Abstract
Lipopolysaccharide (LPS) is a critical virulence determinant in Pasteurella multocida and a major antigen responsible for host protective immunity. In other mucosal pathogens, variation in LPS or lipooligosaccharide structure typically occurs in the outer core oligosaccharide regions due to phase variation. P. multocida elaborates a conserved oligosaccharide extension attached to two different, simultaneously expressed inner core structures, one containing a single phosphorylated 3-deoxy-d-manno-octulosonic acid (Kdo) residue and the other containing two Kdo residues. We demonstrate that two heptosyltransferases, HptA and HptB, add the first heptose molecule to the Kdo1 residue and that each exclusively recognizes different acceptor molecules. HptA is specific for the glycoform containing a single, phosphorylated Kdo residue (glycoform A), while HptB is specific for the glycoform containing two Kdo residues (glycoform B). In addition, KdkA was identified as a Kdo kinase, required for phosphorylation of the first Kdo molecule. Importantly, virulence data obtained from infected chickens showed that while wild-type P. multocida expresses both LPS glycoforms in vivo, bacterial mutants that produced only glycoform B were fully virulent, demonstrating for the first time that expression of a single LPS form is sufficient for P. multocida survival in vivo. We conclude that the ability of P. multocida to elaborate alternative inner core LPS structures is due to the simultaneous expression of two different heptosyltransferases that add the first heptose residue to the nascent LPS molecule and to the expression of both a bifunctional Kdo transferase and a Kdo kinase, which results in the initial assembly of two inner core structures.
Pasteurella multocida is a gram-negative bacterial pathogen that causes disease in a wide range of mammals and birds. P. multocida is classified into serogroup A, B, D, E, or F based on capsular composition and is further classified into 16 Heddleston serotypes based on lipopolysaccharide (LPS) antigens. The type of disease that each strain can cause correlates to some degree with the type of capsule that is expressed, with serogroup A and occasionally serogroup F strains associated with fowl cholera outbreaks in birds, serogroup B and E strains with bovine hemorrhagic septicemia, and toxigenic serogroup D strains with atrophic rhinitis in pigs (5).
The expression of wild-type LPS is critical for the progression of fowl cholera (13). The structure of the LPS has been determined for two serotype A:1 fowl cholera strains, VP161 and X73, and the genome-sequenced strain Pm70 (28-30). The LPS expressed by P. multocida is similar to the LPS (often designated lipooligosaccharide) expressed by other mucosal pathogens, including species within the Neisseria and Haemophilus genera that have mono- and oligosaccharide extensions to the core structure but lack O-antigen polysaccharide repeating units (9, 10, 22, 26, 30). The LPS structures expressed by the three P. multocida strains VP161, X73, and Pm70 all share common regions but differ in their oligosaccharide extension and side branches (28-30). Unusually, two LPS glycoforms were expressed simultaneously by each strain, which differed in the inner core region but retained the same primary oligosaccharide extension (shown for VP161 in Fig. 1). Semiquantitative analysis of the LPS expressed by each strain showed that most glycoforms contained a single phosphorylated 3-deoxy-d-manno-octulosonic acid (Kdo) molecule that was frequently substituted with a phosphoethanolamine (PEtn) residue attached to the phosphate group (glycoform A). Furthermore, this single Kdo glycoform was shown to have a second glucose residue attached to the proximal heptose (Fig. 1). The alternative inner core LPS structure contained two unphosphorylated Kdo residues and lacked the second glucose residue on the proximal heptose (glycoform B).
FIG. 1.
LPS structure of P. multocida VP161 and predicted enzymes for selected biosynthetic steps. Two core types are observed, glycoforms A and B. P. multocida X73 has identical LPS glycoform structures, with the exception of an additional PEtn moiety attached to each terminal galactose residue. Glc, glucose; Hep, heptose; Gal, galactose; PCho, phosphocholine; Kdo, 3-deoxy-d-mannooctulosonate; P, phosphate.
To determine which genes were involved in the biosynthesis of each glycoform, we mutated two genes predicted to encode heptosyltransferases that attach the first heptose residues to Kdo and a gene predicted to encode a Kdo kinase required for phosphorylation of the lipid A-Kdo1. The LPS structures expressed by these mutants were then elucidated, and the mutants were tested for their ability to cause disease in chickens, allowing us to determine for the first time whether expression of both glycoforms was required for virulence in the natural host. Finally, we assessed the role of LPS structure in conferring resistance to antimicrobial peptides which are predicted to be important components of the chicken innate immune system.
MATERIALS AND METHODS
Bacterial strains, plasmids, media, and growth conditions.
The bacterial strains and plasmids used in this study are listed in Table 1. Escherichia coli was grown routinely in Luria-Bertani broth. P. multocida strains were grown in either brain heart infusion (BHI) broth or nutrient broth (Oxoid, Basingstoke, U.K) supplemented with yeast extract (3% wt/vol). Solid media were obtained by the addition of 1.5% agar. When required, media were supplemented with spectinomycin (100 μg/ml), streptomycin (50 μg/ml), kanamycin (50 μg/ml), or tetracycline (2.5 μg/ml for routine culturing or 8 μg/ml for selection of P. multocida transconjugants). For structural studies on LPS isolated from in vitro-grown P. multocida, strains were grown on BHI plates with the appropriate antibiotics for 18 h at 37°C. For structural studies on LPS isolated from in vivo-grown P. multocida, 103 CFU of strain X73 were injected into the breast muscles of 14-week Hy-Line Brown chickens and disease allowed to progress for 14 to 18 h. When birds were showing clear clinical signs of fowl cholera, they were euthanatized by CO2 asphyxiation and 10 ml of blood was recovered by cardiac puncture. Bacteria were purified from the blood as described previously (6). The cells were killed by addition of phenol to 2%, and at 3 h after phenol addition, 1 g of hyaluronidase (Roche Chemicals) was added and stirred for 1 h before cells were harvested by using a Sharples continuous-flow centrifuge.
TABLE 1.
Bacterial strains and plasmids used in this study
| Strain or plasmid | Relevant description | Source or reference |
|---|---|---|
| Strains | ||
| P. multocida | ||
| X73 | Serotype A:1 | 14 |
| VP161 | Serotype A:1, Vietnamese isolate from chickens | 35 |
| AL435 | VP161 carrying a Tn916 insertion in gene PM1417, fully virulent | This study |
| AL838 | VP161 containing pAL99 | This study |
| AL721 | kdkA mutant of AL435 | This study |
| AL839 | AL721 containing pAL99 | This study |
| AL840 | AL721 containing pAL462 | This study |
| AL690 | hptB mutant of AL435 | This study |
| AL844 | AL690 containing pAL99 | This study |
| AL845 | AL690 containing pAL464 | This study |
| AL836 | Nonpolar hptA mutant of AL435 | This study |
| AL846 | AL836 containing pAL99 | This study |
| AL847 | AL836 containing pAL465 | This study |
| E. coli | ||
| DH5α | deoR endA1 gyrA96 hsdR17 (rK− mK+) recA1 relA1 supE44 thi-1 Δ(lacZYA-argFV169) φ80lacZΔM15 F− | Bethesda Research Laboratories |
| Sm10 λ pir | Strain for propagation of pUA826 and its derivatives | 23 |
| Plasmids | ||
| pUA826 | Mob+, R6K replicon, Apr Strr Spcr, single-crossover insertional mutagenesis vector | 7 |
| pUA826tpi | As for pUA826 but containing 240-bp PstI fragment containing P. multocida tpiA promoter region, nonpolar mutagenesis vector | This study |
| pAL99 | P. multocida expression plasmid (Kanr), constitutive tpiA promoter upstream of BamHI | 13 |
| pAL462 | Complete kdkA gene cloned into pAL99 | This study |
| pAL464 | Complete hptB gene cloned into pAL99 | This study |
| pAL465 | Complete hptA gene cloned into pAL99 | This study |
DNA manipulations.
Restriction digests and ligations were performed according to the manufacturers' instructions using enzymes obtained from New England Biolabs (Beverley, MA) or Roche Diagnostics GmbH (Mannheim, Germany). Plasmid DNA was prepared using alkaline lysis (4) and further purified using QIAGEN columns (QIAGEN GmbH, Germany), while genomic DNA was prepared using the cetyltrimethylammonium bromide method (2). PCR amplification of DNA was performed using Taq DNA polymerase or the Expand high-fidelity PCR system (Roche Diagnostics) and purified using the QIAGEN PCR purification kit. The oligonucleotides used in this study are listed in Table 2. DNA sequences were determined on an Applied Biosystems 3730S genetic analyzer and analyzed with Sequencher version 3.1.1 (Gene Codes, Ann Arbor, MI).
TABLE 2.
Oligonucleotides used in this study
| Oligonucleotide | Sequence | Description |
|---|---|---|
| BAP3813 | TTTGGAGTCGACAAGCGCGTATTCTTGGTT | kdkA internal forward primer for mutagenesis |
| BAP3814 | CACTTTGTCGACGTTCACCAAGCTGATGTG | kdkA internal reverse primer for mutagenesis |
| BAP3914 | AAAGGGATCCGTGGCCATTCGCTATCTCTG | kdkA flanking forward primer for expression |
| BAP3915 | TATGTGTCGACCTTATGCTTGATATCCCGC | kdkA flanking reverse primer for expression |
| BAP3721 | TATGTTGTCGACAAGTTGATTGGGTGGTCGAAG | hptB internal forward primer for mutagenesis |
| BAP3722 | TAAATGGTCGACTCCGGTGTCAACAGAAACAA | hptB internal reverse primer for mutagenesis |
| BAP3745 | TAATTGGGATCCAACCGGAATTGGTATTCG | hptB flanking forward primer for expression |
| BAP3746 | TCAAGAGTCGACTGCCGTATAGGCAGTTTCAA | hptB flanking reverse primer for expression |
| BAP3719 | CAAAACGTCGACGCAGTTACTGGCTGGTAT | hptA internal forward primer for mutagenesis |
| BAP3720 | TCAGTAGTCGACAACTTACATTGGGCGAGGA | hptA internal reverse primer for mutagenesis |
| BAP3966 | CACACAGGATCCATTGGACGAGATAATGATGCCG | hptA flanking forward primer for expression |
| BAP3967 | ATAACAGTCGACGGTTCATCATTGAACGTC | hptA flanking reverse primer for expression |
| BAP2782 | GCCCTACACAAATTGGGAGA | pUA826 primer near SalI site |
Construction of P. multocida mutants.
The three P. multocida strains Pm70, X73, and VP161 have all been tested for amenability to mutant construction. To date, insertional inactivation of target genes has routinely been successful only for P. multocida strain VP161. For inactivation of each of the candidate glycosyltransferase genes in P. multocida strain VP161, we used a modification of the previously described single-crossover insertional mutagenesis method which utilizes the λ pir-dependent plasmid pUA826 (7). Internal DNA fragments of kdkA and hptB were amplified by PCR using the primers listed in Table 2, digested with SalI, and ligated into SalI-digested pUA826. As single-crossover insertion of pUA826 is likely to have polar effects on downstream genes, a different vector was constructed for inactivation of the hptA gene, which bioinformatic analysis suggested was cotranscribed with the gene PM1301. We modified pUA826 by insertion of the P. multocida constitutive tpiA promoter between the pUA826 PstI sites located at bp 4789 and 5472 (generating pUA826tpi). Thus, single-crossover insertion of derivatives of this plasmid would allow transcription of downstream genes through initiation of transcription from the P. multocida tpiA promoter located within the integrated plasmid. An internal fragment of hptA was amplified by PCR using the oligonucleotides listed in Table 2, digested with SalI, and ligated into SalI-digested pUA826tpi.
To generate each donor strain, each ligation mix was transformed into E. coli SM10 λ pir cells and the transformants screened for recombinant pUA826 or pUA826tpi. The correct recombinant plasmids were then mobilized into the recipient P. multocida strain AL435 by conjugation. AL435 was used as the recipient strain, as it is resistant to tetracycline (due to integration of Tn916 in gene PM1417), expresses wild-type LPS, and is fully virulent (see Table 4). For each mating, 1 ml of the recipient P. multocida strain AL435 and 1 ml of the donor SM10 λ pir strain, harboring the appropriate recombinant plasmid, were mixed and filtered through a 0.45-μm filter which was then placed on a blood agar plate and incubated at 30°C for 16 h. To recover the bacteria, filters were vortexed in 3 ml of BHI. P. multocida transconjugants were selected by plating the resuspended bacteria onto BHI containing tetracycline, streptomycin, and spectinomycin. Single-crossover insertion of the recombinant plasmid into each of the target genes was confirmed by PCR using flanking primers in combination with a primer located within the vector sequence (Table 2).
TABLE 4.
Virulence of wild-type (VP161) and parent (AL435) strains and LPS mutants as determined by intramuscular challenge of 12-week-old Hy-Line Brown chickens
| Strain (relevant genotype) | Dose (CFU) | Survivala
|
|
|---|---|---|---|
| No. surviving/total | Time (h) | ||
| VP161 | 1.3 × 103 | 0/9 | 14-19 |
| AL435 | 1.7 × 103 | 0/4 | 14-19 |
| AL836 (hptA) | 7 × 102 | 9/9 | NAb |
| AL690 (hptB) | 6 × 102 | 0/9 | 16-19 |
| AL721 (kdkA) | 1 × 103 | 0/4 | 14-20 |
All birds that reached an agreed stage of the infection were euthanatized in accordance with animal ethics requirements.
NA, not applicable.
In trans complementation of mutants.
The complete open reading frame for each gene to be complemented was amplified from P. multocida VP161 genomic DNA using flanking oligonucleotides (Table 2). The amplified DNA fragments were ligated into SalI- and BamHI-digested expression vector pAL99 (Table 1), such that transcription of the gene would be driven by the constitutive P. multocida tpiA promoter. The nucleotide sequence of each of the recombinant plasmids was determined to check fidelity of the cloned gene and then transformed into the corresponding P. multocida mutant, generating the complemented strains listed in Table 1. As a control, the vector pAL99 was transformed separately into each mutant (Table 1).
Assessment of virulence.
All animal experiments were approved by the relevant Animal Ethics Committee. Groups of commercially obtained Hy-Line Brown chickens, aged 12 weeks, were infected with approximately 103 CFU (∼100 times the 50% infective dose) of P. multocida by injection into the pectoral muscle, and the experiment was allowed to proceed for up to 39 h or until each bird was deemed incapable of survival, when birds were euthanatized in accordance with animal ethics requirements. Blood samples for viable counts were obtained from all birds with late clinical signs of infection and from all healthy birds at 22 h postinfection. Postmortem samples of the site of injection were taken from a representative sample of birds within each group and plated onto BHI agar. Recovered bacteria were checked by PCR to confirm the identity of the P. multocida strains present.
Competitive growth assays.
Competitive growth assays were performed as described previously (12) and used to quantify the relative growth rates of the P. multocida strains in vivo. Mutants were identified as attenuated if the relative competitive index (rCI) value was significantly less than 1.0 as determined by statistical analysis using the one sided z test (P < 0.05).
Antimicrobial peptide sensitivity assays.
The chicken antimicrobial peptide fowlicidin-1 (RVKRVWPLVIRTVIAGYNLYAIKKK) (36) was synthesized at 96% purity by Auspep (Parkville, Australia). For peptide sensitivity assays, wild-type and mutant strains were grown to late exponential phase (optical density at 600 nm of 0.65), washed once and then diluted 1/10 in 10 mM phosphate (pH = 7.0), 10% BHI. Equal volumes of washed cells (100 μl) and diluted peptide (0 to 4 μg/ml in 10 mM phosphate [pH 7.0]) were mixed and incubated at 37°C for 1 h. Numbers of surviving bacteria were determined by viable counts on BHI agar.
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and silver staining.
Proteinase K-treated whole-cell lysates were analyzed on a Bio-Rad mini-protein gel apparatus, using sodium dodecyl sulfate-polyacrylamide gel electrophoresis as described previously (18). LPS was then visualized by carbohydrate silver staining (32).
Purification of LPS and LPS-OH.
Cells were killed with 2% phenol, washed four times with water, lyophilized and resuspended in 200 μl of H2O containing 5 μg of proteinase K, and incubated at 37°C for 5 h. The sample was heated to 70°C for 10 min, lyophilized, then dissolved in 200 μl of ammonium acetate buffer (20 mM, pH 7.4) containing 1 μg of RNase and 2 μg of DNase, incubated at 37°C for 5 h, and then lyophilized. The crude LPS-containing samples were O deacylated by dissolving in 200 μl of anhydrous hydrazine and incubating with stirring at 37°C for 1.5 h. Excess hydrazine was destroyed by addition of 5 volumes of ice-cold acetone to the chilled samples and repeated acetone washes. The O-deacylated LPS (LPS-OH) pellet was redissolved in H2O and lyophilized.
MS.
Capillary electrophoresis-electrospray ionization-mass spectrometry (CE-ES-MS) was performed on a crystal model 310 CE instrument (AYI Unicam) coupled to an API 3000 MS (Perkin-Elmer/Sciex) via a microionspray interface (9). A sheath solution (isopropanol-methanol, 2:1) was delivered at a flow rate of 1 μl/min to a low-dead-volume tee (250-μm inner diameter; Chromatographic Specialties). All aqueous solutions were filtered through a 0.45-μm filter (Millipore) before use.
RESULTS
Bioinformatic prediction of inner core LPS biosynthesis genes.
Putative LPS biosynthesis genes in P. multocida VP161 were identified based on their similarity to LPS transferases of known function in other bacteria using both BLAST (1) and the CAZY database of glycosyltransferases (8). Two candidate genes, hptA (annotated on the Pm70 genome as opsX) and hptB (PM1843), had been predicted to encode the 38.4- and 36.3-kDa heptosyltransferases required for the addition of the first heptose residue to either the lipid A-Kdo1-Kdo2 or the lipid A-Kdo2-P core structure (Fig. 1) (30). A conserved domain search (21) showed that both had high levels of identity to heptosyltransferases belonging to pfam01075, the glycosyltransferase family 9, and COG0859, the RfaF group of transferases. An alignment of the two encoded proteins, HptA and HptB, showed only 21% identity. However, HptB has 53% identity with the heptosyl-I-transferase WaaC (RfaC), which adds heptose to the Kdo-Kdo LPS acceptor molecule in Salmonella enterica serovar Typhimurium (27), while HptA has 72% identity to OpsX, a heptosyl-I-transferase in Haemophilus influenzae which adds the first heptose to a singly phosphorylated Kdo acceptor molecule (11).
The protein encoded by the P. multocida VP161 gene kdkA (designated PM1303 on the Pm70 genome) shares 58% identity to the known Kdo kinase KdkA in H. influenzae (34) and belongs to the family of LPS kinases (pfam06293). It is therefore predicted to encode the kinase required for the addition of the phosphate group to the Kdo residue.
Sequencing of the VP161 genome within the regions that encoded HptA, HptB, and KdkA revealed that all genes were very highly conserved and that the genetic organization was identical to the corresponding regions in the annotated Pm70 genome. Each of the candidate genes identified in VP161 was mutated by insertional inactivation, and each mutant was complemented with the full-length gene cloned into the P. multocida expression vector pAL99 (Table 1).
Structural analysis of the mutants expressing modified LPS forms.
CE-MS analyses of the LPS-OH isolated from the mutant strains are summarized in Table 3. The hptA mutant (AL836) produced a fully extended glycoform B (consisting of lipid A-Kdo1-Kdo2), but no fully extended glycoform A was expressed. Additionally, doubly and triply charged ions at m/z 678.4 and 451.6, respectively, indicated an accumulation of lipid A-Kdo1-P truncated LPS (Fig. 2A). In contrast, analysis of the LPS from the hptB mutant (AL690) showed a fully extended glycoform A, and an accumulation of the lipid A-Kdo1-Kdo2 species was observed by virtue of the doubly and triply charged ions at m/z 757.2 and 504.4, respectively (Fig. 2B). The accumulation of the alternative truncated LPS in each of the mutants clearly suggests that HptA cannot utilize lipid A-Kdo1-Kdo2 and that HptB is unable to utilize lipid A-Kdo1-P as an acceptor molecule. The Kdo kinase (KdkA) mutant (AL721) produced a fully extended glycoform B but no glycoform A and no accumulation of truncated LPS (Fig. 2C). These data indicate that an unphosphorylated Kdo on the acceptor molecule is freely available for the addition of a second Kdo residue allowing glycoform B to be produced. When complemented in trans with the appropriate genes, all mutant strains produced both wild-type glycoforms (Table 3).
TABLE 3.
Negative-ion CE-ES-MS data and proposed compositions of LPS-OH from the P. multocida VP161 wild-type and mutant strainsa
| Strain (relevant description) | Observed ions (m/z)
|
Molecular ion (m/z)
|
Relative intensity | Proposed composition | ||||
|---|---|---|---|---|---|---|---|---|
| (M − 4H)4− | (M − 3H)3− | (M − 2H)2− | Observed | Calculated | ||||
| VP161 (wild type) | 992.4 | 2,980.2 | 2,977.6 | 0.1 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH | |||
| 999.6 | 3,001.8 | 2,999.5 | 0.3 | 2PCho, 4Hex, 4Hep, Kdo-P, lipid A-OH | ||||
| 1,033.0 | 3,101.0 | 3,100.7 | 0.3 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn | ||||
| 1,040.7 | 3,125.1 | 3,122.5 | 0.7 | 2PCho, 4Hex, 4Hep, Kdo-P, lipid A-OH-PEtn | ||||
| 1,081.7 | 3,247.7 | 3,245.6 | 1.0 | 2PCho, 4Hex, 4Hep, Kdo-P-PEtn, lipid A-OH-PEtn | ||||
| AL690 (hptB) | 504.4 | 757.2 | 1,516.1 | 1,515.4 | 0.2 | 2Kdo, lipid A-OH-PEtn | ||
| 780.3 | 1,040.4 | 3,124.7 | 3,122.5 | 0.2 | 2PCho, 4Hex, 4Hep, Kdo-P, lipid A-OH-PEtn | |||
| 810.8 | 1,081.6 | 3,247.7 | 3,245.6 | 1.0 | 2PCho, 4Hex, 4Hep, Kdo-P-PEtn, lipid A-OH-PEtn | |||
| AL844 (hptB + vector) | 757.5 | 1,516.1 | 1,515.4 | 0.2 | 2Kdo, lipid A-OH-PEtn | |||
| 1,040.8 | 3,124.7 | 3,122.5 | 0.3 | 2PCho, 4Hex, 4Hep, Kdo-P, lipid A-OH-PEtn | ||||
| 1,081.7 | 1,622.7 | 3,247.7 | 3,245.6 | 1.0 | 2PCho, 4Hex, 4Hep, Kdo-P-PEtn, lipid A-OH-PEtn | |||
| AL845 (hptB, complemented) | 1,033.0 | 3,101.0 | 3,100.7 | 0.2 | 2PCho, 3Hex, 4Hep, 2Kdo, Lipid A-OH-PEtn | |||
| 1,040.7 | 1,561.2 | 3,124.7 | 3,122.5 | 0.3 | 2PCho, 4Hex, 4Hep, Kdo-P, lipid A-OH-PEtn | |||
| 1,081.8 | 1,622.8 | 3,247.7 | 3,245.6 | 1.0 | 2PCho, 4Hex, 4Hep, Kdo-P-PEtn, lipid A-OH-PEtn | |||
| AL836 (hptA) | 616.8 | 1,235.6 | 1,252.2 | 0.1 | Kdo-P, lipid A-OH (− H2O) | |||
| 451.6 | 678.4 | 1,358.8 | 1,375.2 | 0.5 | Kdo-P-PEtn, lipid A-OH (− H2O) | |||
| 1,033.4 | 1,549.9 | 3,102.6 | 3,100.6 | 1.0 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn | |||
| AL846 (hptA + vector) | 616.8 | 1,235.6 | 1,252.2 | 0.1 | Kdo-P, lipid A-OH (− H2O) | |||
| 451.6 | 678.3 | 1,358.8 | 1,375.2 | 0.5 | Kdo-P-PEtn, lipid A-OH (− H2O) | |||
| 774.7 | 1,033.2 | 1,550.3 | 3,102.6 | 3,100.6 | 1.0 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn | ||
| AL847 (hptA, complemented) | 1,033.3 | 3,101.0 | 3,100.7 | 0.5 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn | |||
| 1,040.7 | 1,561.2 | 3,124.7 | 3,122.5 | 1.0 | 2PCho, 4Hex, 4Hep, Kdo-P, lipid A-OH-PEtn | |||
| 1,081.5 | 1,622.4 | 3,247.7 | 3,245.6 | 0.9 | 2PCho, 4Hex, 4Hep, Kdo-P-PEtn, lipid A-OH-PEtn | |||
| AL721 (kdkA) | 743.8 | 992.3 | 2,979.9 | 2,977.6 | 0.2 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH | ||
| 775.0 | 1,033.1 | 3,103.1 | 3,100.6 | 1.0 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn | |||
| 825.5 | 1,101.2 | 3,306.3 | 3,303.6 | 0.2 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn-P-PEtn | |||
| AL839 (kdkA + vector) | 1,033.1 | 1,550.3 | 3,103.1 | 3,100.6 | 1.0 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn | ||
| AL840 (kdkA, complemented) | 1,033.3 | 3,101.0 | 3,100.7 | 0.2 | 2PCho, 3Hex, 4Hep, 2Kdo, lipid A-OH-PEtn | |||
| 1,040.7 | 1,561.3 | 3,124.7 | 3,122.5 | 0.5 | 2PCho, 4Hex, 4Hep, Kdo-P, lipid A-OH-PEtn | |||
| 1,081.7 | 1,622.8 | 3,247.7 | 3,245.6 | 1.0 | 2PCho, 4Hex, 4Hep, Kdo-P-PEtn, lipid A-OH-PEtn | |||
Average mass units were used for calculation of molecular weights based on the following proposed composition: lipid A, 952.00; Hex, 162.15; Hep, 192.17; Kdo, 220.18; Kdo-P, 300.13; PEtn, 123.05; PCho, 165.05.
FIG. 2.
Negative-ion CE-ES-MS of the LPS-OH from AL836 (hptA) (A), AL690 (hptB) (B), and AL721 (kdkA) (C), showing the relative abundances of truncated and full-length LPS. The inset within each panel shows a carbohydrate silver stain of proteinase K-treated whole-cell lysates with parent AL435 (left lane) and the relevant LPS mutant (right lane).
Virulence of the mutants expressing modified LPS forms.
To determine if any of the mutants were capable of causing fowl cholera, virulence assays were performed. All birds except those injected with the hptA mutant succumbed to fowl cholera infection and were euthanatized (Table 4). Blood samples taken from all birds were plated onto BHI to determine levels of bacteremia. Blood samples taken from birds injected with VP161, AL435, AL690, and AL721 showed greater than 104 CFU/ml, but no bacteria were recovered from the blood of birds injected with the hptA mutant (AL836). Similarly, muscle was sampled at the site of injection from birds in all groups, and all birds except those from the AL836 group were positive for P. multocida.
The hptA gene is located upstream of PM1301, which encodes a putative ABC transporter. To rule out the involvement of PM1301 in the observed phenotype, the mutant was complemented in trans with an intact copy of hptA and tested by competition assay. The complemented mutant (AL847) was able to grow in vivo with an average rCI value of 0.73, whereas neither the mutant with vector (AL846) nor the hptA mutant (AL704) could grow (rCI = 0.0). Therefore, HptA activity is essential for P. multocida virulence.
Structural analysis of LPS expressed by P. multocida during in vivo growth.
Based on negative-ion CE-ES-MS data obtained for the P. multocida strains VP161 (this study) and X73 (28), glycoform A is the dominant structure, comprising 70 to 80% of the total LPS produced by both strains. To determine if both glycoforms observed from in vitro-grown P. multocida were also expressed by the bacteria during growth in the host, chickens were infected with P. multocida strain X73. This serotype A:1 strain is closely related to VP161 and expresses very similar LPS structures (28), but it has the unique ability to grow to very high numbers (up to 5 × 1010 CFU/ml) in the blood of chickens, thus allowing analytical quantities of LPS to be isolated. Chickens were injected intramuscularly with P. multocida and the infection allowed to proceed until the birds showed late signs of fowl cholera infection. P. multocida was harvested from the blood and LPS isolated from the bacteria. Structural analysis of the LPS isolated from in vivo-grown bacteria showed that the type and amount of each LPS glycoform were the same as those observed from LPS isolated from in vitro-grown cells (data not shown) (28).
Sensitivity of LPS mutants to the antimicrobial peptide fowlicidin-1.
To determine if P. multocida LPS mutants were more susceptible than the parent strain to chicken antimicrobial peptides, we performed bactericidal assays with the peptide fowlicidin-1 and compared the sensitivities of the parent strain AL435, the heptosyltransferase mutants (AL690 and AL836), and the Kdo kinase mutant (AL721) (Fig. 3). Both of the heptosyltransferase mutants were more sensitive to the effects of fowlicidin-1 than either parent strain or the Kdo kinase mutant. However, there was no significant difference in sensitivity between the two heptosyltransferase mutants, AL690 and AL836.
FIG. 3.
Sensitivities of the P. multocida parent strain (AL435) and of LPS mutants AL836 (hptA), AL690 (hptB), and AL721 (kdkA) to the action of fowlicidin-1. Bacterial survival was determined by direct colony counts after incubation with various concentrations of synthetic fowlicidin-1 for 1 h at 37°C. Numbers are the mean percent survival for three replicates, and error bars are ±1 standard deviation.
DISCUSSION
The P. multocida fowl cholera-causing isolates VP161 and X73, when propagated in vitro in BHI, express two LPS glycoforms containing either a single Kdo residue, which we have designated glycoform A, or two Kdo residues, which we have designated glycoform B (28, 29). The Kdo1 residue in glycoform A has a phosphate residue at the 4 position to which, in the majority of cases, a PEtn residue is attached. Glycoform A also has a glucose residue attached to the 6 position of the first heptose moiety (Fig. 1). In contrast, glycoform B contains two Kdo residues that lack the phosphate or PEtn residues and does not have the second glucose residue (Fig. 1). The expression of two very different Kdo region structures while the same oligosaccharide extension is retained is unusual. Other bacteria that express LPS lacking O polysaccharide include Campylobacter jejuni, H. influenzae, Neisseria gonorrhoeae, and Neisseria meningitidis. Typically, changes in structure occur in the outer core oligosaccharide extension and branch chains which lead to diversity in the LPS structure (reviewed in reference 24). Changes in LPS structure in these bacteria are often the result of phase variation caused by slipped-strand mispairing of homopolymeric tracts or multiple tandem repeats in DNA that results in the switching on or off of key LPS genes (16, 33). However, we have examined the annotated Pm70 genome and the sequenced LPS regions from a number of other P. multocida strains and have found no evidence of any genes that are subject to this type of phase variation.
Bioinformatic analysis of the genome of P. multocida strain Pm70 revealed two encoded proteins with similarity to known heptosyl-1-transferases. One of these, HptB (Pm1843 in Pm70), shares identity with WaaC, a transferase required for the addition of the first heptose residue to the lipid A-Kdo-Kdo acceptor molecule in bacteria belonging to the Enterobacteriaceae family (24, 27). The second transferase, HptA (annotated in Pm70 as OpsX), is similar to OpsX in H. influenzae, a novel type of heptosyltransferase that transfers the first heptose residue onto a Kdo-lipid A but only if the Kdo residue is phosphorylated (11). A search of all available bacterial genomes at NCBI revealed that only Mannheimia haemolytica, Mannheimia succiniciproducens, and Actinobacillus pleuropneumoniae harbor the genes for both types of heptosyltransferase I; these species are also members of the Pasteurellaceae family. Of these, only M. haemolytica has been shown to express more than one inner core LPS glycoform (19). The LPS structures of A. pleuropneumoniae strains examined so far produce only one inner core glycoform containing a single Kdo residue (31), while the composition of LPS expressed by M. succiniciproducens is unknown.
In addition to the identification of HptA and HptB as the predicted heptosyl-1-transferases, the gene PM1303 was also identified as encoding the putative Kdo kinase (KdkA) required for phosphorylation of the Kdo1 residue in P. multocida VP161 (Fig. 1). To confirm the roles of HptA, HptB, and KdkA in LPS biosynthesis, we constructed mutants with mutations in each of the identified genes in strain VP161. Inactivation of hptA produced a mutant strain (AL836) that expressed glycoform B, but approximately 37% of the LPS consisted of the highly truncated species lipid A-Kdo1-P (Table 3; Fig. 2A). This is consistent with hptA encoding the transferase required for the addition of the first heptose residue to the lipid A-Kdo1-P core. The function of HptA was confirmed with structural data on the complemented mutant, AL847, which showed restoration of full-length LPS for both glycoforms.
Inactivation of hptB produced a mutant strain (AL690) that expressed both full-length glycoform A and a truncated LPS species, estimated at 15% of the total product, consisting of lipid A-Kdo1-Kdo2 (Table 3; Fig. 2B). Complementation of this mutant with a functional hptB gene restored expression of the entire glycoform B. These data support the hypothesis that hptB encodes the heptosyltransferase required for the addition of the first heptose to the lipid A-Kdo1-Kdo2 core.
Thus, HptB and HptA are both heptosyl-1-transferases, but each transferase exclusively recognizes different acceptor molecules, with HptA transferring heptose to lipid A-Kdo1-P and HptB transferring heptose to lipid A-Kdo1-Kdo2. Virulence data showed that the hptB mutant, AL690, was still capable of causing lethal infection in chickens (Table 4). However, the hptA mutant, AL836, was fully attenuated. These data suggested that either glycoform A was critical for growth in vivo or AL836 was unable to survive in the host because of the amount of truncated LPS present on the surface; MS analysis indicated that the fully attenuated hptA mutant expressed more than twice the amount of truncated LPS as the virulent hptB mutant did (Fig. 2; Table 3).
To determine if expression of only glycoform B was sufficient for P. multocida growth in vivo, we constructed a Kdo kinase (kdkA) mutant. This kinase was predicted to add the phosphate residue to the first Kdo residue and thus should be essential for the synthesis of glycoform A. Structural analysis of the LPS expressed by the Kdo kinase mutant (AL721) indicated that it produced only glycoform B and not glycoform A, confirming the function of KdkA as the Kdo kinase (Table 3; Fig. 2C). Moreover, the Kdo kinase mutant expressed no truncated LPS, confirming that the lipid A-Kdo1 structure acts as an acceptor molecule both for the addition of a second Kdo molecule and for phosphorylation by KdkA.
We predict that P. multocida, like E. coli, expresses a bifunctional Kdo transferase (KdtA) that can attach both the first and second Kdo residues to lipid A (3). Because of the predicted bifunctional nature of KdtA, we hypothesize that in wild-type P. multocida the Kdo kinase (KdkA) competes with the Kdo transferase (KdtA) for access to the lipid A-Kdo1 acceptor molecule, and the predominance of glycoform A (70 to 80% of total LPS) suggests that phosphorylation of the first Kdo residue (by KdkA) occurs at a higher rate than the addition of the second Kdo residue by KdtA. Once phosphorylated, the lipid A-Kdo1 is not recognized by the Kdo transferase as an acceptor molecule. However, in the kdkA mutant, which lacks a functional Kdo kinase, the Kdo transferase has unhindered access to the lipid A-Kdo1, allowing all of the substrate to be converted to lipid A-Kdo1-Kdo2, the acceptor molecule recognized by the heptosyltransferase HptB. In order to confirm the function of KdtA in P. multocida, we attempted to construct a mutant with an insertion in the kdtA gene, but were unsuccessful. This is not surprising, as the lack of a functional Kdo transferase would most likely result in a nonviable mutant that lacks any LPS on its surface (24).
Although both the hptA mutant (AL836) and the kdkA mutant (AL721) expressed full-length LPS glycoform B, the kdkA mutant was fully virulent while the hptA mutant was fully attenuated (Table 4). The virulence data obtained for the kdkA mutant indicate that P. multocida VP161 can grow in vivo while expressing only glycoform B LPS. Therefore, the attenuation observed in the hptA mutant is probably due to the expression of a large amount of truncated LPS that makes the bacteria more vulnerable to host defense mechanisms than strains expressing only full-length LPS.
Antimicrobial peptides are a critical component of the innate immune systems of higher organisms (17) and are known to interact with LPS. It has been demonstrated in other bacterial species that truncation or alteration of the LPS alters resistance to antimicrobial peptide activity (20, 25). A number of chicken antimicrobial peptides have been characterized, including members of the β-defensin (15) and cathelicidin (36) families. In this study, the resistance of each of the P. multocida LPS mutants to the cathelicidin fowlicidin-1 was determined. We found that although the hptB mutant could cause disease and the hptA mutant could not, bactericidal assays showed both the heptosyltransferase mutants were more sensitive to fowlicidin-1 than the parent strain AL435. Therefore, the profound difference in the abilities of these mutants to cause disease in the host does not correlate directly with sensitivity to the lytic peptide fowlicidin-1. Other antimicrobial peptides, in addition to the complement membrane attack complex and Toll-like receptors on macrophages and heterophils, are also likely to play a role in host defense. Moreover, the Kdo kinase mutant (AL721) was only slightly sensitive to fowlicidin compared to the parent strain (AL435) but was significantly more resistant than the hptA mutant. These two mutants both express full-length glycoform B, but AL836 also expresses a large amount of truncated LPS, indicating that susceptibility of the LPS mutant strains to fowlicidin-1 is directly related to the expression of truncated LPS.
It is interesting that P. multocida can cause disease in the host with only the expression of glycoform B, yet both glycoforms are expressed simultaneously in vitro. To determine if the expression of both glycoforms also occurred in vivo, we isolated bacteria from the blood of chickens infected with P. multocida strain X73 and analyzed the in vivo-expressed LPS by MS. The ratio and type of LPS expressed at a late stage of infection in chickens were very similar to those of LPS expressed in vitro. It is not clear whether P. multocida can cause disease with expression of only glycoform A, as we have been unable to construct a mutant that produced full-length glycoform A and no other LPS species. Such a mutant would require the addition of only the first Kdo residue, and we predict that the Kdo transferase, KdtA, is bifunctional, transferring both the Kdo1 and Kdo2 molecules. Therefore, we have no mechanism to prevent the addition of Kdo2 while allowing the transfer of Kdo1. It is probable that neither glycoform A nor B is deleterious for the bacteria growing systemically within the host, as both are clearly expressed in this niche in vivo, but perhaps the expression and regulation of two different glycoforms allow for P. multocida survival in different environments within the host. We are currently investigating this possibility.
Acknowledgments
We thank Marietta John for excellent technical assistance. We also thank Perry Fleming for bacterial growth and Jacek Stupak for CE-MS.
This work was funded in part by grants from the Australian Research Council, Canberra, Australia, and the Australian Poultry Cooperative Research Centre.
Editor: V. J. DiRita
Footnotes
Published ahead of print on 21 May 2007.
REFERENCES
- 1.Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-Blast: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl (ed.). 1995. Current protocols in molecular biology, vol. 1. John Wiley & Sons, Inc., New York, NY.
- 3.Belunis, C. J., and C. R. Raetz. 1992. Biosynthesis of endotoxins. Purification and catalytic properties of 3-deoxy-d-manno-octulosonic acid transferase from Escherichia coli. J. Biol. Chem. 267:9988-9997. [PubMed] [Google Scholar]
- 4.Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Boyce, J., R. Y. C. Lo, I Wilkie, and B. Adler. 2004. Pasteurella and Mannheimia, p. 385-396. In C. L. Gyles, C. O. Thoen, J. F. Prescott, and G. Songer (ed.), Pathogenesis of bacterial infections of animals. Blackwell Publishing, Oxford, United Kingdom.
- 6.Boyce, J. D., I. Wilkie, M. Harper, M. L. Paustian, V. Kapur, and B. Adler. 2002. Genomic scale analysis of Pasteurella multocida gene expression during growth within the natural chicken host. Infect. Immun. 70:6871-6879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cardenas, M., A. R. Fernandez de Henestrosa, S. Campoy, A. M. Perez de Rozas, J. Barbe, I. Badiola, and M. Llagostera. 2001. Virulence of Pasteurella multocida recA mutants. Vet. Microbiol. 80:53-61. [DOI] [PubMed] [Google Scholar]
- 8.Coutinho, P. M., and B. Henrissat. 1999. Carbohydrate-active enzymes: an integrated database approach., p. 3-12. In H. G. Gilbert, G. Davies, B. Henrissat and B. Svensson (ed.), Recent advances in carbohydrate bioengineering. The Royal Society of Chemistry, Cambridge, United Kingdom.
- 9.Cox, A. D., H. Masoud, P. Thibault, J. R. Brisson, M. van der Zwan, M. B. Perry, and J. C. Richards. 2001. Structural analysis of the lipopolysaccharide from the nontypable Haemophilus influenzae strain SB 33. Eur. J. Biochem. 268:5278-5286. [DOI] [PubMed] [Google Scholar]
- 10.Gamian, A., M. Beurret, F. Michon, J. R. Brisson, and H. J. Jennings. 1992. Structure of the L2 lipopolysaccharide core oligosaccharides of Neisseria meningitidis. J. Biol. Chem. 267:922-925. [PubMed] [Google Scholar]
- 11.Gronow, S., W. Brabetz, B. Lindner, and H. Brade. 2005. OpsX from Haemophilus influenzae represents a novel type of heptosyltransferase I in lipopolysaccharide biosynthesis. J. Bacteriol. 187:6242-6247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Harper, M., J. D. Boyce, I. W. Wilkie, and B. Adler. 2003. Signature-tagged mutagenesis of Pasteurella multocida identifies mutants displaying differential virulence characteristics in mice and chickens. Infect. Immun. 71:5440-5446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Harper, M., A. D. Cox, F. St Michael, I. W. Wilkie, J. D. Boyce, and B. Adler. 2004. A heptosyltransferase mutant of Pasteurella multocida produces a truncated lipopolysaccharide structure and is attenuated in virulence. Infect. Immun. 72:3436-3443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Heddleston, K. L., and P. A. Rebers. 1972. Fowl cholera. Cross immunity induced in turkeys with formalin-killed in-vivo-propagated Pasteurella multocida. Avian Dis. 16:578-586. [PubMed] [Google Scholar]
- 15.Higgs, R., D. J. Lynn, S. Gaines, J. McMahon, J. Tierney, T. James, A. T. Lloyd, G. Mulcahy, and C. O'Farrelly. 2005. The synthetic form of a novel chicken beta-defensin identified in silico is predominantly active against intestinal pathogens. Immunogenetics 57:90-98. [DOI] [PubMed] [Google Scholar]
- 16.Jennings, M. P., D. W. Hood, I. R. Peak, M. Virji, and E. R. Moxon. 1995. Molecular analysis of a locus for the biosynthesis and phase-variable expression of the lacto-N-neotetraose terminal lipopolysaccharide structure in Neisseria meningitidis. Mol. Microbiol. 18:729-740. [DOI] [PubMed] [Google Scholar]
- 17.Jenssen, H., P. Hamill, and R. E. Hancock. 2006. Peptide antimicrobial agents. Clin. Microbiol. Rev. 19:491-511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685. [DOI] [PubMed] [Google Scholar]
- 19.Logan, S. M., W. Chen, A. Aubry, M. A. Gidney, S. Lacelle, F. St. Michael, R. Kuolee, M. Higgins, S. Neufeld, and A. D. Cox. 2006. Production of a d-glycero-d-manno-heptosyltransferase mutant of Mannheimia haemolytica displaying a veterinary pathogen specific conserved LPS structure; development and functionality of antibodies to this LPS structure. Vet. Microbiol. 116:175-186. [DOI] [PubMed] [Google Scholar]
- 20.Loutet, S. A., R. S. Flannagan, C. Kooi, P. A. Sokol, and M. A. Valvano. 2006. A complete lipopolysaccharide inner core oligosaccharide is required for resistance of Burkholderia cenocepacia to antimicrobial peptides and bacterial survival in vivo. J. Bacteriol. 188:2073-2080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Marchler-Bauer, A., and S. H. Bryant. 2004. CD-Search: protein domain annotations on the fly. Nucleic Acids Res. 32:W327-331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Michon, F., M. Beurret, A. Gamian, J. R. Brisson, and H. J. Jennings. 1990. Structure of the L5 lipopolysaccharide core oligosaccharides of Neisseria meningitidis. J. Biol. Chem. 265:7243-7247. [PubMed] [Google Scholar]
- 23.Miller, V. L., and J. J. Mekalanos. 1988. A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. J. Bacteriol. 170:2575-2583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Raetz, C. R., and C. Whitfield. 2002. Lipopolysaccharide endotoxins. Annu. Rev. Biochem. 71:635-700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ramjeet, M., V. Deslandes, F. St Michael, A. D. Cox, M. Kobisch, M. Gottschalk, and M. Jacques. 2005. Truncation of the lipopolysaccharide outer core affects susceptibility to antimicrobial peptides and virulence of Actinobacillus pleuropneumoniae serotype 1. J. Biol. Chem. 280:39104-39114. [DOI] [PubMed] [Google Scholar]
- 26.Schweda, E. K., A. C. Sundstrom, L. M. Eriksson, J. A. Jonasson, and A. A. Lindberg. 1994. Structural studies of the cell envelope lipopolysaccharides from Haemophilus ducreyi strains ITM 2665 and ITM 4747. J. Biol. Chem. 269:12040-12048. [PubMed] [Google Scholar]
- 27.Sirisena, D. M., K. A. Brozek, P. R. MacLachlan, K. E. Sanderson, and C. R. Raetz. 1992. The rfaC gene of Salmonella typhimurium. Cloning, sequencing, and enzymatic function in heptose transfer to lipopolysaccharide. J. Biol. Chem. 267:18874-18884. [PubMed] [Google Scholar]
- 28.St. Michael, F., J. Li, and A. D. Cox. 2005. Structural analysis of the core oligosaccharide from Pasteurella multocida strain X73. Carbohydr. Res. 340:1253-1257. [DOI] [PubMed] [Google Scholar]
- 29.St. Michael, F., J. Li, E. Vinogradov, S. Larocque, M. Harper, and A. D. Cox. 2005. Structural analysis of the lipopolysaccharide of Pasteurella multocida strain VP161: identification of both Kdo-P and Kdo-Kdo species in the lipopolysaccharide. Carbohydr. Res. 340:59-68. [DOI] [PubMed] [Google Scholar]
- 30.St. Michael, F., E. Vinogradov, J. Li, and A. D. Cox. 2005. Structural analysis of the lipopolysaccharide from Pasteurella multocida genome strain Pm70 and identification of the putative lipopolysaccharide glycosyltransferases. Glycobiology 15:323-333. [DOI] [PubMed] [Google Scholar]
- 31.St. Michael, F. S., J. R. Brisson, S. Larocque, M. Monteiro, J. Li, M. Jacques, M. B. Perry, and A. D. Cox. 2004. Structural analysis of the lipopolysaccharide derived core oligosaccharides of Actinobacillus pleuropneumoniae serotypes 1, 2, 5a and the genome strain 5b. Carbohydr. Res. 339:1973-1984. [DOI] [PubMed] [Google Scholar]
- 32.Tsai, C. M., and C. E. Frasch. 1982. A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119:115-119. [DOI] [PubMed] [Google Scholar]
- 33.Weiser, J. N., J. M. Love, and E. R. Moxon. 1989. The molecular mechanism of phase variation of H. influenzae lipopolysaccharide. Cell 59:657-665. [DOI] [PubMed] [Google Scholar]
- 34.White, K. A., S. Lin, R. J. Cotter, and C. R. Raetz. 1999. A Haemophilus influenzae gene that encodes a membrane bound 3-deoxy-d-manno-octulosonic acid (Kdo) kinase. Possible involvement of Kdo phosphorylation in bacterial virulence. J. Biol. Chem. 274:31391-31400. [DOI] [PubMed] [Google Scholar]
- 35.Wilkie, I. W., S. E. Grimes, D. O'Boyle, and A. J. Frost. 2000. The virulence and protective efficacy for chickens of Pasteurella multocida administered by different routes. Vet. Microbiol. 72:57-68. [DOI] [PubMed] [Google Scholar]
- 36.Xiao, Y., Y. Cai, Y. R. Bommineni, S. C. Fernando, O. Prakash, S. E. Gilliland, and G. Zhang. 2006. Identification and functional characterization of three chicken cathelicidins with potent antimicrobial activity. J. Biol. Chem. 281:2858-2867. [DOI] [PubMed] [Google Scholar]