| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
PATHOGENICITY AND VIRULENCE |
1Microbiology, DSTL, CBS Porton Down, Salisbury, Wiltshire SP4 0JQ, UK 2Department of Medical Biochemistry and Molecular Biology, Institute of Biomedicine, University of Turku, Kiinamyllynkatu 10, 20520 Turku, Finland
Correspondence P. C. F. Oyston pcoyston{at}dstl.gov.uk
Received 14 August 2002 Accepted 20 December 2002
All strains of Yersinia pestis examined have been found to lack an O-antigen. In other members of the Enterobacteriaceae, the rough phenotype often results in attenuation. However, Y. pestis is the aetiological agent of bubonic plague. In evolving from the ancestral enteropathogenic Yersinia pseudotuberculosis, and with the development of an arthropod-vectored systemic pathogenesis, smooth LPS production is not necessary for Y. pestis virulence and the metabolic burden has been alleviated by inactivation of the O-antigen biosynthetic operon. To investigate this, Y. pestis strain KIM D27 was transformed with a plasmid carrying the operon encoding the O-antigen of Yersinia enterocolitica O : 3. Expression of the O-antigen could be detected in silver-stained gels. The receptor for bacteriophage 
YeO3-12 has been shown to be O-antigen, and infection by this bacteriophage results in lysis of Y. enterocolitica O : 3. Expression of the O-antigen in Y. pestis conferred sensitivity to lysis by YeO3-12. The O-antigen-expressing clone was shown to be as virulent in mice by the intravenous route of challenge as the rough wild-type. Assays showed no alteration in the ability of Y. pestis to resist lysis by cationic antimicrobial peptides, serum or polymyxin.
LPS is the major component of the Gram-negative outer membrane and the important role it plays in bacterial survival has been well documented (Reeves, 1994; Bayston & Cohen, 1990). Yersinia pestis, the causative agent of bubonic and pneumonic plague (Perry & Fetherston, 1997), is unusual when compared with other pathogenic members of the Enterobacteriaceae in that it does not possess smooth LPS; instead, its outer membrane is composed mainly of lipo-oligosaccharide (LOS). The LPS of the other pathogenic Yersinia species, Yersinia pseudotuberculosis and Yersinia enterocolitica, have been shown to contain an O-antigen essential for virulence, the expression of which is regulated by temperature (al-Hendy et al., 1992; Zhang et al., 1997; Karlyshev et al., 2001; Mecsas et al., 2001). The genetic basis for the lack of O-antigen in Y. pestis has been shown to be the presence of five pseudogenes in the gene cluster directing the biosynthesis of O-antigen (Skurnik et al., 2000; Prior et al., 2001). In agreement with the genetic data, LOS isolated from Y. pestis has been shown to lack O-antigen (Prior et al., 2001; Hitchen et al., 2002). Characterization of the Y. pestis LOS showed the core to possess two distinct molecular species of LOS, differing in terminal galactose or heptose (Hitchen et al., 2002; Vinogradov et al., 2002), and to contain Kdo, like other LPS molecules that have been studied (Ellwood, 1960; Hartley et al., 1974). The ability of Gram-negative bacteria to avoid serum-mediated killing would normally be facilitated by complement binding to the O-antigen of LPS. In other pathogenic bacteria that possess LOS, such as Neisseria and Haemophilus species, the LOS is sialylated, which confers serum resistance on the bacteria (Mandrell et al., 1992; Vogel et al., 1999), possibly by masking sugar residues that are reactive with the C3 component of complement (Estabrook et al., 1997). The role of LOS in the ability of Y. pestis to avoid serum-mediated killing is not immediately apparent from the structure reported (Hitchen et al., 2002), since it lacks both an O-antigen and a sialylated core.
Y. pestis is considered to have evolved from Y. pseudotuberculosis serotype O : 1b 1500–20 000 years ago (Achtman et al., 1999; Skurnik et al., 2000). Y. pseudotuberculosis is a gastrointestinal pathogen and possesses smooth LPS with an O-antigen, which is an essential virulence determinant (Karlyshev et al., 2001). All strains of Y. pestis examined to date are rough. This implies that possession of an O-antigen is somehow detrimental to the lifestyle of the flea-transmitted systemic pathogen. However, Y. pestis is known to infect by the oral route, indicating that the loss of O-antigen is not linked directly to the change in lifestyle of the pathogen (al-Hendy et al., 1991). It is possible that the biochemical burden placed on the organism as a result of O-antigen expression would be disadvantageous. Alternatively, expression of an O-antigen may interfere with the action of surface-exposed proteins required for virulence and, thus, O-antigen expression would be attenuating in vivo.
In this study, we have expressed an O-antigen in Y. pestis in order to determine whether the recombinant organism is attenuated for systemic virulence.
METHODS
Bacterial strains, plasmids, chemicals and growth conditions.
The strains used in this study are listed in Table 1. Y. pestis was routinely cultured on Blood base II (BAB) agar (Oxoid; 40 g l–1) containing 0.25 % (w/v) haemin or in BAB broth. Y. enterocolitica O : 3 and Escherichia coli were grown on Luria–Bertani (LB) medium. Ampicillin and tetracycline were used at respective final concentrations of 55 and 20 µg ml–1 as required. All chemicals were purchased from Sigma-Aldrich. Unless otherwise stated, plasmid extractions, restriction enzyme digestions, DNA ligations and transformations were performed by standard procedures (Sambrook et al., 1989) using enzymes provided by Roche.
|
Plasmid pAY100 (Table 1) carried ampicillin- and tetracycline-resistance markers and the O-antigen operon of Y. enterocolitica 6471/76-c O : 3 (al-Hendy et al., 1991). As tetracycline is an antibiotic used in therapy of plague, the tetracycline-resistance marker was inactivated by digestion with SalI and BamHI and filling of the generated ends with the Klenow fragment of DNA polymerase I (Sambrook et al., 1989). The blunt ends were then religated and the plasmid transformed in E. coli JM109. Ampicillin-resistant colonies were screened for tetracycline sensitivity. The isolated plasmid was designated pAY100.1.
PCR for confirmation of the presence of pAY100.1.
Oligonucleotide primers were designed from the sequence of pBR322 (JP3, 5'-AAATTGCTAACGCAGTCAGG; JP4, 5'-TGGAGTGGTGAATCCGT TAG) and the wbbX gene (JP1, 5'-ATGTTGGCACGAATAATG; JP2, 5'-TAGCGCTTCCATTACAGC) (GenBank accession no. Z18920) and synthesized by Cruachem Ltd. The PCR was performed for 30 cycles of 94 °C for 30 s, 50 °C for 30 s and 72 °C for 1 min.
Gel electrophoresis, silver staining and proteinase K minipreparations.
Glycine gel electrophoresis was performed with the buffers of Laemmli (1970) using a 12.5 % separating gel with a 4.5 % stacking gel. Aliquots (10 µl) of each sample were electrophoresed for approx. 2 h at 100 mV in the Mini-protean II slab system (Bio-Rad). Proteinase K minipreparations of LPS were produced and gels were silver-stained according to the method of Chart (1994).
Bacteriophage lysis assays.
Bacteriophage YeO3-12 (Pajunen et al., 2000, 2001) was propagated by adding a small volume of bacteriophage to 100 µl of an overnight culture of Y. enterocolitica O : 3 diluted in 5 ml fresh broth for 3 h at room temperature and filtered with a 0.22 µm filter. The filtrate contained the bacteriophages, which were stable for several months when stored at 4 °C.
E. coli and Y. enterocolitica strains were cultured overnight and serially diluted in sterile PBS (pH 7.4). To 100 µl aliquots of each diluted culture was added 10 µl of a suspension of 1 x 106 p.f.u. bacteriophage ml–1 and 4 ml molten top agar (tryptone soy agar solidified with 0.7 % agarose; Oxoid) supplemented with antibiotic as required. This was poured onto BAB plates and incubated at ambient temperature overnight. The plates were examined for the development of plaques. Alternatively, the bacteria were mixed with top agar, poured onto the plate and 10 µl undiluted bacteriophage was then dropped onto the centre of the plate.
For lysis experiments using Y. pestis, bacteria were grown on fresh LB agar containing 100 µg ampicillin ml–1. A single colony was used to inoculate 5 ml tryptone soy broth (TSB) supplemented with 100 µg ampicillin ml–1 in a 15 ml Falcon tube, which was incubated at 26 °C for 30 h with gentle rocking to obtain a culture in middle to late exponential phase. A 150 µl aliquot of the culture was mixed with 3 ml molten soft agar (TSB supplemented with 133 µg ampicillin ml–1, solidified with 0.4 % Bacto agar, cooled to 50 °C) and poured onto lambda agar plates (10 g Bacto tryptone l–1, 2.5 g NaCl l–1, 15 g Bacto agar l–1 and 150 µg ampicillin ml–1). The plates were incubated at 26 °C for 90 min and 20 µl aliquots of dilutions of bacteriophage YeO3-12 stock were then pipetted onto the surface. The plates were incubated overnight at 26 °C and examined for plaque formation.
Determination of virulence in mice.
The median lethal doses (MLD) of Y. pestis carrying either pBR322 or pAY100.1 were assessed by intravenous challenge of groups of six female 6-week-old BALB/c mice (Charles River Laboratories) with serial dilutions of exponential phase broth cultures grown at 28 °C. The MLD was determined by the method of Reed & Muench (1938). Prior to challenge and daily after that, the mice were dosed subcutaneously with 50 µl 100 mg ampicillin ml–1 (Intervet) to stabilize the plasmids. Humane end-points were strictly observed and animals deemed incapable of survival were killed humanely by cervical dislocation. Spleens were removed from all culled and some dead mice, homogenized in BAB broth and plated on BAB with and without supplementation with ampicillin to confirm maintenance of plasmids.
Determination of MICs.
Y. pestis KIM carrying either pBR322 or pAY100.1 was grown in BAB broth statically at 28 °C for 18 h followed by 4 h at 37 °C. Cultures were diluted 1 : 100 in fresh BAB broth for MIC assays using polymyxin B, cecropin P1 and mastoparan at final concentrations of 100–0.78 µg ml–1 and guinea pig serum at 10–0.078 % (v/v) (Sigma) in 96-well polypropylene microtitre plates. The MIC was determined as the lowest concentration of the compound that did not allow visible growth after 24 h at 28 °C. These assays were performed in triplicate and significance was determined by Student's t-test in Microsoft Excel.
RESULTS
Expression of LPS in Y. pestis
Y. pestis KIM was transformed with plasmid pAY100.1, expressing the O-antigen from Y. enterocolitica O : 3. Transformants were selected on BAB/ampicillin and the presence of the plasmid was confirmed by PCR. Expression of the O-antigen in E. coli JM109 and Y. enterocolitica O : 3 allows infection by the bacteriophage YeO3-12 and subsequent lysis, which can be observed as plaque formation in bacterial lawns. Plaques were also observed when the bacteriophage plaque assay was performed with Y. pestis KIM/pAY100.1. Proteinase K minipreparations were prepared from Y. enterocolitica O : 3 or Y. pestis KIM carrying either pBR322 or pAY100.1 and these were separated by SDS-PAGE and visualized by silver staining (Fig. 1). The preparation from Y. enterocolitica showed an O-antigen smear rather than a typical LPS ladder, which was also observed for the Y. pestis KIM/pAY100.1 sample. This smear is typically produced by this O-antigen, which does not form the ladders seen after electrophoresis of LPS preparations from other bacteria (Skurnik et al., 1995, 1999). The profile of Y. pestis KIM/pBR322 did not show the O-antigen smear. In some of the preparations (Fig. 1; lanes 1 and 2), there appeared to be some resistant protein bands present, and these were not removed even after prolonged proteinase K digestion. The extract from Y. pestis KIM/pBR322 (Fig. 1; lane 3) shows the typical profile of Y. pestis, with no O-antigen being expressed while the lipid A and core bands are visible.
|
Effect of O-antigen expression on virulence in the mouse model of infection
The effect of expressing the O-antigen in Y. pestis on the virulence of the organism was evaluated. Y. pestis KIM is capable of killing mice when administered by the intravenous route. The MLDs of the recombinant Y. pestis KIM/pAY100.1 and the control strain of Y. pestis KIM/pBR322 were compared. To stabilize the plasmid in vivo, mice were dosed subcutaneously with veterinary-grade ampicillin for the course of the experiment. The MLD for Y. pestis KIM/pBR322 was calculated to be 199.78 c.f.u., while that for Y. pestis KIM/pAY100.1 was calculated as 221 c.f.u. Plating of spleen homogenates taken from mice that had died showed that the plasmids had been stably maintained in Y. pestis KIM.
Effect of O-antigen expression on resistance of Y. pestis to lysis
The effect of a range of surface-active compounds was evaluated against Y. pestis expressing the O-antigen versus the wild-type control. The MIC of polymyxin against both strains was 250 µg ml–1 (Fig. 2a). Absolute inhibition of growth was not achieved with either of the cationic antimicrobial peptides, cecropin and mastoparan, although, at higher concentrations, there was some inhibition of growth (Fig. 2b). There was no difference in the susceptibility of Y. pestis KIM carrying plasmids pAY100.1 or pBR322 to these peptides. Y. pestis KIM/pBR322 was unable to grow in the presence of 0.3125 % guinea pig serum, was partially inhibited at one dilution step lower and was resistant to 0.078 % serum (Fig. 2c). The expression of the O-antigen by Y. pestis did not result in any difference in sensitivity to serum killing.
|
DISCUSSION
We have expressed the O-antigen of Y. enterocolitica in Y. pestis strain KIM. Expression of the O-antigen was proven by observation of the O-antigen smear observed after SDS-PAGE analysis, which is the expected profile of this O-antigen (Ogasawara et al., 1985). The undigested protein bands observed in some of the samples have been reported previously for Y. enterocolitica O : 3 (al-Hendy et al., 1992). Bacteriophage YeO3-12 was able to lyse Y. pestis expressing the O-antigen, showing that the O-antigen was expressed by Y. pestis and was exposed correctly to the phage, allowing initiation of phage infection.
Many virulence factors have been described for Y. pestis (Perry & Fetherston, 1997), but one of the surprising aspects of this pathogen is that it does not possess an O-antigen, its outer membrane instead containing LOS (Hitchen et al., 2002). Y. pestis is the only species of the pathogenic Yersinia to lack an O-antigen (Brubaker, 1991). The LPS O-antigen has been shown to be essential for virulence of other pathogenic Yersinia species (al-Hendy et al., 1992; Zhang et al., 1997; Karlyshev et al., 2001; Mecsas et al., 2001) as well as for other members of the Enterobacteriaceae, as it is required for resistance to complement lysis, cationic antibacterial peptides and phagocytosis (Porat et al., 1987; Makela et al., 1988). In Y. enterocolitica O : 3, the O-antigen is required for resistance to killing by the alternative complement pathway (Wachter & Brade, 1989). A rough mutant of Y. enterocolitica O : 3 was resistant to bacteriophage YeO3-12 lysis and showed reduced virulence in mice (al-Hendy et al., 1992). The expression of an O-antigen in Y. pestis KIM had no significant effect on virulence following intravenous challenge of mice. Virulence was neither enhanced, which would have indicated a role in vivo, nor reduced, which would have been an indication of metabolic burden.
In Klebsiella pneumoniae, loss of LPS O-antigen side-chain expression had no effect on the ability of the pathogen to cause pneumonia (Cortes et al., 2002). Rather, the capsule was shown to play a key role in resistance to complement deposition and alveolar macrophage killing. Y. pestis possesses a capsule composed of F1-antigen (Baker et al., 1952), which is a 17 kDa protein expressed at 37 °C but not at lower temperatures (Bennett & Tornabene, 1974). However, possession of the F1-antigen capsule does not apparently act to compensate for lack of an O-antigen on LPS, as F1-defective strains are still virulent (Welkos et al., 1995). Alternative mechanisms for the avoidance of complement lysis by Y. pestis have been proposed based on degradation of the membrane attack complex. Y. pestis expresses Pla, a plasminogen activator at 37 °C with coagulase activity at lower temperatures (McDonough & Falkow, 1989). The Pla protein is capable of degrading proteins other than plasminogen, including C3, but inactivating the pla gene has variable effects on virulence and the role of Pla is still undetermined to a large extent (Perry & Fetherston, 1997). LPS isolated from the surface of Y. pestis, Y. enterocolitica and Y. pseudotuberculosis and incorporated into liposomes rendered the liposomes resistant to complement lysis (Porat et al., 1995). This may mimic the effect seen in Neisseria gonorrhoeae, where rough LPS confers resistance to complement lysis by aberrant deposition of the membrane attack complex (Joiner et al., 1983). However, this effect was not confirmed by other studies, which suggested that the O-antigen of Y. enterocolitica did indeed play a role in resistance to complement-mediated killing (Wachter & Brade, 1989; al-Hendy et al., 1992; Skurnik & Toivanen, 1993).
Not only does Y. pestis cause bubonic plague, resulting from the bite of an infected flea, it also causes pneumonic plague following aerosol transmission, and orally acquired infections have been reported both in humans (Christie et al., 1980) and, experimentally, in mice (Butler et al., 1982; Sebbane et al., 2001). It has been postulated that the O-antigen may be associated with functions related to the oral route of infection that may be irrelevant in the case of vector-mediated transmission (al-Hendy et al., 1992). For example, in Vibrio cholerae, LPS is required for colonization of the small intestine (Nesper et al., 2001). However, the KIM strain used in this study is virulent only by the intravenous route and not by other peripheral routes (Une & Brubaker, 1984), so it was not possible to determine directly whether possession of an O-antigen had an effect on oral virulence.
In Salmonella enterica var. Typhimurium, loss of O-antigen increased susceptibility to the peptide magainin 2, although not to the defensin NP-1 (Groisman, 1994). Resistance to cationic antimicrobial peptides is complex, and Y. pestis is highly resistant to those that have been examined (Bengoechea et al., 1998; Hitchen et al., 2002). Expression of the O-antigen did not appear to affect the charge of the bacterial surface in an adverse way, as it did not render the strain susceptible to lysis by cationic antimicrobial peptides.
In conclusion, Y. pestis appears to compensate for loss of the O-antigen so that it is highly resistant to many of the stresses and threats that it may encounter. Expression of the O-antigen did not improve the pathogen's ability to resist lysis by serum and had no impact on virulence by the intravenous route. Non-expression of O-antigen would represent a huge saving in energy, as synthesis of such a major macromolecule would be a significant biochemical burden. In addition, an added benefit would be the loss of a molecule used as a receptor by many bacteriophages. However, it must be remembered when interpreting the results that the O-antigen used in this study is different from that which the progenitor strain of Y. pestis would have expressed, and the effect of O-antigen expression may be affected by variations in the O-antigen structure.
ACKNOWLEDGEMENTS
ACKNOWLEDGEMENTS
P. C. F. O. and J. L. P. have contributed equally to the scientific experimental data presented in this paper. We would like to thank Dave Rawkins for technical support. The work of M. S. and S. K. has been supported by the Academy of Finland.
REFERENCES
YeO3-12, specific for Yersinia enterocolitica serotype O : 3, is related to coliphages T3 and T7. J Bacteriol 182, 5114–5120.YeO3-12 of Yersinia enterocolitica serotype O : 3. J Bacteriol 183, 1928–1937.This article has been cited by other articles:
|
|
 |
|
  S.-s. Zhang, C. G. Park, P. Zhang, S. S. Bartra, G. V. Plano, J. D. Klena, M. Skurnik, B. J. Hinnebusch, and T. Chen Plasminogen Activator Pla of Yersinia pestis Utilizes Murine DEC-205 (CD205) as a Receptor to Promote Dissemination J. Biol. Chem., November 14, 2008; 283(46): 31511 - 31521. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Zhang, M. Skurnik, S.-S. Zhang, O. Schwartz, R. Kalyanasundaram, S. Bulgheresi, J. J. He, J. D. Klena, B. J. Hinnebusch, and T. Chen Human Dendritic Cell-Specific Intercellular Adhesion Molecule-Grabbing Nonintegrin (CD209) Is a Receptor for Yersinia pestis That Promotes Phagocytosis by Dendritic Cells Infect. Immun., May 1, 2008; 76(5): 2070 - 2079. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. M. Galvan, M. A. S. Lasaro, and D. M. Schifferli Capsular Antigen Fraction 1 and Pla Modulate the Susceptibility of Yersinia pestis to Pulmonary Antimicrobial Peptides Such as Cathelicidin Infect. Immun., April 1, 2008; 76(4): 1456 - 1464. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Tong, D. Zhou, Y. Song, L. Zhang, D. Pei, Y. Han, X. Pang, M. Li, B. Cui, J. Wang, et al. Pseudogene accumulation might promote the adaptive microevolution of Yersinia pestis J. Med. Microbiol., March 1, 2005; 54(3): 259 - 268. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Sebbane, C. O. Jarrett, J. R. Linkenhoker, and B. J. Hinnebusch Evaluation of the Role of Constitutive Isocitrate Lyase Activity in Yersinia pestis Infection of the Flea Vector and Mammalian Host Infect. Immun., December 1, 2004; 72(12): 7334 - 7337. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Zhou, Y. Han, Y. Song, Z. Tong, J. Wang, Z. Guo, D. Pei, X. Pang, J. Zhai, M. Li, et al. DNA Microarray Analysis of Genome Dynamics in Yersinia pestis: Insights into Bacterial Genome Microevolution and Niche Adaptation J. Bacteriol., August 1, 2004; 186(15): 5138 - 5146. [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 |
) was included in each experiment. (a) Growth of Y. pestis KIM/pBR322 in the absence (
) and presence (
) of polymyxin and Y. pestis KIM/pAY100.1 in the absence (
) and presence (
) of polymyxin. (b) Growth of Y. pestis KIM/pBR322 alone (
) or cecropin (