|
|
||||||||
1Defence Science & Technology Laboratories (Dstl), Biomedical Sciences, Porton Down, Salisbury, Wiltshire SP4 0JQ, UK 2Health Protection Agency (HPA), Porton Down, Salisbury, Wiltshire SP4 0JG, UK
Correspondence Richard Thomas rjthomas{at}dstl.gov.uk
Received February 25, 2004
Accepted April 29, 2004
Intervention in bacterial adhesion to host cells is a novel method of overcoming current problems associated with antibiotic resistance. Antibiotic-resistant strains of bacteria that cause respiratory tract infections are a problem in hospitals and could be used in bioterrorist attacks. A range of bacterial species was demonstrated to attach to an alveolar epithelial (A549) cell line. In all cases, cell surface oligosaccharides were important in attachment, demonstrated by reduced adhesion when A549 cells were pre-treated with tunicamycin. Bacillus anthracis and Yersinia pestis displayed a restricted tropism for oligosaccharides compared to the environmental, opportunistic pathogens, Pseudomonas aeruginosa, Burkholderia cenocepacia, Burkholderia pseudomallei and Legionella pneumophila. The compound with the greatest anti-adhesion activity was p-nitrophenol. Other generic attachment inhibitors included the polymeric saccharides (dextran and heparin), GalNAcß1-4Gal, GalNAcß1-3Gal, Galß1-4GlcNAc and Galß1-3GlcNAc. Burkholderia pseudomallei attachment was particularly susceptible to oligosaccharide inhibition. Combinations of such compounds may serve as a novel generic therapeutics for respiratory tract infections.
| INTRODUCTION |
|---|
|
|
|---|
Attachment to cell surfaces is one of the key steps in bacterial pathogenesis, often involving oligosaccharides located on the host cell surface and bacterial adhesins (Karlsson et al., 1992; Ofek & Sharon, 1990). The terminal di- or trisaccharide units of these oligosaccharides may be used to inhibit these interactions, preventing attachment and hence disease (Zopf & Roth, 1996). The theory has been proven in vitro with a range of bacterial species. For example, lacto-N-neotetraose and asialoganglioside-GM1 (asialo-GM1) inhibit attachment of S. pneumoniae to alveolar epithelial cells (Tong et al., 1999). Similarly 3-sialyllactose inhibits Helicobacter pylori attachment to gastric epithelial cells (Simon et al., 1997). Asialoganglioside-GM1 and -GM2 are used as receptors by a number of respiratory pathogens, which specifically bind their terminal GalNAcß1-3Gal and GalNAcß1-4Gal moieties (Krivan et al., 1988a, b, 1991). The principle has also been proven to be valid in vivo. The symptoms of pneumococcal pneumonia were alleviated in a rabbit model by administration of lacto-N-neotetraose or its
2-3- and
2-6-sialylated derivatives (Idänpään-Heikkilä et al., 1997). Aerosolized dextran significantly reduced murine pneumonia caused by P. aeruginosa (Bryan et al., 1999). Mannose and globotetraose (Gal
1-4Gal terminal moiety) inhibited Escherichia coli urinary tract infection in BALB/c mice (Edén et al., 1982). This study aimed to determine the potential of oligosaccharides to inhibit the attachment to alveolar epithelial cells of a range of pathogens that can cause pneumonic disease.
| METHODS |
|---|
|
|
|---|
Organisms and growth conditions.
Bacillus cereus ATCC 52522, Yersinia enterocolitica ATCC 35669 and Yersinia pseudotuberculosis ATCC 29833T were obtained from the American Type Culture Collection. P. aeruginosa PAK, Legionella pneumophila CORBY and E. coli O127 : H6 were kind donations from Dr J. Boyd (University of Calgary, Canada), Dr G. Blum-Oehler (Universität Würzburg, Germany) and Dr H. Smith (PHLS, Colindale, UK), respectively. Burkholderia cenocepacia 9091, Burkholderia pseudomallei K96243, Yersinia pestis GB, Salmonella typhimurium LT2 and Bacillus anthracis UM23CL2 (pXO pXO2) were from the site culture collection. The bacteria were routinely cultured on agar plates prior to broth inoculation. P. aeruginosa, Burkholderia cenocepacia, Burkholderia pseudomallei, Bacillus cereus and Bacillus anthracis were grown on LuriaBertani (LB) agar plates at 37 °C for 24 h. L. pneumophila was cultured on buffered charcoal yeast extract (BCYE) medium supplemented with 10 % (v/v) Legionella BCYE supplement (Oxoid) for 72 h at 37 °C in an atmosphere of 5 % (v/v) CO2. Y. pestis was cultured on blood agar base (BAB) at 28 °C for 48 h. To obtain cell densities of 108 c.f.u. ml1 for the adhesion studies, single colonies were inoculated into broth cultures as described. P. aeruginosa, Burkholderia cenocepacia, Burkholderia pseudomallei, Bacillus cereus, Bacillus anthracis, E. coli, S. typhimurium, Y. enterocolitica and Y. pseudotuberculosis were grown in LB broth for 16 h at 37 °C. L. pneumophila was grown for 24 h at 37 °C in nutrient broth supplemented with 10 % (v/v) Legionella BCYE supplement in an atmosphere of 5 % (v/v) CO2. Y. pestis was grown in BAB broth at 28 °C for 16 h. All broth cultures were incubated statically. Stocks were stored at 80 °C in the respective base broth supplemented with 10 % (v/v) glycerol.
Culture of alveolar type II pneumocytes.
Adherent A549 cells were cultured in tissue culture flasks containing Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10 % (v/v) fetal calf serum and 2 mM glutamine. Cells were incubated at 37 °C in an atmosphere of 5 % (v/v) CO2 and subcultured 1 : 3 every 34 days. For the purpose of adhesion assays, the cells were rinsed twice with PBS before incubation in the presence of trypsin/0.05 % (w/v) EDTA. After removal of the trypsin solution, the cells were resuspended in fresh medium, seeded in 96-well microtitre plates to a density of 50 000 cells per well and incubated overnight as described previously to produce confluent monolayers ready for the adhesion assay. In experiments that required reduced expression of surface oligosaccharides the A549 cells were cultured in the presence of 0.55 µg tunicamycin ml1 for 96 h prior to completion of the adhesion assay.
Adhesion assay.
Confluent monolayers in 96-well microtitre plates were washed with PBS. Non-specific binding was blocked by incubation for 1 h at 37 °C with 0.5 % (w/v) BSA before rinsing twice with PBS. Bacteria at a density of 107 c.f.u. ml1 were incubated for 1 h at room temperature in a 1 : 1 ratio with various saccharide concentrations (Table 1). The relevant concentrations were prepared in tissue culture medium. Controls were always performed with no saccharide. A 100 µl volume of the bacteria-saccharide solution was added to the A549 cells and incubated for 1 h at 37 °C. Non-adherent bacteria were removed by rinsing five times with PBS. Cells were lysed by incubation for 30 min at 37 °C with a 0.1 % (v/v) Triton X-100 solution. Serial dilutions were performed in PBS and 100 µl volumes were plated in duplicate on to the organism-specific plates and incubated accordingly. The percentage inhibition was calculated by the equation: (controltest/control) x 100 %.
|
Cytotoxicity assay.
The CellTitre 96 non-radioactive cell proliferation assay (Promega) was used for all cytotoxicity assays in accordance with the manufacturer's instructions.
Statistical analysis.
All experiments were performed as three or five independent assays. Assays were performed in triplicate per experimental run. Results are expressed as the mean value and standard deviations (SD) are provided where appropriate. P-values were calculated using paired t-tests.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
|
It is interesting that the enteric yersiniae were much less adherent to the alveolar cell line than Y. pestis. This implies the expression of adhesins specific to Y. pestis that may mediate pneumonic disease. Adhesins have been identified in Y. pestis. The pH 6 antigen is expressed at pH 6 and 37 °C, forming fibrils that bind asialo-GM1, asialo-GM2 and lactosylceramide (Lindler & Tall, 1993; Payne et al., 1998). The plasminogen activator is an outer membrane serine protease that has also been shown to bind laminin, heparin and globoseries glycolipids that contain terminal GalNAcß1-3Gal moieties (Kienle et al., 1992; Lähteenmäki et al., 1998). Recently, the Y. pestis genome has been shown to possess two homologues to the Erwinia HecA adhesin (Rojas et al., 2002). It is unknown whether the homologues are expressed and functional. Other possible adhesins include the capsule (F1 antigen) and lipopolysaccharides (LPS) (Straley, 1993).
Oligosaccharides on the surface of A549 cells mediate attachment
Bacteria often attach to the oligosaccharide moiety of glycoproteins and glycolipids located in the host cell membrane (Karlsson et al., 1992; Ofek & Sharon, 1990). Tunicamycin is a specific inhibitor of glycosylation in eukaryotic cells, and can be used to determine the effects of glycosylation on bacterial adhesion (Cundell & Tuomanen, 1994). Tunicamycin inhibited the attachment of each bacterial species examined to A549 cells to varying degrees: Bacillus anthracis, 37.9 ± 5.7 %; Burkholderia cenocepacia, 88.4 ± 6.2 %; Burkholderia pseudomallei, 66.6 ± 1.2 %; L. pneumophila, 83.5 ± 5.1 %; P. aeruginosa, 45.9 ± 9.7 % and Y. pestis, 64.9 ± 4.0 %. The evidence indicates that oligosaccharides located on the A549 cell surface are involved in the attachment of these pathogens. The finding that tunicamycin treatment only inhibits P. aeruginosa attachment by
46 % is surprising, given the range of adhesins it expresses and saccharide structures it can bind (Tables 1 and 3; Bargouthi et al., 1996; Bryan et al., 1999; Krivan et al., 1988b; Ramphal et al., 1991). Tunicamycin inhibits the biosynthesis of high mannose and complex-type glycoproteins (Cundell & Tuomanen, 1994), however, low level glycosylation could still occur, resulting in attachment of P. aeruginosa to saccharides unaffected by tunicamycin treatment. The low value obtained for Bacillus anthracis is consistent with the evidence that few saccharide structures inhibited attachment to a large degree. Additionally, tunicamycin tends to inhibit the synthesis of high mannose oligosaccharides, and mannose was a poor inhibitor of Bacillus anthracis in comparison to the other bacteria tested (Table 2).
|
|
Oligosaccharide inhibition of respiratory pathogen binding
A panel of oligosaccharides was screened for their ability to inhibit the attachment of six bacterial species associated with respiratory infection (Bacillus anthracis, Burkholderia cenocepacia, Burkholderia pseudomallei, L. pneumophila, P. aeruginosa and Y. pestis; Tables 1 and 2). Incubation with oligosaccharides did not adversely affect A549 cells as indicated by the absence of any cytotoxic effects and microscopic examination (data not shown). The range of saccharide structures that inhibited the attachment of certain bacterial species was surprising and probably reflected the fact that environmental opportunistic pathogens (P. aeruginosa, Burkholderia cenocepacia, Burkholderia pseudomallei and L. pneumophila) would require a range of adhesins to allow attachment to environmental surfaces. In contrast, the more specialized pathogens (Bacillus anthracis and Y. pestis) bind a narrower range of saccharides due to the reduced requirement for attachment to environmental substrata.
Large polymeric saccharides such as dextran, dextran sulphate and heparin were particularly adept at preventing adhesion of P. aeruginosa, Burkholderia cenocepacia, Burkholderia pseudomallei or L. pneumophila. This has been documented previously with dextran in P. aeruginosa (Bargouthi et al., 1996; Bryan et al., 1999) and Burkholderia cenocepacia (Chiu et al., 2001). Heparin has been shown to reduce P. aeruginosa colonization of contact lenses (Duran et al., 1993) and also to inhibit the adhesion of bacteria to epithelial cells (Plotkowski et al., 2001) and basement membrane collagen IV (Tsang et al., 2003). The polymeric saccharides are less effective at inhibiting adhesion of Bacillus anthracis and Y. pestis with the exception of dextran sulphate in Bacillus anthracis. It is likely that Bacillus anthracis attachment is inhibited more effectively by dextran sulphate than dextran because of the negative charge of this polysaccharide. That heparin is both negatively charged and a poor inhibitor may indicate that saccharide structure is also important for interaction with Bacillus anthracis adhesins.
The sulphated polymeric saccharides, heparin and dextran sulphate may have therapeutic benefits other than inhibition of bacterial attachment. Pathogenic yersiniae utilize a type III secretion system (Ysc) to deliver toxins and enzymes (Yops) into host cells. Exogenous heparin and dextran sulphate interact with LcrG from Y. enterocolitica and decrease the level of YopE translocation into host cells (Boyd et al., 1998). Similar interactions may occur in Y. pestis. Nebulized heparin has been shown to thin the mucus layer in Burkholderia-cenocepacia-colonized cystic fibrosis patients due to electrostatic interactions. In addition, heparin had an anti-inflammatory effect, reducing sputum and sera levels of IL6 and IL8 (Ledson et al., 2001). It has been reported that bacterial attachment can be mediated by hydrophobic interactions that can be interfered with by hydrophobic aromatic compounds such as p-nitrophenol (Falkowski et al., 1986). In this study, hydrophobic interactions were important for attachment because all of the bacteria, with the exception of Bacillus anthracis, were strongly inhibited by p-nitrophenol.
The sialylated saccharides, 3-sialyllactose and 6-sialyllactose, were good inhibitors of Burkholderia cenocepacia, L. pneumophila and P. aeruginosa attachment (Table 1). In contrast, mannosylated saccharides, with the exception of mannose and Man
1-2Man, were less effective inhibitors. The fucosylated saccharides, 2-fucosyllactose and 3-fucosyllactose only inhibited Burkholderia cenocepacia attachment and this was dependent on the Galß1-4Glc region rather than just the terminal Fuc
1-2Gal moiety. The effectiveness of digalactosyl saccharides (Gal
1-3Gal, Gal
1-4Gal and Galß1-4Gal) was linkage-dependent. Burkholderia cenocepacia and L. pneumophila preferred the
-linkage over the ß-linkage. In contrast, P. aeruginosa attachment was not inhibited by any digalactosyl saccharides. It has been demonstrated previously that the Gal
1-4Gal-containing glycolipids, digalactosylceramide and globotriosylceramide, can bind Burkholderia cenocepacia (Sylvester et al., 1996). Other common trends were observed. Disaccharides containing Gal linked to either GalNAc or GlcNAc by a ß1-3 or ß1-4 linkage were more inhibitory than if present with the corresponding ß1-6 linkage. This was particularly the case for P. aeruginosa. The observation that GalNAcß1-4Gal was amongst the best anti-adhesion compounds for each bacterium is interesting. This disaccharide is a component of the gangliosides, asialo-GM1 and asialo-GM2, demonstrated previously to bind a range of pathogens that cause respiratory disease (Table 3). For most of these bacteria GalNAcß1-4Gal has been determined as the minimum structure required for binding. Therefore, with the exception of Mycoplasma pneumoniae (Krivan et al., 1989), Mycobacterium tuberculosis, Coxiella burnetti and Chlamydia psittaci, all of the major causes of bacterial respiratory infection in humans have been demonstrated to bind the GalNAcß1-4Gal moiety in asialo-GM1. This raises the possibility of the creation of anti-adhesion compounds that can inhibit a range of bacterial pathogens and warrants further investigation.
Oligosaccharide specificity in relation to bacterial adhesins
The particular adhesins involved in the interactions with these saccharides are largely unknown. This highlights a significant gap in the understanding of the interaction of these bacteria with host cells. However, the literature contains details of various adhesins that may be involved in interactions with the respiratory epithelium. Burkholderia cenocepacia expresses cable pili (Cbl) that allows attachment to cytokeratin 13 (Sajjan et al., 2000). However a Cbl mutant was demonstrated to bind to A549 cells at a similar level to the wild-type strain (Tomich & Mohr, 2003), indicating that other mechanisms mediate attachment to this cell line. Pili have been observed that react with antibodies raised against the pili protein (PilA) from P. aeruginosa (Kuehn et al., 1992). Such pili may allow attachment to cell surface oligosaccharides. Indeed, the P. aeruginosa PilA protein has been demonstrated to bind GalNAcß1-4Gal (Schweizer et al., 1998), and may have a similar tropism in Burkholderia cenocepacia. Interestingly, Burkholderia pseudomallei adhesion was inhibited by
90 % by a broad range of saccharide structures (Table 2). Galactose, glucose and N-acetylgalactosamine (1 mg ml1) have been demonstrated previously to inhibit attachment, along with asialo-GM1 (12.5 µg ml1) confirming the binding action of this glycolipid (Gori et al., 1999). In contrast to this study, mannose was not shown to inhibit attachment, however the concentration used was 25-fold less. It is well established that asialo-GM1 and asialo-GM2 are host cell receptors for Burkholderia pseudomallei (Kanai et al., 1997; Gori et al., 1999), however, the adhesins they bind have not been characterized. Possible adhesins identified include the capsule, LPS and flagellum (Atkins & Oyston, 2003). The completion of the genome sequence will lead to the identification of further adhesin candidates such as pili and non-pilus adhesins (www.sanger.ac.uk/Projects/B_pseudomallei). The flagellum has been demonstrated to mediate adhesion to amoebae and subsequent invasion (Inglis et al., 2003). The structure the flagellum attaches to, and whether a similar strategy is used to attach and invade human cells, are unknown.
Little is known about adhesion of vegetative Bacillus anthracis cells to human tissues. Putative adhesins include the S-layer proteins (EA1 and Sap), teichoic acids, the poly-
-D-glutamic acid capsule and fibronectin- and collagen-binding proteins (Ariel et al., 2003; Ezzell & Welkos, 1999; Fouet et al., 1999). The carbohydrate specificities displayed in this study (Table 2) were not due to interactions with the capsule, because the strain used lacked the pXO2 plasmid that encodes this virulence factor. Rather the S-layer and other putative adhesins, such as teichoic acid, are the likely candidates for such interactions. It cannot be discounted that expression of the capsule would alter the inhibition profile of this bacterium. Similarly, it should be recognized that in the event of pneumonic disease, such as would occur in the case of bioterrorism, initial interactions with the host cell would involve the B. anthracis spore prior to germination (Cieslak & Eitzen, 2000; Leggiadro, 2000). The spore may bind to entirely different structures from the vegetative cell and requires further investigation. The exosporium is a hexagonal lattice with filamentous appendages and glycoproteins (Fox et al., 2003; Mock & Fouet, 2001; Sylvestre et al., 2002) that may provide a mechanism for spore attachment to host cells prior to macrophage internalization.
This investigation details the potential of oligosaccharides to provide generic therapeutic benefits by inhibiting the attachment of a range of bacterial pathogens based on their pathological location, in this case the respiratory tract. Oligosaccharide inhibition of Burkholderia pseudomallei attachment to alveolar epithelial cells is particularly striking. Some of these compounds will be the subject of an in vivo investigation to determine the potential therapeutic benefit during intranasal and aerosol infections in a murine melioidosis model. As a generic therapy, the theory could be extended to include viruses. Sialylated glycoconjugates and sulphated dextrans have been shown to inhibit the attachment of various respiratory viruses to cell lines (Hosoya et al., 1991; Neyts et al., 1995; Suzuki et al., 1992; Suzuki et al., 2001). Theoretically, mixtures composed of saccharides such as 3-sialyllactose, polymeric saccharides and GalNAcß1-4Gal could afford protection against a spectrum of bacterial and viral respiratory pathogens, and may offer a novel prophylaxis.
| ACKNOWLEDGEMENTS |
|---|
|
|
|---|
| REFERENCES |
|---|
|
|
|---|
4)ß-D-Gal analogs. Can J Microbiol 44, 307311.This article has been cited by other articles:
![]() |
M. Drasbek, G. Christiansen, K. R. Drasbek, A. Holm, and S. Birkelund Interaction between the P1 protein of Mycoplasma pneumoniae and receptors on HEp-2 cells Microbiology, November 1, 2007; 153(11): 3791 - 3799. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. P. Anisimov and K. K. Amoako Treatment of plague: promising alternatives to antibiotics. J. Med. Microbiol., November 1, 2006; 55(Pt 11): 1461 - 1475. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Behling-Kelly, H. Vonderheid, K. S. Kim, L. B. Corbeil, and C. J. Czuprynski Roles of Cellular Activation and Sulfated Glycans in Haemophilus somnus Adherence to Bovine Brain Microvascular Endothelial Cells Infect. Immun., September 1, 2006; 74(9): 5311 - 5318. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. A. Boddey, C. P. Flegg, C. J. Day, I. R. Beacham, and I. R. Peak Temperature-Regulated Microcolony Formation by Burkholderia pseudomallei Requires pilA and Enhances Association with Cultured Human Cells Infect. Immun., September 1, 2006; 74(9): 5374 - 5381. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Thomas and T. Brooks Attachment of Yersinia pestis to human respiratory cell lines is inhibited by certain oligosaccharides. J. Med. Microbiol., March 1, 2006; 55(Pt 3): 309 - 315. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |