Genetic analysis of virulence factors of Mannheimia (Pasteurella) haemolytica A1

Sixteen 3-week-old calves were intratracheally inoculated with Mycoplasma bovis. Follow-up consisted of regular bronchoalveolar lavages (BALs) and clinical examinations. Animals were slaughtered from 4 to 21 days after inoculation. Counts were made of the mycoplasmas and other bacteria systematically isolated from the BAL liquids and lung lobes after slaughter. On the 6th day, spectinomycin 20mg/kg was given intramuscularly in three repeated doses at 24h intervals to six randomly chosen calves. All animals had developed a persistent M. bovis infection with a maximum BAL count on the 6th day (start of treatment). Co-occuring Pasteurella multocida infection was found in most animals with a maximum rate on the 14th day. The extent of lung surface lesions varied widely (0-64%) but was greater in the later slaughtered calves. Average counts of M. bovis and P. multocida in the BAL liquids were lower in treated calves than in untreated ones but the difference was not statistically significant. However, M. bovis and P. multocida counts in the lungs of the treated group were significantly lower than in the untreated group (p=0.003 and 0.009, respectively).     

Investigations on the species specificity of Mannheimia (Pasteurella) haemolytica serotyping

Eleven serotypes (1, 2, 5-9, 12-14 and 16) have been demonstrated within Mannheimia haemolytica. Subsequent serotyping of 166 Mannheimia haemolytica-like strains, genetically and phenotyphically distinct from Mannheimia haemolytica, and isolated from ruminants, pigs, hares and rabbits showed that 13.2% were typeable, 19 of which were serotype 11 representing strains now being classified as M. glucosida. In addition, three strains belonged to serotypes 6, 9 and 16, respectively. Additionally, the serotyping results of 98 (P.) haemolytica-like isolates from non-ruminant sources collected by the UK Veterinary Investigation Centres during the period 1982-1996 were investigated. None of these isolates have been kept, making further genetic characterization impossible. Among these isolates, 25.5% were typeable representing serotypes 1, 2, 3, 5, 6, 9, 10, 13 and 15. Substantial evidence has been reported indicating that M. haemolytica-like isolates from non-ruminant sources represent species different from M. haemolytica. The present investigation underlines that serotyping does not represent a reliable method for the identification of M. haemolytica or M. glucosida. These observations emphasize that extended phenotypic and genetic characterization is necessary for the proper identification of these organisms.     

Serotyping of Mannheimia (Pasteurella) haemolytica isolates from the upper Midwest United States.

Mannheimia (Pasteurella) haemolytica biotype A serotype1 (A1) is the primary bacterial agent responsible for the clinical signs and pathophysiologic events in bovine pneumonic pasteurellosis. The goal of this study was to determine the prevalence of other serotypes of M. haemolytica biotype A organisms obtained from the upper Midwest diagnostic laboratories. A total of 147 M. haemolytica isolates were collected from Minnesota, South Dakota, and Michigan. Isolates were tested against M. haemolytica antisera obtained from the National Animal Disease Center, Ames, Iowa. Results indicated that M. haemolytica serotype 1 represented approximately 60%, serotype 6 represented 26%, and serotype 2 represented 7% of the total examined isolates. In addition, 7% of the isolates were serotype 9, 11, or untypable. This finding suggests that M. haemolytica serotypes other than serotype 1 can be isolated from the lung lesions of diseased cattle and seem to be capable of causing the pathologic changes observed in the lung with pneumonic pasteurellosis.     

Survival of Mannheimia (Pasteurella) haemolytica in tracheobronchial washings of sheep and cattle

The growth, morphology and long-term survival of a representative isolate of Mannheimia haemolytica serotypes A1 and A2 were monitored in ovine and bovine tracheobronchial washings. Both strains survived for at least 244 days in ovine tracheobronchial washings and 156 days in bovine tracheobronchial washings. The addition of fresh washings at these times prompted an increase for serotype A2 but no change in viability for serotype A1 in ovine tracheobronchial washings and an increase for both serotypes in bovine tracheobronchial washings. When growth and survival was compared using tracheobronchial washings from ruminant and non-ruminant species there was a trend towards longer survival in ruminant fluids.Long-term survival was associated with temporary or permanent change from normal size colonies to 'micro-colonies' on sheep blood agar. Subculture allowed reversion to normal colony morphology. Analysis showed these micro-colonies to consist of chains of elongated bacteria. M. haemolytica serotype A2 was more robust in its ability to withstand nutrient deprivation for long periods of time. These survival mechanisms may have important implications for pathogenesis.     

Relationships between clinical and pathological signs of disease in calves infected with Mannheimia (Pasteurella) haemolytica type A1.

Respiratory disease was induced in 16 calves, and the terminal clinical signs of disease and postmortem pathological observations were recorded. By Spearman's correlation coefficient, the respiratory rate, rectal temperature and clinical scores of the calves were significantly correlated with the extent of lung consolidation. The respiratory rate was the clinical sign most consistently correlated with the other clinical and pathological signs of respiratory disease.     

Characterization of Mannheimia (Pasteurella) haemolytica leukotoxin interaction with bovine alveolar macrophage beta2 integrins

Mannheimia (Pasteurella) haemolytica, the etiologic agent of bovine pneumonic mannheimiosis, produces an exotoxic leukotoxin. The leukotoxin (LktA) is a member of the RTX (repeats in toxin) family of bacterial cytolysins and is distinguished from other toxins by its unique target cell specificity to ruminant leukocytes occurring through binding to a specific receptor. We have demonstrated previously that the beta2 integrin LFA-1 is a receptor for LktA in bovine leukocytes and is involved in leukotoxin-induced biological effects. However the subunits within LFA-1 involved in binding to LktA, and post-binding signaling leading to cellular activation have not been well characterized. The purpose of our study was to pinpoint these precise subunits on bovine alveolar macrophages and to characterize their interaction with LktA. The results in this study indicate that although LktA can efficiently bind to the CD18 subunit of both LFA-1 and Mac-1, post-binding signaling events including elevation of intracellular calcium and CD18 tail phosphorylation are only observed through LFA-1. Furthermore, LktA also binds to the CD11a subunit of LFA-1. LktA binding to CD11a could be inhibited by a small molecule inhibitor of the I(inserted)-domain, the major ligand binding interface on CD11a. I-domain inhibition significantly blunts LktA-induced intracellular calcium elevation and tyrosine phosphorylation of the CD18 tail. Based on our results we suggest that LFA-1 serves as the functional leukotoxin receptor on bovine alveolar macrophages.     

beta(2) integrin Mac-1 is a receptor for Mannheimia haemolytica leukotoxin on bovine and ovine leukocytes

Pneumonia caused by Mannheimia haemolytica is an important disease of cattle (BO), domestic sheep (DS, Ovis aries) and bighorn sheep (BHS, Ovis canadensis). Leukotoxin (Lkt) produced by M. haemolytica is cytolytic to all leukocyte subsets of these three species. Although it is certain that CD18, the beta subunit of beta(2) integrins, mediates Lkt-induced cytolysis of leukocytes, whether CD18 of all three beta(2) integrins, LFA-1 (CD11a/CD18), Mac-1 (CD11b/CD18) and CR4 (CD11c/CD18), mediates Lkt-induced cytolysis of BO, DS and BHS leukocytes remains a controversy. Based on antibody inhibition experiments, earlier studies suggested that LFA-1, but not Mac-1 and CR-4, serves as a receptor for M. haemolytica Lkt. PMNs express all three beta(2) integrins, and they are the leukocyte subset that is most susceptible to Lkt. Therefore we hypothesized that all three beta(2) integrins serve as the receptor for Lkt. The objective of this study was to determine whether Mac-1 of BO, DS and BHS serves as a receptor for Lkt. cDNAs for CD11b of BO, DS and BHS were transfected into a Lkt-non-susceptible cell line along with cDNAs for CD18 of BO, DS and BHS, respectively. Transfectants stably expressing BO, DS or BHS Mac-1 specifically bound Lkt. These transfectants were lysed by Lkt in a concentration-dependent manner. Increase in intracellular [Ca(2+)](i) was observed in transfectants following exposure to low concentrations of Lkt indicating signal transduction through secondary messengers. Collectively, these results indicate that Mac-1 from these three species serves as a receptor for M. haemolytica Lkt.     

Tilmicosin does not inhibit interleukin-8 gene expression in the bovine lung experimentally infected with Mannheimia (Pasteurella) haemolytica.

The expression of the interleukin-8 (IL-8) gene was examined by in situ hybridization in lung tissues from calves experimentally infected with Mannheimia (Pasteurella) haemolytica and treated with tilmicosin. Interleukin-8 mRNA expression was detected in alveolar areas, particularly along interlobular septa, in the lumen, and in the epithelial cells of some bronchioles. In lesional lung tissues from animals that had received tilmicosin, we found large areas with limited inflammation. There was no staining for IL-8 mRNA in these areas. In contrast, in strongly inflamed areas, the same patterns and intensities of staining for IL-8 mRNA were detected in tilmicosin- and sham-treated animals. We conclude that tilmicosin does not affect the expression of IL-8 mRNA in tissue showing microscopic signs of inflammation. Together with previous reports, this supports the view that the pro-apoptotic properties of tilmicosin on neutrophils do not compromise the host defense mechanisms required to control the infection.     

Pasteurella haemolytica bacteriophage: identification, partial characterization, and relationship of temperate bacteriophages from isolates of Pasteurella haemolytica (biotype A, serotype 1)

Pasteurella haemolytica (biotype A, serotype 1) isolates (n = 15) from the upper respiratory tract of clinically normal cattle, as well as from lung lesions from cases of fatal bovine pasteurellosis, were examined for the presence of bacteriophage after irradiation with UV light. Treatment of all P haemolytica isolates with UV irradiation resulted in lysis of bacteria due to the induction of vegetative development of bacteriophages. The extent of growth inhibition and bacterial lysis in irradiated cultures was UV dose-dependent. Bacterial cultures exposed to UV light for 20 s reached peak culture density between 60 and 70 minutes after irradiation; thereafter, culture density declined rapidly, so that by 120 minutes, it was approximately 60% of the original value. When examined ultrastructurally, lytic cultures from each isolate revealed bacteriophages with an overall length of approximately 200 nm and that appeared to have a head with icosahedral symmetry and a contractile tail. Cell-free filtrate from each noninduced bacterial isolate was inoculated onto the other bacterial isolates in a cross-culture sensitivity assay for the presence of phages lytic for the host bacterial isolates. Zones of lysis (plaques) did not develop when bacterial lawns grown from the different isolates were inoculated with filtrates from the heterologous isolates.     

Alterations in bovine platelet function and acute phase proteins induced by Pasteurella haemolytica A1.

Platelet function was assessed by aggregometry in 10 Holstein calves before and after exposure to Pasteurella haemolytica (biotype A, serotype 1) by intrabronchial challenge. At 24 h after exposure the platelets had become more reactive to stimulation with known platelet agonists such as adenosine diphosphate (ADP) and platelet-activating factor (PAF) and the platelet aggregates that formed were more resistant to disaggregation. The activation of platelets was an early response in the challenged calves as platelet function had returned to pretreatment levels 72 h after exposure to the bacteria while the acute phase reactant proteins, haptoglobin and fibrinogen, were approaching their peak values and alpha 2-macroglobulin levels had also risen significantly (P < 0.05) at this time. The plasma levels of these proteins were still elevated and albumin levels were depressed 6 d post-treatment. At post-mortem all calves exhibited pneumonic tissue damage. When P. haemolytica leukotoxin was added directly to bovine platelet suspensions both spontaneous aggregation and an increase in the aggregation response to ADP and PAF stimulation were observed. The morphological appearance of the platelet aggregates exhibited the typical pattern for bovine platelets with 2 distinct zones of cells being visible within each aggregate. One zone contained platelets in which the cytoplasmic granules were still evident and the other zone contained irregularly shaped platelets devoid of granular content. In the latter zone, discrete gaps, or pores, were evident in the plasma membrane of numerous platelets. This pore formation is characteristic of leukotoxin action and is not observed in ADP or PAF induced aggregates.     

Pharmacokinetic and pharmacodynamic integration and modelling of marbofloxacin in calves for Mannheimia haemolytica and Pasteurella multocida

The pharmacokinetics (PK) and pharmacodynamics (PD) of marbofloxacin were established in calves for six strains of each of the pneumonia pathogens Mannheimia haemolytica and Pasteurella multocida. The distribution of marbofloxacin into inflamed (exudate) and non-inflamed (transudate) tissue cage fluids allowed comparison with the serum concentration–time profile. To establish the PD profile, minimum inhibitory concentration (MIC) was determined in Mueller–Hinton broth (MHB) and calf serum.Moderately higher MICs were obtained for serum compared to MHB. An initial integration of PK–PD data established C_max/MIC ratios of 45.0 and AUC_24h/MIC values of 174.7 h, based on serum MICs, for both bacterial species. Using bacterial time-kill curves, generated ex vivo for serum marbofloxacin concentrations, PK–PD modelling established three levels of growth inhibition: AUC_24h/MIC ratios for no reduction, 3 log₁₀ and 4 log₁₀ reductions in bacterial count from the initial inoculum count were 41.9, 59.5 and 68.0 h for M. haemolytica and 48.6, 64.9 and 74.8 h for P. multocida, on average respectively. Inter-strain variability for 3 log₁₀ and 4 log₁₀ reductions in bacterial count was smaller for P. multocida than for M. haemolytica. In conjunction with literature data on MIC₉₀ values, the present results allowed prediction of dosages for efficacy for each organism for the three levels of growth inhibition.     

Vaccination of neonatal colostrum-deprived calves against Pasteurella haemolytica A1.

Colostrum-deprived Holstein calves were vaccinated at 2 and 4 wk of age with a Pasteurella haemolytica A1 culture supernatant vaccine to determine whether active immune responses and protection could be induced in this age group in the absence of maternal antibodies. All calves responded to vaccination with high titers of IgM antibodies to capsular polysaccharide within 1 wk of primary vaccination. Mean titers of IgG1 and IgG2 antibodies to this antigen increased significantly by 2 wk after secondary vaccination, but peak antibody titers were low. All of the vaccinated calves seroconverted with production of leukotoxin-neutralizing antibodies, but peak antibody titers were low. Vaccinated calves experienced considerable lung damage after experimental challenge, but survival rate, clinical scores, and percent lung involvement were significantly better than those of control (placebo-injected) calves.     

Effect of experimental infection of cattle with bovine herpesvirus-1 (BHV-1) on the ex vivo interaction of bovine leukocytes with Mannheimia (Pasteurella) haemolytica leukotoxin

Mannheimia (Pasteurella) haemolytica A1 produces an extracellular leukotoxin (LKT) that is reported to bind the beta(2)-integrin CD11a/CD18 (LEA-1) on ruminant leukocytes. LKT binding induces activation, and subsequent cytolysis, of these cells. It is well known that active viral infection greatly increases the susceptibility of cattle to pasteurellosis. To better understand the mechanism by which this occurs, we investigated the effects of experimental in vivo infection of cattle with bovine herpes virus-1 (BHV-1) on the ex vivo interaction of bovine leukocytes with the M. haemolytica LKT. In this study, we demonstrated that active BHV-1 infection increased the expression of the beta(2)-integrin CD11a/CD18 (as defined by the mAb BAT75) on bovine peripheral blood neutrophils, enhanced the binding of LKT to bronchoalveolar lavage (BAL) leukocytes and peripheral blood neutrophils, and increased the killing of BAL leukocytes and peripheral blood leukocytes by LKT. In addition, BHV-1 greatly increased the number of BAL, resulting in many more LKT-responsive cells being present in the lungs. These findings might explain in part the increased susceptibility of BHV-1 infected cattle to pneumonic pasteurellosis.     

Pasteurella haemolytica A1 and Bovine Respiratory Disease: Pathogenesis

The severe fibrinonecrotic pneumonia associated with pneumonic pasteurellosis usually results from colonization of the lower respiratory tract by Pasteurella haemolytica biotype A, serotype 1(A1). Despite recent research efforts, the authors lack a detailed understanding of the interactions and host response to P. haemolytica in the respiratory tract. The authors hypothesize that management and environmental stress factors or viral infection alters the upper respiratory tract (URT) epithelium allowing P. haemolytica to colonize the epithelium. Once the URT is colonized, large numbers of organisms enter the lung where they interact with alveolar macrophages. Endotoxin, released from the bacteria, crosses the alveolar wall where it activates pulmonary intravascular macrophages, endothelium, neutrophils, lymphocytes, platelets, complement, and Hageman factor leading to complex interactions of cells and mediators. It is the progression of this inflammatory response with neutrophil influx that is ultimately responsible for the pulmonary injury. Leukotoxin is a major virulence factor of P. haemolytica that allows it to survive by destroying phagocytic cells. At subcytolytic concentrations it may also enhance the inflammatory response by activating cells to produce mediators and release reactive oxygen metabolites and proteases.     

Construction and analysis of a Mannheimia haemolytica A1 luxS mutant

Mannheimia haemolytica A1 is the causative agent of bovine pneumonic pasteurellosis, a major cause of sickness, death, and economic loss to the feedlot cattle industry. M. haemolytica A1 produces autoinducer-2 (AI-2) like molecules that are capable of inducing quorum sensing system 2 of Vibrio harveyi. This interspecies quorum sensing system has been shown to regulate the expression of virulence genes in several pathogenic bacteria. The protein central to the production of AI-2 is LuxS. To determine if quorum sensing is involved in the regulation of virulence genes in M. haemolytica A1, a luxS mutant was constructed by replacing luxS with a cat cassette. This mutant was verified by PCR analysis, Southern hybridization, as well as its inability to induce bioluminescence in the V. harveyi reporter strain. RT-PCR analysis showed there was no difference in leukotoxin (lktC) mRNA levels, however there were increased mRNA levels of putative virulence associated genes, transferrin binding protein B (tbpB), adhesin (ahs) and capsule biosynthesis (nmaA). Electron microscopy showed that the level of encapsulation in the mutant is higher than the parent. Additionally, the mutant was slightly more adherent to bovine tracheal cells than the parent. In vitro competition assays showed the mutant out-competed the parent under iron-restricted conditions. However, in a calf challenge, the parent was the dominant isolate recovered.     

Pharmacokinetic/pharmacodynamic integration and modelling of amoxicillin for the calf pathogens Mannheimia haemolytica and Pasteurella multocida

The antimicrobial properties of amoxicillin were determined for the bovine respiratory tract pathogens, Mannheima haemolytica and Pasteurella multocida. Minimum inhibitory concentration (MIC), minimum bactericidal concentration (MBC) and time‐kill curves were established. Pharmacokinetic (PK)/pharmacodynamic (PD) modelling of the time‐kill data, based on the sigmoidal E_max equation, generated parameters for three levels of efficacy, namely bacteriostatic, bactericidal (3log₁₀ reduction) and 4log₁₀ reduction in bacterial counts. For these levels, mean AUC_(0–24 h)/MIC serum values for M. haemolytica were 29.1, 57.3 and 71.5 h, respectively, and corresponding values for P. multocida were 28.1, 44.9 and 59.5 h. Amoxicillin PK was determined in calf serum, inflamed (exudate) and noninflamed (transudate) tissue cage fluids, after intramuscular administration of a depot formulation at a dosage of 15 mg/kg. Mean residence times were 16.5 (serum), 29.6 (exudate) and 29.0 h (transudate). Based on serum MICs, integration of in vivo PK and in vitro PD data established maximum concentration (C_max)/MIC ratios of 13.9:1 and 25.2:1, area under concentration–time curve (AUC_0–∞)/MIC ratios of 179 and 325 h and T>MIC of 40.3 and 57.6 h for P. multocida and M. haemolytica, respectively. Monte Carlo simulations for a 90% target attainment rate predicted single dose to achieve bacteriostatic and bactericidal actions over 48 h of 17.7 and 28.3 mg/kg (M. haemolytica) and 17.7 and 34.9 mg/kg (P. multocida).     

A multiplex polymerase chain reaction assay for the identification of Mannheimia haemolytica, Mannheimia glucosida and Mannheimia ruminalis

The objective of this study was to design a multiplex PCR assay to identify Mannheimia haemolytica, Mannheimia glucosida and Mannheimia ruminalis. The multiplex PCR included primer sets HP, amplifying a DNA region from an unknown hypothetical protein, Lkt and Lkt2, amplifying different regions of the leukotoxinD gene, and 16S to amplify universal bacterial sequences of the 16S rRNA gene. Based on positive amplification, isolates were delineated as M. haemolytica (HP, Lkt, 16S), M. glucosida (HP, Lkt, Lkt2, 16S), or M. ruminalis (HP, 16S). The validity of the assay was examined against 22 reference strains within the family Pasteurellaceae and 17 field isolates (nasal) that had been collected previously from feedlot cattle and tentatively identified as M. haemolytica based on morphology and substrate utilization. Additionally, 200 feedlot cattle were screened for M. haemolytica using multiplex PCR. Forty-four isolates from 25 animals were identified as M. haemolytica. The PCR assay positively identified all M. haemolytica, as confirmed by phenotypic tests and clustering based upon cellular fatty acid methyl ester (FAME) profiles. Selected nasal isolates that exhibited evidence of haemolysis, but were M. haemolytica-negative based on PCR, were also confirmed negative by phenotypic and FAME analyses. The multiplex PCR assay required no additional phenotypic tests for confirmation of M. haemolytica, within the group of bacteria tested.     

Pharmacodynamics of amoxicillin against Mannheimia haemolytica and Pasteurella multocida and pharmacokinetic/pharmacodynamic (PK/PD) correlation in sheep

Silicone-made tissue cages were implanted in sheep. Blood serum (SBS) and tissue cage fluid (TCF) samples were collected after amoxicillin intravenous and intramuscular administrations, at the dose of 15 mg/kg. Amoxicillin pharmacodynamics were studied in an artificial culture medium, SBS and TCF with use of a Mannheimia haemolytica and a Pasteurella multocida strain. A concentration-independent antimicrobial activity of amoxicillin was confirmed for levels higher than 0.79–1.75 × MIC. This result favored the use of the percentage of the 24 h dosing interval during which drug levels remain above MIC as the appropriate pharmacokinetic/pharmacodynamic index. The subsequent correlation revealed that intravenous administration could be considered effective against “deep” infections caused by bacteria with MICs < 1 μg/mL or “shallow” infections caused by bacteria with MICs < 0.1 μg/mL. Intramuscular administration could be safely considered effective against both “deep” and “shallow” infections when the MICs of the targeted pathogens are lower than 1 μg/mL.