Isolation of Pasteurella haemolytica leukotoxin mutants.

Two mutants of Pasteurella haemolytica A1 that do not produce leukotoxin were isolated. Following mutagenesis, colonies were screened with antiserum by a filter assay for absence of the secreted leukotoxin. The two mutants both appeared to produce normal amounts of other antigens, as judged by reactivity with polyclonal serum from an animal with pasteurellosis, and were not altered in beta-hemolytic activity as seen on blood agar plates. There was no evidence of either cell-associated or secreted leukotoxin protein when Western blots (immunoblots) were carried out with the polyclonal serum or with a monoclonal antibody directed against the leukotoxin. Southern blots revealed that both mutants show the wild-type restriction pattern at the leukotoxin locus, although the strain with the lktA2 mutation showed differences in other regions of the chromosome on analysis by pulsed-field gel electrophoresis. The strain with the lktA2 mutation grew more slowly than did the wild-type strain, while the strain with the lktA1 mutation was indistinguishable from the wild-type strain in its growth properties. The strain with the lktA1 mutation should be valuable in determining the role of the leukotoxin in virulence as well as in identifying other virulence factors of P. haemolytica.     

Binding of Pasteurella haemolytica leukotoxin to bovine leukocytes.

Pasteurella haemolytica is the principal bacterial pathogen in the bovine respiratory disease complex. This organism produces an exotoxin (referred to as leukotoxin) during logarithmic-phase growth that is a potent leukocyte-modulating agent. At low concentrations, it activates neutrophils and mononuclear phagocytes to release inflammatory mediators, while at the same time making these cells destined to undergo apoptotic cell death. At higher concentrations, the toxin causes rapid swelling and loss of cell viability. In this study, we demonstrated that toxin binding can be directly evaluated by flow cytometry with biologically active biotinylated leukotoxin. Leukotoxin binding was blocked by the addition of a neutralizing anti-leukotoxin monoclonal antibody and was not detected when bovine leukocytes were incubated with culture filtrates from a mutant strain of P. haemolytica that does not produce biologically active leukotoxin. In addition, treatment of bovine leukocytes with protease K eliminated subsequent binding of leukotoxin, suggesting that there is a protein on the leukocyte surface that is either a leukotoxin binding site or is required for stabilization of leukotoxin binding. We did not detect binding of biotinylated leukotoxin to porcine or human leukocytes, which have been reported previously to be resistant to the lytic effects of the leukotoxin. These findings suggest that there may be a specific binding site for P. haemolytica leukotoxin on bovine but not on porcine or human leukocytes and that it might be involved in the activation and lytic activities of the leukotoxin.     

Hemolytic activity of the Pasteurella haemolytica leukotoxin.

A Pasteurella haemolytica mutant incapable of producing leukotoxin was created by allelic replacement. Concentrated culture supernatants from wild-type P. haemolytica, but not from the mutant, contained the 102-kDa leukotoxin protein and lysed bovine lymphoma cells and sheep erythrocytes. Wild-type P. haemolytica demonstrated the typical beta-hemolytic phenotype on sheep and rabbit blood agar, whereas the mutant did not.     

Neutralizing monoclonal antibodies to Pasteurella haemolytica leukotoxin affinity-purify the toxin from crude culture supernatants.

The leukotoxin of Pasteurella haemolytica is a major virulence factor of the organism. It is an unstable protein which has proven very difficult to purify using traditional techniques. Hybridomas secreting monoclonal antibodies (mAbs) to P. haemolytica leukotoxin were derived from spleen cells of a mouse immunized with crude culture supernatant. Five hybridomas secreting mAbs specific for the leukotoxin were stabilized. Each of the mAbs reacted with a protein of approximately 100 kDa in toxic culture supernatants, and two of them completely neutralized the toxin in vitro. Affinity chromatography of crude culture supernatant on a column prepared with one of the neutralizing mAbs resulted in the isolation of biologically active toxin.     

Identification and characterization of the Pasteurella haemolytica leukotoxin.

The identification and chromatographic characterization of the leukotoxin of Pasteurella haemolytica is described. The toxin, which has an apparent native molecular weight of greater than 400,000 as judged by gel exclusion chromatography, has a 105-kilodalton (105K) polypeptide as its major protein component. The proteolytic degradation of the 105K polypeptide could be correlated with the loss of toxin activity in aging cultures of P. haemolytica. Antisera raised against purified 105K polypeptide neutralized toxin activity. A 3.9-kilobase-pair fragment of the P. haemolytica genome cloned into a plasmid vector resulted in the production of intracellular toxin in Escherichia coli host cells. The restriction map of this clone shows significant overlap with the map of a previously reported leukotoxin clone (R. Y. C. Lo, P. E. Shewen, C. A. Strathdee, and C. N. Greer, Infect. Immun. 50:667-671, 1985). Finally, antisera raised against the 105K species labeled the P. haemolytica cell surface in a nonuniform, punctate manner.     

Antigenic and virulence properties of Pasteurella haemolytica leukotoxin mutants.

Antigenic properties of two mutants of Pasteurella haemolytica, strains 59B0071 and 59B0072, that do not produce detectable leukotoxin were investigated. Western blot (immunoblot) analysis with a number of polyclonal sera from animals recovering from pasteurellosis revealed that both mutants secreted a variety of antigens that were also present in cultures of several wild-type strains. These antigens ranged from about 100 to 15 kDa. Mutant strain 59B0071 was found to be totally deficient in leukotoxin, as judged not only by Western blotting but also by cytotoxicity assays with bovine lymphoma (BL-3) cells or bovine polymorphonuclear cells as targets. The mutant strain 59B0071 had normal levels of a secreted sialylglycoprotease, however. When strains were tested for virulence in goat and cattle challenge experiments, a reduction in mortality and lung lesions was observed with the mutant 59B0071 in comparison with results obtained with wild-type strains. These results are consistent with an important role for leukotoxin in P. haemolytica virulence and suggest that leukotoxin-negative mutants may be useful tools in the investigation of other virulence properties involved in P. haemolytica infections.     

Activation of bovine neutrophils by partially purified Pasteurella haemolytica leukotoxin.

In this study we developed a new method for the partial purification of Pasteurella haemolytica leukotoxin by size-exclusion high-performance liquid chromatography. The partially purified leukotoxin had a molecular weight of 104,000, as estimated by using sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and reacted on an immunoblot with an antileukotoxin monoclonal antibody. As expected, high concentrations of the leukotoxin were inhibitory or lethal to bovine neutrophils. Incubation of bovine neutrophils with diluted leukotoxin, however, resulted in significant neutrophil activation that was comparable in magnitude to that obtained with standard activating agents such as opsonized zymosan or zymosan-activated serum. Dilute leukotoxin (1:128 to 1:8,192 dilutions) stimulated an oxidative burst (luminol-dependent chemiluminescence) by bovine neutrophils that was comparable in magnitude to that obtained with opsonized zymosan. Preincubation with leukotoxin did not significantly prime the neutrophils for an enhanced oxidative burst when they were then exposed to opsonized zymosan as a second stimulus. Dilute leukotoxin (1:100 to 1:1,000 dilutions) also stimulated cytoskeletal alterations in bovine neutrophils, as measured by a significant shape change response. Preferential release of secondary granule constituents (lactoferrin) occurred when neutrophils were incubated with 1:100 to 1:500 dilutions of leukotoxin. Significant release of primary granules, as measured by beta-glucosaminidase activity, was not observed except at low dilutions (1:20) of leukotoxin that resulted in significant release of cytosolic constituents (i.e., lactate dehydrogenase activity). The neutrophil-activating activity of the leukotoxin was heat labile, unaffected by polymyxin B, and abrogated by a leukotoxin-neutralizing monoclonal antibody. These data indicate that P. haemolytica leukotoxin, like the closely related Escherichia coli hemolysin, is a potent neutrophil-activating agent. Leukotoxin-stimulated release of neutrophil oxygen intermediates and granule constituents may contribute to the intense inflammation that characterizes bovine pulmonary pasteurellosis.     

Activation of bovine neutrophils by Pasteurella haemolytica leukotoxin is calcium dependent.

In this study, we used the fluorescent probe Fluo-3 to show that an increase in cytosolic free calcium, [Ca2+]i, occurred when suspensions of bovine neutrophils were incubated with sublethal concentrations of P. haemolytica leukotoxin. This increase in [Ca2+]i was dependent on the concentration of leukotoxin present in the medium and, at a given concentration of leukotoxin, dependent on the external calcium concentration. The calcium channel blocker verapamil and the beta-adrenergic antagonist propranolol inhibited leukotoxin-stimulated Ca2+ gain, as did a neutralizing antileukotoxin monoclonal antibody. As reported previously, incubation of bovine neutrophils with partially purified leukotoxin stimulated a vigorous luminol-dependent chemiluminescence response (LDCL). The present study shows that LDCL stimulation was dependent on the presence of extracellular calcium and was inhibited by the addition of verapamil and propranolol. These data indicate that bovine neutrophils exhibit a considerable increase in cytoplasmic free calcium when they are incubated with P. haemolytica leukotoxin in the presence of external calcium. They also provide evidence that an increased [Ca2+]i is required for functional activation of the bovine neutrophil oxidative burst by P. haemolytica leukotoxin.     

Enhancement of neutrophil-mediated injury to bovine pulmonary endothelial cells by Pasteurella haemolytica leukotoxin.

In this study, we used an in vitro coculture system to determine which virulence factor from Pasteurella haemolytica A1 was responsible for augmenting bovine polymorphonuclear neutrophil (PMN)-mediated killing of bovine pulmonary artery endothelial cells (BPAEC). A 51Cr release cytotoxicity assay was used as a measure of BPAEC killing. The mechanisms associated with this BPAEC killing were also studied. Our results demonstrated that the leukotoxin and not the lipopolysaccharide from P. haemolytica was responsible for augmenting the PMN-mediated killing of BPAEC. Furthermore, this augmented killing was related to the stimulation of PMNs by the leukotoxin. Killing of BPAEC by leukotoxin-stimulated PMNs was diminished in the presence of the H2O2 inactivator, catalase. The membrane-permeant H2O2, hydroxyl radical (HO.) scavenger 1,3-dimethyl-2 thiourea, and the HO. scavenger dimethyl sulfoxide but not the myeloperoxidase inhibitor sodium azide attenuated this BPAEC killing. Pretreatment of BPAEC with a 21-aminosteroid (U74500A), a potent iron chelator-antioxidant, provided the most effective protection against BPAEC killing induced by leukotoxin-stimulated PMNs. These data were compatible with the concept that the H2O2 generated by leukotoxin-stimulated PMNs interacts with intracellular iron in the endothelial cell to form highly reactive HO.. We suggest that HO. may be a key factor in BPAEC killing. Furthermore, since the elastase-specific inhibitor N-methoxy-succinyl-Ala-Ala-Pro-Val-chloromethyl ketone (CMK) also attenuated BPAEC killing and both CMK and 1,3-dimethyl-2 thiourea functioned additively in protecting against BPAEC killing, we conclude that both HO. and elastase may jointly contribute to BPAEC killing induced by leukotoxin-stimulated PMNs. This study broadens our understanding of how leukotoxin-stimulated PMNs injure lung endothelial cells and provides new insight into the pathogenesis of bovine pneumonic pasteurellosis.     

Nucleotide sequence of the leukotoxin genes of Pasteurella haemolytica A1.

A 4.4-kilobase-pair DNA fragment coding for the leukotoxin of Pasteurella haemolytica A1 has been isolated, and its nucleotide sequence has been determined. Two open reading frames, designated lktC and lktA, coding for proteins of 19.8 and 101.9 kilodaltons, respectively, were identified. Expression of the two genes in minicell-labeling experiments resulted in the production of the predicted proteins LKTC and LKTA. By using an antiserum against the soluble antigens of P. haemolytica A1 in Western blot (immunoblot) analysis of total cellular proteins from the Escherichia coli clones, LKTA was identified as an additional antigenic protein. Results from subcloning of the DNA fragment suggested that expression from both lktC and lktA is required for leukotoxin activity, indicating that the leukotoxin of P. haemolytica A1 is encoded by two genes. A comparison of the organization and the DNA sequence of the leukotoxin genes with those of the E. coli alpha-hemolysin genes showed a significant degree of homology between the two loci. This analysis suggested that the leukotoxin genes of P. haemolytica A1 and the E. coli alpha-hemolysin genes may have evolved from a common ancestor and that the two toxins may share similar activities or functional domains or both.     

Simple visual assay for determination of Pasteurella haemolytica cytotoxin neutralizing antibody titers in cattle sera.

A simple visual assay is described for determining the capacity of bovine serum to neutralize the cytotoxin produced by Pasteurella haemolytica serotype 1. The test was reproducible from day to day with different target cell populations and cytotoxin preparations. Cytotoxin neutralization titers obtained by the visual assay were comparable to those determined by the trypan blue exclusion and 51Cr-release methods. The visual assay was used to measure neutralization titers of bovine sera obtained from vaccination experiments and fractions of purified serum obtained by gel filtration. The major advantages of the visual assay over other assays are that it is rapid, inexpensive, and does not use radioisotopes. It also does not require specialized equipment, making it adaptable to most laboratories.     

Influence of Pasteurella haemolytica A1 crude leukotoxin on bovine neutrophil chemiluminescence.

Pasteurella haemolytica A1 crude leukotoxin (25%, vol/vol) rapidly diminished the bovine neutrophil chemiluminescence response to opsonized zymosan. This inhibition was neither prevented nor reversed by 75 mM sucrose. Dilute leukotoxin did not directly stimulate neutrophil chemiluminescence nor did it alter the chemiluminescence response of the neutrophils to opsonized zymosan.     

Pasteurella haemolytica leukotoxin induces bovine leukocytes to undergo morphologic changes consistent with apoptosis in vitro.

Infection of the bovine lung with Pasteurella haemolytica results in an acute respiratory disorder known as pneumonic pasteurellosis. One of the key virulence determinants used by this bacterium is secretion of an exotoxin that is specific for ruminant leukocytes (leukotoxin). At low concentrations, the leukotoxin can activate ruminant leukocytes, whereas at higher concentrations, it inhibits leukocyte functions and is cytolytic, presumably as a result of pore formation and subsequent membrane permeabilization. We have investigated the possibility that the activation-inhibition paradox is explained in part by leukotoxin-mediated apoptosis (i.e., activation-induced cell death) of bovine leukocytes. Incubation of bovine leukocytes with P. haemolytica leukotoxin caused marked cytoplasmic membrane blebbing (zeiosis) and chromatin condensation and margination, both of which are hallmarks of apoptosis. The observed morphologic changes in bovine leukocytes were leukotoxin dependent, because they were significantly diminished in the presence of an anti-leukotoxin monoclonal antibody. In addition, bovine leukocytes incubated with culture supernatant from a mutant strain of P. haemolytica that does not produce any detectable leukotoxin failed to exhibit the morphologic changes characteristic of cells undergoing apoptosis. These observations may represent an important mechanism by which P. haemolytica overwhelms host defenses, contributing to the fibrinous pleuropneumonia characteristic of bovine pasteurellosis.     

Pasteurella haemolytica leukotoxin inhibits mitogen-induced bovine peripheral blood mononuclear cell proliferation in vitro.

In this study we demonstrate that partially purified Pasteurella haemolytica leukotoxin inhibits the proliferative response of bovine peripheral blood mononuclear cells (PBMC) to mitogens in vitro. Inhibition of PBMC proliferation did not appear to be due to cell death. Addition of a neutralizing anti-leukotoxin monoclonal antibody restored a normal proliferative response.     

Molecular chimerization of Pasteurella haemolytica leukotoxin to interleukin-2: effects on cytokine and antigen function.

A chimeric recombinant protein composed of the lktA gene product from Pasteurella haemolytica fused to bovine interleukin-2 (IL-2) was made. The LKT-IL-2 chimera was compared with recombinant bovine IL-2 with regard to the ability to induce proliferative responses and LAK cell activity in bovine peripheral blood mononuclear cells in vitro. In both instances, chimerization had no effect on IL-2 activity. Similarly, the LKT component was unaffected in its ability to induce an effective immune response after immunization. The adjuvant properties of IL-2 have been established in a number of models, and this effect was tested by using the chimera. A multiple-injection protocol of LKT-IL-2 was compared with single-dose administration of LKT. The results obtained indicate that while there was no increase in specific antibody production, the IL-2 component of the chimera may be able to affect antigen-specific proliferation, as assessed by limiting-dilution analysis. Use of cytokine-antigen chimeras may provide a valuable antigen-adjuvant formulation that is simple to produce and purify and thus have economic advantages over conventional preparations. Furthermore, chimerization will also ensure that the adjuvant acts at the same site as the antigen, thus optimizing immunostimulatory activity.     

Lymphocyte function-associated antigen 1 is a receptor for Pasteurella haemolytica leukotoxin in bovine leukocytes

Pasteurella (Mannheimia) haemolytica leukotoxin (Lkt) causes cell type- and species-specific effects in ruminant leukocytes. Recent studies indicate that P. haemolytica Lkt binds to bovine CD18, the common subunit of all beta2 integrins. We designed experiments with the following objectives: to identify which member of the beta2 integrins is a receptor for Lkt; to determine whether Lkt binding to the receptor is target cell (bovine leukocytes) specific; to define the relationships between Lkt binding to the receptor, calcium elevation, and cytolysis; and to determine whether a correlation exists between Lkt receptor expression and the magnitude of target cell cytolysis. We compared Lkt-induced cytolysis in neutrophils from control calves and from calves with bovine leukocyte adhesion deficiency (BLAD), because neutrophils from BLAD-homozygous calves exhibit reduced beta2 integrin expression. The results demonstrate for the first time that Lkt binds to bovine CD11a and CD18 (lymphocyte function-associated antigen 1 [LFA-1]). The binding was abolished by anti-CD11a or anti-CD18 monoclonal antibody (MAb). Lkt-induced calcium elevation in bovine alveolar macrophages (BAMs) was inhibited by anti-CD11a or anti-CD18 MAb (65 to 94% and 37 to 98%, respectively, at 5 and 50 Lkt units per ml; P < 0.05). Lkt-induced cytolysis in neutrophils and BAMs was also inhibited by anti-CD11a or anti-CD18 MAb in a concentration-dependent manner. Lkt bound to porcine LFA-1 but did not induce calcium elevation or cytolysis. In neutrophils from BLAD calves, Lkt-induced cytolysis was decreased by 44% compared to that of neutrophils from control calves (P < 0.05). These results indicate that LFA-1 is a Lkt receptor, Lkt binding to LFA-1 is not target cell specific, Lkt binding to bovine LFA-1 correlates with calcium elevation and cytolysis, and bovine LFA-1 expression correlates with the magnitude of Lkt-induced target cell cytolysis.     

Enzyme-linked immunosorbent assay for detection of serum antibodies to Pasteurella haemolytica cytotoxin (leukotoxin) in cattle.

An enzyme-linked immunosorbent assay (ELISA) was developed for detection of bovine serum antibodies to the cytotoxin (leukotoxin) of Pasteurella haemolytica. A partially purified, cytotoxic, and immunogenic protein obtained from supernatants of logarithmic-phase P. haemolytica was used as the ELISA antigen. Preadsorption of sera with various cytotoxic, somatic, and capsular antigen preparations demonstrated that the assay was specific for anticytotoxin antibodies. ELISA anticytotoxin titers had a strong, significant correlation to cytotoxin-neutralizing-antibody titers. The ELISA, however, was more rapid and allowed for greater numbers of samples to be run than did the neutralization technique. ELISA anticytotoxin titers were high in cattle vaccinated with a live P. haemolytica vaccine, whereas unvaccinated cattle and cattle receiving a P. haemolytica bacterin had low ELISA anticytotoxin titers. A significant positive correlation between ELISA titers and resistance to experimental bovine pneumonic pasteurellosis was present.     

Expression of the Pasteurella haemolytica leukotoxin is inhibited by a locus that encodes an ATP-binding cassette homolog.

Multicopy and single-copy chromosomal fusions between the Pasteurella haemolytica leukotoxin regulatory region and the Escherichia coli beta-galactosidase gene have been constructed. These fusions were used as reporters to identify and isolate regulators of leukotoxin expression from a P. haemolytica cosmid library. A cosmid clone, which inhibited leukotoxin expression from multicopy and single-copy protein fusions, was isolated and found to contain the complete leukotoxin gene cluster plus additional upstream sequences. The locus responsible for inhibition of expression from leukotoxin-beta-galactosidase fusions was mapped within these upstream sequences, by transposon mutagenesis with Tn5, and its DNA sequence was determined. The inhibitory activity was found to be associated with a predicted 440-amino-acid reading frame (lapA) that lies within a four-gene arginine transport locus. LapA is predicted to be the nucleotide-binding component of this transport system and shares homology with the Clp family of proteases.     

Correlation of Pasteurella haemolytica leukotoxin binding with susceptibility to intoxication of lymphoid cells from various species

Pasteurella haemolytica, the causative agent of shipping fever pneumonia in cattle, produces a leukotoxin (LKT) which lyses ruminant leukocytes with high efficiency but is reputed to not affect leukocytes from nonruminant species. In this study, we tested the supposition that LKT binding correlates positively with susceptibility to intoxication of susceptible isolated bovine lymphocytes and lymphoma tissue culture cells (BL3 cells) and negatively with reputed nonsusceptible equine, porcine, and canine lymphocytes and human lymphoid tissue culture cells (Raji cells). Bovine lymphocytes and BL3 cells were highly susceptible to LKT intoxication, exhibiting both substantial increase in intracellular Ca(2+) concentration and marked leukolysis. Exposure of reputed LKT-nonsusceptible porcine lymphocytes and Raji cells to LKT caused a slightly increased intracellular Ca(2+) concentration but no leukolysis. No LKT effect was detected for equine and canine lymphocytes. LKT bound to lymphoid cells from all species tested. Intact 102-kDa LKT was recovered from exposed isolated lymphoid cell membranes. Pro-LKT acylation was not required for LKT binding to BL3 cells. LKT binding was rapid, with maximal binding occurring by 3 min, and was proportional to the LKT concentration in the range 0.04 to 4.0 microg/ml. For this LKT concentration range, BL3 cells bound more LKT than did porcine lymphocytes or Raji cells, suggesting that LKT binds to BL3 cells with higher affinity than to porcine lymphocytes or Raji cells. Above 4.0 microg/ml, LKT demonstrated saturable binding to BL3 cells. Neutralizing anti-LKT monoclonal antibody (MAb) MM601 diminished LKT binding to BL3 by 36% while decreasing leukolysis by 81%. In contrast, MM601 did not diminish LKT binding to Raji cells. Pretreatment of target cells with 120 microg of protease K per ml diminished LKT binding to BL3 cells by 75%, with only a 25% decrease in leukolysis. However, pretreatment with 150 microg of protease K per ml abolished the remaining 25% of LKT binding and 75% leukolysis. Therefore, P. haemolytica LKT binds rapidly to susceptible and to reputed nonsusceptible lymphoid cells. LKT binding resulting in species-specific leukolysis was characterized by high affinity, inhibition by MAb MM601, and relative resistance to protease K pretreatment of lymphoid cells. Two types of LKT binding to lymphoid cells are proposed. High-affinity binding leads to efficient leukolysis. In some lymphoid cells from reputed LKT-nonsusceptible species, low-affinity LKT binding may cause a low-efficiency increase in the intracellular Ca(2+) concentration without leading to leukolysis.     

Efficacy of recombinant leukotoxin in protection against pneumonic challenge with live Pasteurella haemolytica A1.

The recombinant leukotoxin (rLKT) of the bacterium Pasteurella haemolytica A1 was examined for its ability to protect cattle from experimental challenge with logarithmic-phase P. haemolytica. Six different vaccines were utilized in the experiment: P. haemolytica culture supernatant, P. haemolytica culture supernatant enriched with rLKT, rLKT alone, P. haemolytica culture supernatant enriched with Escherichia coli supernatant not containing LKT, E. coli supernatant alone, and phosphate-buffered saline. rLKT alone showed no protective capacity against development of clinical signs of respiratory disease or against development of postmortem lung lesions after experimental challenge. It was, however, shown to enhance the efficacy of the culture supernatant vaccine and decrease clinical signs and pneumonic lesions. The complexity of protective immunity in this disease is emphasized in this study, and, although LKT is an important virulence factor of the organism, an immune response to LKT alone does not protect animals against disease.