacteria and potential contaminations by Gram-negative bacterial species were checked by Gram staining and microscopy. Until the extraction, the bacteria were frozen at –20 °C. LTA Purification—The bacteria underwent butanol extraction and hydrophobic interaction chromatography as described (17Morath S. Geyer A. Hartung T. J. Exp. Med. 2001; 193: 393-397Crossref PubMed Scopus (373) Google Scholar). Briefly, after resuspension in 0.05 m citrate buffer at pH 4.7, bacteria were disrupted by sonification for 15 min. The bacterial lysate (30 ml) was mixed with an equal volume of n-butanol (Merck, Darmstadt, Germany) under stirring for 20 min at room temperature. Centrifugation at 17,200 × g for 40 min resulted in a two-phase system, and the lower, aqueous phase was collected before the addition of fresh citrate buffer for a second extraction. This re-extraction was performed twice, and the three aqueous phases were pooled and lyophilized. After resuspension of the sample in 35 ml of chromatography start buffer (15% n-propanol in 0.1 m ammonium acetate, pH 4.7), it was centrifuged at 26,900 × g for 60 min and filtered (0.2 μm). The supernatant was subjected to hydrophobic interaction chromatography (HIC) on an octyl-Sepharose column (2.5 × 11 cm) using a linear gradient from 15% to 60% n-propanol in 0.1 m ammonium acetate (pH 4.7). LTA Dealanylation—7.9 mg of LTA Fp23 in Tris buffer (1.25 m in MilliQ, pH 8.5) was stirred at 500 rpm for 24 h at room temperature. After lyophilization the sample was purified by HIC as described above. The absence of d-alanine in the LTA Fp23 was confirmed by NMR analysis. Phosphate Determination—This method is based on the formation of phosphomolybdenum blue from phosphate. 50 μl of a phosphate standard (0.65 mm, Sigma), and 100 μl of each fraction was mixed with 200 μl of ashing solution (2 m H2SO4 and 0.44 m HClO4) and incubated in open polypropylene vials (Eppendorf, Hamburg, Germany) at 145 °C for 2.5 h. Then 1 ml of reducing solution was added containing 3 mm ammonium molybdate, 0.25 m sodium acetate, and 1% ascorbic acid. After 2 h at 45 °C, 250 μl of each sample was transferred to flat-bottom, ultrasorbant 96-well plates (Nunc, Wiesbaden, Germany), and the absorption was measured in an ELISA Reader (Rainbow, Tecan, Crailsheim, Germany) at 700 nm. The LTA containing fractions were pooled, and the endotoxin contamination of the LTAs was assessed by the kinetic Limulus amoebocyte lysate assay (Charles River, Charleston, SC). Whole Blood Incubation—Human whole blood incubations were performed as described previously (23Hartung T. Wendel A. Altex. 1995; 12: 70-75PubMed Google Scholar). Briefly, human blood was drawn from healthy volunteers into heparinized S-monovettes® (Sarstedt, Nümbrecht, Germany) and diluted 5-fold in RPMI 1640 medium (Biochrom, Berlin, Germany). The following stimuli and inhibitors were used: 50 μl of each chromatography fraction, which had been evaporated and resuspended in 50 μl of 0.9% NaCl, LPS from Salmonella abortus equi (Sigma), LTA from S. aureus (LTA Sa), LTA from Streptococcus pneumoniae strain R6 (LTA R6) and strain Fp23 (LTA Fp23), which all were isolated in-house by n-butanol extraction, Zymosan A from Saccharomyces cerevisiae (Sigma), and polymyxin B (Sigma) and Limulus antilipopolysaccharide factor LAL-F (a generous gift from F. Jordan, Charles River). The final volume was adjusted to 500 μl. Incubations were carried out in open polypropylene vials overnight for 22 h at 37 °C and 5% CO2. The pelleted blood cells were then resuspended by gentle shaking and were centrifuged at 400 × g for 2 min. The cell-free supernatants were stored at –80 °C until cytokine measurement by ELISA. Isolation of Human Neutrophilic Granulocytes and Determination of the Myeloperoxidase Activity—Neutrophils were obtained with a Percoll gradient (BD Biosciences) as described (24von Aulock S. Morath S. Hareng L. Knapp S. van Kessel K.P. van Strijp J.A. Hartung T. Immunobiology. 2003; 208: 413-422Crossref PubMed Scopus (60) Google Scholar). After the centrifugation of heparinized blood at 270 × g for 20 min, the resulting buffy coat was mixed with 0.72% dextran T500 (Sigma) in 50 ml of phosphate-buffered saline. 30 min later the erythrocytes had sedimented. The supernatant contained leukocytes, which were centrifuged at 850 × g for 7 min and washed once with HEPES buffer. The cells were transferred to a discontinuous Percoll/HEPES buffer gradient (density 1.093, 1.088, 1.072, and 1.059) and spun for 15 min at 450 × g. The neutrophil band was transferred to a new tube and was washed once with HEPES buffer at 850 × g for 7 min. Afterward differential cell counts were determined with a Pentra60 (ABX Diagnostics, Montpellier, France), and the cells were diluted to 5 × 106 cells/ml with RPMI 1640 medium (Biochrom) and 10% autologous serum. Finally the neutrophilic granulocytes were plated to 96-well culture plates (5 × 105 cells/well, Greiner, Nürtingen, Germany) and were incubated with different stimuli for 22 h at 37 °C and 5% CO2. 50 μl of a potassium buffer (50 mm, pH 6.0) containing EDTA (10 mm) and 0.5% (w/v) hexadexylammonium bromide (both Sigma) were added to the supernatants of the incubations. Myeloperoxidase from human leukocytes (Sigma) served as the standard. After the addition of 3,3′,5,5′-tetramethylbenzidine (Sigma), the activity of the myeloperoxidase could be determined. The reaction was stopped by the addition of H2SO4, and the absorption was measured at 450 nm. Isolation of Murine Bone Marrow Cells—C3H/HeJ mice, characterized by a non-functional TLR4, and the corresponding wild-type mouse strain C3H/HeN were purchased from Charles River Laboratories. TLR2-deficient mice kindly provided by Tularik (South San Francisco, CA) and the corresponding wild-type mice (129Sv/B57BL/6) were bred in the animal facilities of the University of Konstanz and genotyped. Mice were killed by terminal pentobarbital anesthesia (Narcoren, Merial, Halbergmoos, Germany). The humeri and femurs of the mice were lavaged with 10 ml of ice-cold sterile phosphate-buffered saline (Invitrogen). The lavages were transferred to siliconized glass tubes (Vacutainer, Bioscience, Heidelberg, Germany), and bone debris was removed. After one centrifugation step the cell counts were determined. The cells were diluted to 5 × 106/ml with RPMI 1640 supplemented with 10% heat-inactivated fetal calf serum (Biochrom), 100 IU/ml penicillin/streptomycin (both from Biochrom), and plated to 96-well culture plates for stimulation with different stimuli. After 22 h at 37 °C, 5% CO2 supernatants were frozen at –80 °C until cytokine measurement by ELISA. Cytokine Measurement—Cytokines released by human whole blood were measured by in-house sandwich ELISA using commercially available antibody pairs and recombinant standards. Monoclonal antibody pairs against human TNF, IL-8, and IFNγ were purchased from Endogen (Perbio Science, Bonn, Germany), against human IL-1β and G-CSF from R&D (Wiesbaden, Germany), and against human IL-10 from BD Biosciences. Recombinant standards for TNF and IL-1β were kind gifts from S. Poole (National Institute for Biological Standards and Control, Herts, UK), rIL-8 from PeproTech (Tebu, Frankfurt, Germany), rIFNγ from Thomae (Biberach, Germany), rG-CSF from Amgen (Thousand Oaks, CA), and rIL-10 from BD Biosciences. The release of murine TNF by bone marrow cells was measured with the DuoSet-kit from R&D. Assays were carried out in flat-bottom, ultrasorbant 96-well plates. The secondary biotinylated antibodies were detected by horseradish-peroxidase-conjugated streptavidin (BIOSOURCE, Camarillo, CA), and 3,3′,5,5′-tetramethylbenzidine (Sigma) was used as substrate. The reaction was stopped with 1 m H2SO4, and the absorption was measured in an ELISA reader at 450 nm with a reference wavelength of 690 nm. NMR Spectroscopy—NMR spectra were obtained on a Bruker Avance 600 spectrometer at 300 K using sample tubes with a 5-mm outer diameter. Spectra were measured for solutions in D2O using sodium 3-trimethylsilyl-3,3,2,2-tetradeuteropropanoate as an internal standard for 1H NMR (δH 0.00 ppm) and acetone for 13C NMR (δC 30.02 ppm). For 31P NMR spectra 2% phosphoric acid was used as external standard (δP = 0.00 ppm). Two-dimensional homonuclear DFQ-COSY, TOCSY, NOESY, and ROESY experiments and two-dimensional heteronuclear HMQC and HSQC (1H-13C) were performed using standard Bruker pulse programs. TOCSY, NOESY, and ROESY experiments were performed in the phase-sensitive mode using mixing times of 100 ms in the TOCSY and 200-ms spinlock for NOESY and ROESY. Statistics—Statistical analysis was performed using the GraphPad Prism program (GraphPad Software, San Diego, CA). Data are shown as means ± S.E. For statistical analysis of two groups of non-parametric data, Wilcoxon matched pairs test (human data) and unpaired Student's t test (mouse data) were used. Repeated-measure analysis of variance was assessed using the one-way analysis of variance test followed by the Bonferroni multiple comparison test (Figs. 3 and 13). A p value <0.05 was considered significant. In the figures *, **, and *** represent p values <0.05, <0.01, and <0.001, respectively. Cytokine levels are given per milliliter blood, i.e. corrected for the dilution factor 5 in the 20% blood incubation.FIGURE 13Comparison of cytokine induction by LTA from S. pneumoniae R6, Fp23, and dealanylated LTA Fp23. Human whole blood from eight healthy volunteers was stimulated with the given concentrations of either LTA for 22 h. TNF release was measured in the cell-free supernatants by ELISA. Data are means ± S.E. *, p < 0.05; **, p < 0.01; and ***, p < 0.001 (repeated-measure analysis of variance with Bonferroni multiple comparison test).View Large Image Figure ViewerDownload Hi-res image Download (PPT) LTA Purification by Hydrophobic Interaction Chromatography—After the butanol extraction of S. pneumoniae the aqueous phase was subjected to hydrophobic interaction chromatography. The resulting fractions were screened for phosphate content. For both strains of S. pneumoniae we obtained high phosphate contents in the flow-through fractions (fractions 1–19), representing DNA and proteins that had not bound to the chromatography column, and a second peak from fraction 43 to 50, corresponding to the fractions that yield LTA upon extraction of S. aureus (Fig. 1, A and B). We screened the fractions for IL-8 induction in human whole blood. The resulting profile of the IL-8-induction by the fractions from strain R6 (Fig. 1A) was congruent with that of the Fp23 strain fractions as shown in Fig. 1B, whereas in both cases the peaks were broader than the phosphate peaks (pooled fractions: 30–57) as expected due to the higher sensitivity of the human whole blood assay compared with the phosphate measurement. The screening profiles shown are representative for two extractions of the R6 strain and nine extractions of the Fp23 strain. Exclusion of LPS Contamination in LTA Fp23 and LTA R6 Preparations—To exclude an LPS contamination we used different inhibitors and the Limulus amoebocyte lysate (LAL) test. Polymyxin B is an inhibitor for negatively charged molecules like LPS. LAL-F inhibits endotoxin-induced signals more specifically than polymyxin B by binding to the Lipid A portion (25Kellogg T.A. Lazaron V. Wasiluk K.R. Dunn D.L. Shock. 2001; 15: 124-129Crossref PubMed Scopus (11) Google Scholar). As expected, LPS-induced IL-8 release was reduced to 25 and 2% by Polymyxin B and LAL-F, respectively, showing the efficiency of the inhibitors. The cytokine release induced by LTA S. aureus, LTA Fp23, and LTA R6, respectively, were influenced neither by polymyxin B nor by LAL-F (Fig. 2). The results from these experiments were confirmed by the LAL showing for all LTA extractions an LPS contamination below 2 endotoxin units/mg of LTA, i.e. <200 pg LPS per mg. Neither LTA Fp23 nor LTA R6 Activate the Release of Myeloperoxidase by Neutrophilic Granulocytes—Knowing that LTA from S. aureus does not induce myeloperoxidase release from neutrophilic granulocytes, we investigated whether this also holds true for pneumococcal LTA. Zymosan, a well known activator of myeloperoxidase release, was used as a positive control. LPS alone was able to stimulate the neutrophilic granulocytes. Myeloperoxidase release was not induced by LTA, and LTA had no effect on the zymosan spike (Fig. 3). Pneumococcal LTA Is TLR2- and Not TLR4-dependent—To test the TLR dependence of the pneumococcal LTAs, we stimulated bone marrow cells from TLR2+/+ and TLR2–/– mice (Fig. 4A). As expected 1 ng/ml LPS induced similar amounts of TNF in cells from wild-type and TLR2–/– mice. However, 10 μg/ml S. aureus LTA, a known TLR2 agonist, exclusively induced measurable TNF amounts in cells from TLR2 wild-type mice. The stimulation of TLR2+/+ wild-type cells with LTA Fp23 and LTA R6 (each 10 μg/ml) resulted in comparable TNF release, whereas the cells from TLR2–/– mice did not respond to these stimuli, showing TLR2 dependence of the pneumococcal LTA. The results of the stimulation of TLR4 wild-type cells and cells with a non-functional TLR4 receptor (Fig. 4B) showed no difference in the cytokine release of both cell types after stimulation with LTA Fp23 and LTA R6. LTA Fp23 Is as Potent as LTA from S. aureus in Inducing Different Cytokines—We compared the cytokine release induced by LTA from S. aureus and LTA Fp23. Human whole blood was stimulated with increasing concentrations of either LTA (from 10 ng/ml to 10 μg/ml), and the cytokine release was measured by ELISA. Comparing the TNF release, the LTAs were equipotent at each concentration (Fig. 5A). From the same samples we measured further cytokines and chemokines. With 10 μg/ml there was a high IL-8 (300 ng/ml) and IL-1 induction (10 ng/ml) by both LTAs. The G-CSF release (2000 pg/ml) was similar to the TNF induction (1600 pg/ml), and there was only a very low IFNγ (600 pg/ml) and IL-10 (200 pg/ml) release (data not shown). The induced cytokine release by the LTA from S. aureus at 1 μg/ml was compared with the corresponding cytokine induction by LTA Fp23 at 1 μg/ml. Fig. 5B shows that LTA Fp23 induced all measured cytokines and chemokines at nearly the same level or even at significantly higher levels compared with LTA from S. aureus. LTA R6 Is a Weaker Cytokine Inducer than LTA Fp23—Next we compared the LTA from the two pneumococcal strains. Both LTAs were able to stimulate TNF release, but LTA Fp23 showed a significantly higher TNF induction compared with LTA R6 (Fig. 6A). We observed the same lower activity of LTA R6 on the level of further cytokines. At the concentration of 1 μg/ml the LTA R6 induced only 50% of the G-CSF and IL-8 release compared with LTA Fp23. For TNF, IL-1, IL-10, and IFNγ the cytokine release by LTA R6 was reduced to 20–30% of that of LTA Fp23 (Fig. 6B). NMR Analysis of LTA R6 Indicates N-Acetyl-d-galactosamine Substituent at the Ribitol—Fischer et al. characterized LTA structures isolated from different Gram-positive bacteria, including S. pneumoniae strain R6, by chemical degradation methods (12Behr T. Fischer W. Peter-Katalinic J. Egge H. Eur. J. Biochem. 1992; 207: 1063-1075Crossref PubMed Scopus (122) Google Scholar, 26Klein R.A. Hartmann R. Egge H. Behr T. Fischer W. Carbohydr. Res. 1994; 256: 189-222Crossref PubMed Scopus (17) Google Scholar, 27Klein R.A. Hartmann R. Egge H. Behr T. Fischer W. Carbohydr. Res. 1996; 281: 79-98Crossref PubMed Scopus (22) Google Scholar, 28Fischer W. Behr T. Hartmann R. Peter-Katalinic J. Egge H. Eur. J. Biochem. 1993; 215: 851-857Crossref PubMed Scopus (168) Google Scholar). Yet, ch