Search tips
Search criteria 


Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. 2009 November; 191(22): 6950–6959.
Published online 2009 September 18. doi:  10.1128/JB.00787-09
PMCID: PMC2772478

Structural and Genetic Basis for the Serological Differentiation of Pasteurella multocida Heddleston Serotypes 2 and 5 [down-pointing small open triangle]


Pasteurella multocida is classified into 16 serotypes according to the Heddleston typing scheme. As part of a comprehensive study to define the structural and genetic basis of this scheme, we have determined the structure of the lipopolysaccharide (LPS) produced by P. multocida strains M1404 (B:2) and P1702 (E:5), the type strains for serotypes 2 and 5, respectively. The only difference between the LPS structures made by these two strains was the absence of a phosphoethanolamine (PEtn) moiety at the 3 position of the second heptose (Hep II) in M1404. Analysis of the lpt-3 gene, required for the addition of this PEtn residue, revealed that the gene was intact in P1702 but contained a nonsense mutation in M1404. Expression of an intact copy of lpt-3 in M1404 resulted in the attachment of a PEtn residue to the 3 position of the Hep II residue, generating an LPS structure identical to that produced by P1702. We identified and characterized each of the glycosyltransferase genes required for assembly of the serotype 2 and 5 LPS outer core. Monoclonal antibodies raised against serotype 2 LPS recognized the serotype 2/5-specific outer core LPS structure, but recognition of this structure was inhibited by the PEtn residue on Hep II. These data indicate that the serological classification of strains into Heddleston serotypes 2 and 5 is dependent on the presence or absence of PEtn on Hep II.

Pasteurella multocida is a gram-negative pathogen that causes serious diseases in animals and humans. It is the causative agent of fowl cholera (7), hemorrhagic septicemia in cattle (9), atrophic rhinitis in pigs (6), and dog and cat bite infections in humans (28).

P. multocida isolates may be grouped serologically based on capsular antigens into five serogroups—A, B, D, E, and F—using a passive hemagglutination test with erythrocytes sensitized with capsular antigen. Structural information is available for the capsular polysaccharides synthesized by serogroups A (hyaluronic acid) (22), D (heparin) (10), and F (chondroitin) (10). The genes involved in biosynthesis of the capsules have been identified for all five serogroups (27), and capsule is a critical virulence factor for serogroups A (8) and B (3).

Lipopolysaccharide (LPS) is also an important virulence factor in P. multocida (13) and can be used for the identification of strains, with two main somatic typing systems reported (14, 17). The Namioka system is based on a tube agglutination test and is able to recognize 11 serotypes (17), whereas the Heddleston system uses a gel diffusion precipitation test and can recognize 16 serotypes; the Heddleston system is currently the preferred method (14). Current classification of P. multocida strains combines capsular typing with Heddleston somatic typing. Strains are given a designation in which the first letter indicates the capsular group and the number designates the Heddleston LPS serotype (e.g., A:1 indicates a strain that is capsular group A and LPS serotype 1). LPS produced by each of the 16 Heddleston serotype strains has been examined previously for sugar content and reactivity with LPS antisera (21). The LPS isolated from serotype 2 and 5 strains was virtually identical in sodium dodecyl sulfate-polyacrylamide gel electrophoresis migration profile (19), sugar composition, and serological reactivity with anti-LPS antibodies (21). Interestingly, serotypes 2 and 5 were the only serotypes found to elaborate two isomers of heptose in their LPS, namely l-glycero-d-manno-heptose (ld-Hep) and d-glycero-d-manno-heptose (dd-Hep) (21). The aims of this study were to determine whether the LPS molecules made by these two serotypes were structurally distinct and to compare the LPS structures with those previously determined for P. multocida serotypes 1 and 3 (24-26). Furthermore, we identified the transferase genes responsible for the assembly of the outer core LPS structure in each of these strains and characterized the function of each glycosyltransferase.


Bacterial strains and growth conditions.

The strains and plasmids used in this study are shown in Table Table1.1. P. multocida strains were grown at 37°C with aeration in brain heart infusion (BHI) broth (Oxoid). For culture on solid media, 1.5% agar was added to either BHI or nutrient broth (Oxoid) containing 0.3% yeast extract or strains were grown on chocolate agar. For structural studies of LPS, strains were ultimately grown in 24 liters of BHI in a 28-liter NBS fermentor as described previously (4).

Strains, plasmids, and primers used in this study

DNA manipulations.

P. multocida genomic DNA was prepared by the cetyltrimethyl ammonium bromide method (2). PCR amplification of DNA was performed using Taq DNA polymerase or the Expand high-fidelity PCR system (Roche), and PCR products were purified using the QIAquick PCR purification kit (Qiagen). Restriction digests and ligations were performed according to the manufacturers' instructions using enzymes obtained from NEB or Roche. The genes required for the synthesis of the serotype 2 or 5 outer core LPS structure were identified between the conserved flanking genes priA and fpg. The nucleotide sequence of this region in strains M1404 and P1702 was determined by double-stranded sequencing of amplified PCR fragments using the Applied Biosystems 3730S genetic analyzer. Sequencing chromatograms were analyzed and the LPS loci assembled using Sequencher version 3.1.1 (GenCodes, Ann Arbor, MI) or Vector NTI advance version 10 (Invitrogen, Carlsbad, CA). DNA and protein database comparisons were made using the BLAST program (1).

Recombinant expression of LPS transferases in P. multocida.

To characterize the function of each of the LPS transferases, we cloned the genes encoding the candidate transferases into the P. multocida expression vector pAL99 (Table (Table1).1). For construction of all recombinant plasmids except pAL588, the appropriate region was amplified by PCR from genomic DNA isolated from M1404 (for amplification of nctA and hptF) or X-73 (for lpt-3) using oligonucleotides with engineered restriction sites and cloned into pAL99 to generate plasmids pAL284, pAL580, and pAL581 (Table (Table1).1). For construction of pAL588, the amplified DNA encoding gatD was inserted into the SalI site of pAL581, downstream of nctA and hptF. Therefore, plasmid pAL588 contains all three genes, nctA, hptF, and gatD, under the control of the constitutive tpi promoter present in pAL99. The nucleotide sequence of each of the recombinant plasmids was determined to check fidelity of the cloned genes, and then each plasmid was transformed into the appropriate P. multocida strain. For complementation experiments using lpt-3, plasmids pAL99 or pAL284 were used to transform P. multocida M1404. For heterologous expression experiments, the plasmids pAL99, pAL580, pAL581, and pAL588 were separately used to transform the VP161 gatA mutant (AL725) (Table (Table1),1), which has a truncated LPS structure terminating at the Hep IV residue (4).

Structural analysis of LPS.

O-deacylated LPS (LPS-OH) and core oligosaccharide were prepared from purified LPS as described previously (4). Sugars were determined as their alditol acetate derivatives, and linkage analysis was determined following methylation analysis by gas-liquid chromatography mass spectrometry (MS) as described previously (25). Capillary electrophoresis electrospray MS (CE-ES-MS) and nuclear magnetic resonance (NMR) spectroscopy experiments were performed as described previously (25).

Analysis of LPS by carbohydrate silver stain and immunoblotting.

Proteinase K-treated whole-cell lysates (WCLs) were analyzed by polyacrylamide gel electrophoresis, and the gels were either stained for carbohydrate or transferred to nylon membrane by Western blotting as described previously (12). For immunoblots, the monoclonal antibodies (MAbs) raised against M1404 LPS (18) were used at a 1/10,000 dilution followed by sheep anti-mouse immunoglobulin-horseradish peroxidase conjugate (Millipore Corporation, Billerica, MA) at a 1/1,000 dilution.

Nucleotide sequence accession numbers.

The GenBank accession numbers for the nucleotide sequences of the LPS outer core loci in strains M1404 and P1702 are GQ444331 and GQ444330, respectively. The accession numbers for nucleotide sequences of lpt-3 from strains X-73, M1404, and P1702 are GQ444334, GQ444332 and GQ444333, respectively.


Structural analyses.

Sugar analyses of the LPS derived from P. multocida Heddleston type strains M1404 (serotype 2) and P1702 (serotype 5) identified the presence of glucose (Glc), galactose (Gal), l-glycero-d-manno-heptose (ld-Hep), d-glycero-d-manno-heptose (dd-Hep), and N-acetyl-glucosamine (GlcNAc) in the approximate ratio of 2:1:4:1:1. These data are consistent with previously published analyses indicating that LPS from serotype 2 and 5 strains contains the isomer dd-Hep (21).

The LPS-OH was analyzed by CE-ES-MS (Table (Table2).2). For M1404 (serotype 2), major amounts of the triply charged ions at m/z 966.63− and 1,007.83− were observed as well as a minor amount of triply charged ion at 912.73−. For P1702 (serotype 5), large amounts of triply charged ions were observed at m/z 967.13− and 1,008.33− with less significant amounts of triply charged ions observed at m/z 954.03− and 1,049.43−. The ions observed at m/z 966.63− and 967.13− in samples derived from strains M1404 and P1702, respectively, correspond to a glycoform composition of HexNAc, 3Hex, 5Hep, Kdo-P, and lipid A-OH (Table (Table2).2). The observed ions at m/z 1,007.83− and 1,008.33− (serotypes 2 and 5, respectively) correspond to the above composition, but with the addition of a single phosphoethanolamine (PEtn) residue. However, the ion at m/z 1,049.43−, observed only in the sample derived from P1702 (serotype 5), corresponds to HexNAc, 3Hex, 5Hep, Kdo-P, and lipid A-OH with the addition of two PEtn residues. Small amounts of ions corresponding to the loss of a single hexose residue were also observed for both M1404 and P1702 (m/z 912.73− and 954.03−, respectively).

Negative-ion CE-ES-MS data and proposed compositions of LPS-OH and core oligosaccharide from P. multocida Heddleston serotypes 2 and 5

MS analysis of the core oligosaccharide confirmed that the major glycoform produced by strain M1404 is HexNAc, 3Hex, 5Hep, and Kdo (m/z 934.22−), whereas P1702 made significant amounts of two glycoforms: one corresponding to the glycoform produced by M1404 (m/z 934.42−) and one with an additional PEtn residue in the core oligosaccharide (m/z 996.02−) (Table (Table2).2). Taken together, the LPS-OH and core oligosaccharide data clearly indicate that only P1702 expresses an LPS glycoform with a PEtn residue attached to the oligosaccharide component of the LPS.

In order to determine the linkage pattern of the LPS molecule, methylation analysis was performed on the core oligosaccharide isolated from strain M1404. These analyses revealed the presence of terminal Glc, terminal Gal, 6-substituted Glc, terminal ld-Hep, 2-substituted ld-Hep, 2-substituted dd-Hep, 6-substituted ld-Hep, and 4-substituted HexNAc in approximately equimolar amounts and smaller amounts of 3,4,6-trisubstituted ld-Hep and 3,4-disubstituted ld-Hep residues. Similar results were obtained from core oligosaccharide from P1702 (serotype 5), with slightly smaller amounts of the 2-substituted ld-Hep residue being observed. The identification of a 2-substituted dd-Hep residue corroborated and extended the sugar analysis data. Taken together, the sugar, methylation, and MS analyses suggested that the type strains belonging to serotypes 2 and 5 differed only with the nonstoichiometric addition of a PEtn residue to the core oligosaccharide of P1702 (serotype 5) (Fig. (Fig.11).

FIG. 1.
Schematic representation of the LPS structure produced by the Heddleston M1404 and P1702 strains (A), the genetic organization of the genes encoding the transferases responsible for synthesis of the outer core section of the LPS molecules (B), and the ...

In order to elucidate the exact locations and linkage patterns of the oligosaccharide, NMR studies were performed. 1H-NMR spectra of core oligosaccharide from P1702 gave an indication of the presence of PEtn by virtue of signals at ~3.30 ppm consistent with proton resonances for the ethanolamine moiety adjacent to the amino functionality, while in similar spectra of core oligosaccharide from M1404, these signals were absent. The inner core glycose linkages for both M1404 and P1702 LPS were identical to the inner core structures published for P. multocida strains Pm70, VP161, and X-73 (24-26). The assignments of 1H resonances and linkages of these inner core residues of the oligosaccharide samples were achieved by homonuclear correlation spectroscopy (COSY), total COSY (TOCSY), and nuclear Overhauser effect spectroscopy (NOESY) experiments (Table (Table3),3), with reference to the structurally related core oligosaccharide from P. multocida strains Pm70, VP161, and X-73. Since the assignments for M1404 and P1702 were virtually identical, only the assignments for M1404 are detailed. In addition to the residues of the conserved inner core structure, additional heptose, N-acetyl-glucosamine, and galactose residues were identified from spin systems arising from their anomeric 1H resonances at 5.20, 4.55, and 4.47 ppm, respectively (Fig. (Fig.2).2). An inter-NOE connectivity from the anomeric proton of the heptose residue at 5.20 ppm to a resonance at 4.20 ppm was identified (Fig. (Fig.3).3). This resonance was subsequently assigned as the proton at the 6 position of the Hep IV residue by virtue of 13C-1H heteronuclear single quantum coherence (HSQC) and 13C-1H HSQC-TOCSY spectroscopy experiments (data not shown). This assignment is also consistent with the methylation analysis data which identified a 6-linked ld-Hep residue. The resonance at 4.55 ppm was identified as having the gluco configuration by virtue of the characteristic spin system in a TOCSY experiment (Fig. (Fig.2)2) and as an amino sugar by virtue of the 13C chemical shift for the 13C resonance of the C-2 carbon correlating to the H-2 1H-resonance, being at 56.1 ppm, consistent with a nitrogen-substituted carbon atom. An inter-NOE connectivity from the anomeric proton of the N-acetyl-glucosamine residue to a resonance at 4.17 ppm was identified (Fig. (Fig.3).3). This resonance was subsequently assigned as the proton at the 2 position of the Hep V residue. Additionally, interanomeric NOE connectivities consistent with a 2-substituted heptose residue were also observed between the GlcNAc and Hep V anomeric protons (Fig. (Fig.3).3). The resonance at 4.47 ppm was identified as having the galacto configuration by virtue of the characteristic spin system in a TOCSY experiment (Fig. (Fig.2),2), and an inter-NOE connectivity from the anomeric proton to a resonance at 3.73 ppm was subsequently identified as the H-4 1H-resonance of the N-acetylglucosamine residue (Fig. (Fig.3).3). The NOE connectivities for the outer core residues were consistent with the linkages identified in the methylation analysis and confirmed the linkage pattern of the outer core oligosaccharide. A 13C-1H HSQC experiment was performed in order to determine the 13C chemical shifts, and by comparison to previously published data, these were assigned as detailed in Table Table3.3. NMR analysis of the core oligosaccharide from P1702 was virtually identical to that described for M1404, except for the region of the TOCSY spectra for the spin systems from the Hep II residues, which revealed an additional resonance with an up-field shift for the H-1 and H-2 resonances for the Hep II residue in M1404 when compared to the same region for P1702 (Fig. (Fig.4).4). Additionally, when a longer mixing time (150 ms) was adopted in the TOCSY experiment, it was possible to access the H-3 1H-resonance at 4.39 ppm, consistent with the presence of PEtn at the 3 position of the Hep II residue in P1702. MS analysis and 1H-NMR data indicated that there was a PEtn residue in the core oligosaccharide of P1702, and this was corroborated with the assignment of characteristic signals for a PEtn residue in the 13C-1H HSQC experiment. Confirmation of the 3 position of Hep II as the location of PEtn substitution of P1702 was obtained from 31P-1H-HSQC and 31P-1H-HSQC-TOCSY experiments on the oligosaccharide sample. The HSQC experiment identified a cross-peak from the phosphorus signal to the proton resonance at 4.39 ppm which had been assigned to the 3 position of the Hep II residue, and this was confirmed and extended in the HSQC-TOCSY experiment, which revealed the H-2 and H-1 proton resonances of Hep II at 4.30 and 5.87 ppm, respectively (data not shown). Therefore, combined structural analyses identified that the only difference between the LPS produced by the serotype 2 and 5 type strains was the nonstoichiometric addition of PEtn at the 3 position of Hep II in P1702. Therefore, strains from these two Heddleston serotypes express structurally very similar, but nevertheless distinct, LPS molecules.

FIG. 2.
Anomeric region of the 1H-NMR two-dimensional TOCSY spectra (excluding Hep II) of the purified core oligosaccharide from the Heddleston serotype 2 strain M1404. Residues and resonances are as indicated. The spectrum was obtained at 25°C and is ...
FIG. 3.
Anomeric region of the 1H-NMR two-dimensional NOESY spectra (excluding Hep II) of the purified core oligosaccharide from the Heddleston serotype 5 strain P1702. Residues and inter-NOE connectivities for residues in the extension from Hep I are as indicated. ...
FIG. 4.
The region of the 1H-NMR two-dimensional TOCSY spectra for the spin systems from the Hep II residues of the purified core oligosaccharide from the Heddleston serotype 5 strain P1702 (a) and the Heddleston serotype 2 strain M1404 (b). Residues and resonances ...
1H- and 13C-NMR chemical shiftsa for the core oligosaccharide from P. multocida Heddleston serotypes 2 and 5

Genetic analyses.

In previous studies, we identified and characterized the glycosyltransferases required for the addition of each sugar to the Heddleston 1 VP161 LPS (4, 11, 13). In the present study, we have shown that the Heddleston 2 and 5 type strains (M1404 and P1702, respectively) express an inner core structure from Kdo to Hep IV that is the same as previously reported for the P. multocida strains Pm70 (serotype 3), X-73 (serotype 1), and VP161 (serotype 1), with the exception of the presence or absence of a PEtn residue on the 3 position of the HepII residue (Fig. (Fig.1)1) (24-26).

The genes required for assembly of the outer core section of the LPS produced by the P. multocida strains VP161 (serotype 1) and Pm70 (serotype 3) are located within a single region of the genome between the conserved priA and fpg genes, which are not involved in LPS biosynthesis (4). Determination of the nucleotide sequence of the corresponding region in strains M1404 (serotype 2) and P1702 (serotype 5) identified four open reading frames encoding putative glycosyltransferases, designated gatD, nctA, hptF, and hptE (Fig. (Fig.1).1). The nucleotide sequence of this region was highly conserved between the two strains (>95% identity).

The protein encoded by hptE showed high identity (91%) to the VP161 heptosyltransferase HptE (4), and we predict that this protein is an HptE ortholog and that the hptE gene encodes the heptosyltransferase required for the addition of ld-Hep to the 6 position of the β-Glc. The gene located upstream of hptE, designated hptF, encodes a protein with significant identity to heptosyltransferases within the glycosyltransferase family 9. HptF is therefore predicted to be the heptosyltransferase responsible for the addition of the dd-Hep residue to the 6 position of the Hep IV in strains belonging to serotypes 2 and 5. The gene nctA encodes a protein that shares identity with transferases belonging to the glycosyltransferase family 2, including 29% identity to the galactosyltransferases WbnI from Shigella dysenteriae and WbqE from Escherichia coli. Glycosyltransferases belonging to family 2 donate UDP-glucose, UDP-N-acetyl-galactosamine, GDP-mannose, or CDP-abequose, to a range of substrates. Based on similarity of NctA to transferases within the glycosyltransferase family 2, we predicted that it was required for the addition of N-acetylglucosamine to the 2 position of the dd-Hep residue in the outer core of the LPS made by serotype 2 and 5 strains. The last gene in the region, designated gatD, encodes a protein belonging to the family 25 group of glycosyltransferases and shares 53% and 47% identity, respectively, with the predicted galactosyltransferase PM1141 in P. multocida strain Pm70 and Lic2B in Haemophilus influenzae. Therefore, we predicted that it was required for the transfer of the terminal Gal residue onto the 4 position of the GlcNAc in the Heddleston 2 and 5 LPS structure.

The LPS produced by the Heddleston serotype 1 and 2 strains (VP161 and M1404, respectively) is identical up to and including the Hep IV residue (Fig. (Fig.1).1). To determine the function of each of the Heddleston 2 LPS transferases, we cloned each gene into a P. multocida expression vector in an order predicted to be required for the progressive assembly of the outer core (Table (Table1).1). Thus, pAL580 contained only hptF, pAL581 contained both nctA and hptF, and pAL588 contained nctA, hptF, and gatD. Each of these constructs, as well as empty vector pAL99, was transformed into the P. multocida VP161 gatA mutant (AL725) (Table (Table1)1) that expresses an LPS molecule truncated at Hep IV (Fig. (Fig.1).1). Structural analyses of the LPS made by AL725 and AL807 [AL725(pAL99)] confirmed that these strains express LPS that terminates at Hep IV (Table (Table4).4). However, AL1086, harboring hptF on plasmid pAL580, assembled LPS with an additional heptose residue. Similarly, AL1088, which expressed both HptF and NctA from plasmid pAL581, produced LPS with additional heptose and N-acetyl-hexosamine residues. The full-length serotype 2 LPS molecule was successfully assembled in AL1131, containing plasmid pAL588 and expressing all three serotype 2/5-specific transferases, HptF, NctA, and GatD (Table (Table4).4). Therefore, the heterologous expression of these serotype 2/5 transferases in AL725 confirmed that HptF is the α-1,6-dd-heptosyltransferase, NctA is the β-1,2 N-acetylglucosaminyltransferase, and GatD is the terminal β-1,4 galactosyltransferase. Interestingly, MS analysis also revealed the presence of phosphocholine (PCho) in the AL725 strain containing the full serotype 2 outer core extension (AL1131), and tandem MS (MS/MS) analysis confirmed that the PCho addition was to the terminal hexose residue of the core OS (Fig. (Fig.5).5). This was not unexpected for heterologous expression of LPS in strain AL725 (VP161 gatA mutant) as we have shown previously that in the parent strain VP161 the terminal galactose residues on the LPS are decorated with PCho (12). In AL725 expressing the full-length serotype 2 LPS structure (AL1131), the terminal galactose residue can also act as an acceptor for the endogenous PCho transferase (Fig. (Fig.55).

FIG. 5.
CE-ES-MS/MS spectrum of core oligosaccharide from P. multocida strain AL1131. The positively charged ion at m/z 10172+ was selectively hit in an MS/MS experiment, revealing diagnostic ions for the Heddleston 2 outer core oligosaccharide of Hex-HexNAc-Hep. ...
Negative-ion CE-ES-MS data and proposed compositions of LPS-OH from the heterologous expression of serotype 2 glycosyltransferases in AL725 and PEtn transferase (Lpt-3) expression in M1404

Structural data clearly showed that the only difference between the LPS made by the serotype 2 and 5 type strains was the addition of a PEtn residue to the 3 position of Hep II to ~40% of the LPS molecules produced by P1702 (serotype 5). The gene encoding the PEtn transferase required for this addition was predicted to be lpt-3 (annotated as dcaA in the Pm70 genome), based on similarity to the gene lpt-3 required for the addition of PEtn to the second heptose residue in the inner core of Neisseria meningitidis LPS (16). The lpt-3 gene was intact in serotype 5 strain P1702, but it carried a mutation at nucleotide 282 in serotype 2 strain M1404 resulting in a stop codon (data not shown). To confirm the role of lpt-3 in the addition of the PEtn residue to the second heptose, we complemented M1404 with an intact copy of the P. multocida lpt-3 gene provided in trans on the expression plasmid pAL99 (strain AL1101) (Table (Table1).1). CE-MS analyses of LPS-OH derived from strain AL1101 revealed the presence of an additional PEtn residue compared to the vector control strain AL1100 (Table (Table44).

To determine the immunological significance of the PEtn residue attached to the 3 position of Hep II of the serotype 5 strain P1702, we analyzed the specificity of a panel of MAbs raised against M1404 LPS (Fig. (Fig.6)6) (18). In immunoblotting experiments, two MAbs, T1C6 and T6B2 (18), were identified that reacted strongly with the LPS in M1404 (serotype 2) WCLs but only weakly with P1702 (serotype 5) WCLs and not at all with VP161 (serotype 1) WCLs (Fig. (Fig.6).6). Furthermore, while both MAbs reacted strongly with M1404, this reactivity was significantly diminished when the same MAbs were incubated with M1404 expressing PEtn on the Hep II of the LPS (AL1101) and therefore was identical in structure to the PEtn-decorated LPS glycoform produced by the serotype 5 strain P1702 (Fig. (Fig.6).6). Thus, the addition of PEtn to Hep II of the LPS structure, as observed in P1702 and AL1101, significantly inhibits the binding of these MAbs.

FIG. 6.
Immunoblot analysis of the P. multocida strains examined in this study using MAbs T1C6 (A) and T6B2 (B). WCLs were derived from P. multocida strain VP161 (serotype A:1); the VP161 gatA mutant (AL725); AL725 containing pAL99 (AL807); AL725 expressing HptF ...

We also tested the reactivity of the MAbs T1C6 and T6B2 against WCLs derived from the P. multocida strain VP161 gatA mutant (AL725) and its recombinant derivatives expressing the M1404 LPS glycosyltransferases HptF, NctA, and GatD. Neither MAb reacted with WCL from AL725 or AL725 expressing only HptF (AL1086). Therefore, both MAbs require an LPS molecule extended beyond Hep V for recognition. The MAb T1C6 reacted strongly with WCL from AL725 expressing two of the serotype 2 transferases, HptF and NctA (AL1088), indicating that T1C6 requires the addition of a GlcNAc residue for epitope recognition, but does not require the terminal Gal residue. In contrast, MAb T6B2 bound very weakly to WCL from AL725 expressing only two of the serotype 2 transferases but strongly to WCL of AL725 expressing HptF, NctA, and GatD (AL1131), indicating that MAb T6B2 binds optimally to a fully extended serotype 2 LPS structure.


In this study, we have determined the structures of the oligosaccharides derived from the LPS of the serotype 2 and 5 P. multocida type strains M1404 and P1702. We have characterized three novel glycosyltransferases required for the assembly of the LPS outer core in both M1404 and P1702. The LPS produced by each of these strains contains the same inner core structure previously identified in the LPS of other P. multocida strains as well as strains of Actinobacillus pleuropneumoniae (23) and Mannheimia haemolytica (5, 15). The presence of a PEtn moiety at the 3 position of the Hep II in ~40% of LPS glycoforms assembled by P1702, as evidenced by CE-ES-MS analysis of the core oligosaccharide (Table (Table2),2), is the only difference between the LPSs produced by the two type strains and as such is predicted to be responsible for their original serological distinction (14). Indeed, expression in strain AL1101 of serotype 2 LPS decorated with PEtn at the Hep II residue strongly inhibited reactivity of two MAbs raised against serotype 2 LPS, indicating that the presence or absence of PEtn is immunologically important. However, not all Heddleston typing sera can distinguish between these strains (P. Blackall, personal communication) and this reactivity is almost certainly due to their high structural similarity. Indeed, PEtn is not always present on the triheptose unit of P1702 (serotype 5) LPS, and in those instances where it is absent, the LPS is identical to that produced by serotype 2. The PEtn transferase responsible for the addition of PEtn to Hep II has been identified in this study as Lpt-3, an ortholog of DcaA in strain Pm70. P. multocida strain VP161 does not elaborate a PEtn at this position (25), and, like the Heddleston 2 strain analyzed in this study, has a nonsense mutation present within lpt-3 (our unpublished data).

The LPS structures made by the serotype 2 and 5 type strains M1404 and P1702 both contained the terminal trisaccharide β-Gal-(1-4)-β-GlcNAc-(1-2)-α-dd-Hep attached at the 6 position to the ld-Hep IV residue. These data confirmed that M1404 and P1702 express both ld- and dd-heptose isomers, as observed previously (21). Indeed, there are both ld- and dd-Hep residues in the outer core oligosaccharide, which is unusual. The LPS molecules produced by strains from P. multocida serotypes 1 and 3 have only ld-Hep isomers in the LPS (24-26), and the related veterinary pathogens A. pleuropneumoniae (23) and M. haemolytica (5, 15) have only dd-Hep isomers in the outer core but ld-Hep isomers in the inner core.

The activities of the predicted α1,6-heptosyltransferase (HptF), the β-1,2 N-acetylglucosaminyltransferase (NctA), and the terminal β-1,4 galactosyltransferase (GatD) were confirmed by heterologous expression in a P. multocida VP161 (serotype 1) gatA mutant (AL725). Furthermore, the immunological reactivity of two MAbs (raised against M1404 LPS) with the recombinant strains harboring the serotype 2 glycosyltransferases was assessed. Neither MAb recognized AL725 expressing LPS extended by just one serotype 2-specific sugar, HepV, but both reacted strongly with the recombinant AL725 that assembled LPS with the terminal trisaccharide β-Gal-(1-4)-β-GlcNAc-(1-2)-α-dd-Hep, identical to the LPS outer core produced by serotype 2 and 5 strains. Interestingly, only the T1C6 MAb reacted strongly with LPS extended by just β-GlcNAc-(1-2)-α-dd-Hep. Thus, each MAb recognizes different epitopes within the LPS outer core region; T6B2 recognizes an epitope that includes the terminal Gal residue while T1C6 recognizes an epitope that does not require the presence of the Gal residue. The binding of both MAbs was strongly inhibited by the presence of PEtn on the HepII residue, located in the triheptose inner core region of the LPS. These data indicate that the correct conformation of the epitopes recognized by the MAb requires both the involvement of the outer core and the serotype 2-specific triheptose unit or that the presence of PEtn on the triheptose unit alters the conformation of epitopes located only within the outer core region. Interestingly, the binding of each of the MAb to the fully extended serotype 2 structure appeared unaffected by the addition of PCho to the terminal galactose residue.

This study has established both the structural and genetic basis for the serological differentiation of Heddleston serotypes 2 and 5. We have shown that the LPS glycoforms produced by both type strains are identical, except for the nonstoichiometric addition of PEtn to the Hep II residue of the LPS assembled by the serotype 5 type strain, P1702. Thus, a proportion of the serotype 5 LPS molecules are identical to those made by the serotype 2 type strain, M1404. However, the addition of the PEtn residue to Hep II is clearly important for antibody recognition, as the presence of a PEtn residue in this position strongly inhibited binding of two MAbs raised against serotype 2 LPS. It would be of great value to produce a vaccine that protects against strains belonging to both serotypes 2 and 5 as most hemorrhagic septicemia isolates belong to these serotypes. These results suggest that antibodies against P1702 LPS may protect livestock against both serotype 2 and 5 infections. We are currently pursuing studies in an effort to ascertain the importance of PEtn residues in relation to the virulence of P. multocida; this includes the Hep II-linked PEtn residue of Heddleston serotype 5 LPS.


We thank Perry Fleming for bacterial growth.

This work was in part funded by the Australian Research Council, Canberra, Australia.


[down-pointing small open triangle]Published ahead of print on 18 September 2009.


1. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. H. 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. [PMC free article] [PubMed]
2. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1987. Current protocols in molecular biology. Greene Publishing Associates and Wiley-Interscience, New York, NY.
3. Boyce, J. D., and B. Adler. 2000. The capsule is a virulence determinant in the pathogenesis of Pasteurella multocida M1404 (B:2). Infect. Immun. 68:3463-3468. [PMC free article] [PubMed]
4. Boyce, J. D., M. Harper, F. St. Michael, M. John, A. Aubry, H. Parnas, S. M. Logan, I. W. Wilkie, M. Ford, A. D. Cox, and B. Adler. 2009. Identification of novel glycosyltransferases required for assembly of the Pasteurella multocida A:1 lipopolysaccharide and their involvement in virulence. Infect. Immun. 77:1532-1542. [PMC free article] [PubMed]
5. Brisson, J.-R., E. Crawford, D. Uhrin, N. H. Khieu, M. B. Perry, W. B. Severn, and J. C. Richards. 2002. The core oligosaccharide component from Mannheimia (Pasteurella) haemolytica serotype A1 lipopolysaccharide contains L-glycero-D-manno- and D-glycero-D-manno-heptoses: analysis of the structure and conformation by high-resolution NMR spectroscopy. Can. J. Chem. 80:949-963.
6. Chanter, N. 1990. Molecular aspects of the virulence of Pasteurella multocida. Can. J. Vet. Res. 54(Suppl.):S45-S47. [PubMed]
7. Christensen, J. P., and M. Bisgaard. 2000. Fowl cholera. Rev. Sci. Tech. OIE Rev. 19:626-637. [PubMed]
8. Chung, J. Y., I. Wilkie, J. D. Boyce, K. M. Townsend, A. J. Frost, M. Ghoddusi, and B. Adler. 2001. Role of capsule in the pathogenesis of fowl cholera caused by Pasteurella multocida serogroup A. Infect. Immun. 69:2487-2492. [PMC free article] [PubMed]
9. De Alwis, M. C. L. 1992. Haemorrhagic septicaemia—a general review. Br. Vet. J. 148:99-112. [PubMed]
10. DeAngelis, P. L., N. S. Gunay, T. Toida, W.-J. Mao, and R. J. Linhardt. 2002. Identification of the capsular polysaccharides of type D and F Pasteurella multocida as unmodified heparin and chondroitin, respectively. Carbohydr. Res. 337:1547-1552. [PubMed]
11. Harper, M., J. D. Boyce, A. D. Cox, F. St. Michael, I. W. Wilkie, P. J. Blackall, and B. Adler. 2007. Pasteurella multocida expresses two lipopolysaccharide glycoforms simultaneously, but only a single form is required for virulence: identification of two acceptor-specific heptosyl I transferases. Infect. Immun. 75:3885-3893. [PMC free article] [PubMed]
12. Harper, M., A. Cox, F. St. Michael, H. Parnas, I. Wilkie, P. J. Blackall, B. Adler, and J. D. Boyce. 2007. Decoration of Pasteurella multocida LPS with phosphocholine is important for virulence. J. Bacteriol. 189:7384-7391. [PMC free article] [PubMed]
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. [PMC free article] [PubMed]
14. Heddleston, K. L., J. E. Gallagher, and P. A. Rebers. 1972. Fowl cholera: gel diffusion precipitin test for serotyping Pasteurella multocida from avian species. Avian Dis. 16:925-936. [PubMed]
15. 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. [PubMed]
16. Mackinnon, F. G., A. D. Cox, J. S. Plested, C. M. Tang, K. Makepeace, P. A. Coull, J. C. Wright, R. Chalmers, D. W. Hood, J. C. Richards, and E. R. Moxon. 2002. Identification of a gene (lpt-3) required for the addition of phosphoethanolamine to the lipopolysaccharide inner core of Neisseria meningitidis and its role in mediating susceptibility to bactericidal killing and opsonophagocytosis. Mol. Microbiol. 43:931-943. [PubMed]
17. Namioka, S. 1978. Pasteurella multocida—biochemical characteristics and serotypes, p. 272-292. In T. Bergen and J. R. Norris (ed.), Methods in microbiology. Academic Press, London, United Kingdom.
18. Ramdani, and B. Adler. 1991. Opsonic monoclonal antibodies against lipopolysaccharide (LPS) antigens of Pasteurella multocida and the role of LPS in immunity. Vet. Microbiol. 26:335-347. [PubMed]
19. Rimler, R. B. 1990. Comparisons of Pasteurella multocida lipopolysaccharides by sodium dodecyl sulfate-polyacrylamide gel electrophoresis to determine relationship between group B and E hemorrhagic septicemia strains and serologically related group A strains. J. Clin. Microbiol. 28:654-659. [PMC free article] [PubMed]
20. Rimler, R. B. 1996. Passive immune cross-protection in mice produced by rabbit antisera against different serotypes of Pasteurella multocida. J. Comp. Pathol. 114:347-360. [PubMed]
21. Rimler, R. B., P. A. Rebers, and M. Phillips. 1984. Lipopolysaccharides of the Heddleston serotypes of Pasteurella multocida. Am. J. Vet. Res. 45:759-763. [PubMed]
22. Rosner, H., H. D. Grimmecke, Y. A. Knirel, and A. S. Shashkov. 1992. Hyaluronic acid and a (1-4)-β-D-xylan, extracellular polysaccharides of Pasteurella multocida (Carter type A) strain 880. Carbohydr. Res. 223:329-333. [PubMed]
23. St. Michael, F., J. R. Brisson, S. Larocque, M. A. 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. [PubMed]
24. St. Michael, F., J. Li, and A. D. Cox. 2005. Structural analysis of the core oligosaccharide from Pasteurella multocida strain X-73. Carbohydr. Res. 340:1253-1257. [PubMed]
25. 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. [PubMed]
26. 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. [PubMed]
27. Townsend, K. M., J. D. Boyce, J. Y. Chung, A. J. Frost, and B. Adler. 2001. Genetic organization of Pasteurella multocida cap loci and the development of a multiplex capsular PCR typing system. J. Clin. Microbiol. 39:924-929. (Erratum, 39:2378.) [PMC free article] [PubMed]
28. Weber, D. J., J. S. Wolfson, M. N. Swartz, and D. C. Hooper. 1984. Pasteurella multocida infections. Report of 34 cases and review of the literature. Medicine (Baltimore) 63:133-154. [PubMed]
29. 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. [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)